The Scrum Field Guide: Agile Advice For Your First Year And Beyond (Addison Wesley Signature Series) Guide 2nd Edition 2016

User Manual:

Open the PDF directly: View PDF PDF.
Page Count: 475 [warning: Documents this large are best viewed by clicking the View PDF Link!]

AboutThiseBook
ePUBisanopen,industry-standardformatforeBooks.However,supportofePUBand
itsmanyfeaturesvariesacrossreadingdevicesandapplications.Useyourdeviceorapp
settingstocustomizethepresentationtoyourliking.Settingsthatyoucancustomizeoften
includefont,fontsize,singleordoublecolumn,landscapeorportraitmode,andfigures
thatyoucanclickortaptoenlarge.Foradditionalinformationaboutthesettingsand
featuresonyourreadingdeviceorapp,visitthedevicemanufacturer’sWebsite.
Manytitlesincludeprogrammingcodeorconfigurationexamples.Tooptimizethe
presentationoftheseelements,viewtheeBookinsingle-column,landscapemodeand
adjustthefontsizetothesmallestsetting.Inadditiontopresentingcodeand
configurationsinthereflowabletextformat,wehaveincludedimagesofthecodethat
mimicthepresentationfoundintheprintbook;therefore,wherethereflowableformat
maycompromisethepresentationofthecodelisting,youwillseea“Clickheretoview
codeimage”link.Clickthelinktoviewtheprint-fidelitycodeimage.Toreturntothe
previouspageviewed,clicktheBackbuttononyourdeviceorapp.
THESCRUMFIELDGUIDE
SECONDEDITION
AgileAdviceforYourFirstYearandBeyond
MitchLacey
Boston•Columbus•Indianapolis•NewYork•SanFrancisco•Amsterdam•CapeTown
Dubai•London•Madrid•Milan•Munich•Paris•Montreal•Toronto•Delhi•Mexico
City
SaoPaulo•Sidney•HongKong•Seoul•Singapore•Taipei•Tokyo
Manyofthedesignationsusedbymanufacturersandsellerstodistinguishtheirproducts
areclaimedastrademarks.Wherethosedesignationsappearinthisbook,andthe
publisherwasawareofatrademarkclaim,thedesignationshavebeenprintedwithinitial
capitallettersorinallcapitals.
Theauthorandpublisherhavetakencareinthepreparationofthisbook,butmakeno
expressedorimpliedwarrantyofanykindandassumenoresponsibilityforerrorsor
omissions.Noliabilityisassumedforincidentalorconsequentialdamagesinconnection
withorarisingoutoftheuseoftheinformationorprogramscontainedherein.
Forinformationaboutbuyingthistitleinbulkquantities,orforspecialsalesopportunities
(whichmayincludeelectronicversions;customcoverdesigns;andcontentparticularto
yourbusiness,traininggoals,marketingfocus,orbrandinginterests),pleasecontactour
corporatesalesdepartmentatcorpsales@pearsoned.comor(800)382-3419.
Forgovernmentsalesinquiries,pleasecontactgovernmentsales@pearsoned.com.
ForquestionsaboutsalesoutsidetheU.S.,pleasecontactinternational@pearsoned.com.
VisitusontheWeb:informit.com/aw
LibraryofCongressControlNumber:2015955194
Copyright©2016MitchellG.Lacey
Allrightsreserved.PrintedintheUnitedStatesofAmerica.Thispublicationisprotected
bycopyright,andpermissionmustbeobtainedfromthepublisherpriortoanyprohibited
reproduction,storageinaretrievalsystem,ortransmissioninanyformorbyanymeans,
electronic,mechanical,photocopying,recording,orlikewise.Forinformationregarding
permissions,requestformsandtheappropriatecontactswithinthePearsonEducation
GlobalRights&PermissionsDepartment,pleasevisitwww.pearsoned.com/permissions/.
ISBN-13:978-0-13-385362-9
ISBN-10:0-13-385362-4
TextprintedintheUnitedStatesonrecycledpaperatRRDonnelleyinCrawfordsville,
Indiana.
Firstprinting,December2015
Thisbookisdedicatedtotwoteams;Thefirstteamismyfamily.
Mywife,Bernice,andmykids,Ashley,Carter,andEmma—withouttheir
supportandconstantlyasking“areyoudoneyet?”thisbookwouldnot
behere.Theykeptmefocusedandsupportedmethroughout.
ThesecondteamisthegroupofguysfromtheFalconprojectwhile
atMicrosoft.JohnBoal,DonavanHoepcke,BartHsu,MikePuleio,
MonLeelaphisut,andMichaelCorrigan(ourboss),thankyouforhaving
thecouragetoleapwithme.Youguysmadethisbookareality.
Contents
ForewordbyJeffSutherland
ForewordbyKennethS.Rubin
Preface
Acknowledgments
AbouttheAuthor
Chapter1Scrum:Simple,NotEasy
TheStory
Scrum
WhatIsScrum?
ImplementingScrum
WhenIsScrumRightforMe?
ChangeIsHard
KeystoSuccess
References
PARTIGETTINGPREPARED
Chapter2GettingPeopleonBoard
TheStory
TheModel
ChangeTakesTime
EstablishaSenseofUrgency
FormaPowerfulGuidingCoalition
CreateaVision/PaintaPictureoftheFuture
CommunicatetheVision
EmpowerOtherstoActontheVision
PlanforandCreateShort-TermWins
ConsolidateImprovements
InstitutionalizeNewApproaches
KeystoSuccess
BePatient
ProvideInformation
References
WorkConsulted
Chapter3UsingTeamConsultantstoOptimizeTeamPerformance
TheStory
TheModel
EstablishingaTeamConsultantPool
BuildingYourTeam
KeystoSuccess
Accountability
Experiment
BeCautiousofOverloading
PlanforPotentialDowntime
TeamConsultantsAreNotaReplacementforDedicatedTeams
Reference
WorksConsulted
Chapter4PredictingTeamVelocity
TheStory
TheModel
TheProblemwithHistoricalData
SheddingLightonBlindEstimation
WaitandSee(UseRealData)
TruncatedDataCollection
KeystoSuccess
References
Chapter5ImplementingtheScrumRoles
TheStory
TheModel
ChoosingRoles
MixingRoles
When,NotIf,YouDecidetoMixRolesAnyway
KeystoSuccess
Chapter6DeterminingSprintLength
TheStory
TheModel
ProjectDuration
ProductOwnerandStakeholders
ScrumTeam
DeterminingYourSprintLength
BeWarned
BeyondtheQuiz
KeystoSuccess
SprintsLongerthanOneMonth
ExtendingSprintLength
Reference
Chapter7HowDoYouKnowYou’reDone?
TheStory
TheModel
Introduction
Brainstorming
Categorization
SortingandConsolidation
CreationandPublishing
Whatabout“Undone”Work?
KeystoSuccess
References
Chapter8TheCaseforaFull-TimeScrumMaster
TheStory
TheModel
KeystoSuccess
RemoveImpediments/ResolveProblems
BreakUpFights/ActasTeamMom
ReportTeamData
FacilitateandHelpOutWhereNeeded
EducatetheOrganizationandDriveOrganizationalChange
InSummary
References
WorkConsulted
PARTIIFIELDBASICS
Chapter9WhyEngineeringPracticesAreImportantinScrum
TheStory
ThePractices
ImplementingTest-DrivenDevelopment
Refactoring
ContinuousIntegrationandMoreFrequentCheck-Ins
PairProgramming
AutomatedIntegrationandAcceptanceTests
KeystoSuccess
NotaSilverBullet
StartingOut
GettheTeamtoBuyIn
DefinitionofDone
BuildEngineeringintoProductBacklog
GetTraining/Coaching
PuttingItTogether
References
WorksConsulted
Chapter10CoreHours
TheStory
TheModel
ColocatedTeams
DistributedandPart-TimeTeams
KeystoSuccess
Chapter11ReleasePlanning
TheStory
TheModel
SketchaPreliminaryReleasePlan
ConsiderDegreeofConfidence
MaintainingtheReleasePlanthroughouttheProject
ProjectCost
DeterminingtheEndGame
KeystoSuccess
CommunicateUpFrontandOften
UpdatetheReleasePlanafterEverySprint
TrytoDotheHighest-PriorityItemsFirst
RefineEstimatesonBiggerItems
DeliverWorkingSoftware
Reference
Chapter12DecomposingStoriesandTasks
TheStory
TheModel
SettingtheStage
StoryDecomposition
TaskDecomposition
KeystoSuccess
References
WorksConsulted
Chapter13KeepingDefectsinCheck
TheStory
TheModel
KeystoSuccess
AdditionalInformation
References
WorkConsulted
Chapter14SustainedEngineeringandScrum
TheStory
TheModel
DedicatedTimeModel
DataGatheredoverTime
DedicatedTeamModel
KeystoSuccess
CycleDedicatedMaintenanceTeamMembers
RetrofitLegacyCodewithGoodEngineeringPractices
IntheEnd
References
Chapter15TheSprintReview
TheStory
TheModel
PreparingfortheMeeting
RunningtheMeeting
KeystoSuccess
TakeTimetoPlan
DocumentDecisions
AskforAcceptance
BeBrave
WorksConsulted
Chapter16Retrospectives
TheStory
ThePractice
GiveRetrospectivesTheirDueDiligence
PlananEffectiveRetrospective
RuntheRetrospective
KeystoSuccess
ShowThemtheWhy
BuildaGoodEnvironment
HoldThemWhenYouNeedThem
TreatRetrospectivesLiketheFirst-ClassCitizensTheyAre
References
PARTIIIFIRSTAID
Chapter17FacilitatingaProductiveDailyScrum
TheStory
TheModel
TimeofDay
StartandEndonTime
ExposeHiddenImpediments
EndwiththeBeginninginMind
KeystoSuccess
KeeptheMeetingCadence
Stand—Don’tSit
WorkasaTeam
BePatient
Chapter18TheFourthQuestioninScrum
TheStory
TheModel
KeystoSuccess
Reference
Chapter19KeepingPeopleEngagedwithPairProgramming
TheStory
TheModel
PromiscuousPairing
Micro-Pairing
KeystoSuccess
References
Chapter20AddingNewTeamMembers
TheStory
TheModel
TheExercise
KeystoSuccess
AccepttheDropinVelocity
ChooseWisely
RiskyBusiness
References
Chapter21WhenCulturesCollide
TheStory
TheModel
KeystoSuccess
ControlYourOwnDestiny
WorkwithWhatYouHave
StaytheCourse
References
WorksConsulted
Chapter22SprintEmergencyProcedures
TheStory
TheModel
RemoveImpediments
GetHelp
ReduceScope
CanceltheSprint
KeystoSuccess
References
PARTIVADVANCEDSURVIVALTECHNIQUES
Chapter23SustainablePace
TheStory
TheModel
ShortenIterations
MonitorBurndownCharts
IncreaseTeamTime
KeystoSuccess
References
Chapter24DeliveringWorkingSoftware
TheStory
TheModel
CoreStory
NumberofUsers
StartwiththeHighestRiskElement
ExpandandValidate
KeystoSuccess
ChangeinThinking
Rework
FocusonEnd-to-EndScenarios
WorkConsulted
Chapter25OptimizingandMeasuringValue
TheStory
TheModel
Stories
Taxes
Spikes
TechnicalDebt
OtherPotentialCategories
StructuringtheData
UsingtheData
KeystoSuccess
EducateStakeholders
WorkwithStakeholders
DetermineTrendsandPatterns
Reference
WorksConsulted
Chapter26Up-FrontProjectCosting
TheStory
TheModel
FunctionalSpecifications
UserStories
EstimatingStories
PrioritizingStories
DeterminingVelocity
DerivingCost
BuildtheReleasePlan
KeystoSuccess
References
Chapter27DocumentationinScrumProjects
TheStory
TheModel
WhyDoWeDocument?
WhatDoWeDocument?
WhenandHowDoWeDocument?
DocumentinginanAgileProject
StartingProjectswithoutExtensiveDocumentation
KeystoSuccess
References
Chapter28OutsourcingandOffshoring
TheStory
TheModel
ConsidertheTrueCosts
DealingwithReality
KeystoSuccess
ChoosetheRightOffshoreTeam
AllocatetheWorkintheLeastPainfulWay
StickwiththeScrumFramework
BuildaOne-TeamCulture
BePreparedtoTravel
HaveaProject/TeamCoordinator
NeverOffshoreWhen…
References
WorkConsulted
Chapter29PrioritizingandEstimatingLargeBacklogs—TheBigWall
TheStory
TheModel
Team
Stakeholders
KeystoSuccess
PreplanningIsEssential
FocusDiscussionsandSetTimeLimits
UseaParkingLotforUnresolvableDisagreements
BringExtraCards/PaperforStoriesCreatedintheRoom
TheBigWallwithRemote/DistributedTeams
RemindEveryoneThatThingsWillChange
References
Chapter30WritingContracts
TheStory
TheModel
TraditionalContractsandChangeOrders
Timing
RangesandChanges
KeystoSuccess
CustomerAvailability
AcceptanceWindow
Prioritization
TerminationClauses
Trust
References
PARTVWILDERNESSESSENTIALS
Chapter31DrivingtoDonethroughCollaboration
TheStory
TheModel
TaskPoker
PairProgramming
LimitOpenWorkItems
Two-WeekSprints
CreatingVisibilitywithaTaskBoard
KeystoSuccess
EachVoiceIsHeard
SharedUnderstandingoftheWork
EveryTeamMemberIsInvestedintheOutcome
Don’tAverageTaskEstimates
AvoidExcessivelyGranularTaskEstimates
ScrumIsBuiltonTeamwork
References
Chapter32HowStoryPointsRelatetoHours
TheStory
TheModel
FearFactor
WideRanges
It’sHow,NotHowLong
KeystoSuccess
CollectingtheRightData
UsingDatatoMakeImprovements
REFLECTonStoryPoints
References
Chapter33ImmersiveInterviewingandHiring
TheStory
TheModel
Forecast
HirefortheRightReasons
CostofaBadHire
Skills,Competencies,orBoth
HowtoHire
CandidateScreening
KeystoSuccess
BuildaRepeatableHiringProcess
FocusonCompetencies,NotonQuestions
SkillsAreEasytoLearn,CompetenciesAreNot
FindPeopleStrongerthanYou
UnderstandtheCostsandInvestHeavily
References
WorkConsulted
Chapter34AligningIncentiveswithOutcomes
TheStory
TheModel
SettheFocus
AlignaroundCustomerSatisfaction
PrioritizeandShift
SideBenefits
KeystoSuccess
IntegrateSalesandDevelopment
StopSacrificingPeopleandQuality:HaveaPrioritizedProjectPortfolio
AddDedicatedTeams,NotRandomPeople
Organization-wideCoordination
ExecutiveSupport
Reference
Chapter35RiskManagementinScrum
TheStory
TheModel
CustomerRisk:ProductOwner
SocialRisk:ScrumMaster
TechnicalRisk:Development(Core)Team
IntheEnd…
KeystoSuccess
LetGo
LetAgile
WorksConsulted
AppendixScrumFramework
TheRoles
ScrumMaster
ProductOwner
DevelopmentTeam
TheArtifacts
TheProductBacklog
TheSprintBacklog
TheBurndown
TheMeetings
PlanningMeeting
DailyScrum
SprintReview
SprintRetrospective
PuttingItAllTogether
Index
Foreword
byJeffSutherland
MitchandIhaveworkedtogetherformanyyearstrainingdevelopersinScrum.Studying
thisbookcanhelpusersovercomethebiggestchallengesthathaveoccurredinrecent
yearsasagilepractices(75percentofwhichareScrum)havebecometheprimarymodeof
softwaredevelopmentworldwide.
TenyearsaftertheAgileManifestowaspublished,someoftheoriginalsignatoriesand
alargergroupofagilethoughtleadersmetatSnowbird,Utah,thistimetodoa
retrospectiveontenyearsofagilesoftwaredevelopment.Theycelebratedthesuccessof
theagileapproachtoproductdevelopmentandreviewedthekeyimpedimentstobuilding
onthatsuccess.Andtheycametounanimousagreementonfourkeysuccessfactorsfor
thenexttenyears.
1.Demandtechnicalexcellence.
2.Promoteindividualchangeandleadorganizationalchange.
3.Organizeknowledgeandimproveeducation.
4.Maximizevaluecreationacrosstheentireprocess.
Let’sseehowMitch’sbookcanhelpyoubecomeanagileleader.
DemandTechnicalExcellence
ThekeyfactordrivingtheexplosionoftheInternetandtheapplicationsonsmartphones
hasbeendeployingapplicationsinshortincrementsandgettingrapidfeedbackfromend
users.Thisprocessisformalizedinagilitybydevelopingproductsinshortsprints,always
amonthorlessandmostoftentwoweeksinlength.WeframedthisissueintheAgile
Manifestobysayingthat“wevalueworkingsoftwareovercomprehensive
documentation.”
TheTenYearAgileRetrospectiveoftheManifestoconcludedthatthemajorityofagile
teamsarestillhavingdifficultydevelopingproductsinshortsprints(usuallybecausethe
management,thebusiness,thecustomers,andthedevelopmentteamsdonotdemand
technicalexcellence).
Engineeringpracticesarefundamentaltosoftwaredevelopment,and17percentof
ScrumteamsimplementScrumwithXPengineeringpractices.ThefirstScrumteamdid
soin1993beforeXPwasevenborn.Itisonlycommonsensetoprofessionalengineers.
MitchsaysinthefirstchapterthatheconsiderscertainXPpracticestobemandatory—
sustainablepace,collectivecodeownership,pairprogramming,test-drivendevelopment,
continuousintegration,codingstandards,andrefactoring.Thesearefundamentalto
technicalexcellence,andthe61percentofagileteamsusingScrumwithoutimplementing
thesepracticesshouldstudyMitch’sbookcarefullyandfollowhisguidance.Neglectingto
usethesemandatoryXPpracticesisthereasontheydonothaveshippablecodeattheend
oftheirsprints!
Mitch’sbookcontainsmuchmoreguidanceontechnicalexcellence,andagileleaders,
whethertheyareinmanagementorengineering,needtodemandthetechnicalexcellence
thatMitcharticulatessowell.
PromoteIndividualChangeandLeadOrganizationalChange
Agileadoptionrequiresrapidresponsetochangingrequirementsalongwithtechnical
excellence.ThiswasthefourthprincipleoftheAgileManifesto—“respondtochangeover
followingaplan.”However,individualsadaptingtochangeisnotenough.Organizations
mustbestructuredforagileresponsetochange.Ifnot,theypreventtheformationof,or
destroy,high-performingteamsbecauseoffailuretoremoveimpedimentsthatblock
progress.
MitchstepsthroughtheHarvardBusinessSchoolkeysuccessfactorsforchange.A
senseofurgencyisneeded.Changeisimpossiblewithoutit.Agileleadersneedtoliveit.
Aguidingcoalitionforinstitutionaltransformationisessential.Agileleadersneedtomake
suremanagementiseducated,trained,onboard,andparticipatingintheScrum
implementation.
Creatingavisionandempoweringothersisfundamental.Arbitrarydecisionsand
commandandcontrolmandateswillkillagileperformance.Agileleadersneedtoavoid
thesedisastersbyplanningforshort-termwins,consolidatingimprovements,removing
impediments,andinstitutionalizingnewapproaches.Agileleadersneedtobepartof
managementormusttrainmanagementaswellasengineering,andMitch’sbookcanhelp
youseewhatyouneedtodoandhowtodoit.
OrganizeKnowledgeandImproveEducation
Alargebodyofknowledgeonteamsandproductivityisrelativelyunknowntomost
managersandmanydevelopers.Mitchtalksabouttheseissuesthroughoutthebook.
SoftwareDevelopmentIsInherentlyUnpredictable
FewpeopleareawareofZiv’slaw:Softwaredevelopmentisunpredictable.Thelarge
failurerateonprojectsworldwideislargelyduetolackofunderstandingofthisproblem
andtheproperapproachtodealwithit.Mitchdescribestheneedtoexpectandadaptto
constantchange.Thestrategiesinthisbookhelpyouavoidmanypitfallsandremove
manyblockstoyourScrumimplementation.
UsersDoNotKnowWhatTheyWantuntilTheySeeWorkingSoftware
Traditionalprojectmanagementerroneouslyassumesthatusersknowwhattheywant
beforesoftwareisbuilt.ThisproblemwasformalizedasHumphrey’slaw,yetthislawis
systematicallyignoredinuniversityandindustrytrainingofmanagersandprojectleaders.
Thisbookcanhelpyouworkwiththisissueandavoidbeingblindsided.
TheStructureoftheOrganizationWillBeEmbeddedintheCode
AthirdexampleofamajorproblemthatisnotgenerallyunderstoodisConway’slaw:The
structureoftheorganizationwillbereflectedinthecode.Atraditionalhierarchical
organizationalstructurenegativelyimpactsobject-orienteddesign,resultinginbrittle
code,badarchitecture,poormaintainabilityandadaptability,alongwithexcessivecosts
andhighfailurerates.MitchspendsalotoftimeexplaininghowtogettheScrum
organizationright.Listencarefully.
MaximizeValueCreationAcrosstheEntireProcess
Agilepracticescaneasilydoubleortripletheproductivityofasoftwaredevelopmentteam
iftheproductbacklogisreadyandsoftwareisdoneattheendofasprint.Thisheightened
productivitycreatesproblemsintherestoftheorganization.Theirlackofagilitywill
becomeobviousandcausepain.
LackofAgilityinOperationsandInfrastructure
Assoonastalentandresourcesareappliedtoimproveproductbacklog,theflowof
softwaretoproductionwillatleastdouble—andinsomecasesbefivetotentimeshigher.
Thisexposesthefactthatdevelopmentoperationsandinfrastructurearecrippling
productionandmustbefixed.
LackofAgilityinManagement,Sales,Marketing,andProduct
Management
Atthefrontendoftheprocess,businessgoals,strategies,andobjectivesareoftennot
clear.Thislackofclarityresultsinaflatordecayingrevenuestreamevenwhen
productionofsoftwaredoubles.
Forthisreason,everyoneinanorganizationneedstobeeducatedaboutandtrainedon
howtooptimizeperformanceacrossthewholevaluestream.Agileindividualsneedto
leadthiseducationalprocessbyimprovingtheirabilitytoorganizeknowledgeandtrain
thewholeorganization.
TheBottomLine
ManyScrumimplementationsmakeonlyminorimprovementsandfinditdifficultto
removeimpedimentsthatembroiltheminconstantstruggle.Workcanbebetterthanthis.
Allteamscanbegood,andmanycanbegreat!Workcanbefun,businesscanbe
profitable,andcustomerscanbereallyhappy!
Ifyouarestartingout,Mitch’sbookcanhelpyou.Ifyouarestrugglingalongtheway,
thisbookcanhelpyouevenmore.Andifyouarealreadygreat,Mitchcanhelpyoube
greater.Improvementneverends,andMitch’sinsightistrulyhelpful.
—JeffSutherland
ScrumInc.
Foreword
byKennethS.Rubin
In1988,IwasthefirstemployeehiredbyParcPlaceSystemsaftertheSmalltalkresearch
teamwasspunoutofXeroxPARC(PaloAltoResearchCenter).Ourmissionwasto
commercializetheuseofobject-orientedtechnology.Duringthoseearlydaysofthe
object-technologymovement,weoftendiscussedwritingasortofrecipebooktohelp
companiesgetstartedwithobjecttechnology.Theideawastocollectthemostimportant
situations/issuesthatwesawcompaniesencounteringandpresentthemasasetofpatterns
orrecipesintheformat,“Ifyoufindyourselfinthissituation,trydoingthefollowing….”
Weneverdidwritethatparticularbook.
Fast-forwardover20yearstotheeraofagiledevelopment,andMitchLaceyhas
writtenhisownrecipebook:afieldguideonthetopicofScrum.Init,heshareshiswealth
ofScrumexperiencewithcompaniesthataregearinguptouseScrumorthosethatare
stillinthenascentstagesofapplyingit.
IfirstmetMitchin2007,shortlyafterIbecametheveryfirstmanagingdirectorofthe
worldwideScrumAlliance.Atthattime,MitchwasalreadyaCertifiedScrumTrainer
(CST)andhadbeenapplyingScrumforsomenumberofyearsbothinsideMicrosoft,
wherehewasfirstexposedtoScrum,andlaterwithbothlargeandsmallcompaniesasa
Scrumtrainerandcoach.
IcouldtellfromourfirstmeetingthatMitchwaspassionateabouthelpingpeoplebe
successfulwithScrum.Atthatfirstencounter,hetookouthislaptopandstartedwalking
methroughdatahehadbeencollecting.Hisgoalwastoreinforceanecdotalsuccess
storieswithrealdatadrawnfromhisexperiences.Inhindsight,thisexchangewas
foreshadowingforwhatwastobecomeTheScrumFieldGuide.
Whenyoureadthisbook,youwillexperiencewhatIdidduringthat2007conversation
andinnumerousdiscussionsanddebatesIhavehadwithMitcheversince—thathehasa
keenabilitytocollectandanalyzereal-worldexperiencesandsynthesizetheminto
actionableadvice.Youwillbenefitfromthisadviceineachandeverychapter!Each
chapterbeginswithastorythatcapturestheculminationofMitch’sexperiencesona
specifictopic.Ifindthistechniquetobeveryeffective,sinceIlikeagoodstoryandMitch
isquiteaneffectivestoryteller.Butdon’tjuststopafterreadingachapter’sintrostory!
Followingeachstoryisanexpansionandinterpretationoftheconceptspresentedinthe
storythatamplifytheadvicebeingoffered.
OnethingyouwillnoticewhilereadingthisbookisthatMitchisn’ttryingtoteachyou
theory.He’stryingtosaveyourbacon.Eventhemajorsectionnamesconveythisintent:
“GettingPrepared,”“FieldBasics,”“FirstAid,”“AdvancedSurvivalTechniques,”and
“WildernessEssentials.”Atruefieldguideindeed!
Youalsohavethebenefitofreadingthesecondeditionofthisbook.Mitchhasnotbeen
sittingstillsincethefirsteditionwaspublishedin2012.Thisneweditionincludesupdated
versionsofmanyoftheoriginalchaptersaswellasfivecompletelynewchaptersinthe
newsection“WildernessEssentials.”Thesechaptersaddresstopicsasdiverseas“getting
todone,”therelationshipofstorypointstohours,techniquesforinterviewingandhiring,
howtoalignincentiveswithoutcomes,andtheall-importanttopicofmanagingrisk
duringagiledevelopment.Ifyouarealreadyfamiliarwiththefirstedition,Iamsureyou
willbenefitfromthechangesandadditionsMitchhasmadetothislatestedition!
ItisanhonortohaveMitchandTheScrumFieldGuide,SecondEdition,inthesame
bookseriesasmyEssentialScrumbook:TheMikeCohnSignatureSeries.Ihavehigh
regardfortheauthorsandbooksinthisseries.Theyallpassmylitmustest:“WouldI
recommendabookintheseriestomyclientswithoutreservations?”Iamhappytosay
thatIcandefinitelyrecommendMitch’sbook!
And,forthosereaderswhoarefamiliarwithmybook,EssentialScrum:APractical
GuidetotheMostPopularAgileProcess,Iamconfidentthatyouwillfindmybookand
thisbooktobegoodcomplementstooneanother,even(andespecially)wherethetwo
differonparticularissuesorapproaches.And,apparently,manyothersagree!Aquick
lookonAmazon.comatthefirsteditionofTheScrumFieldGuideshowsthatthebook
mostfrequentlypurchasedalongwithitismyEssentialScrumbook.So,likefinewine
andcheese,Ihopeyouenjoythepairing!
—KennethS.Rubin
ManagingPrincipal,Innolution,LLC,
andauthorofEssentialScrum:
APracticalGuidetotheMostPopularAgileProcess
Preface
WelcometotheSecondEditionofTheScrumFieldGuide
WhenImentionedthatIwasthinkingofupdatingtheoriginalFieldGuide,mywife
questionedmysanity.Afterall,sheremindedme,thefirstbookhadnearlydonemein.
YetasIreflectedonmyfirsteffort,IfeltthatnotonlydidIhavemoretosay,butIalso
wantedtoadjustsomeofthecontentI’dalreadypublished.Putsimply,Iwantedto
refactor,addsomenewfeatures,andreleasea2.0version.Sohereweare.
Youshouldstillreadthisbookinthesamewayyoudidthefirstedition:Pickachapter
thataddressesaproblemyouseeinyourcompany—andreadit.Thengoapplymyadvice
andseewhathappens.
Agileisajourney.I’velearnedagreatdealsincethefirsteditionwaspublishedin
2012.Ifyouhavereadthisbookbefore,youwillimmediatelyseethatIhaveaddednew
ideasandconceptstotheoriginalchapters.Manychaptershavebeenrewrittenbymore
than80percent;othersbyonly10percent.Youwillseeanewsection:PartV,
“WildernessEssentials,”containingadditionaltipsforthefield,inspiredbymyfirsthand
experienceworkingwithorganizationsaroundtheglobe.Thesenewchaptersinclude
managingrisks,interviewing,thefallacyofgettingitrightthefirsttime,andmore.
HowThisBookCametoBe
WhenmydaughterEmmawasborn,Ifeltoutofmydepth.Weseemedtobeatthe
doctor’sofficemuchmorethanwehadbeenwithourotherchildren.Ikeptaskingmy
wife,“Isthisnormal?”Onenight,Ifoundmywife’scopyofWhattoExpecttheFirstYear
onmypillowwithanotefromher:“Readthis.You’llfeelbetter.”
AndIdid.Knowingthateverythingwewereexperiencingwasnormalformychild,
evenifitwasn’ttypicalformeorobservedbefore,mademefeelmoreconfidentand
secure.ThiswasaroundthesametimeIwasstartingtoexperimentwithScrumandagile.
AsIencounteredobstaclesandranintounfamiliarsituations,IbegantorealizethatwhatI
reallyneededwasaWhattoExpectbookforthefirstyear(andevenbeyond)ofScrum
andXP.
Theproblemis,unlikeaWhattoExpectbook,Ican’ttellyouexactlywhatyourteam
shouldbedoingorworryingaboutduringmonths1to3or9to12.Teams,unlikemost
children,don’tdevelopatapredictablerate.Instead,theyoftentumble,stumble,and
bumbletheirwaythroughtheirfirstyear,takingtwostepsforwardandonestepbackas
theylearntofunctionasateam,adoptagileengineeringpractices,buildtrustwiththeir
customers,andworkinanincrementalanditerativefashion.
Withthisinmind,Ichosetostructurethisbookwithmoreofan“I’vegotapainhere,
whatshouldIdo”approach.I’vecollectedstoriesaboutteamsI’vebeenapartofor
witnessedintheirfirstyearofagilelife.AsIcontinueddownmyagilepath,Inoticedthat
thestoriesandthepatternsincompanieswereusuallysimilar.Iwouldimplementanidea
inonecompanyandtweakitforthenext.Inrepeatingthisprocess,Iendedupwitha
collectionofreal-worldsolutionsthatInowcarryinmyvirtualtoolbelt.Inthisbook,I
sharewithyousomeofthemostcommonpainsandsolutions.Whenyourteamishurting
orintrouble,youcanturntothechapterthatmostcloselymatchesyoursymptomsand
find,ifnotacure,atleastawaytorelievethepain.
TheScrumFieldGuide,SecondEdition,ismeanttohelpyoufine-tuneyourown
implementation,navigatesomeoftheunfamiliarterrain,andmoreeasilyscalethehurdles
weallencounteralongtheway.
WhoShouldReadThisBook
IfyouarethinkingaboutgettingstartedwithScrumoragile,areatthebeginningofyour
journey,orhavebeenatitayearorsobutfeellikeyou’vegottenlostalongtheway,this
bookisforyou.I’mofficiallytargetingcompaniesthatarewithinsixmonthsofstartinga
projecttothosethatareayearintotheirimplementation—an18-monthwindow.
Thisisabookforpeoplewhoarepragmatic.Ifyouarelookingfortheoryandesoteric
discussions,grabanotherofthemanyexcellentbooksonScrumandagile.If,ontheother
hand,youwantpracticaladviceandrealdatabasedonmyexperiencerunningprojects
bothatMicrosoftandwhilecoachingteamsandconsultingatlargeFortune100
companies,thisbookfitsthebill.
HowtoReadThisBook
Thebookisdesignedforyoutobeabletoreadanychapter,inanyorder,atanytime.
Eachchapterstartsoutwithastorypulledfromateam,company,orprojectthatIworked
onorcoached.Asyoumightexpect,I’vechangedthenamestoprotecttheinnocent(and
eventheguilty).Onceyoureadthestory,whichwilllikelysoundfamiliarinsome
fashion,Iwalkyouthroughthemodel.ThemodeliswhatIuseinthefieldtohelpaddress
theissuesevidentinthestory.Someofthemodelsmightfeeluncomfortable,oryoumight
believetheywon’tworkforyourcompany.Iurgeyoutofighttheinstincttoignorethe
adviceortomodifythemodel.Tryitatleastthreetimesandseewhathappens.Youmight
besurprised.Attheendofeachchapter,Isummarizethekeystosuccess—thosefactors
thatcaneithermakeorbreakyourimplementation.
Thisbookisorganizedinfiveparts.
PartI,“GettingPrepared,”givesyouadviceongettingstartedwithScrum,helping
yousetupforsuccess.IfyouarejustthinkingaboutScrumorarejustbeginningto
useit,startthere.
PartII,“FieldBasics,”discussesitemsthat,onceyougetstarteddowntheagile
path,helpyouoversomeoftheinitialstumblingblocksthatteamsandorganizations
encounter.Ifyou’vegottenyourfeetwetwithScrumbutarerunningintoissues,
youmightwanttostarthere.
PartIII,“FirstAid,”iswhereIdealwithsomeofthelarger,deeperissuesthat
companiesface,suchasaddingpeopletoprojectsorfixingdysfunctionaldaily
standupmeetings.Thesearesituationsyou’lllikelyfindyourselfinatonepointor
anotherduringyourfirstyear.Thesechaptershelpyoutriageandtreatthesituation,
allowingyourteamtoreturntoahealthystate.
PartIV,“AdvancedSurvivalTechniques,”containsaseriesofitemsthatteams
seemtostrugglewithregardlessofwheretheyareintheiradoption—thingssuchas
costingprojects,writingcontacts,andaddressingdocumentationinagileandScrum
projects.
PartV,“WildernessEssentials,”containschaptersthatfocusonoverlooked,yetjust
ascostly,problemsthatmostorganizationsfacewhentheyareinthemiddleoftheir
agileadoption,suchasmanagingrisks,interviewing,gettingitrightthefirsttime,
andmore.
IfyouarestartingfromscratchandhavenoideawhatScrumis,I’veincludedashort
descriptionintheappendixtohelpfamiliarizeyouwiththeterms.Youmightalsowantto
dosomemorereadingonScrumbeforedivingintothisbook.
WhyYouShouldReadThisBook
Regardlessofwhereweareonouragilejourney,weallneedafriendlyreminderthatwhat
weareexperiencingisnormal,somesuggestionsonhowtodealwithissues,andafew
keysforsuccess.Thisbookgivesyouallthatinaformatthatallowsyoutoreadonlythe
chapteryouneed,anentiresection,orthewholething.Itsreal-lifesituationswillresonate
withyou,anditssolutionscanbeappliedbyanyteam.Turnthepageandreadthestories.
Thisfieldguidewillbecomeatrustedcompanionasyouexperiencethehighsandlowsof
ScrumandExtremeProgramming.
SupplementalMaterialforThisBook
Throughoutthisbook,youmayfindyourselfthinking,“IwishIhadatoolor
downloadabletemplatetohelpmeimplementthatconcept.”Inmanycases,youdo.Ifyou
gotohttp://www.mitchlacey.com/supplements/,youwillfindalistofvariousfiles,
images,spreadsheets,andtoolsthatIuseinmyeverydayScrumprojects.Whilesomeof
theinformationisrefined,mostofthestuffisprettyraw.Why?Formyprojects,Idon’t
needittobepretty;Ineedittobefunctionalandtowork.Whatyougetfrommywebsite
supplementswillberaw,true,andfromthetrenches.
Acknowledgments
WhenIfirsthadtheideaforthisbook,itwasraw.LittledidIknowthatIwasattempting
toboiltheocean.Mywife,Bernice,keptmegrounded,asdidmykids.Withouttheir
strength,thisbookwouldnotbeheretoday.
DavidAnderson,WardCunningham,andJimNewkirkwereallinstrumentalinhelping
meandmyfirstteamgetoffthegroundatMicrosoft.Eachofthemworkedthereatthe
timeandcoachedusthroughsomeroughperiods.Istilllookbackatmynotesfroman
earlysessionwithWard,withaquestionhighlightedthatread,“Can’twejustskipTDD?”
Eachofthesethreepeoplehelpedturnourteamofmisfitsintosomethingthatwasreally
special.David,Ward,andJim—thankyou.
IthankMikeCohnandhisauthorsforacceptingthistitleintotheMikeCohnSignature
Series.Itisquiteanhonortobeamongsomeofthebestagileauthorsontheplanet.
Icouldnothavedonethiswithoutthehelpofmywife,BerniceLacey,andthebest
editorontheplanet,RebeccaTraeger.Bothputincountlesshoursediting,keepingmeon
track,keepingmefocused,andhelpingmeturnmyrawthoughtsandwordsintocohesive
chapters.
Iwouldalsoliketoonceagainthankthefollowingfriends,eachofwhomhelpedcraft
thisbookintowhatitistoday.Everyonelistedherehasgivenmeinvaluablefeedbackand
contributedmanyhourseitherlisteningtomeformulatethoughtsorreadingearlydrafts.I
cannotthankeachofyouenough,includingTiagoAndradeeSilva,AdamBarr,artistTor
Imsland,BrentBarton,MartinBeechen,ArloBelshee,JelleBens,JohnBoal,Jedidja
Bourgeois,StephenBrudz,BrianButton,SharonButton,MikeCohn,JimConstant,
MichaelCorrigan,ScottDensmore,EstherDerby,SteinDolan,MarcFisher,Paul
Hammond,BillHanlon,ChristinaHartikainen,ChristianHassa,MartinaHiemstra,Jim
Highsmith,LizHill,DonavanHoepcke,BartHsu,WilhelmHummer,RonJeffries,Lynn
Keele,ClintonKeith,JamesKovaks,BenLinders,RockyMazzeo,SteveMcConnell,Jeff
McKenna,BrianMelton,AdeMiller,RaulMiller,JimMorris,JimNewkirk,Jacob
Ozolins,MichaelPaterson,BartPietrzak,DavePrior,PeterProvost,MichaelPuleio,Scott
Robinson,RenéRosendahl,KenSchwaber,TammyShepherd,LisaShoop,Michele
Sliger,TedSt.Clair,JeffSutherland,GaylynThompson,IsaacVaron,BasVodde,and
BradWilson.
I’dalsoliketothanktheteamatAddison-Wesley,includingElizabethRyan,ChrisZahn
andChrisGuzikowski.Iappreciateallthatyourteamdid.Also,CarolLallierandKim
Arney—copyeditorandproductionprojectcoordinator—thankyou.Youcaughtsomany
thingsthatIoverlookedandmadethisveryeasy.
Booksdon’tjustpopoutofyourheadandontopaper.They,likemostprojectsI’veever
encountered,aretrulyateameffort.ThepeopleIhavementioned(andlikelyafewthatI
forgot)havelistenedtome,toldmewhereIwasgoingastray,givenmeideasto
experimentwithonmyteamsandwithclients,andbeenthereformewhenIneeded
reviews.IimaginetheyareasgladasIamthatthesecondeditionofthisbookisfinallyin
print.Ihopethatafteryoureadthis,youtoowilljoinmeinthankingthemforhelpingto
makethisguideareality.
AbouttheAuthor
MitchLacey,founderofMitchLacey&Associates,Inc.,helpscompaniesreachfull
potentialbybuildingperformingorganizationsthroughtheadoptionofagilepractices,
includingScrumandXP.
Mitchisaself-describedtechnerdwhostartedhistechnologycareerin1991at
AccoladeSoftware,acomputergamingcompany.Afterworkingasasoftwaretest
engineer,atestmanager,adeveloper,andatavarietyofotherjobsinbetween,hesettled
onhistruecalling—projectandprogrammanagement.
Mitchwasaformallytrainedprogrammanagerbeforeaddingagiletohisprojecttool
beltin2005.HefirstpracticedagileatMicrosoftCorporation,wherehesuccessfully
releasedcoreenterpriseservicesforMSNandassistedindrivingagileadoptionacross
multipledivisions.
Mitch’sfirstagileteamwascoachedbyWardCunningham,JimNewkirk,andDavid
Anderson.HehasworkedextensivelyinallScrumrolesonavarietyofprojects.Today,
withdecadesofexperienceunderhisbelt,Mitchcontinuestodevelophiscraftby
experimentingandpracticingwithbothleadershipandprojectteamsatmanydifferent
organizations.
Mitch’srich,practicalexperienceandhispragmaticapproacharetrustedbymany
companiesincludingAdobeSystems,AeraEnergy,Bio-Rad,EchoStar,Microsoft,Oracle,
Qualcomm,SalemHospital,SAP,Sony,andmore.HeisaCertifiedScrumTrainer(CST),
aPMIProjectManagementProfessional(PMP),andanAgileCertifiedPractitioner
(ACP).
Mitchhaspresentedatavarietyofconferencesworldwide,chairedAgile2012and
Agile2014,andservedontheboardofdirectorsfortheScrumAllianceandtheAgile
Alliance.
Formoreinformation,visitwww.mitchlacey.com,whereyouwillfindMitch’sblogas
wellasavarietyofarticles,tools,andvideosthatwillhelpyouwithyourScrumandagile
adoption.HecanalsobefoundonTwitterat@mglaceyandbyemailat
mitch@mitchlacey.com.
Chapter1.Scrum:Simple,NotEasy
Scrumiselegantlydeceptive.Itisoneoftheeasiestframeworkstounderstandyetoneof
thehardestframeworkstoimplementwell.Isay“implementwell”becauseScrum’s
inherentsimplicitycanseduceteamsintothinkingitiseasytodo,wheninrealityitcan
takeyearstodoitwell.Scrumseemstogoagainstthecommonwisdomacquiredthrough
many,manyyearsofwaterfalldevelopment.Therefore,itstandstoreasonthatitwilltake
awhiletounlearnbadhabitsandadjusttoanewreality.
IntheappendixofthisbookIexplainthemechanicsofScrum.Ifyouarenotfamiliar
withScrumandhowitworks,Isuggestyoustartthere.Ifyoualreadyhavesome
backgroundonwhatScrumis,youknowthatthemechanicsarefairlystraightforward.So
straightforward,infact,thatmanymistakenlybelievethey“getit”andthinktheycan
beginimmediatelymodifyingScrumtobetterfittheirsituation.Toooftentheyinstead
findthemselveslostorwoundedandinneedofassistance,whichiswherethisbook
comesin.ThestorythatfollowsillustrateshowquicklythingsfallapartwhenScrumis
implementedwithoutunderstandingandwithoutasolidgroundinginthecoreagile
conceptsthatmakeScrumwork.
TheStory
JeffwasasanagilecoachhelpingteamsadoptScruminsidealargesoftwarecompany.
Oneday,hereceivedanemailfromSuzy,aprogrammanagerinhisdivision.
“Jeff,pleasehelp.We’vebeendoingScrumforaboutsixmonthsandourcodequality
isn’timprovingthewayI’dlike.Ithinkweneedyoutocomeandtalktousaboutpair
programming.NextMondayisthebeginningofourplanningweek.Canyoucomethen?”
Jeffsatbackinhischair.Talkingaboutpairprogrammingwasrelativelysimple.He
couldbringhisfriendJulie,agreatdeveloperandseasonedagilepractitioner.Noproblem.
Yettwowordsfromtheemailkeptrepeatinginhishead.Planningweek?Scrumcallsfor
twosprintplanningmeetingsthatlastnomorethanfourhourseach.Andthisteamtooka
week?Somethingtoldhimtheyhadmoretoworkonthanjustpairprogramming.Itwas
goingtobeaninterestingMonday.
OnMonday,JeffandJuliearrivedatthemeetingandfoundSuzyandherteamofeight
intheconferenceroomtogether.AfterSuzyintroducedthem,JeffandJuliespokebriefly
abouttheirexperience,thenbegantoaskaboutthecodequalityissuesSuzyhadtoldthem
about.
Theteamwasquicktoanswer.Theleadtester,Mike,wentfirst:“Thereasonourcode
ispoorqualityisbecausewehavenotimetotest.Thedeveloperswritethecoderightup
untilthelastdayduringourone-monthsprint.Thecodingandtestingsprintissupposedto
beforcodingandtesting.Butourtestingiseithershovedrighttotheendofthesprintorit
overflowsintotheintegrationsprint.”
Julieinterrupted.“I’msorry,Mike,didyousaytheintegrationsprint?”Shelookedat
Suzy,whonodded.
“Oh,Ididn’texplainourmodifications,didI?”saidSuzy.“See,weknowthatScrum
callsforareleaseeveryfourweeks,butthat’sjustnotpossibleforthekindofworkwedo.
Imean,beforetryingScrum,wetriedtoreleasequarterly,andthatwasacomplete
disaster!SowhatwedidwaschangeScrumtofitourprocessandrealitybetter.”Suzy
wenttothewhiteboardandbegantowrite.
“Firstwehaveoursprintplanningweek,followedbyourone-monthactualsprint,
wherethedeveloperswritethecodeandthetesterswritethetestcases.Afterthat,wedo
ourintegration,thenwedeploy.Ofcourse,Iusuallyaddaweekofbuffer,too,justin
case,”saidSuzy.
Whenshefinished,shehadjustaboutfilledupthewhiteboard.
JeffandJulielookedateachotherandthenbackatSuzy.Therestoftheteamlooked
bored.Jeffaskedtheobviousquestion:“Suzy,areyoursprintsreallyeightornineweeks
long?”
“That’sright,”saidSuzy.“Yousoundsurprised.Iknowit’snotout-of-the-boxScrum,
butitworksforus.I’mthinkingwemayneedtoaddanotherweek,youknow,forwriting
specsandtestplans.We’rehavingtroublefittingthatin.Rightnow,wedothatduringour
bufferweek,butIreallyhatetocutintoourcushion.”
“Okay,we’llcomebacktothat,”saidJulie.ShelookedatJeff,whoheldhishandsupas
iftosay,“Whatchagonnado?”Juliecontinued,“Mike,youweresayingthatyoudon’t
havetimetotest.”
WyattspokeupbeforeMikecouldanswer,“Don’tlistentoMike.Theyhaveplentyof
timetotest.We’retheoneswhoneverhaveenoughtime.Webustiteverysprinttogetas
muchcodingdoneaswecan.Sowhatifittakesusallfourweeks?That’showlongit
takes.”WyattlookedatMikeandcontinued,“Allyouandtherestofthetestershavedone
sincewestartedusingScrumiscomplainthatyouhavenotime.MaybeScrumisthe
problem.”
JeffandJulieexchangedglances.
Suzyinterjected,“Team,comeon,we’renotheretocomplainaboutScrum.We’rehere
togetourcodequalityup.”Shepausedandtookadeepbreath.“I’vebeensayingthisfor
sixmonths,”shesaidtoJeffandJuliewithaquickrollofhereyes.
Jeffnodded.“Icantellyou’refrustrated.AndIcantellWyattandMikearefrustrated.
CanIexplorethisabitwiththeteam?SeeifIcangettotherootoftheproblem?”Suzie
noddedenthusiastically.
Jeffbeganwithhisfirststandardassessmentquestion.“Okay.BothScrumandExtreme
Programming—orXP—callfordailycheckpoints.InScrumthesearethedailyscrum
meetings.Howwouldyourateyourdailyscrummeetings?”Jeffaskedtheteam.
Mikelaughed.“Dailymeeting?Areyoukidding?Wedon’thavetimeforthose.We
meettwiceaweekforanhour.Andthat’sbadenough.”
“Okay,Mike.Tellmeaboutthosemeetings,”saidJeff.
“Well,firstoff,wesaythesamethingeverysingletime.Developerssaytheyare
workingontasks,andwesaywearebuildingtestcases.Bignews,there.Thenwespend
20minutesorsotriagingourbuglist,whichboilsdowntogoingthroughthelistand
saying‘thisbugisperthedesign’or‘we’llfixthatoneinthenextsprint.’Ofcourse,we
neverdofixthem.It’sjustabigdysfunctionalmess,”saidMike.
Suzywasnowbecomingvisiblyangry,soJeffgavehertheopportunitytospeak.
“Thanks,Mike.Suzy,whatareyourthoughts?”askedJeff.
“Mike’srightaboutonething.Dailymeetingswon’tworkforus.Wehavetodothe
meetingeveryotherday.Myscheduleistoobusytomeeteveryday,andsometeam
membersareonotherprojects.Iknowit’snotideal,butit’swhatweneedtodo.What
getsmeisthattheteamfightsmeonevenhavingthemtwiceaweek—theythinkitistoo
muchoverhead.YouheardthewayMikewastalkingjustthen.Theyallcomplainabout
themeetings,theschedule,thelackoftime!ButIcan’thelpthefactthatmanagementis
pushingustoreleasemoreoften.Plus,likeIkeeptellingthem,it’smyproject.Iamsetting
theplan,andIamsettingthestructure.Tellthem,Jeff.I’mtheScrumMaster.Theyshould
dowhatIsay,right?”demandedSuzy.
Jeffgaveanoncommittalshrugandbithistongue.Hewasstartingtorealizehowlittle
theyunderstoodaboutScrum.HeglancedatJulieandgaveheroneofthosetheydon’tget
itlooks.Julieacknowledgeditwithaslightnod.Jeffcontinuedwithhisassessment.
“Okay,Ihearyou.Let’snotgetaheadofourselves.Itsoundslikeonepotentialissue
mightbethatyourdailyscrumsarenotdaily,notproductive,andlasttoolong.Wecanfix
that.Let’sskipthatfornowandbringourthoughtprocessupalevel.Tellmewhat
broughtyoutoeight-weeksprints.”
Wyattspokeup.“I’vebeenhereovertenyearsandbelieveme,I’veseeneverynewfad
thereis.Theycomeandtheygo.ButIdranktheKool-Aidthistimeandreallythought
Scrummightbedifferent.Whatajoke!Thiswholethingstartedbecausemanagementwas
pushingustoreleasemorequickly.Wehadgottendowntoquarterlyreleases—goingfrom
yearlytoquarterlyhurt,letmetellyou,butweweredoingit.Butthatwasn’tquick
enoughformanagement,wasit?”Wyattpausedandlookedaroundtheroomfor
confirmation.
Hecontinued,“So,SuzyandIweresittingatlunchonedayandhappenedtomeetup
withacolleagueofminewhoworksinanotherdivision.Hetoldushisteamwasusing
Scrumandhowtheyweredeliveringeveryfourweeksandhowhappythepeoplewere.
Hesaidtheirqualitywasthroughtheroof.Themanagementteamhadnotbeenthishappy
inyearsandthecustomerswereecstatic.Thisguywaslikeme,askeptic,youknow?SoI
thought,ifhesaysitworks,itmustwork.SuzyandIspenttheafternoonwithhim,
learningallaboutScrum.Seemedsimpleenough,butithadsomeproblems.Firstoff,
thosedailymeetings.Whohastimeforthat?Simplefixwastomakethemtwiceweekly.
Then,itwasobviousthataone-monthcyclewouldnotworkforus—afterall,wecould
barelydoaquarterlycycle,sowedecidedtodoubleitateightweeks.Fromthere,we
brokeourworkflowdownintoeightweeks.Itwasjustamatterofshrinkingeverything
down.Afterall,Scrumisjustanotherincrementalprocess.”
JeffandJulieglancedateachotheragain.
Wyattcontinued,“Iseethatlook.ButI’mtellingyou,weknowourteamandour
product.There’snowayScrumwouldworkforusthewayitcomesoutofthebox.So,we
didwhatanysoftwareteamwoulddo—wecustomizedittosuitourneeds.Thewaywe
haveitstructuredfitsbestwithhowwe’vedonethingsinthepast.”
“Right,”confirmedSuzy.“Imean,Scrumisjustawayforprojectmanagerstomanage
thework,onlyshorter.”
Jeffsatback.Theassessmentquestionshehadpreparedweren’tdesignedforasituation
likethis.Hewasn’tsurewhattosayatthispoint.Juliesteppedintohelp,“Doyouhavea
commondefinitionofdone,Wyatt?”
“Surewedo.Wehaveourdesignreviewmeetingafterthefirstweek.Then,attheend
ofweekfive,wehaveourcodecompletemilestone.Attheendofintegration,wearetest
andintegrationcomplete.Oncewehitthesemilestones,wereleasetothelivesite.It’s
reallynotthathardtounderstand,”saidWyatt.
“No,it’ssimple.Igetit,”saidJuliereassuringly.“So,team,Wyatt,Mike,andSuzy
haveexplainedyourprocess.ThewayIunderstandityouhavetwice-weeklydaily
scrums,aneight-orsometimesnine-weeksprint,andhavecheckpointsatcertainkey
stagestoletyouknowyouaredone.Howisitworkingforyou?Areyouhavingfun?Is
yourcodequalityalotbetter?”askedJulie.
“Well,it’snotbad,”saidWyatt.
“Notbad?”saidSuzy.“It’sawful.”
“Well,it’snotourfaultit’sawful!Wetrytotest,but,likeIsaid,there’sjustnotime!”
screamedMike.Therestoftheteamstaredatthetable.Theyweren’tgettinginvolvedin
thisfight.
“I’mnotblamingyou,Mike,thoughIcould.Nope.TherealproblemisScrum,”said
Wyatt.“It’sastupidprocessanditdoesn’twork.”
“Notthatagain,”saidSuzy.“Howmanytimesdowehavetohavethisargument?We
gothroughthisateveryretrospective.”
“Retrospective?”interjectedWyatt.“That’sjustafancynameforatwo-daybitch
session.Retrospectivesdon’tchangeanything.Scrumdoesn’tchangeanything.Itakethat
back.Itdoeschangeonething—itmakesusallmiserable.”
“Wyatt,thiswasyouridea,too.WhydidyousignuptodoScrumifallyouweregoing
todowassabotageit?”askedSuzyheatedly.
Jeffstoodup.Itwastimetostopaskingquestionsandstartgettingthisteamtofacethe
truth.
“Look,it’snotanyone’sfaultandit’snotScrum’sfaulteither.Tome,andI’msureJulie
agrees,itsoundslikeyou’renotreallydoingScrum.Whatyou’redoingiswhatyou’ve
alwaysdone,onlysqueezedintoaneight-weekcycle.You’rejustcallingitScrum.”
WyattandSuzystartedtoargue,butJulieputherhanduptostopthem.“Letmeaskone
question.Thisisforthewholeteam.Wyatt,Suzy,Mike,pleasedon’tanswer.”Julie
lookeddownthetableateachteammemberbeforecontinuing,“Hasthisnewwayof
workingmadeyourlivesmiserableorwasitalwayslikethis,butjustmaybenotas
obvious?”askedJulie.
Everyonelookeddown,theirchinstotheirchestsindeepthought.
“Itsuckedbefore,too,”saidavoicefromtheback.
“Yeah,”saidanotherteammember.“Ijustneverknewhowmuchitsuckedbefore.”
Therewasadeadsilenceintheroomasself-realizationstartedtodawnonthem.
Wyattsighedandsaid,“You’reright.Itwasnotanybetterbefore;itjustwasn’tas
obvious.Weonlyfeltthepaineveryquarter.Nowwefeelitevery8weeks.”
Mikechimedin,“Youknow,lookingbackonthelastsixmonths,wehaveuncovereda
lotofstuffthatwecouldandshouldfix,butwehaven’tdoneanythingaboutit.Not
really.”
Suzyinterrupted.
“Guys,I’mreallytired.Canwepostponethisforafewdays,reconvenenextweek?”
sheasked.
Theteamnodded.Theyweretired.
Suzyleftthemeetingknowingthingswerebad.Shespenttheweekendandtheearly
partofthenextweekthinkingabouthowtomovethingsforward.SheinvitedJeffand
Juliebackforanothermeeting,oneshehopedwouldsetanewtonefortheteam.
“I’msorry,everyone,”shesaidtoopenupthemeeting.“Iknewthingswererough,but
Ididn’tknowitwasthisbad.IoriginallyaskedJeffandJulietohelpuswithpair
programmingbecauseIthoughtthatwouldfixourqualityissues.ClearlyIhavebeen
blindtoourrealneedsandforthatIapologize.We’vegoneaboutthisallwrong.It’snot
Scrumthathasfailed;wehavefailedtouseScrum.I’dliketoaskyouallifwecanstart
over.IthinkJeffandJuliecanhelpusdothat,”saidSuzy.
Wyattnoddedandlookedattheteam,“IknowIcanbeajerk.I’vebeenheresolong
sometimesIthinkIowntheplace,butIdon’t.IknowIcangetbetter.I’llstop
complainingandgiveitarealchancetowork,butonlyononecondition.”
“What’sthat?”askedJeff.
“Thatwedoitforreal,”saidWyatt.“Nomorecustomizations.Andweneedacoach,
someonewhocanshowuswhattodo.Thishasn’tbeennearlyaseasyasitlooked.”
Mikelookedup.“Ihaveacondition,too.Wefixtheproblems.Reallyfixthem,not
blameeachotherforthem.”HeheldoutahandtoWyatt.“Thinkwecandoit?”
WyattlookedatMike,lookedathishand,andthenshookit,“Ithinkweareupforthe
challenge.IfJeffwillstillhelpus,thatis,”hesaidplayfully,raisinghiseyebrowsatJeff.
Everyonelaughed—thefirstlaughthey’dhadasateaminquiteawhile.Itwasastart.
Theywentaroundthetable,allverballycommittingtotryingScrum,forrealthistime.
JeffandJulieleftthemeetingashorttimelater,feelinglikethey’daccomplishedquitea
lotbutknowingtherewasmuchmorelefttodo.
“Whatareyougoingtodonow,Jeff?”askedJulie.
“ThefirstthingI’mgoingtodoistoteachthemwhatScrumis—itsvalues,the
framework,theshiftinmindsetthattheyaregoingtoneedtogothrough,”saidJeff.
“Don’tforgettoshowthemhowitmanagesriskandhelpsidentifytheproblems,”said
Julie.
“Exactly.I’llstartwiththebasicsandworkmywaythroughtheobstaclesastheypop
up.It’llbeastruggleattimes,buttheycandoit.Nopain,nogain,right?”
Scrum
Scrumseemsprettybasic.Butwhatmanypeopledon’tunderstandisthat,todoitwell,
youmustfundamentallychangethewayyoudevelopsoftware.Andthat’snoteasy.
You’regoingtostruggle.You’regoingtoencounterobstacles.Theteaminthischapter’s
storyfoundthatoutthehardway,andifyou’vepickedupthisbook,you’veprobably
discoveredittoo.Tohelpyouunderstandhowsomethingsosimplecanbesohard,the
nextsectionlooksmorecloselyatScrum.
WhatIsScrum?
JeopardyhasalwaysbeenoneofmyfavoriteTVgameshows.Ikeepwishingforthe
softwaredevelopmenteditionwithcategorieslikeMethodologiesandFrameworks,
CommonCausesofSoftwareFailure,FamousSoftwareArchitects,orStupidQuotesfrom
SmartPeople.Icanthinkofdozensofquestionsthatwouldfillthesecategories,suchas,
“Hesaid,‘Benicetonerds.Chancesareyou’llendupworkingforone.’”Ienvisiona
questionunderMethodologiesandFrameworkstobe,“Namedafterarugbyterm,this
projectmanagementframeworkdeliversworkingsoftwareeverytwotofourweeks.”One
commonanswerthatwouldlikelybeaccepted,intheformofaquestionofcourse,would
be“Whatis…Scrum?”
SowhatisScrumreally?Scrumisnotamethodologyorsetofengineeringpractices.It
is,instead,alightweightframeworkdesignedtomanagesoftwareandproduct
development.KenSchwaberandJeffSutherland[SCHWABER01]describeScrumlike
this:
Scrum(n):Aframeworkwithinwhichpeoplecanaddresscomplex,adaptive
problems,whileproductivelyandcreativelydeliveringproductsofthehighest
possiblevalue.Scrumis:
Lightweight
Simpletounderstand
Extremelydifficulttomaster
Scrumisaprocessframeworkthathasbeenusedtomanagecomplexproduct
developmentsincetheearly1990s.Scrumisnotaprocessoratechniqueforbuilding
products;rather,itisaframeworkwithinwhichyoucanemployvariousprocessesand
techniques.Scrummakescleartherelativeefficacyofyourproductmanagementand
developmentpracticessothatyoucanimprove.
Scrumreliesonafixedcadenceofiterativecyclescalledsprints.Eachsprintbegins
withaplanningmeetingandendswithademonstrationofpotentiallyshippableproduct.
Scrumischaracterizedbyahighleveloffeedbackandtransparency,bothwithintheteam
andoutsideit.Itsshortcyclesandcollaborativenaturemakeitidealforprojectswith
rapidlychangingand/orhighlyemergentrequirements.
Scrumisbuiltonfivecorevaluesandhasthreedistinctroles,threeartifacts,andthree
(orfour)meetings.MoredetailonthemechanicsofScrummaybefoundintheappendix,
“ScrumFramework.”
ImplementingScrum
WhileScrummightappearsimpletoimplement,itisactuallychallenging.Why?Because
doingsorequiresmorethanjustputtingtherightmechanicsintoplaceandthenpressinga
button.ImplementingScrumcorrectlyrequiresteamstobewillingtomakethefollowing
changes:
DevelopanunderstandingofScrum’sunderlyingvalues.
Undergoanoftenenormousshiftinmindset.
Planforchangetooccur—andadaptwhenitdoes.
Dealwithnewlyexposedoremergingissues.
Incorporateagileengineeringpractices.
ScrumIsBuiltonValues
Anyframeworkworthusingisbuiltonprinciplesandvalues.Eachoftheoriginalagile
practices—XP,Scrum,dynamicsystemsdevelopmentmethod(DSDM),Crystal,and
feature-drivendevelopment(FDD)—aswellasKanbanandLean,hasasetofcorevalues.
Thesevaluesguidedecisionmaking,provideclarityintimesofambiguity,andmost
importantly,communicatewhyagileteamsdowhattheydo.Asyoureadinthischapter’s
story,theteammemberswereattemptingtogothroughthemotionsofScrum,buttheydid
notunderstandwhy.Whattheylackedwasaclearunderstandingofthevaluesbehindthe
Scrumframework:focus,respect,commitment,courage,andopenness[SCHWABER02].
Focus
Tofocusmeanstoconcentrate,todirectattentiononsomething.InScrum,theteammust
havefocusifitistoaccomplisheverythingthatneedstobedonetodeliverapotentially
releasableincrementoffunctionality.Focusmeansworkingononeprojectatatime.It
maymeanhavingdedicated“teamtime,”wheretheentireteamisoffemail,instant
messaging,mobilephones,andmeetings.Focusisdoingwhateverittakestoallowthe
teamtoconcentrateonthedeliveryathandfortheentirelengthofanygivensprint.
Respect
Asthesayinggoes,respectisearned,notgranted.InScrum,thisisespeciallytrue.
Respectforone’steammates,orlackthereof,canmakeorbreakaproject.High-
performingScrumteamstrusteachotherenoughtoadmittoobstacles.Theyhavefaith
thatwhenoneteammembercommitstoatask,thatteammemberwillfollowthrough.
ThereisnousversusthemonatrueScrumteam.Inthestorythatopenedthischapter,
testersanddeveloperswereclearlyatoddsandhadlittlerespectforeachother.Theteam
waslucky,though,inthatthepeoplehadnotyetlostallrespectforeachother,as
evidencedbytheirabilitytocometogetherintheend.
Commitment
Acommitmentisapledgeorpromise,anobligationtodeliver.Commitmentsshouldnot
bemadelightly—theyshouldbemadewithasmuchinformationaspossible.Theteam
reconfirmsitscommitmenttotheorganizationandtoeachotherduringeachsprint
planningmeeting.Attheendofsprintplanning,eachteammembershouldhavethesame
levelofunderstandingaboutwhattheteamisforecastingitcanaccomplishduringthat
sprint.
Courage
Courageistheabilitytofacedifficultyinspiteofyourfears.Alleviatingthesefearsisone
ofthebestwaysteamsandorganizationscanhelpteammembersbecourageous.Ateam
thatdemonstratesunderstandinginthefaceoffrankdiscussioncoupledwithleadership
thathasprovenitwilllistentobadnewsobjectivelyhelpgiveindividualsthecourageto
speaktheirminds.Remember,whenteamslackthecouragetodowhattheyfeelisright,
therightthingwilllikelynotbedone.
Openness
Opennessenablesteamstobereceptivetonewideas.Nowhereisateam’sopennessmore
apparentthaninthesprintretrospectives.Havingthewillingnesstoreceivenewideas,
perceptions,andwaysofthinkinghelpsdeveloplearningorganizationsandhigh-
performingteams.
ScrumRequiresaShiftinMindset
AlbertEinsteinsaid,“Wecan’tsolveproblemsbyusingthesamekindofthinkingwe
usedwhenwecreatedthem.”Oneofthebiggestimpedimentstoasuccessfuladoptionof
Scrum,ortoanyagilemethodforthatmatter,istheinabilitytohaveashiftinmindset,the
inabilitytouseanewwayofthinkingtosolveproblems.Asaresult,Scrumandallagile
practicesrequireacertaindegreeofopenmindedness,atleastforthefirstthreetosix
months.WhenIworkedonmyfirstScrumproject,ittooknearlyayearbeforeItruly
understoodwhatScrumwasabout.
Duringthatyear,IlearnedthatScrumwasapowerfulbutpotentiallydangeroustool.
DidyoueverseethatoldshowHomeImprovement?Themaincharacter,Tim“The
Toolman”Taylor,wouldalwaysgetintotroublewithashinynewtoolwhenhefailedto
takesafetyprecautions,riggedittobeusedinwaysitwasneverintended,orsimplybit
offmorethanhecouldchew.Scrumisthesameway.Ifnotoperatedaccordingto
instructions,especiallyatfirst,Scrumcanmakeyourprojectgohorriblywrong,fast.Too
manyteamstakearudimentaryunderstandinganddecidetheyknowbetter;theirsituation
isdifferent;they’vegotthisthingfiguredout.
Takemyadvice:GettoknowScrumbeforeyoudecideitneedstobecustomized.Useit
asintended:“out-of-the-box,”ifyouwill.Spendtimelearningasmuchasyoucanabout
it.Setasidesomeroominyourbraintoletthisknowledgegrow,tosteeplikeacupoftea.
TomysoftwareaudienceIsay,setasidesomememoryaddressspaceinyourbraintolet
thisknowledgesit.Donot—Irepeat,donot—trytocombineScrumwithothertoolsyou
arefamiliarwithjustyet;nowisnotthetime.Onlywhenyouhavemasteredonetoolcan
youlearntouseitwithotherssuccessfully.Aboveall,havethestrengthanddisciplineto
giveitafairshot,even(andespecially)whenitisdifficultandchallenging.Youmaybe
surprisedathowlittleyouneedtochangeScrumandhowmuchyouendupchanging
yourmind.YoumightthinkthisgoesagainstthefirstvalueoftheAgileManifesto,
Individualsandinteractionsoverprocessesandtools.Onthecontrary,individualsand
interactionsenableyoutolearntouseScrum(oranyotheragilepractice),andstickingto
practicingitbythebookhelpsyoumakeinformeddecisionsonwhatworks.
ScrumTakestheShortestPath,NottheSetPath
Thelongestpaththroughaprojectisknownasthecritical,orrate-limiting,path.Teams
buildaplanaroundthecriticalpathandfollowtheplan,frompointAtopointB.Along
thispath,issuesandproblemssurface:thestakeholdersmightnotknowexactlywhatthey
wantedwhentheplanwasbuilt,ornewdetailsmightberevealedduringdevelopment,
businessgoalsmightchangeastheproductmovesthroughitslifecycle,andtheproject
mightneedtorespondtointernalandexternalforces,suchasacompetitor’srelease.
Theseareallthingsthathappenonnearlyeveryproject;softwareprojectsarenotunique
inthis.
Withatraditionalplanningmethod,whentheseissuesariseteamsarestillforcedtogo
fromAtoB,makingsacrificesalongtheway—sacrificesinquality,functionality,and
customersatisfaction.OnceteamsfinallyreachpointB,theybegintotriagethebaggage
theyhaveaccumulatedandbeginbuildingtheplantogettoC,theplacetheydiscovered
thattheyneededtoreachbutcouldnot,simplybecausetheplanwouldnotallowthemto
(seeFigure1-1).
FIGURE1-1Thetraditionalplanningmethod
Scrumacknowledgesthatissuesliketheonesidentifiedabovearegoingtohappen—
theyareafactoflife.Insteadofbuildingtheperfectplantomitigatethem,Scrum
embracesthefactthattheseissueswilloccurandcontainsbuilt-inmechanismstoallow
forthemtobehandledastheysurfaceduringtheproject.WithScrum,teamsdon’thaveto
makeapitstopatB;theycangodirectlytoC.
Scrumdoesthisbyschedulingdaily,weekly,andmonthlycheckpointswithpeopleat
variouslevels(team,management,customers)toensureteamsareontrackandbuilding
therightfunctionalitytomeetthevision.Insteadofbettingonanall-encompassing
requirements-gatheringphase,Scrumteamsgatherrequirementsovertime,adjustingand
refiningthemasnewinformationisrevealedduringdevelopment.Becauseteamscando
this,theprojectisallowedtoadapttotheneedsofthebusinessandexternalmarket
conditions,providingtheshortestpathtosuccess(seeFigure1-2).
FIGURE1-2Scrumplanningmethod
Successdoesnotcomewithoutsomecost,though.Intheperfect-planmodel,
companieshavetheperception(oftenfalse)ofknowingtheexactdatetheprojectwillbe
done.Thisperceptionishardtogiveup.Inreality,though,eitherthescheduleslips,
features(whethertheyarethemostimportantfeaturesornot)arecuttomeettheschedule,
orqualitysufferstosqueezeinasetamountoffunctionalitybyapredetermineddate.In
theend,allthesesacrificescostthebusinessintime,quality,andmoney.
Scrumteamsdon’tpromisetodeliverasetamountoffunctionalitybyacertaindate.
Theyvaryeithertheenddateorthesetoffunctionality.Onsomeprojects,teamspromise
todeliveranapproximateamountoffunctionalitybyasetdate.Theydothisbylockingin
thecost(whatthecustomercanspendpersprint)andschedule(whentheprojectwillend
andhowmanysprintswillberun)upfront.Theythenestimatehowmanyfeaturesthey
cancompletewithinthoseconstraints.BecauseScrumprojectsalwaysdothehighest-
priorityworkfirst,anyfeaturesthataren’tcompletedbytheendoftheprojectarethe
featuresthatareofrelativelylowvalueandlikelylessimportanttotheoverallsuccessof
theproject.
Onotherprojects,wherefeaturesaremoreimportantthanasetcompletiondate,teams
mightlockinfeaturesbutallowthedate(andconsequentlythecost)tofluctuate.They
promisetodeliverasetamountoffunctionalitybyanapproximatedateordaterange.
Whenchangesoccur(newfeaturesareaddedorexistingfeaturerequestsarebetter
understood),thecustomerunderstandsthathemustmakeachoicebetweenreceivingthat
functionalityintheagreed-upontimeorextendingthescheduletoaccommodatethe
desiredfunctionality.
Thisisalargedeviationfromatraditionalapproach,wherethefeaturesarecapturedup
frontandthentheteamsareaskedtoestimatethetimeandcostconstraintsneededto
deliverthem,oftenonlytohavethecostandtimeconstraintsreducedwhilebeingaskedto
committothesamefeatureset.Figure1-3summarizesthedifferencesbetweenthetwo
approaches.Itisalsoahugeshiftinmindset,whichiswhytrustinandalignmentwiththe
valuesdiscussedinthischapteraresoimportant.
FIGURE1-3Contrastingtheplanningmethods
ScrumUnearthsIssues
Scrumexposesissues,thosethatwerelongagosweptundertherugandforgottenand
thosethatsomepeopledidnotknowexisted.Italsoexposesnewissues.Theseissuesare
notlimitedtocodingorteamwork.Forexample,oneofthemostcommonquestionsthatI
seesurfaceis,“Howshouldwecompensateourpeople?”Anotheris,“Ittakeshowlongto
runtheacceptancetestsuite?”Inatraditionalmodel,adevelopermightberewardedon
thebasisofgettinghiscodingdoneontimeandatacertainqualitybar.Howdoesthis
maptoacross-functionalteam,wheretherearenoindividualmetricsonfunctionalareas?
ItisorganizationalnormslikethisthatScrumchallenges,forcingmanagementtomake
hardchoices:addresstheissuesorignorethem.ThebenefitofScrum,however,isthat
oncethesebehaviorsandpatternsbecomevisible,itisdifficulttoignorethem.
ScrumShouldMarry
Scrumisaprojectmanagementframework.Assuch,ittellsyouhowtomanageyour
project,butitdoesnotcontainthespecificengineeringpracticesthatenableyoutodeliver
potentiallyshippablesoftwareeverytwotofourweeks.Forthat,youneedScrum’sPrince
Charming:ExtremeProgramming(XP).WhileScrumandXPhaveagreatdealin
common(e.g.,XPhastheplanninggame;Scrumhasplanningmeetings),XPcontains
somedifferentpractices,suchascontinuousintegration,testfirst,andpairprogramming,
thatcomplementScrumbeautifully.
WhileimplementingScrumbyitselfhelpsteams,marryingScrumandXPyields
tremendousresults.That’snottosaythatyoushouldjumpintoXPpracticesbeforeyou
haveasolidScrumfoundation—becarefulaboutintroducingtoomuchchangeallatonce.
Onceyourteamisexperiencedwiththeroles,artifacts,andmeetingsofScrum,theyare
readytostartintegratingXPpractices,butbeware—doingagileprojectmanagementwhile
usingtraditionalengineeringpracticesisoftenavolatilemix,causingmoreproblemsthan
itattemptstosolve.Thismaytakeaday,aweek,oramonth;italldependsontheteam’s
toleranceforchange.However,onceXPisintroduced,thetruemagicoccurs.Formoreon
theXP’sengineeringpractices,seeChapter9,“WhyEngineeringPracticesAreImportant
inScrum.”
TheXPpracticesthatIconsidermandatoryfortheprojectsthatIworkonareas
follows:
Sustainablepace—Worka40-hourworkweekandconsistentlyworkat85percent
output(seeChapter23,“SustainablePace,”formore).
Collectivecodeownership—Thewholeteamworksontheentirecodebase.There
isnosingleowner(seeChapter21,“WhenCulturesCollide,”formore).
Pairprogramming—Pairprogrammingistheartoftwopeopleworkingtogether
onthesamepieceofcodeunderdevelopment.Theycollaborateonthebestwayto
solvetheproblems(seeChapter19,“KeepingPeopleEngagedwithPair
Programming,”formore).
Test-drivendevelopment—Alsoknownastestfirst,thismeanswritingthetest
beforethecodewiththetypicalmodelofredtoindicateafailingtest,greento
indicateapassingtest,andthenrefactortomakethecodebetter.Alwayswork
towardgreen,noexceptions(seeChapter19formore).
Continuousintegration—CIenablesteamstohaveacontinuouspictureofhowthe
codebaseisbehavingandperforming.Checkinyourcode,ataminimum,daily.
Strivetogohomeinagreenstateeveryday,justlikehavingeachtestintest-driven
development(TDD)begreen.
Codingstandards—Oftenoverlooked,nothavingcodingstandardswreakshavoc
oncollectivecodeownershipaseachteammemberimplementshisorherown
personalstyle.Createthem,publishthem,andhangthemonthewalltoremind
everyone.
Refactoring—Thisistheartofchangingthecodewithoutchangingtheintentofthe
code,resultinginsimplerandmoreconsistentcode.Constantlyrefactoringwill
enabletheteamtostayalertandfresh.
Potentiallyshippable—thesetwowordsoftencauseteamsmanyheadachesand,asmy
friendChrisSterlingwroteinablogpost,areoneofthe“forgottenScrumelements”
[STERLING].Totrulyreachashippablestate,teamsneedtochangenotonlytheirproject
managementprocessbutalsothewaytheydevelopsoftware.Otherwise,alltheissues
presentedinthischapter’sstory—poorquality,notimetotest,andinabilitytodeliver—
willremainandwillsinktheproject.IntegratingXPpracticesisthebestwayI’vefoundto
bringabouttrueandlastingchange.
WhenIsScrumRightforMe?
Withallofthisinformationinmind,youcanseethatimplementingScrumismuchmore
complicatedthanitmightappear.Soshouldyoueventry?Itdepends.
Ihaveseveralfriendswhoworkintheconstructionindustry.Ioftenaskthemtoassist
meonsmallhouseprojectsinreturnfordinner,andtheyarehappytoobligeme.One
thingInoticeeverytimetheycomeoveristheplethoraoftoolstheyhaveontheirbelts.I
askedJoachimonceifhecarriesastandardsetorifhecustombuildshisbeltforeachjob
hiscrewworkson.Hetoldmehealwayshashisbasics(atapemeasure,pencil,andsafety
glasses),butalmosteverythingelsedependsonthejob.Ifheisdoingframingwork,he’ll
loadhisbeltwiththetoolsneededforframing.Whenheisdoingfinishcarpentryortiling,
heusesadifferentset.Atthesametime,though,thetoolsthatarenotonhisbeltare
alwaysclosebyinthetruck.HowdoesthisapplytoScrum?
IviewScrumasatoolinaperson’stoolbelt.Likeanygoodcontractor,apersonshould
useScrumwhenit’sappropriate,whichmightnotbe100percentofthetime.Justasyou
wouldnotuseascrewdrivertosecureanail,youshouldnotuseScrumonaprojectthat
doesnotrequireit.SowhenshouldyouuseScrum?
Softwareprojectsgenerallyfallintofourdifferentareas:simple,complicated,complex,
andanarchy(seeFigure1-4),asadoptedfromRalphStacey’sStrategicManagementand
OrganisationalDynamics,SecondEdition[STACEY].Simpleprojectsaremadeup
primarilyofknowns:Similarprojectshavebeendoneoverandover,soeveryoneagrees
onwhatisneededandthetechnologyiswellunderstood.Simpleprojectsareeasyto
implementandunderstand.
FIGURE1-4Therelationshipbetweenrequirements,technology,andtypeofproject
Asprojectsmovefartheroutfromthesimplezone,however,thingsbecomealittle
fuzzier.Requirementsortechnology—orboth—maynotbeasstableashoped.Perhaps
peoplearenotascertainaboutwhatthefinalproductwilllooklike.Maybethetechnology
isneworuntried.Inthisarea,theareaofcomplicatedandcomplex,projectsencountera
substantialamountofchange.Projectsherearemuchmoredifficulttoimplementand
predict.Then,asprojectsmovebeyondcomplextowardanarchy,theybegintohavetruly
unknownunknowns.Whenprojectsorcompaniesareinthisstate,theyshouldfocuson
fixingtheproblemsthatputthemthereinsteadofbuildingsomethingnew.
Whatthisallmeansisthatifyouareanywherebutinthesimplearea,youwillhave
changeinyourproject—changeinrequirements,technology,orboth.IhaveusedScrum
andXPonsimpleprojects,andmoreoftenthannot,Ifounditoverkill.However,asyou
moveawayfromagreementandcertainty,youwanttouseScrumandXP,asdoingso
helpsyouadapttothechangingbusinessrequirementsthataresuretocome.
ChangeIsHard
Scrumisallaboutchange.Somepeopleembracechangeandsomeavoiditlikethe
plague.Bothresponsesareperfectlynatural.Softwareteamstypicallyresistchangenot
becausetheyareafraidofitbutbecauseteamshavebeenconditionedtoprovidetheright
thingthefirsttime.Softwareteamsequatechangeto“bugs”orfeaturesthattheymissed
whentheyweredevelopingtheproduct—tothem,achangerequestsuggeststheyhave
failedinsomeway.Whattheyoftendon’tunderstandisthatchangeisinevitable.Noone
canpredictthefuture;youcannotbeexpectedtoknowaboutsuddenmarketshiftsora
competitor’snewestreleasebeforetheyhappen.Tobesuccessful,teamsmustlearnto
embracechange.Scrumprovidesthetoolstodoso,butmovingtoaplacewherechangeis
integratedintotheprocessisadifficultjourney.
Theteammembersinthischapter’sstorywantedtouseScrumbutwerefearfulof
leapingintotheunknown.Theytried,invain,toholdontofamiliarpatternsandprevent
mistakesbymodifyingtheprocessandaddingplanningandintegrationsprints.Insteadof
keepingthemfromfailing,thesemodificationsactuallycausedproblemsandmadethings
muchworse.
UnderstandingcommonreactionstochangeiskeytoasuccessfulScrum
implementation.VirginiaSatirwasafamilytherapistwho,throughclinicalstudies,
identifiedacommonpatterninthewaypeopleexperiencechange.Thestagesincludethe
latestatusquo,theintroductionofaforeignelement,chaos,integrationandpractice,and
thenewstatusquo.ThesepatternsofbehaviorareknownasSatir’sStagesofChange
[SATIR],whicharevisuallyrepresentedinFigure1-5.
FIGURE1-5Satir’sStagesofChange
Ifindthatthesestagesarefairlyconstant,nomatterwhatlevelofchangeapersontakes
on.MymothertriedtoquitsmokingwhenIwasgrowingup,andIobservedthese
patterns.Ialsohaveobservedthemwithkidswhentryingtorideabicycleforthefirst
time.ConsiderhowScrumteamsmovethroughthesestages.
LateStatusQuo
Whenpeopleareinthestatusquo,theyarecomfortable.Often,peoplefindexcusesto
stayinthestatusquobecauseitistheworldtheyarefamiliarwith.Peopletendtoignore
theneedforchangeandjustkeepdoingthesamething,onlybetter,harder,ormore
diligently.Anall-too-familiarexampleofthestatusquoisthedeathmarch,whereteams
soldierontowardcertainfailurewhileworkinglongerhoursandmakingunwanted
sacrificestoqualitytotrytomeetadeadline.Eventually,theteammembersbeginto
complainaboutthefutilityofbelievingthatallcanbemadewelliftheyjustkeep
pluggingawayharderandstronger.Atthispoint,acriticalmassstartstogrow,asaselect
fewpeopletireofthewaythingsarebeingdoneormanagementinstitutesanewwayof
working.Thisisthepointatwhichoneormoreforeignelementsareintroduced.
FromForeignElementsandChaostoTransformingIdeas
Aforeignelementisachallenge;itisthethingthattakesyououtofthelatestatusquo.In
somecases,itcanbeaminorelement,suchasusinganewspecificationtemplatefora
project.Inothercases,itcanbeacompletelynewwayofworking,suchastheadoptionof
Scrum.Regardlessofwhichitis,onceintroduced,theforeignelementcannotbeignored,
andchaosensues.Whilethedegreeofchaosdependsonthenatureoftheforeignelement,
inthecaseofalarge-scalechange,itcanfeelasifeverythingfamiliarhasvanished.Even
thosewhoaremostcertainabouttheneedforanewwayofdoingthingsareforcedto
workinawaythatfeelsunfamiliaranduncomfortable.Peopleinthisstatearenervous
anduncertainandoftenfeelvulnerable.Theyareafraidofmakingawrongmove.Asa
result,anxietybuildsandperformancevarieswidely.Asthegroupworksthroughthis
stage,thingswillgetworsebeforetheygetbetter,buttheyindeedwillgetbetter.
Theendofchaoscomeswiththeidentificationofoneormoretransformingideas.What
transformingideaswillendthechaosthatScruminitiallyintroduces?Itvariesfromteam
toteam,buttheoverallthemeisusuallythegeneralthought,“Hey,thisisnotthatbad.We
candothis.”Peoplestarttoseetheconnectionbetweentheforeignelementandtheway
thingscanbeinthefuture.Theconnectionismade.
PracticeandIntegration
Mykidsplaysports.Everytimetheycomplainaboutattendingpractice,Iremindthem
thatsomeoneelseispracticingrightnow,practicingtobeatthemwhentheirtwoteams
meet.Practiceinsoftwaredevelopmentisequallyessential.Withoutpractice,aforeign
elementanditsassociatedtransformingideaswillneverbecomefamiliar,secondnature,
aningrainedhabit.
AsateambeginstolearnthepracticesandunderstandtheprinciplesofScrum,itbegins
torealizeScrum’sbenefits.Asthesebenefitsemerge,theteammembersgradually
embracetheirnewreality.Chaosbeginstofadeastheteammovestowardanew
equilibrium.Trust,relationships,andidentitiesemergeintheteamandinthegroupat
large.Aspeoplelearntoworktogetherinanewway,performancestartstoimprove,often
exceedinglevelsseenpriortothechange.Don’ttaketheseearlygainsasasignthatthe
teamisnow“fine.”Onthecontrary,teamsinthisstageneedmorenurturingandcaring
thantheywilllater.Whileit’struethatpeoplearegainingconfidenceintheirwork,they
maystillfeelvulnerableeventhoughitmaynotbevisible.Thenewfoundsuccessis
tenuousanddelicateandshouldbetreatedappropriately.Still,witheachsuccess,the
prospectsofwhatcanbeaccomplishedstarttoemerge.Formost,itmightfeellike
nirvana;foraselectfew,itmayalwaysfeelsomewhatchaotic.
NewStatusQuo
Astheteammembersbecomeskilledandcomfortablewiththetransformingidea,they
begintooutpacetheirpreviousresults.Theyfeelrestedandhavethesenseof
accomplishmentthatcomesfromgoingthroughatransformationaleventtogether.This
stageenablespeopletoexplorenewwaystoimprove,tofeeltheyhavearighttomake
thingsbetter,andtoworktogethertocreatealearningorganization.Thefreedomto
continuetopracticeinasafeenvironmentiskeyhere.Justasinsports,younevergetso
goodthatyoudon’tneedtopracticeorlearnnewthings.
KeystoSuccess
Asthischapter’sstoryillustrates,manyteamsstrugglewiththeirinitialimplementationof
Scrum,notbecausetheyarereluctantorclosed-mindedbutbecauseScrumrequiresthem
toworkinawaythatfeelsuncomfortableandunfamiliar.Thecharactersinthestorytried
toholdontotheelementsthattheywerefamiliarwith,notknowingthatthoseelements
wereslowingthemdown,keepingthemfromsuccess.TohaveasuccessfulScrum
implementation,teamsmustadheretothebasics.
First,teamsmustunderstandtherules.Scrumisasimpleframeworkforgettingthings
done.Becauseofitssimplicity,peopletendtomodifytheruleswithoutunderstandingthe
implicationsoftheiractions.Thisisdangerous.Justasyouwouldnotdriveacarwithout
knowingtherulesoftheroad,youshouldnotattempttoimplementScrumwithoutagood
dealofstudy.Learningtherulestakestime.UsethistimetounderstandhowScrum
appliestoyou,yourteam,andyourcompany.
Second,teammembersmustlearnthebasemechanics.Teamsandtheirorganizations
needtounderstandhowthingsworkandfeelwithScrum.Ittakestimeforthistohappen.
AsexplainedinSatir’sStagesofChange,movingfromtheknowntotheunknownisa
difficultprocessthattakespatienceandpractice.Itisakintolearningtorideabicycle.
PeoplecannotrideLeTourdeFrancewhentheyarefiveyearsold;instead,theywobble
andweaveandoftenfalldown.It’spracticethatmakespeoplesureandsteady.Give
peopletimetolearnthebasicsandhelpthemalongtheway.Makesureyoukeeptheir
feelingsinmindandunderstandthatasthebasemechanicsarebeingpracticed,people
maybefragile,skeptical,andanxious.Theseinsecuritieswillbeovercomewithtime,but
onlyifpeoplecanpractice.
Third,giveittime.Learninganewwayofworkingtakesmorethanaweekend
workshop.I’vefoundthatthemajorityofScrumteamsandcompaniesimplementing
Scrumneedatleastthreemonthstolearnthebasicsandtostartrolling.Somecandoitin
lessandsometakelonger,butonaverage,planforthreetosixmonths,dependingonthe
groupbeingtransitioned.
Fourth,don’tadoptScrummidstream.Ihaveseencountless“Scrumfailures,”where
teamsimplementedScruminthemiddleofadevelopmentcycle,oftenasaresultof
managementdecidingthatScrumisthemagicbulletthatwillsolveallthecompany’s
problems.Management,seeingthepotentialbenefitsthatcomewithahigh-performing
team,expectstheteamtobecomeexpertsatusingScrumovernight.Asyoucanimagine,
thisthrowstheteamintoatailspinastheyareforcedtoworkinanewfashionwhilean
immovablependingdeadlinelooms.Theensuingchaoscausesdelayafterdelayandgreat
sacrificesinquality.IrecommendthatyouimplementScrumatthebeginningofanew
project.Inthemeantime,youcancreateatransitionbacklogofthethingsyouneedto
haveinplaceandworkingtoensurethebestScrumadoptionpossible.
Last,ensuretimeisallocatedforcontinuouslearning.Toooftenteamsignore
retrospectives,consideringthemeitheraluxuryorawasteoftime.Teamswhoskipthe
retrospectiveorgothroughthemotionswithoutapplyinganyofthelearningtofuture
sprintsmostoftenfallbackintotheiroldhabits.Why?Becausetheyarenotsettingaside
timetolearn.LearningisoneofthekeyaspectsofScrumthatcannotbesacrificed.The
teamcannotmovetowardahigh-performingstatewithoutit.
Throughallthesetechnicalandpsychologicalchallenges,justremembertostickwithit.
LearningScrumisnodifferentfromlearninganythingnew.Therewillbegreat
satisfactiononemomentandgreatfrustrationatanother.Educateyourselfandkeepan
openmind.RemembertoothatifNASAcouldputamanonthemoonwithacomputer
systeminsidetheApollo11thathadsomewherebetween36kand72kofmemory,wecan
implementScrum.
References
[SATIR]Satir,Virginia,JohnBanmen,JaneBerber,andMariaGomori.1991.TheSatir
Model:FamilyTherapyandBeyond.PaloAlto,CA:ScienceandBehaviorBooks,pp.
98–119.
[SCHWABER01]Schwaber,Ken,andJeffSutherland.2013.“TheScrumGuide.”
http://www.scrumguides.org(accessed28October2015).Alsoavailableat
http://www.mitchlacey.com/resources/official-scrum-guide-current-and-past-versions.
[SCHWABER02]Schwaber,Ken,andBeedle,Mike.2002.AgileSoftwareDevelopment
withScrum.UpperSaddleRiver,NJ:PrenticeHall,pp.148–154.
[STACEY]Stacey,Ralph.1996.StrategicManagementandOrganisationalDynamics,
SecondEdition.London:Pitman.
[STERLING]Sterling,Chris.GettingAgilewebsite.
http://www.gettingagile.com/2009/07/15/the-forgotten-scrum-elements/(accessedon3
July2010).
PartI:GettingPrepared
Chapter2.GettingPeopleonBoard
M.ScottPeckwrote,“Thetruthisthatourfinestmomentsaremostlikelytooccurwhen
wearefeelingdeeplyuncomfortable,unhappy,orunfulfilled.Foritisonlyinsuch
moments,propelledbyourdiscomfort,thatwearelikelytostepoutofourrutsandstart
searchingfordifferentwaysortrueranswers.”
Asmuchasteamsmightcomplain,peopleareoftenall-too-comfortablewiththeirruts
andtheirunhappiness.Afterall,asbadastheircurrentsituationmaybe,atleastthey
knowtheterrainandrules.Scrumoffersawayout,butitbringswithitanew,unknown,
unfamiliarwayofworking.Whenfacedwiththechoicebetweentheknownandthe
unknown,peopleareoftenhesitantorevendownrightreluctanttochange.Itisnotuntil
immensepressureisapplied,untiltheircurrentrutitselfbecomestoopainful,thatthey
escapetoabetterfuture.
GettingpeopletotakeachanceonaScrumprojectrequiresmorethanjusta“Scrumis
Good”campaign.Ittakesahealthydoseofpersuasion,asteadyincreaseinpressure,anda
wholelotofpatience.
TheStory
RyanwasanenthusiasticprojectmanagerworkinginthecorporateITdepartment.
Historically,hehadagoodtrackrecord,buthislastprojecthadrequiredasignificant
architecturalchange,causingadelayofnearlysixmonths.Bythetimetheprojecthad
ended,Ryanwasreadytofindanewwaytowork.Likemanyofhispeersinthegroup,he
wasfrustratedwithhisteam’sinabilitytodeliverontimeandwiththehighquality
expectedbythebusiness.TheproblemwasthatRyanhadnoideahowtomakethings
better.
Ryanandhisfamilyhadatwo-weekvacationplannedattheircabininMontana.Ryan
decidedthatthetwoweekswouldbebestspentfindingsomenewtoolsforhisproject
manager’stoolbelt.Hepackedasuitcasefullofbooksonprojectmanagement.Ashewas
abouttozipupthebag,hesawabookthathisboss,Steve,hadgivenhimseveralmonths
ago,AgileProjectManagementwithScrumbyKenSchwaber.“Mightaswellthrowthis
intoo,”Ryanthoughtandtosseditinthebagwiththerest.
Ryanburnedthroughthetraditionalprojectmanagementbooks.Hepickedupafew
tips,mainlyoncommunicationplansandriskmanagement,butfoundnothingthatwould
havepreventedthearchitecturalissueonhislastprojectoranyoftheotherproblemshis
teamshadbeenhaving.Reluctantly,hepickedupSchwaber’sbook,grabbedadrink,and
wentouttothedecktoreadit.
Hisinitialskepticismsoongavewaytoexcitement.Hecouldvisualizethepatternsin
hisworkthatdirectlyrelatedtotheonesinthebook,patternsaroundcomplexityand
value-drivenwork.Healsorecognizedtruthsthathehadignoredonhispastteams,such
asactuallybuildingateamofpeopleinsteadofusingacollectionofresources.Themore
Ryanread,themoreinterestedhebecame.Heclosedthebooklaterthatafternoonwitha
greatdesiretolearnmore.Heimmediatelywentinsideandorderedtwomorebookson
agiledevelopment,withnextdayshipping.
Ryanwasnotaworkaholic,butthereadinghehaddonesparkedapassion.Whilehis
wifereadthelatestsummernovelsandhiskidsplayedsoccerintheyard,hespenthis
timeporingoverbooksonlean,agile,XP,Kanban,andScrum.Uponreturningfromhis
vacation,RyanranintoSteve—hisoldfriendandthedivisionmanagerwhohadgiven
himtheScrumbook.
“Steve,I’vegotgreatnews!”Ryanexclaimed.“WhileIwasinMontana,Ireadallthese
books,andIthinkIhaveasolutiontonotonlytheproblemsI’vebeenhavingwithmy
projectsbuttoalltheprojectissueswehavebeenhavinginourdivision!”
“That’salottogetoutofafewbooks,Ryan,”saidSteve,smiling.“Butcomeoninand
tellmeaboutit.”
RyanandStevespentthenextthreehourstalkingabouteverythingRyanhadread.Then
RyanaskedthequestionthathadbeeninhismindthewholewaybackfromMontana:
“YouknowtheMichaelsonprojectwehavecomingup?”
“Yes…,”saidSteve,withahintofcuriosityinhisvoice.
“WouldyouletmetrythisScrumstuffonthatproject?”Ryanasked.“Rememberthat
objectiveyoulaiddownforthegrouptofindawaytoincreaseoveralloperational
efficiencyby50percentwithoutincreasingtheoverallgroupsize?Ithinkthisisoneway
tohelpachievethatgoal.AndIthinkitcanbedonewithinthisyearlikeyouhadhoped.”
Stevesatquietlyforaminute,thinking,andthenreplied,“Ryan,Ithinkyouareright.
TheMichaelsonprojectwillbeagoodtrialforthis,andIagreethatScrumcouldhelp
increaseoverallefficiencywithoutgrowingtheoverallsizeofthegroup.”
“Great!”exclaimedRyan.“Myhunchwascorrect!NowcanI—”
“Wait,”Steveinterrupted.“Now,don’tgettooexcited.Youdon’tknowthis,butwhenI
workedatMegaco,IactuallyranseveralScrumprojectswithgreatsuccess.Ihave
managedlargeorganizationsandteamsthathavebeenabletoimprovebymorethan50
percent,soIknowthiscanwork.Yourbiggestchallengewillbegettingstarted,whichis
moreworkthanyourealize,especiallygettingpeopleonboardwithScrum.”
“SoIcanhaveateamtotrythisout?”askedRyan.
“No,”saidSteve.“Youcanhavetheproject,butyou’llneedtobuildyourownteam.I
wantyoutofindotherteammemberswhoarewillingtoworkthiswaybeforeIgivemy
finalapproval.Iwillsupportyouanywayyouneedmeto,andyouhavefreereigntodo
whatyouseefit,withinreason,togettheprojectoffthegroundandthepeopleonboard.I
willbeyoursponsorandwilllettheotherseniormanagersknowwhatwearedoing.
Remember,noteveryonewillhavetheenthusiasmyouhaveforthis,sotreadlightly.”
“I’llworkonrecruitingmyteam.Idohaveonelastquestionforyou,”saidRyan.“If
youknewaboutScrum,whyhaven’tyoualreadyintroduceditorExtremeProgramming
tothegroup?”
Stevepaused.“WhenIwasatMegaco,Iwasverymuchlikeyou.Enthusiastic,
outgoing,veryin-your-facewiththisagilestuff.Ipusheditfromthetopdownonmy
peoplethere,andtheyreactedbyrebelling.Ieventuallywonmostofthemover,butnot
withoutalotofhardshipandattrition.Thatexperiencemademerealizethatthisneedsto
beabottom-upapproach.Iputthegoaloutthereinthebeginningoftheyearhoping
someonewouldgrabontoitandsuggestanagilesolutionlikeScrumorXP,orboth.”
BothRyanandStevesatthere,lettingSteve’scommentsinkin.
“Thatmakessense,”saidRyan.“I’llkeepyoupostedonmyprogress.”
“Remember,I’mheretohelp,”saidSteve.“Asyoursponsorforthiseffort,it’smyjob
tohelpclearbiggerissuesandprovideguidance.Don’tbeafraidtocometomefor
advice.”
Ryanleftfeelingexcitedbutalittleworriedabouttheideaofgettingpeopleonboard
withhisideas.Hespentafewdaysconsideringhisapproach.
ThefirstpersonRyanmetwithwasDaniel,adeveloperinhismid-fortieswhohadbeen
atthecompanyformorethan12years.Danielwasmethodicalinhisapproachandvery
goodathisjob,butlatelyhehadbecomecynicalandvocallycriticalofthewaythegroup
wasrun.
“Daniel,I’dliketotalktoyouaboutaprojectIamstarting.It’stheMichaelsonproject
and—”Ryanbegan.
“Whyareyoutalkingtomeaboutthis?”Danielinterrupted.“Youshouldtalktomy
manager.Hehandlesmytimeandwilltellmewhattoworkonnext.”
Ryanfeltalittleperturbedthathewascutoffmidsentence,butheforgedon.
“Icandothat,Daniel.IthoughtIwouldtalktoyoufirsttoseeifyouhadaninterest
beforetalkingtoAlanaboutthis.Stevesaiditwouldbeagoodideaformetotalkto
peopledirectlytobuildtheteamandyou—”
Ryanwascutoffagain.
“Steve?OurbigbossSteve?”askedDaniel.
“Yes,”saidRyan.“Yousee,thisMichaelsonproject,it’sarealinterestingproject.My
ideaistousethisprojectmanagementframeworkcalledScrumtorunit.We’lluse
ExtremeProgrammingpracticesliketest-drivendevelopmentandpairprogramming,”
Ryansaidwithexuberance.
“Pairwhat?ExtremeScrum?Idon’tgetit,”statedDanielbluntly.
“Pairprogramming,youknow,wheretwopeopleworkonthesamecodeatonce.And
Scrumisaprojectmanagementframeworkthatallowsusto—”saidRyan,beforebeing
cutoffyetagain.
“Ryan,listen.Ihearyou’reaniceguyandall,butthisisn’tforme.Ihavenointerestin
thisScrumstuff,andImostcertainlywillnottrypairprogrammingortestwhatever.
Sorry,butno,”saidDaniel.
“But,Daniel,ifyou’djustlisten,you’dbeamazedathow—”Ryanstarted.
“Look,Ryan,Idon’tknowhowtobeclearer.Theanswerisno.I’mnotinterested!”
exclaimedDaniel.
Ryanleft,discouraged.Still,hehadanotherchanceattheoneo’clockmeetinghehad
scheduledwithMeghan.
Atoneo’clock,RyanwenttoMeghan’soffice,butMeghanwasn’tthere.Whereisshe?
Ryanwondered.Hesatandwaitedinheroffice.Shefinallycamein15minuteslaterand
seemedsurprisedtoseehimsittingthere.
Ryangreetedher,“HiMeghan,Iwasjustabouttogiveuponyou.I’mherebecauseI
wantedtotalktoyouaboutthisnewprojectI’mdoing.We’regoingtouseScrumand—”
Meghancuthimoff.
“Ryan,Ineedtocomecleanwithyou.Danielhasalreadywarnedmeaboutwhatyou
wanttotalkabout.IcanalreadytellyouI’mnotinterested.Honestly,Iwaskindofhoping
youwouldn’tbehereifIshoweduplate.Ihatesayingno,butI’mnotsayingyes.”
Ryanhunghisheadandsighed.Meghanrelented.
“Okay.Fifteenminutes.”
Ryangotrighttothepoint.Heenthusiasticallydescribedwhattheprojectwasabout,
therevelationhehadoverhisvacation,andhowStevesaidhecouldbuildateamtodo
thisworkandtrysomethingnew.HewasverycleartoemphasizethathehadSteve’sfull
support.Itwasapproachinghalfpastone.
“That’ssomegoodstuff,Ryan.Icanseehowitmightworkonawebprojector
somethingsmall,buttheMichaelsonprojectishuge!Itrequiresalotofarchitectureand
design,whichweneedtodointhebeginningtoplanoutthework.Idon’tthinkwhatyou
areproposingwillwork.And,I’mwithDaniel,Idon’twanttopairprogramordo
testing,”saidMeghan.
“It’snotexactlytesting.It’stest-drivendevelopment—”Ryanbegan.Meghanquickly
cuthimoff.
“Letmeputitanotherway.I’mnotdoingthis.Ever.It’snotforme.It’snotforour
division.Andifyouwanttokeepyourjob,itshouldn’tbeforyoueither.There’snowin
foranyofusintryingsomethingsoextreme,”Meghanexplained.“I’msorry.I’vegotto
gotomynextmeeting,”shefinished,gettingupandleavingRyansittinginheroffice.
Ryansattherefor15minuteswithhisheadinhishands,wonderingwhyMeghanand
Danielweresoresistant.Whatwashedoingwrong?
ByFriday,hehadtwomoremeetingswithtwodifferentpeople.Noluck.Hewentto
seeStevelaterthatafternoontodiscusshislackofprogress.
“Ryan,”Stevebegan,“IhadaquicktalkwithMeghanandDanielyesterdayafteryou
left.I’vegotsomeadviceforyou,asafriend.Youcan’tbein-your-facewithpeople.Let
themcomearoundthewayyoudid,slowly.Theyneedtoreadaboutit.Understandit.
Havetimetomakeuptheirminds.Givethembooksorsomeofthosepapersyouread
overyourvacation.Whenpeopletrynewthings,wehavetograduallybringthemaround
andgivethemtimetogetcomfortable.”
“You’reright,you’reright,”saidRyan.“I’vegoneaboutthisthewrongway.I’mgoing
totakeyouradviceandtakeastepback.Gradually.Logically.Gotit.”
RyanspentthenexttwoweeksslippingMeghanandDanielcopiesofchaptersfromhis
favoritebooksandresearchpapers.Hehighlightedkeypartsandmadenotesonthecopies
astowhyhelikedthemandhowhethoughtthosesectionscouldapplytotheteamhewas
tryingtobuild.
OneMondaymorning,Meghandroppedbyhisoffice.
“Okay,Ryan,”shebegan.“I’mstartingtoseeyourpoint.I’mwillingtohelpwithone
condition.I’vegotkidsandIcannotaffordtolosemyjob,notinthiseconomy.I’mreally
goingoutonaledgehere,andIneedtobesurethatStevewillseethisthrough,”said
Meghan.
RyanassuredherthathewouldhaveStevegiveheracall.Laterthatday,heonceagain
droppedbySteve’soffice.Heupdatedhimonhisprogresssofar,addingthatMeghanwas
willingtojointheteambutneededsomereassurancesfromStevebeforecommitting.
“I’lltalktoher,”saidSteve.“Sothat’sone.Howareyougoingtoconvinceeveryone
else?”
“Well,Ididsomeresearch,anditturnsoutwehavetwoexpertsinXPhereatthe
company,JimandWard.Iaskedthemiftheywouldbewillingtohaveadiscussionwith
ourdivisionaboutwhatthisagilestuffisallabout,”saidanervousRyan.
“Goodidea,”saidSteve.“Whatdoyouneedmetodo?”
ThrilledthatStevelikedtheidea,Ryanlaidouthisplanforthetalk.“Atthediscussion,
I’dlikeyoutobethefirstonetotalk.Reiterateyourgoalsforthedivisionandthenabout
whatI’mtryingtodo.Soundgood?”
“Oh,absolutely!”saidSteve,smiling.
Overthecourseofthenextweek,Ryanputupflyersaroundthebuilding,setupthe
logistics,andgaveJimandWardsomebackgroundonthegroup.Meghanhelpedgetthe
engineeringpeopletocommitandsuggestedtoRyanthattheyorderlunchtoensurea
crowd.“Freefoodisastrongincentive,”shesaid.
Thedayofthemeetingarrivedatlast.Ryanlookedoutatthecrowdandsmiled.More
than90percentofthedivisionhadshownup,muchmorethanhehadexpected.Steve
spokefirst,explainingthegoalshewastryingtomeetandRyan’sefforttofindbetter
waystobuildsoftwaresystems.ThenitwasWardandJim’sturn.
MostoftheattendeesknewwhoJimandWardwereandwhattheyhadaccomplished.
WardandJimspentacoupleofhoursdescribingtheprinciples,values,andpracticesof
XP.Theaudiencelistenedattentivelyandhadplentyofquestions.Evenafterthe
discussionhadofficiallyended,nearly50peoplestayedtotalkmore;itwasthreeo’clock
beforethelastpersonleft.Throughitall,WardandJimhappilyansweredeveryone’s
questionsandprovidedgentlereassurancethatScrumandXPwerejustanotherwayto
approachsoftwaredevelopment—modernengineeringpractices,ifyouwill.Thelast
persontoleave?Daniel.HeapproachedSteveandRyan.
“Well,I’vegottosay,theseguyswereprettypersuasive.I’mstillnotcompletelysold,
butifStevethinksthisisimportant,I’mwillingtodoitononecondition,”saidDaniel.
“What’sthat,Daniel?”askedSteve.
“Thatifthingsgowrong,Iwon’ttakethefall,”saidDaniel,lookingSteveintheeye.
“Firstofall,it’snotifthingsgowrong,Daniel,it’swhen.MyfirstcluethatScrumand
XPareworkingwillbewhenyoufail—howcouldIblametheteamfordoingtheright
thing?”saidSteve.
“I’mnotfollowing.Whatdoyoumean,whenwefail?”askedDaniel.
“I’mtalkingaboutoneofthebestthingsagileprocessesgiveteams—theabilitytofail
early.Iknowtheteamwillmessup;what’simportantiswhentheymessup.Iftheteam
waitsuntiltheendoftheprojecttofail,ashashappenedonourotherprojects,we’re
goingtohaveallthesameproblemswe’vehadbefore.Ontheotherhand,ifthisteamfails
inthefirstthreemonthsandlearnsfromit,theprojecthasamuchbetterchanceof
success.AskRyan.Hewisheshehadfailedearlyonhislastprojectbecauseifhehad
knownabouthisarchitecturalissuessooner,hecouldhavequicklyfixedtheproblemsand
avoidedalltheend-of-projectpain.Failingisn’ttheproblem;failinglateis.
“Asfarastakingthefallgoes,Iknowthattherewillbestumblingblocks,andI’m
willingtogiveyousomeleewaytofigureouthowtoworkinanewway,butIexpect
everyoneontheteamtobefullycommitted.Yousucceed—andfail—asateam,”
cautionedSteve.
“Soit’sallofusornoneofus,andyouwantustofail,aslongaswedoitearly.Wow.I
actuallythinkI’mgoingtosayyes,”saidDaniel.“Ijustneedalittlemoretimetofigureit
alloutinmyhead—really,itscaresthebejeezusoutofme.”
“Metoo,”saidRyan.
ThiscaughtDanieloffguard.
“ButRyan,you’retheguypushingthiscrap.Whatdoyoumeanitscaresyou?”asked
Daniel.
“Areyoukidding?Ithinkit’stherightthingtodo,butthatdoesn’tmeanI’mnot
scared.Ihaveavisionofhowthiscanwork,ofhowwecanhelpachievethegoallaidout
bySteve.Butoncewemakethisleap,likeStevesays,weareinthistogether,asateam—
succeedorfail,it’sallonus,”saidRyan.
“Wellsaid,Ryan,”saidSteve.
“Letmenoodleonthis,”saidDanielasheturnedtowalkbacktohisoffice.
“Notabadshowingtoday,Ryan.Howdoyoufeel?”askedSteve.
“Thisissohard,”saidRyan.“Gettingthesepeopletochangetheirmindsis
excruciatinglytedious.Iwantedtobeupandrunningbynow,buthereweare,nearlyfour
weeksintoit,andIstilldon’thaveateam.It’sprettyfrustrating,”saidRyan.
“Hanginthere.Todaywasagreatmeetingforourdivision.Peoplewantedtohear
aboutthis,andIthinkyoumadealotofprogress.Still,don’tbesurprisedifthereare
peoplewhowon’tchangetheirminds,nomatterwhatyoudoorsayorgivethemtoread,”
saidSteve.
Overthenextthreemonths,RyanworkedwithMeghanandDaniel,calmingtheirfears
andwinningthemover.They,inturn,wonoverafewmorepeopletofillouttheteam.
Theywereallsurprisedattheresistancefrommiddlemanagement,whowereunsurehow
theywoulddoperformancereviewsandmonitorwhowasworkingonwhatiftheirpeople
wereonacross-functionalteam.Stevewasabighelpwiththemanagers,reassuringthem
thathehaddonethisbeforeandthathewouldhelpthemworkthroughanyproblems.
Finally,afterfourmonths,theteamwasreadytostarttheproject.
“Phew!”saidRyantoSteveoverbeersonaFridaynight.“Ifinallydidit.”
“Nowcomesthehardpart,”saidSteve.“Youhavetofigureitallout,experiment,and
learn.”
“Oh,that’llbeeasyafterthesepastfourmonths,”saidRyan.
“Sureitwill,Ryan,”saidStevewithasmile.“Sureitwill.”
TheModel
Inthisstory,Ryanwascomingoffalargeprojectthathadrequiredanarchitecturalchange
lateinthedevelopmentcycle.Ryanwasworriedthatfailinglatehadbecomeanemerging
trendofhisgroup.Desperateforanswers,heboughtandreadseveralbooksonproject
managementthatdidn’ttellhimmuchhehadn’talreadytried.Asalastresort,hereada
bookonScrum,giventohimbySteve,theheadofthegroup.Hisinitialskepticismgave
waytoexcitement.Hereadmorebooksonotheragilepractices.Themorehelearned,the
moreconvincedhebecamethatthiswasthewayforward.
ConvincingothersprovedmuchmoredifficultforRyan.Ryan’speersmethis
newfoundenthusiasmwithfear,anger,andsuspicion.OneproblemwasthatRyan’slast
projectdidnotinspiretrustwithhiscolleagues.AnotherobstaclewasRyan’sdelivery
style.ByalmostattackingpeoplewithhiszealforScrum,Ryanwasmakingpeoplefeel
defensivebeforethey’devenhadachancetohearwhathehadtosay.
Steve,Ryan’ssponsorofthischangeeffort,helpedhimrealizeitwashisemotional
methodofpersuasionthatwasalienatingpeople.HeadvisedRyantoadoptamuchslower,
fact-basedapproach.HeexplainedthatwhileRyanhadtheluxuryofreadingandthinking
aboutScrumforweeksbeforecomingtoadecision,hewasexpectingotherstojumpon
thebandwagononnomorethanhissay-so.
Inresponse,Ryanstoppedhisactivecampaignandtookagentlerbutmoretime-
consumingapproach.Hehandedoutpapersandchapterstogivepeoplethefacts.He
involvedSteveasavocalsponsortohelpsoothepeople’sfearsandalsotohelpthem
understandtheneedtomakechanges.Finally,hereachedoutwithinthecompanynetwork
tofindreputablepeopletoadvocateanagileapproach,sothatitwasn’tjusthisadvicethat
peopleweretaking.Evenafterallthat,itwasstillseveralmonthsbeforehehadafullteam
readytobeginanewproject.
BringingpeopleonboardwithScrum,especiallyinabottom-upmanner,oftentakes
time,persuasion,andpatience.
ChangeTakesTime
Changeishard.Thestatusquo,asbadasitmaybe,isatleastknown.Venturingintothe
unknownrequiresquitealeapoffaith.Ryanwasabletomakethisleapbecauseitwas
cleartohimthatthepathhewasonwouldonlyleadtomorefailure.Stevewaswillingto
changebecausehehadseenScrumworkbefore.
Steveprovidedaclearvision:hisdesiretoincreaseoperationalefficiencyby50
percent.Ryanhadhisownsenseofurgency:Hewantedtohavemoresuccessfulprojects
andfeelbetterabouthisownjobperformance.Neitherwassurehowtoaccomplishhis
goals,butbothwereopentothepossibilitiesinherentinachangetoScrum.
Ina1995articleintheHarvardBusinessReview,JohnKotter[KOTTER]outlinedan
eight-stepmodelforaddressingchange.Asummaryofthatmodelisshowninthe
followinglist:
1.Establishasenseofurgency.
2.Formapowerfulguidingcoalition.
3.Createavision.
4.Communicatethevision.
5.Empowerotherstoactonthevision.
6.Planforandcreateshort-termwins.
7.Consolidateimprovementsandproducestillmorechange.
8.Institutionalizenewapproaches.
BothRyanandStevewerefollowing(albeitunknowingly)someofKotter’s
approaches.Whatfollowsaresomeconcretestepsyoucantaketohelpeasepeople’sfears
ofwhatlifewillbelikewithScrum.
EstablishaSenseofUrgency
Steve’sgoalwastoget50percentmoreefficientwithoutincreasingtheoverallgroupsize
(e.g.,addingpeople),andhewanteditdonewithintheyear.Steveissueditasachallenge
tothedivision,encouragingthemtofindwaystomakeithappen.Thiswasoneofthe
contributingfactorsthatpromptedRyantolookforwaystoaddnotonlytohisown
projectmanagementtechniquesbutalsototheperformanceanddeliveryoftheentire
divisionaswell.
Toconvincepeopletotrysomethingnew,youneedtogivethemareasontomakethe
change.Issuingachallengeorestablishingasenseofurgencyhelpspeopleunderstand
whyachangeisnecessary.Keepinmind,however,thatthechallengemustbereachable
andthesenseofurgencyreal.Peoplewillrisetoarealchallengebutwillshrugoffone
thattheyperceiveasimpossibleorfalse.
FormaPowerfulGuidingCoalition
Asponsorisapersoninanorganizationwhoultimatelyhastheresponsibilityforthe
successoftheprojectandforbreakingdownbarriersanddysfunctionwithinthe
organization.Inthischapter’sstory,Stevewasthesponsor.Stevehadhopedthatthe
divisionwouldturntoagileprinciplesandpracticesbuthadalsorealizedthatthedivision
wouldberesistanttoatop-downimplementationofScrum.Asaresult,hedecidedtowait
forsomeonetotakehischallengeofincreasingoperationalefficiency,whichwouldresult
inabottom-up,grassrootsapproach.Oncethegrassrootsmovementbegan,Steve’sjob
wastwofold:ClearthepathforRyanandalsoworkwithmiddlemanagementtohelpthem
understandwhattheirroleswouldbe.WhileRyanwasbuildinghisteam,Stevewas
workingbehindthescenestobuildacoalitionofsupportersthatwouldultimatelyenable
Ryantobesuccessfulinachievinghisgoal.
Agoodsponsorissomeoneintheorganizationwhobelievesinthechangeandhasthe
authoritytoworkwiththeleadershipteamandthemiddlemanagerstocommunicatea
vision,explainwhyitisimportant,andestablishatimelineforeffectingthechange.The
sponsorshouldalsobeabletohelpexplaintomanagerswhattheirroleswillbeinthenew
processandsolicittheirhelpinmakingithappen.Whennecessary,asponsorshouldalso
intervenewithpotentialsaboteurstoalleviatetheirfearsandconcerns.
CreateaVision/PaintaPictureoftheFuture
Beforetryingsomethingnew,peoplefirstneedtounderstandwhatthefuturewilllooklike
andhowtheywillfit.Inthischapter’sstory,Ryanexpectedpeopletojumponboard
basedonlyonhisenthusiasm.Hewasforcedtostepbackandtakeadifferentapproach.
HeremindedpeopleofSteve’sgoal,thattheyhadbeenaskedtoratchetuptheoutput,and
startededucatingthemaboutthemechanismshehadfoundtoachievetheirgoals:Scrum
andXP.HealsominedthecompanynetworkforpeoplewhohadtriedScrumandXPand
achievedsuccess.HefoundJimandWard,twodistinguishedpeopleintheagilefieldwho
happenedtoworkathiscompany.Afterabriefintroductionaboutwhathewantedtodo,
RyangotWardandJimtocometohisdivisiontogiveatalkaboutXP.Peoplestayed
aroundafterthemeetingtotalkaboutwhattheyheard,andtheybegantovisualizehowa
newapproachmightworkforthem.
CommunicatetheVision
Therearemanyapproachesyoucantakewhenintroducingpeopletoagileprinciplesand
practices.First,provideliteratureforthemtoread—books,whitepapers,researchpapers,
andthelike.Theseresourcesareagreatwaytohelppeopleunderstandwhathasbeen
tried,whattheissuesare,andthesuccessesandfailuresthatothershavehad.Literature
alsohelpsseparateemotionsfromthedata,enablingpeopletoreachtheirown
conclusions.
Noteveryonelearnsbyreading,though,soalsoconsidersociallearningmechanisms.
Createabookclub,havebrownbaglunchsessionswherepeopleinthegrouptalkabouta
topicpassionatetothem,participateinwebcastsandlocalusergroups,and,ofcourse,if
otherteamsyouknowofaredoingScrumorXP,askifyoucanwatchtheirreviewsor
planningmeetings.
Next,useyournetwork.Thereareoftenpeople,bothinsideandoutsideyourcurrent
company,whoareveryfamiliarwithScrum.Reachouttothoseinyourlocalareaand
becomegenuinelyinterestedinthem.Invitethemtopresenttoyourgroupabouttheir
experiences,howtheygotstarted,andtechniquesforsuccess.Eveniftherearenoexperts
inyourlocalarea,it’sstillhelpfultobringinoutsiderstosharetheirstories.Afterall,
peoplewilloftenlistentosomeonethattheydon’tknowandtrusttheirwords,evenwhile
disregardingpeopletheydoknow,whoaresayingthesamething.
EmpowerOtherstoActontheVision
Steve,asthedirectmanagerofthegroupandRyan’sprojectsponsor,wasalsothecreator
ofthevisionandtheonebuildingthesupportingcoalition.Assuch,hewasableto
empowerRyantodowhatneededtobedone.Hisbacking,ineffect,gaveRyanabadge
andagun.Ryanaskedforandwasgiventheauthoritytofindaprojectonwhichhecould
trythenewprocess.ThechallengeforRyanwasassemblingateamofwilling
participants.Steve’sbackinggaveRyanbothcredibilityandafootinthedoor.Fromthere,
however,itwasuptoRyantouseinfluencetobuildhisteam.
EvenafterpeoplewereconvincedthatScrummightbeagoodidea,theywerestill
hesitanttocommit.Meghangaveabunchofexcuses,butherrealfearwasthatshewould
failandloseherjob.Shehadagoodreputationanddidn’twanttoloseit.Thedecisionfor
Danielalsocamedowntowhetherornothe’dbesuccessful.Asthesponsor,makesure
youclearthewayforteammemberstoaddressanyissuesthatsurfaceontheteamorin
theproject.Reassurethemthattheycandowhattheyneedtodotoresolvetheirown
issueswithimpunity.
If,evenwithstrongbacking,peoplearereluctanttocommit,considerofferingthema
trialrun.Inatrialrun,theteammemberswouldhaveadesignatedperiodoftime(perhaps
threemonths)inwhichtheywouldagreetoputforththeirbesteffort.Attheendofthe
trial,theycouldeithercontinuewithScrumorwalkawayclean.Runningatrialhelpsto
removethefearoftheunknown,creatingacomfortzonewherepeoplecantrynewthings
withoutrepercussions.
Atrialruncanalsobeawaytoprovetotheteam,theskeptics,andtherestofthe
organizationthatthenewmethodologydoesn’tjustworkintheory,butitworksin
practice,inthiscompany,andonaspecifictypeofproject.Itcanalsoprovideinsights
intothingsthatdon’tworkinitiallyandhowthesechallengescanbeovercome.Thedata
fromatrialruncanbeusedtohelpbuildacoalitionofsupporters.Overtime,fearwillbe
replacedwithconfidence.
PlanforandCreateShort-TermWins
Scrumisdesignedtoprovideshort-termwins.WhentransitioningteamstoScrum,help
themunderstandthatthepotentiallyshippableproductincrementisnotintendedto
highlightwhattheyhaveyettofinish;onthecontrary,itistohighlightwhattheteamhas
accomplishedandtogetcustomerfeedback.Itisalwaysbettertofailearlywhenthe
investmentissmallthanitistofaillatewhentheinvestmentislarge.Whenshowing
workingsoftware,alwaysremembertobetransparentandhonest—beupfrontaboutboth
thegoodandthebadoftheprojectandembracethechangesthatarebeingmade.
Short-termwinsalsohelpbuildthetrustbetweentheteamandthestakeholders.Trustis
fragileinthebeginning.Showingworkingsoftware,evensmallbitsofit,willhelp
increasetheconfidenceoftheteammembersandthestakeholders.
ConsolidateImprovements
Helppeoplerealizethattheywillnotseea100percentincreaseinanythingoutofthe
gate;ittakestime.Justasaweightliftercannotexpecttostartwitha300-poundweight
butmustbuilduptoit,teamscan’texpecttosuddenlybepowerfulandstrong.Thatkind
ofchangetakestraining,practice,andtime.
Astheprojectprogresses,newpracticeswillbecomeingrainedhabits,makingthework
itselfeasier.Atthesametime,improvementswillgetharderandsmaller.Newchampions
ofScrumwillemergeaspeoplefindthewaytheyareworkingnowtobeasignificant
improvement.
InstitutionalizeNewApproaches
Iliketothinkofinstitutionalizednewapproacheslikeachangeindietorexercise.Atfirst
it’spainful.Asyouprogress,though,youachieveshort-termwinsandgetbetterandbetter
byfindingwhatworksforyou.Eventuallyyouwonderhowyoueverdidwhatyouusedto
doasthenewwayjustfeelsright.Agileapproachesarenodifferent.
Astheconsolidatedimprovementsbecomecommonplace,teammemberswillbeginto
wonderwhatitwasliketoworkinanon-agileenvironmentwheretheyonlyshippedonce
ayear.Whenthishappens,askthesechampionstospawnoffandcreatenewteams,
furtherinstitutionalizingtheframeworkacrosstheorganization.
KeystoSuccess
Evenwhenyoudoalltherightthings,winningpeopleoverischallenging.IttookRyan
fourmonthstobuildateam;itmighttakeyoujustaslongtobringpeopleonboard.Itall
dependsonthecultureofthecompany,themanagement,andofcourse,thepeople
themselves.Bottom-upeffortstakeawhiletospread.Themostimportantkeystosuccess
arepatienceandinformation.
BePatient
Oneofthehardestthingstodoisletotherscometoarealizationyou’vealreadyreached.
ButasStevelearnedathisformercompany,pushingasolutiondownpeople’sthroatsis
equallychallenging.Planttheseedsbutbewillingtoletthemgrow.Peoplelearnat
differentpacesandusingdifferentstyles.Whatconvincedyoumaynotworkfor
everyone.Askyourself,“HowwouldIfeelifthesepeoplewereaskingofmewhatIam
askingofthem?”Often,youranswertothisquestionwillcauseyoutochangeyour
approach.
Findoutwhatotherpeople’striggersare.Listenwhenothersexplainthattheydon’t
understand,areafraid,orarereluctanttochange.Askquestionsdesignedtoelicitanswers
thatwillallowyoutouncoverandaddressunderlyingconcerns.Trynottoletwhatyou
perceiveasalackofprogressfrustrateyou.Thingsofgreatvaluetakeafairamountof
time.
ProvideInformation
Ryangavepeoplehighlightedchapters,researchpapers,andcasestudiestohelpprovehis
case.Bepreparedtogiveyourowncolleaguessomeharddatatoovercometheirinitial
reluctance.Ryanalsoarrangedforpeopletotalktotrustedindividualswhohadsucceeded
withanagileapproach.Youshoulddothesame.Trainingisanothergreatoptionfor
helpingopenpeople’smindstoanewwayofthinking.
Rememberthatthoughthecurrentrutmaybepainful,mostpeopleareunwillingto
moveunlesstheytrulybelievethereisabetterpath.Yourjobistoclearlyexplainthe
alternativeyouareproposingandtowaitpatientlywhilepeoplemakeuptheirminds.In
theend,theonlypersonyoucancontrolisyourself.Throughoutthechallengingprocess
ofgettingpeopleonboard,keepyourheadup.Youmayhearnomoreoftenthanyes.Stay
focused.Staypositive.Staystrong.Thisischallengingstuff.
Reference
[KOTTER]Kotter,JohnP.2007.“LeadingChange:WhyTransformationEffortsFail.”
HarvardBusinessReview73(2):59.
WorkConsulted
Schwaber,Ken.2004.AgileProjectManagementwithScrum.Redmond,WA:Microsoft
Press.
Chapter3.UsingTeamConsultantstoOptimizeTeam
Performance
InTheFifthDiscipline,PeterSengedefinedlearningorganizationsas“organizations
wherepeoplecontinuallyexpandtheircapacitytocreatetheresultstheytrulydesire,
wherenewandexpansivepatternsofthinkingarenurtured,wherecollectiveaspirationis
setfree,andwherepeoplearecontinuallylearningtolearntogether”[SENGE].Creatinga
teamstructurewherepeoplecangrowtheirskillswhileaccomplishingbusinessgoalsisa
challenge.
Datashowthatthebestteamsizesarebetweenfiveandninepeople,allofwhomare
fullydedicatedtoaprojectforthedurationoftheproject,andwhoworktogetherina
cross-functionalwaytodeliverworkingsoftwareattheendofeverysprint[PUTNAM].
Whatmanycompaniesfind,however,isthattherearen’tenoughgoodpeopletogo
around.Creatingoneortwoteams?Sure.Creatingawholecompanyoffullycross-
functional,dedicatedteams?Muchmoredifficult.
Somepeoplehavegreatskillsbutdon’tnecessarilymakegoodteammembers.Others
areinsuchhighdemandthatit’shardtoholdthemononeteamforthelengthofthe
project.Inlightoftheserealities,howcanyoustructureateamthatisdedicatedtothe
project,containsalltheskillsyouneedtogetthejobdone,andprovidesopportunitiesfor
personalgrowth?
ThiswastheproblemfacingRebecca,aprojectmanageratamid-sizedcompany.Her
solutionisonethathasworkedatmanycompaniesandinmanysituations.Implementing
it,however,requiresanopenmindandanewwayoflookingatteams.
TheStory
Rebeccaisamanagerinamedium-sized(300+)softwarecompany.Today,asshedoes
everysingleFriday,sheissittingintheresourceplanningmeeting.Andtoday,asitdoes
everysingleFriday,themanagementteamisreviewingwhateveryoneisworkingon,
whichprojectsarecomingup,andhowtojugglekeysubjectmatterexpertsamong
competingprojects.WejoinRebeccainthemiddleofthismeeting.She’sboredandis
passingthetimedoodlingonhernotepad.Throughthehazeofherdoodlingandthebuzz
thataccompaniesthebusinessofshufflingpeoplefromprojecttoproject,itsuddenlyhits
her:Shewillneverhaveenoughskilledpeopletocreateadedicatedteamforher
upcomingCougarproject.Shestopsdoodlingandbeginstotunein.Justasshesuspected,
noneofthepeopleshewantsareavailableforadedicatedteam.Whatdoesthatmean?It
meansshe’sgoingtohavetoeithershareateamwithotherprojectsorcreateasecond-rate
teamthatshecanhavealltoherself.
“Ijustneedtochoosethelesserofthetwoevils,”shedecides.“That’sprobablya
sharedteam.Sowhatiftheteamissplitacrossmultipleprojects?IfItimeitright,Ican
cherry-picktheskillplayers.I’dhaveanall-starteam.”
Rebeccastartstogetexcitedaboutthisoption.Glancingattheschedule,sheseesthat
thestardevelopersandtestersareeachavailablefor20hoursperweekduringherproject
—andtheycanbeallhersifshelockstheminnow.Justassheleansforwardtospeakup,
shehearssomeoneshufflingyetanotherpersonoffoneprojecttohelpsaveanother.
Rebeccasitsbackinherseat,recallingherPoodleproject.Onthatproject,shehadMark
andBill,whoweresupposedlyfullydedicatedtoherproject.Yetwhenpushcameto
shove,theyweretheonlypeoplewhocoulddocertainwork,sotheygotcalledoffher
teamtohelpothers.Eventually,theirsplittimecausedherprojecttoslip.“Thatwon’t
work,either,”sheconcludes.“EvenifIgetmyall-stars,intheendtheywilljustbepulled
offmyprojecttohelpsaveothers.Thatwillleavemeoverbudget,understaffed,and
strugglingtomaintainquality.”
“WhatotheroptionsdoIhave?”shewonders.Shepicksupherpenandbegins
doodlingagain.Shebeginstolistthepeopleshecangetfull-time.
Larry:greatatSQLandreallybadatUI[userinterface],butverywillingtolearn
newskills
Scott:wonderfulall-aroundtesterwithsomedevelopmentbackground
John:C#andarchitectureaficionado
Michelle:greatdatamodelingskills,veryopenminded,andworkswellwithothers
Thosepeoplewouldmakeadecentteam,buttheycan’tdoeverythingtheprojectneeds.
“WhoelsecouldIchoose?”Rebeccawonders,lookingupattheresourcechart.“Nobody
I’dwant.ThepeoplewhohavetheskillsIneedandareavailablefull-timearenotteam
players.They’ddomoreharmthangood.1Rebeccatossesherpenonthetable.Nomatter
howsheslicesit,adedicatedteamcan’tcoveralltheprojectneeds.
1.SeeChapter21,“WhenCulturesCollide,”foranexampleofateammemberwhodoesmoreharmthangood.
Suddenlyanideacomestomind.“WhatifIcanhaveboth?”Rebecca’sideaissimple
yetoutsidethemainstreamformostcompanies.Shepicksupherpenanddrawsabig
circlearoundherkeyplayers.“Acoreteam,”shemuses.“Icouldhaveacore,dedicated
team,supplementedwithapoolof‘teamconsultants,’”sherealizes.Theseteam
consultantswouldbecompanyemployeeswhodon’tworkforanyteamfull-timebut
insteadareoncalltosupplytherightskillsattherighttime.Thiscouldwork.Theonly
problemisthatthecompanycurrentlydoesn’thavethatpoolofteamconsultants.That’sa
problemshecansolve.Shespeaksupinthemeeting.
“Everyone,Iapologizeforinterrupting,butI’mnoticingarecurringtrend.Duringthe
last30minutes,we’vehadtomovesevenpeoplefromtheircurrentlyassignedprojectto
anotherprojectindistress.Why?Becausetheyaretheonlypeoplewhomwebelieveare
capableofdoingaparticulartask.Iagreethattheyhavespecialskills,butwheredoes
movingthemleavetheiroriginalprojects?Itleavesthoseprojectsinthelurchandin
dangerofneedingsavingaweekortwofromnow.That’swhere.Andwillmovingthem
worktosavethetroubledprojects?Thetruthis,wedon’tknow,butthegamewe’vebeen
playinghaslikelybeencreatingthetroubledprojectsmorethanfixingthem.”Aheavy
silencefollowsRebecca’swordsaseveryheadturnstolookather.
“Whatothersolutionsdowehave?”asksRebecca’sboss,Dave.“Shiftingpeople
aroundistheonlywaytomanagetheworkwehave.”
Rebeccalooksdownatherhalf-bakedidea,sketchedamongabunchofdoodles.She
forgeson,unabletopassuptheopportunityDave’squestionpresents.
“Well,Iwasthinking.ThemostsuccessfulprojectsIhaveeverrunhavehada
dedicatedteam—”
“That’simpossiblehere,”interruptsDave.
“Iknow.Hearmeout,”saysRebecca.“Traditionally,dedicatedteamshaveallthe
personneltheyneedtobesuccessful,fromdevelopersandtesterstodesignersand
architects.I’mnotaskingforthat.I’msuggestingthatwhatIneedisadedicatedcore
team.Theycandomostoftheday-to-daywork.Forthosespecialtimeswhenweneeda
SQLexpertoranarchitectoraUIdesigner,I’dpullfromapoolofspecialistswhoaren’t
assignedtoanyoneteam.Apoolofinternalconsultantsthatwecouldall”—shegestures
aroundtheroom—“utilizeasneeded.They’dactasconsultantstothecoreteam.Inthe
shortterm,theywoulddo,orhelptheteamdo,thespecializedtasksthatthecoreteam
isn’tcomfortabledoing.Overthelongterm,ifweusepairingandcoachingcorrectly,the
coreteammemberscanpickupnewskillsandtakeonmoreworkthemselves.We
accomplishourbusinessgoalswhileworkingtowardourculturalgoalofbeingalearning
organization.”
Daveshakeshishead.“Rebecca,Iappreciateyoureffortintryingtosolvethisproblem,
butIreallydon’tthinkitwillwork.Wehaveaverycomplexorganizationalstructure.We
needtokeepourorganizationmatrixedsowecansolveissueslikethatwhentheycome
uponourprojects.This‘teamconsultant’roleyouaretalkingaboutjustseemslikeitis
notgoingtowork.HowwillImanagetheirtime?Whatguaranteecanyoumakethatthese
problemswon’tkeephappening?”
Rebeccathinksforamoment.Davedoesn’tseemaversetotheideaasmuchasunsure
ofwhyitwouldworkandhowhewouldmanageit.Shethinksofananalogythatshe
knowswillspeaktoeveryone.
“Okay,everyone—question:Haveyoueverhadyourhouseremodeledorhadanew
housebuilt?”sheasks.Overhalfofthe40peopleintheroomraisetheirhands.“Now,
howmanyofyouwerecompletelysatisfiedwithyourgeneralcontractor?Howmanyof
youhadcontractorswhoshowedupontimeanddeliveredwhattheypromised?”
Everyonelaughs,butnooneraisesahand.
“Yep.That’swhatIthought.Sad,right?Okay.Nowhowmanyofyouhaveusedthat
samecontractoragainafteryourinitialexperience?”Rebeccaasks,eyebrowsraised.
Again,nohands.
“Nobody?Whynot?”sheasks.
Abarrageofanswerscomeout.
“Theybrokeourtrust.”“Theywouldshowuplateornotatall.”“Theydidn’treturnour
callsinatimelyfashion.”“IfeltlikeIhadtohoundthemtogetourworkdone.”“They
wereconstantlyfullofexcuseslike‘thepersontodothetilingisonanotherjobtoday.’”
Thelistgoesonandon.
“Right.Ihearyou.Now,howmanyofyouhavesincehiredspecialists(maybepeople
youmetthroughyourgeneralcontractor)todoajob?”Rebeccacontinues.
“Ihave,”saysIan.“Thetileguy.He’sgreat.Heshowsupwhenhesays,alwayscalls
meback,andisverygoodathiscraft.Infact,hejustwentoutonhisownasan
independent.”
Rebeccaasks,“Doyouthinkthatbeingonhisownmakeshimmoreaccountableto
you?”
“Absolutely,”saysIan.“Hisreputationisatstake.Allhisbusinessisbyreferral.Ihave
senthimtoallmyneighborsforwork,andtheylovehim,too.”
“Doeshemanagehisowntime?”asksRebecca.
“Well,hesurebetter.He’stheonemakingthecommitmentstohiscustomers,andhe’s
theonedoingthework.Itrusthimtomanagehistime,”saysIan.
“Thatisakeypoint,Ian,”saysRebecca.“Forpeopletobeabletofollowthroughon
theirwork,tobefullycommitted,theyhavetomakethecommitmentforthemselves.In
thismeeting,thisresourceplanningmeeting,whatarewedoing?Wearemaking
commitmentsforresources,peopleotherthanourselves.Wearejugglingthemaroundlike
pawns,notconsideringtheirtime,theirpreviouscommitments,ortheirinterests.Weare
simplylookingattheirskillsandapplyingthemtoprojectsthatareinurgentneedofthose
skills.Worse,wedon’ttrustthemenoughtoallowthemtodoitthemselves.”
Sheglancesaroundtheroomagain,“Everyone,we’reactinglikegeneralcontractors,
andwe’resurprisedwhenitdoesn’twork.Ourcustomerstrustustodeliverontime,on
budget,andwiththequalitytheyexpect.What’sgoingtohappenifwedon’tmeettheir
expectations?”
“Weloseourcustomers,”saysavoicefromtheback.
“Right,weloseourcustomers!”exclaimsRebecca.“Topleaseourcustomers,weneed
tohaveemployeeswhoarevestedintheprojects.Thebestwayforustodothisistohave
themsignupforworkandmanagetheirownschedules.We’vesaidwewanttobea
learningorganization,acompanyfilledwithempoweredemployeeswhoarecross-trained
andsatisfiedwiththeirwork.Thismodelseemslikeawaytopleaseourcustomersby
deliveringwhattheyneedwhentheyneedit.Atthesametime,itgivesusawaytofoster
happyemployeeswhowanttostaywiththecompanylongterm.”
Daveisvisiblyflustered.“Rebecca,youcan’tbeserious.Ourdeliveryteamsarenot
capableofmanagingtheirownschedules.That’swhywehaveresourcemanagers.”
“No,Dave,wehaveresourcemanagersbecauseweare,howdidyousayit,ahighly
matrixedorganizationanddon’tpushaccountabilityorcommitmenttothepeopledoing
thework.Thestructurewehaveinplacemakesittoomuchofachallengeandfartoo
bureaucraticforourpeopletotrytomanagetheirowntime;that’swhywecreatedthe
resourcemanagerrole.Ian’stileguywentoutonhisownbecausehewasabletomanage
histimeandbringinworkforhimself.Hemadecommitmentsbasedonhisscheduleand
hisskillset.Wedon’tseeIancommittinghistileguytoinstallwindows.Sure,hemight
beabletoinstallawindowifhereallyneededto,butwhathedoesbestistile,sothat’sthe
servicehesells.Wehavepeopleinourcompanywhoaregreatgeneralists—theylove
doingalittlebitofeverything—andwehavepeopleinourcompanywhoarevery
specialized—greatatdatabasedesignandobject-orienteddesignpatternsorfantasticat
systemandnetworkdesign.Theyliketheirspecializationsandwanttofocusonthem.Our
problemisthatwearemanagingeveryone’stimeforthem.Whenwedothat,ifteamsfail
todeliver,theycanblameusbecausewearetheonestellingthemwhattodo.Theyare
notaccountabletotheirprojects;theyarejustexecutingwork—andwedon’tenablethem
todotheirbestworkbecauseweholdthemback,”saysRebecca.
Daveseemsatalossforwords.
Rebeccagoesinforthekill,“Dave,tellyouwhat,letmerunanexperimentonthe
Cougarproject.Givemelicensetoapproachafewpeopleaboutbeingpartofacoreteam
thatisfullydedicatedtotheproject.Iftheyagreetoworktogetherandwithme,we’llthen
sitdownandidentifysomecandidatestobeourteamconsultants.Iwanttobeabletotell
everyonethatwehaveyoursupportincreatingahybridteamofcoremembersandafew
peoplewhowillfloatbetweenprojects.Inreturn,we’llkeepyouandtherestof
managementwellinformedofhowthistrialisgoing.”
BeforeDavecanobject,Jim,thedirectorofthegroup,speaksup.
“Rebecca,”saysJim,“I’mgoingtoletyoutrythis.Twocaveats:Youwillworkinour
customers’bestinterests,andyouwillbeatestbedfortherestofthecompany.”Helooks
atDave.“She’sright.Toomanyhoursarewastedinthismeeting,andtoomanypeopleare
shuffledfromprojecttoproject.Thisproblemhastostop,andthisisthebestsolutionI’ve
heardtodate.Rebecca,letmeknowwhatyouneedtomakethisasuccess.”
Withthat,Rebeccahaswhatsheneedstoimplementthemodelthatstartedoutasa
doodlebornofboredomanddesperation.
TheModel
Inthischapter’sstory,Rebeccawasdealingwithfrustrationinsidehercompany.Shewas
abletoimplementasolutionforherteamthateventuallyspreadthroughtheentire
company:theuseofteamconsultants.
Ateamconsultantissomeoneinyourorganizationwhoisavailableforsomeamountof
workanddirectlyfillsaskillgapbetweentheteamandtheproject.Teamconsultantsare
notcoreteammembers.Theyhavenoteam,perse.Theychooseinsteadtoofferuptheir
servicesasinternal“gunsforhire,”providingspecificexpertiseasneeded.Team
consultantsoftenworkformultipleteamsandaretypicallyveryspecialized,so
specialized,infact,thattheymightnotmakegoodcoreteammembers.Thatdoesn’tmean
theteamconsultantshoardtheirknowledge,though.Aspartofalearningorganization,
teamconsultantshelpothersbyteaching,givingadvice,coaching—whatevertheteam
needsduringasprint.Andtheteamconsultantsgrow,too,learninghowtoworkaspartof
across-functionalteamandhowtosharetheirexpertisewithothers.
Thoughteamconsultantsoccasionallymaybebookedforthelengthoftheproject,they
typicallydonotworkfull-timeonthatproject.Theyarehiredspecificallytosolvean
immediateproblemortocontributealimitednumberofhourspersprint.Thecoreteam,
ontheotherhand,ismadeupofteammemberswhosetimeisfullydedicatedtothe
project.Collectively,theyareresponsiblefordeliveryofthework.Theyareacross-
functionalteamwhoworkononlyoneproject,eliminatingcontextswitchingand
multitasking.
JustasaScrumteamdoesn’thavetotakeonalltheworkproposedbytheproduct
owner,ateamconsultantdoesnothavetoworkforeveryteamthatasks.However,team
consultantsshouldmakethesego/no-godecisionswithinthecontextofbusinesspriorities,
personalavailability,andtechnicalexpertise.
Implementingtheteamconsultantmodelisatwo-partprocess:establishingateam
consultantpoolandbuildingyourteam.Ifyouarelimitingyourimplementationtoaone-
teamtrial,however,youcanskipthefirstpartuntilyoudecidetoscalethemodel
throughoutyourorganization.
EstablishingaTeamConsultantPool
Establishingapoolofavailableteamconsultantsiseasierthanitmayseem.Itinvolves
thesetwosteps:
Step1.Developandcommunicateatransitionplan.
Step2.Letpeoplechoose.
Step1:DevelopandCommunicateaTransitionPlan
Thefirstthingyouneedtodoisdevelopatransitionplanthatincludesthereasonforthe
change,thevisionofthefuture,andthepathtoachievingyourgoals.
Detailthecommonproblemsthatyouareexperiencingwithinyourorganization.
Communicateasharedvisionastowhyitisimportanttocreatealearning
organization.
Helppeopleunderstandthattheirrolesarenotgoingaway;onthecontrary,theyare
beingenhanced.
Reinforcethefactthatnow,morethanever,individualandteamcommitmentsdrive
customersatisfactionand,ultimately,businesssuccess.
Thenextthingyouneedtodoisaskpeopletosignupaseitherteamconsultantsorcore
teammembersbyacertaindate,whichleadstostep2:Letpeoplechoose.
Step2:LetPeopleChoose
Thepersoninthebestpositiontodeterminewhetherheorsheshouldbeacoreteam
memberorateamconsultantisthepersondoingthework.Communicatethedifferences
inthetworoles,perhapsevenpostingatablesuchasTable3-1thatliststhebenefitsand
downsidesofeachrole.
TABLE3-1Roles,Benefits,andDownsides
Teamconsultantswillworkonmultipleprojectsandwillneedtomanagetheirtime
accordingly.Teamconsultantsoftenchoosethisrolebecausetheyarepassionately
dedicatedtotheircrafts.Thesepeoplehavebeenworkingintheirsubjectmatterforso
longthattheyarethego-topeopleforcertaintypesofwork.
Signinguptobeateamconsultant,however,doesnotmeanthatyougettopractice
yourcraftwithouthelpingothers.Onthecontrary,wheneverpossible,teamconsultants
shouldworktobringuptheskillsoftheentireorganization.Thisiseasierinsomeareas
thanothers.Forinstance,noonewouldexpecttoseeagraphicdesignerteachingdesign
elementstoaten-yearveteranofinfrastructureprojects,butitwouldnotbeunusualtosee
aSQLexperthelpingoutacoreteammemberwhohassomeSQLskillsbutneedstolearn
moretobemosteffectiveinhercurrentproject.Teamconsultantsalsoarenottheretotell
theteamhowtodothework.Instead,theyadviseandinfluencetheteamastowhich
directiontotake.Consultantsworkalongsidetheteambutprovideleadershipaswell.
Atypicalweekforateamconsultantwithatechnicalbackgroundincludescoding,code
reviews,documenting,solvingtechnicalproblemstheteamisstuckon,providing
architecturalguidance,ordoinganythingelsethattheteamasks.Forcreativeservices
suchasgraphicdesignandtechnicalwriting,thetimewilllikelybespentwiththeteam
and/orproductowner,reviewingtheitemsneeded.Technicalwritersespeciallyneedtobe
awareofthe“done”criteriafortheteamandhelptheteamdeliverallrequired
documentationeachsprint.
Itisuptotheteamtoscheduletimedirectlywiththeteamconsultantandtomakesure
theworkisqueuedwhentheconsultantarrives.Ifateamconsultantoverbooksherselfon
toomanyprojects,it’suptohertofixthesituation.Thebenefittheteamconsultantsseeis
thattheycanstayfocusedontheircoreskillsandthattheycanbeinstrumentalinteaching
others,therebygrowingandbuildingthecompanywhileraisingthebaracrosstheboard.
Thepeoplewhomakeupthecoreteam,ontheotherhand,areskilledinmultipleareas.
Theyarenotexpectedtobeexperts;infact,theteamisdesignedtoprovidecross-
functionalbalanceofexpertise.Likeateamconsultant,acoreteammemberisresponsible
forhisorherowntime,butallthistimeisdedicatedtooneteamandoneproject.Core
teammemberscanstayfocusedonasingleprojectthroughoutitslifecycleandcanlearn
newskillsbyworkinghand-in-handwithothercoreteammembersandteamconsultants.
Theabilitytostaywithoneteamandprojectyetlearnnewskillsisanadvantagethat
shouldnotbedownplayed.TwoofthebiggestcomplaintsIhearatcompaniesaroundthe
worldarethefollowing:
1.Ihavetosplitmytime.
2.Ineverlearnnewthings.
Asaresult,peoplelookelsewhereforbetteropportunities.
Onacross-functionalScrumteam,coreteammemberscontinuallylearnnewskillsand
areabletostayfocusedonasingleproject.Assuch,thebusinessprofitsbyhavinga
stableworkforcethatiscontinuallylearning,enablingthebusinesstoremaincompetitive
andresponsivetochange.
BuildingYourTeam
Whetheryouchoosetofirstbuildacompanywidepoolofconsultantsorarejusttryingout
themodelonasmallscale,thefirstthingtoconsiderwhenbuildingateamistheskillsets
youneedfortheproject.Ifyouarenottechnicalenoughtodothistask,askforhelpfrom
acolleaguefamiliarwiththeproject.Afteryou’veidentifiedthetechnicalskills,you
shouldalsoconsidersoftercompetencies,suchascommunicationskills,open-
mindedness,andtheabilitytobeateamplayer.Thesecompetenciesareoftenmuch
hardertodetermine,butarevitaltosuccess.Thelistlookssomethingliketheoneshown
inthefirstcolumnofTable3-2.
TABLE3-2Samplelistofcompetencies,skills,andavailabilitywithsubjectiveanalysis
ofpotentialteammembers
Onceyouhavelistedtheskillsandcompetenciesyouarelookingfor,beginidentifying
thepeopleyouhavetochoosefromandmapthemtoyourwishlist.Includetheir
availabilityaswell.Rankeachindividualonascaleofzerotofivestars.Remember,there
isnothingquantitativeaboutthistask;itispurelysubjective.Youmightwanttoaskfor
helpwithidentifyingyourcandidates,justfortheaddedperspectiveasecondopinioncan
addtotheprocess.Theresultingchart,fullycompletedasshowninTable3-2,willbe
customizedtoyourownpeopleandrequirements;it’samoreformalizedversionofthe
analysisRebecca,fromthestory,didonthefly.
CoreTeam
SupposeTable3-2isthechartthatRebeccacreatesfollowinghermeeting.Lookingat
Table3-2,youcanseerightawaythatbothJohnandMichelleareveryavailable,and
togethertheybringthreekeyskillstothetable:architecture,SQL,anddatamodeling.In
addition,theyarebothconsideredgoodteamplayers.Asithappens,Rebecca’sprojectis
heavyinSQLwork,sothoughMichellehasSQLskills,Rebeccadecidesshewillneed
evenmorehelpinthatarea.LookingatherotherSQLpeople,RebeccacirclesLarryand
Scott;theyarebothmarginallyavailablebutcanprobablybecomefull-timemembers.Her
prospectsforthededicatedteamareJohn,Michelle,Larry,andScott.Asyoucanseeby
lookingatFigure3-1,Rebecca’scoreteamdoesnotcovertheentiresetofskillsneeded
forthisproject.
FIGURE3-1Thecoreteamcan’tcovertheentireproject.
TeamConsultants
Onceyouhaveformedyourcoreteam,meetasagrouptodetermineyourownstrengths,
weaknesses,andgapsinprojectcoverage.Ifyoufindyouhavegapsinskillcoverage,it’s
timetoconsiderwhichteamconsultantsyoumightbeabletoutilize.
Havetheteamreachouttoindividualsintheteamconsultantgroup(orinthe
organizationasawholeifyoudonotyethaveadedicatedpoolofteamconsultants)who
meetthoseneedsandwhotheyfeelwouldworkwellwiththeteamonalimitedbasis.
SupposethatRebecca’steamgetsacommitmentfromMichaeltobeavailableforC#code
reviewsandsomepairinghelponAPIstheteamisnotsureabout.Theteamalsoasks
DavidandStefantobeconsultantsforUIandAjax,respectively.Thefinalteamchart
looksliketheoneinFigure3-2.Noticehowtheteamconsultantsclosethecoreteam’s
coveragegaps.
FIGURE3-2Theteamconsultantsfillthevoidsontheproject.
SizeMatters
Everyonehasaninherentunderstandingthatteamsthataresmallerinsizetendtoproduce
faster.Addingpeoplecomeswiththeadditionalcostoframp-uptime.IntheMythical
ManMonth,FredBrooksputouttheprincipleknownasBrooks’law:“Addingmanpower
toalatesoftwareprojectmakesitlater”[BROOKS].Butisthisalwaysthecase?
Ina1999article,SteveMcConnellsuggeststhat“controlledprojectsareless
susceptibletoBrooks’lawthanchaoticprojects,”becausepeoplecanbesafelyaddeddue
tobettertracking,documentation,anddesign[MCCONNELL].However,even
McConnellacknowledgesthateveryprojectreachesapointwhenaddingpeoplebecomes
counterproductive.Thispotentialforcounterproductivitythenbegsthequestion—howbig
shouldacoreteambeandhowmanyconsultantsshouldithave?
AstudydonebyLawrenceH.PutnamandWareMyersin1998investigatedteamsize
[PUTNAM].TheirfindingswerepublishedintheCutterConsortiumin1998.Putnamand
Myerslookedat491medium-sizedprojects,thosewithbetween35,000and95,000“new
ormodified”sourcelinesofcode.Allwereinformationsystemsandwerecompleted
between1995and1998(thelastthreeyearspriortopublishing).
WhatPutnamandMyersconfirmedwasnotonlythatsmallerteamsoffivetoseven
deliveredinlesstime(seeFigure3-3)butalsothatasignificantincreaseineffortoccurred
whenteamsizesreachednineormore(seeFigure3-4).
FIGURE3-3Averageschedulemonths—smallertakeslesstime
FIGURE3-4Exponentialdevelopmenteffortgrowthwhenteamsizeincreasesbeyond
eight
Whenworkingwithteamconsultants,therightteamsizeshouldbefourtosixcore
teammembers,allowingyoutoaddbetweenoneandthreeconsultantswhilestillkeeping
theteamsizedcorrectly.Remember,consultantswillcomeandgo,butthecoreteam
shouldremainthesame.
CoreTeamsWorkingwithTeamConsultants
Onceyou’vebuiltateamofcoremembersandconsultants,youneedtoestablish
guidelinestoencourageteamwork.Thespecificguidelinesvaryamongorganizationsbut
shouldincludethefollowing:
Duringthecourseofthesprint,theteamandteamconsultantsarejointlyand
equallyresponsiblefordeliveringontheircommitments.
Teamconsultantsshouldshowupontime,bepartofdailymeetingswhenever
possible,andconductthemselvesasteammembersforthelengthofthesprint.
Teamconsultantsareexpertsinoneareabutotherwisenodifferentthanacoreteam
member.Thereisnohierarchyamongcoreteammembersandconsultants.
Ifanyoneoranythingisblockingatask,thetaskshouldbebroughtupatthedaily
scrumandaddressedimmediately.
Remember,too,thatyoucannothaveacoreteammadesolelyofconsultants.The
majorityoftheteamshouldbededicatedtotheprojectfrombeginningtoend.
TeamConsultantsandMeetings
Whenevereconomicallyfeasibleandlogisticallypossible,theteamconsultantsshould
attendallscrummeetingsduringthesprintsforwhichtheyareengaged.Rememberthat
theteamconsultantisthereforaspecificreason,notasanemergencyoverflowperson.
Consultantshavespecifictasksforthesprint,whetherthatistodotheworkdirectly,help
theotherteammembersdotheworkthemselves,orboth.Assuch,theyshouldbeableto
reportonprogressinadailymeetingandshouldhaveinputintoplanningandreviews.
Planningmeetings—Havingtheconsultantsattheplanningmeetingscanadda
newlevelofexpertise.Iftheteamisestimatingandcomesacrossastorythatitis
notasfamiliarwithastheteamconsultant,theconsultant’sinputwouldbe
beneficial.
Dailyscrummeetings—Havingconsultantsintheroomhelpsdriveawarenessand
keepsthemuptospeedonhowtheprojectisprogressing.
Sprintreviewmeetingsandretrospectives—Havingtheconsultantsintheroom
duringthereviewmeetingshowsthecustomereveryonewhoisworkingonthe
project,bothfull-timeandasaconsultant.Havingthemintheretrospectiveis
beneficialbecausetheconsultantsgetabetterunderstandingofteamdynamics,but
rememberthatthetimetheyspendintheretrospectivecomesoutofthetotal
availabilitypooltheyhave.
Ifbudgetaryortimeconstraintsmakesprintmeetingattendanceimpossible,theteam
shouldworkwiththeproductownertodecidehowbesttospendtheconsultant’stime.Ifa
consultantisavailableforonlyeighthoursduringasprintandthetasksheisbeingasked
tocommittowilltakethefulleighthours,itmightbemorevaluabletotheteamand
productownertogetmoreworkdonethantohavetheconsultantinthesprintmeetings.
Ontheotherhand,theteamandproductownermightchoosetodialbackaconsultant’s
taskstohavehimbeapartoftheday-to-daysprintactivities.Whateveryoudecidetodo
withregardtomeetings,thecoreteam,Scrum-Master,consultants,andproductownerall
needtobeawareofthecommitmentsandarrangementsmadebetweentheteam
consultantsandthecoreteam.
KeystoSuccess
Usingteamconsultantscanbeagoodideanotonlybecauseitfillsskillgapsbutalso
becauseitcreatesanenvironmentwhereone’sjobstabilityisdirectlyrelatedtothework
donewithandforothers.Inthischapter’sstory,Rebeccacomparedtheteam’sabilityto
hire(andfire)consultantstoahomeowner’sabilitytohireanindependentcontractorto
makerepairstoahouse.Homeownershirespecializedpeoplewhotheyknoworsuspect
aregoodattheirjobs.Thecontractorsworkhardtopleasethosehomeownerssothatthey
getreferralsandrepeatbusiness.Havingteamconsultantsenablesteamstofollowa
similarmodel.
Accountability
Accountabilityisastrangething.Whileitisnotsomethingthatcanbemandatedor
managed,companiesstilltry.TakeDavefromthischapter’sstory.Hismatrixedsolution
assumedthatpeoplecouldnotself-organize.Rebecca’s,ontheotherhand,pushedthe
responsibilityformanagingone’stimedowntotheindividual.InScrum,thepeoplewho
estimatetheworkdothework.Thesamephilosophyapplieshere—thepeoplewhodothe
workcommittothework,whethertheyareteamconsultantsorcoreteammembers.Team
consultantsareaccountabletothepeopletheysignuptodoworkfor.Coreteammembers
areaccountabletoeachotherandtotheteamconsultantsaswell.
Oneofthekeystomakingthismodelworkistheself-monitoringandbalanceof
accountability.Asyoucanimagine,thismodelquicklyidentifieslowornonperformers.
Forinstance,imaginethatDavidisateamconsultant.Davidfrequentlycommitstoteams
buteitherfailstodeliverontimeorturnsinlow-qualitywork.TheendresultisthatDavid
doesnothavemanyteamswillingtoworkwithhim,eitherbecausehisreputation
precedeshimorbecausetheyhaveworkedwithhimbefore.IfsomeonelikeDavidis
allowedtocontinueatthecompanydespitehispoortrackrecord,itwouldbe
demotivatingfortheotherteamconsultantsandforthecoreteammemberswhowouldbe
forcedtoworkwithhim.Thebettercourseofactionistotreathimliketheunnecessary
overheadhehasbecomeandremovehimfromthepayroll.
Identifyingpeoplewhoareunderperforming,thoughpainful,enablescorrectiveaction
—underutilizedconsultantswilleitherchooseontheirowntoimprovetheirperformance
orleavethecompany,ortheywillbeforcedbymanagementtomakechangesor,again,
leave.Toensurethattheyremainrelevantandemployed,teamconsultantsshouldseekout
feedbackonhowtheyaredoing,theircontributionstotheirteams,andthevaluetheyare
adding.Intheend,withcompaniesconstantlystrivingtobecomemoreefficientand
productive,reducingoverheadandtrimmingfatisagoodthing.
Experiment
Forthismodeltosucceed,itmusthavemanagementsupport.BymanagementsupportI
meanthatthemiddleandseniormanagersnotonlyrealizetheywantaneffective
organizationbutalsohavethecouragetoactuallydosomethingaboutit.Toget
managementandindividualcontributorstogetafeelforthismodel,tryitoutonasmall
scale,asRebeccadidinthischapter’sstory.Chooseyourcoreteamanddeterminewhich
expertswillbeneeded.Gaintheircommitmentandletthemknowapproximatelywhen
theirexpertisewillbeneeded.If,whenthetimecomesforthemtoworkwiththeteam,
theyareunabletobecauseamanagerhascommittedtheexperts(consultants)towork
withanotherteam,youwillneedmanagementtobewillingtostepinandsolvethe
conflict.Thesetypesofblockingissueswillbecommonasanorganizationmovestoward
adoptingteamconsultants.Eventually,theteam’sScrumMasterwillbeabletoremove
thatsortofimpediment,butduringatestrun,theScrumMastermightnothaveenough
authorityorinfluencetomakethingshappen.
BeCautiousofOverloading
WhenIfirsthelpcompaniesimplementthismodel,Inoticethattheteamconsultantshave
atendencytoovercommit—afterall,optimismisontheirside.Thethresholdfor
overloadingseemstobecrossedwhenaconsultanttakesonthefourthproject.Threeis
okay.Fourisusuallytoomany.Ifyouseeconsultantssignupformorethanthreeprojects,
warnthemthattheyneedtoleavesomeavailabilityandflexibilityfortheteamstheyare
servicing.Thereisnothingworseforateamthanbeingstuckorneedingguidanceona
projectandhavingaconsultantwhodoesnothavetimetoaddressituntilthenextsprint.
ThisisablockingissuethataScrumMasterwillneedtoaddress.
PlanforPotentialDowntime
Therewillbetimesintheebbandflowofprojectswhenteamconsultantsandcoreteams
bothgoidle.Sincepeoplehavebeentrainedtobeat100percentutilization,andsuddenly
theyarenot,managementmightpushtobreakupateam,forcingitsmembersinto
potentialteamconsultancy,orworse,togobacktothewaythingswereandhave
specialistsdospecificjobsonprojectsandthenmoveontothenextspecificjob.Don’tlet
themdoit.Takeacuefromcontractors:Keepafewsmallprojectsinreserveforslow
times—projectsthatneverseemtogetdonebecausesomethingelseismoreimportant.
Whendowntimehappens,feedteamsandteamconsultantsthoseprojectstoworkonuntil
thetimingisrighttotakeonanotherlargejob.
TeamConsultantsAreNotaReplacementforDedicatedTeams
SomemightlookatthismodelandthinkthatitcircumventsScrum’scallfordedicated,
cross-functionalteams.Itdoesnot.Whatitdoesdoisprovideteamsandcompaniesthe
flexibilitytostaywithintheboundariesofScrumbyhavingadedicated,cross-functional
teamwhileenablingthatteamtocallonexperts,teamconsultants,tohelpoutasidentified
andneededbytheteam.Itisnotthejobofaresourcemanageroranexecutiveto
randomlyassignpeopletotheteam;thatwillnotwork.Itisthejoboftheteammembers
todetermineandmanagetheirschedules.Ifyoustarttosee“teams”ofconsultants,ones
inwhichthereareonlyoneortwocoreteammembers,thededicated,cross-functional
teamhasbeenlost.Remindthebusinessoftheimportanceofanactualcoreteam
supplementedbyconsultantsandreboot.
Havingadedicatedteamwithalltheskillsnecessarytogettheprojectcompletedis
ideal.Butmostteamswill,atsomepointintheproject,findthemselveswishingfora
particularexpertise.Ifyouhaveinternalpeoplewhocanfillthisgaponatemporarybasis,
bringthemintotheproject.Ifyourorganizationdoesnothaveteamconsultantswhocan
fillthegaps,youmaywanttobringinexternalconsultantsforlimitedengagements.A
teamconsultantcansometimesscorealotofpointsforaprojectthatmightotherwise
strugglewithaskillgap.
References
[BROOKS]Brooks,Frederick.P.,Jr.1996.TheMythicalMan-Month,Anniversary
Edition.Reading,MA:Addison-Wesley.
[MCCONNELL]McConnell,Steve.Personalwebsite.
http://www.stevemcconnell.com/ieeesoftware/eic08.htm(accessedon2January2011).
[PUTNAM]Putnam,L.,andW.Myers.1998.“FamiliarMetricManagement:SmallIs
Beautiful—OnceAgain.”
http://www.qsm.com/sites/www.qsm.com/themes/diamond/docs/fmm_28.pdf.
[SENGE]Senge,P.M.1990.TheFifthDiscipline.NewYork:Doubleday/Currency.
WorksConsulted
Pedler,M.,J.Burgogyne,andT.Boydell.1997.TheLearningCompany:AStrategyfor
SustainableDevelopment,SecondEdition.London:McGraw-Hill.
O’Keeffe,T.2002.“OrganizationalLearning:ANewPerspective.”JournalofEuropean
IndustrialTraining,26(2):130–141.
Chapter4.PredictingTeamVelocity
Softwareteamsareabitlikecarsinthatyoucanmeasuretheirperformanceintermsof
efficiency.Carsmeasureefficiencyintermsoffueleconomy,howfaracarcantravelper
unitofgasoline.Teams,ontheotherhand,measuretheirefficiencyintermsofhowmuch
worktheycancompleteoveracertainperiodoftime.Thismeasureiscommonlyknown
asvelocityandistypicallyexpressedintermsofstorypoints.Simplyput,velocityisthe
averagenumberofstorypointsateamcompletespersprint,asdefinedbytheteam’sdone
criteria.(ReadmoreaboutdonecriteriainChapter7,“HowDoYouKnowYou’re
Done?”).
Isaythatteamsareabitlikecarsbecausetheyaremostlyunlikecars.Teamsdonot
comepackagedwithamanufacturer’sestimateofvelocity.Teams,therefore,mustpredict,
withsomedegreeofconfidence,whattheirfuturevelocitiesmightbe.And,whilepeople
caneasilycomparetwocars’fuelefficiencies,theycannot(andshouldnot)comparethe
efficiencyoftwoteamsbasedontheirrelativevelocities.That’sbecausewhilemilesper
gallon(orkilometersperliter)isanobjective,standardizedmeasure,velocitydependson
eachteam’sownsubjectiveestimates.Sowhileoneteamcansometimesuseanother
team’svelocityasastartingpointforwhatitsfuturevelocitymightbe,youcannotsaythat
oneteaminacompanyissomehowmoreefficientbecauseithasahighervelocitythan
anotherteam.Theteamsmightjustestimatedifferently.
Asyouhavelikelyexperienced,predictingvelocityisadifficulttask,especiallyfor
teamsthatarenewtoeachotherandnewtoanagileframework.Thefollowingstory
describeshowoneScrumteamdealtwiththiscomplexissue.
TheStory
WendiwastheScrumMasterforanewteam.TheteamwasexcitedtostartitsfirstScrum
projectbutwasstymiedbyhowtogivemanagementevenaballparkfigureforhowmuch
workitcouldaccomplishandbywhen.
Wendiaskedtheteamforsuggestions.Oneteammember,Paul,whowasvisibly
frustrated,startedthediscussion,“Howarewesupposedtogivethemvelocityestimates
beforewe’veevenstartedourfirstsprint?Whatarewesupposedtodo?Justpointata
randomspotontheproductbacklogandsayhere’showfarwe’llget?Thatseems
irresponsible.Makingwildguesstimatesiswhathasgottenusintosomuchtroubleinthe
past.Scrumwassupposedtobedifferent.”
“Iagree,Paul,butcanyouimaginewhatwouldhappenifwetellthemwehaveno
idea?Thatwe’lljustletthemknowwhenwefigureitout?”askedhisteammateFreddie.
“Therealityis,ifwedon’tgivethemsomekindofnumber,theywilldoitforus—‘Here’s
yourgoal.Hitit.’We’veallbeendownthatroadbefore.”
“Iwishwehadsomehistoricaldata,”musedIgnacio.“Italkedtoafriendofminewho
isrunningaScrumproject.Histeamlookedatitspastvelocitiesonotherprojectsand
usedthosenumbersasastartingpointforanewproject.Ithoughtaboutusinghisdatato
helppredictourvelocity,butdismissedtheideabecauseeventhoughourteamis
somewhatsimilartohis,we’realsoverydifferent.Ourteamisnew,we’veneverused
Scrumbefore,andourprojecthassomecomplicatingfactors.Idon’tknowhowtoaccount
forallthevariances.”
TheyalllookedatWendi.
Paulasked,“Sowhatarewegoingtodo?”
Wendiwentuptothewhiteboardandwrotedownthethreeideasthathadbeen
proposedandrejected.
“Iknow,Freddie,thatyoursuggestiontotellthem‘we’llletyouknowwhenwefigure
itout’wasmoretongueincheekthananything.ButI’dlikeustoconsideritasaviable
option,”Wendiexplained.“Sothatleavesuswiththreeoptions.Allareproblematic,butit
lookslikewe’vegottochooseoneofthem.So,let’sspendsometimelookingatthem,”
suggestedWendi.“Maybeifwebrainstormsomepositivesandnegatives,wecanfigure
outwhatthebestsolutionwillbe.Let’sstartwithhistoricaldata.Calloutsomepositives
andnegatives.”
Astheteambrainstormedanswers,Wendifilledinthechartshehadbegunonthe
whiteboard.
“Good,”saidWendi.“Let’sdiscussthisabitbeforewemoveontothenextsection.”
“I’llgofirst,”saidIgnacio.“Weareanewteam,sowedon’thavedataofourown.
Ideally,we’dpulldatafromasimilarteaminthecompany,butIdon’tknowofany.”
“Yep,”agreedPaul.“Ican’tthinkofanyteamsthatwecanuseasanexampleeither.
Mostaren’tdoingScrumatall.ThosethatareusingScrumareeitherjustgettingstarted
likeusortheirprojectsaretoodifferentfromours.”
“Wecouldlookatmyfriend’steam,”suggestedIgnacio.“Butwithhimbeingata
differentcompany,itfeelslikeastretch.”
Freddienodded.“Ijustdon’tthinkthisonewillworkforus.Ifwehadbeenworking
togetherforalongtime,andcoulduseourownhistoricaldata,Ithinkit’dbeokay,butto
grabsomerandomteam’snumbersandusethemasourestimatereallyscaresme.”
Everyonenoddedinagreement.
“Ithinkweallagreethathistoricaldatahasfarmorenegativesthanpositives.Let’s
moveontothenextone,blindestimation,”saidWendi,reachingforhermarker.
Theteamquicklycameupwiththeprosandconsofblindestimation.
“Thisisagoodlist,”saidWendi.“Let’stalkaboutacoupleofthese.Whocameupwith
thephrase‘estimatesarecommitments’?”askedWendi.
“Idid,”saidPaul.“Myconcernisthatonceweestimateourvelocity,we’llbeheld
accountabletoitthroughouttheproject.Whatifourinitialestimateistoooptimistic?We
needtobuildinasafetycushion.Takeourestimateandsubtractfive,orsomething.”
Theteamnoddedinagreement.
“LetmemakesureIunderstand.You’resayingthatifourinitialvelocityestimateistoo
high,wewillburnouttryingtohitthathighernumbereachsprint?”askedWendi.
“Exactly!”saidPaul.
“Ihearyou,Paul,”Freddieinterjected.“Butit’snobetterifourestimateistoolow.
They’llsay,‘Withallthattalentonyourteam,you’resayingyoucanonlydothismuch
workpersprint?Whatwillyoudowiththerestofyourtime?’Peoplewillpanicbeforewe
evengetstarted;managementmightevensayweneedtoaddmorepeopletogetthe
velocityup.Then,we’llhavetostartover.”
“Itwouldn’tbethefirsttime,”agreedIgnacio.
“IfIhearyoucorrectly,Freddie,you’reconcernedthatmanagementwilleithertellusto
domore—tohaveahighervelocity—orthatthey’lladdpeopletotheprojectbecausethey
willassumethatwemustneedmorepeopletogetthejobdone,right?”askedWendi.
“Yes,”saidthewholeteam.
“Fairenough,”saidWendi.“Evenifwedon’tgowiththisoption,I’llmakeanoteto
talktomanagementaboutvelocityingeneralbecausethismightbeaconcernnomatter
howwearriveatanestimate.Iwanttomakesurethattheyunderstandexactlywhata
velocityestimatemeansandwhatitdoesn’tmean.”
Wendiwroteanactionitem,Wenditoeducatemanagementaboutvelocity,ontheside
ofthewhiteboard.
“Okay,”Wendicontinued.“Let’slookatanotherofthepoints.Whatdoes‘createsa
falsebaseline’mean?”
“Iwrotethatone,”saidFreddie.“Ifwestartwithawildguessasourbaseline,andthen
dobetterthanweexpect,peoplewilltalkabouthowmuchwemusthaveimproved.Ifwe
doworsethantheso-calledbaseline,peoplewillwonderwhy.Wecantellthemit’sa
guessallwewant,butoncethereisarealnumberwrittendownsomewhere,we’re
comparedagainstit.”
“Iknowexactlywhatyoumean.I’llmakeapointtoincludethattendencyinmytalk
withmanagement,”Wendisaid.
“Evenwitheducation,though,whenitcomesdowntoit,it’sawildguess,astabinthe
dark,”exclaimedPaul.“Ifwe’regoingtobeagile,Iwantustofollowtheprinciples.
Pluckinganumberoutoftheairandpromisingtodeliveritmightmakeeveryonefeel
better,butnotforlong.It’sanemptypromiseandwillcreateabadenvironmenttotryto
makeScrumwork.”
“Ihearyou,Paul,”saidIgnacio.“Butthefactis,thatmightbewhatwehavetodo,at
leastthistime.Onournextproject,we’llhavesomehistoricaldataofourowntowork
with.”
Therestoftheteamnoddedinreluctantagreement.
“Well,Idoknowsomewaystomakeblindestimationalittlelessofashotinthedark,”
saidWendi.“Butbeforewegiveupandstartguessing,let’stakealookatmypersonal
favorite,‘waitandsee.’BythatImeanwepolitelybutfirmlyrefusetogivemanagement
anestimatenow.Instead,wepromiseamuchmoreaccuratenumberafterwe’verunthree
sprintsandhavesomerealdata.Icantellyoumyideasonthepositives.”
Wendiwrotethreepositivesontheboard.“So,youguystellmewhyitwon’twork.”
Theteamquicklypointedouttheflaws.
“Thoughts?”askedWendi.
“I’dlovetogowith‘waitandsee,’”saidIgnacio,“butIdon’tthinkit’srealistic.There
isnowaymanagementwillletuswaitthreesprintstogivethemanestimate.”
“And,”Freddieadded,“eventhoughhavingrealdatawouldallowustohavean
accurateandreliablereleaseplan,we’reanewteam.Itwilltakesometimeforustoget
uptospeed.I’mnervousaboutwhatouractualvelocitywillbeduringthefirstfew
sprints.Whatifit’sreallylow?”
“Itisarisk,”saidWendi.“ButIthinkit’sourbestoptionifwewanttosucceed.Wejust
havetheonetinyobstacleofgettingmanagementbuy-in,”saidWendi,smilingathow
clearlyfacetiousherlaststatementwas.
“We’llnevergetthisbythem!”Paulexclaimed.“Theywillrequireustoprovidedata
upfront.Therehastobeanotheroption.”
“Ifsomeonehassomethingbetterthanthethreewe’vealreadydiscussed,I’dloveto
hearit,”saidWendi,lookingaroundtheroom.
Theteamwasquiet.
Wendicontinued,“We’vewalkedthroughouroptions.Thetruthis,wejustdon’thave
enoughinformationatthistimetobeable,withanyconfidence,todetermineourlikely
velocity.Wecouldgivemanagementanumber,butitwouldbeawildguess.Let’stry
somethingasboldasweare.Let’sshowthemtheoptions,walkthemthroughtheprosand
cons,andexplainthatwewanttostartthisprojectlikeweintendtocompletethisproject:
transparentlyandhonestly.
Theteamlookedatherwithhesitation.
“But,”Wendiwenton.“Andthisabigbut.Wewillpromisethemthatwe’llpostour
observedvelocity,andassoonaswehaveenoughhistoricaldataofourown—three
sprintsworth,let’ssay—we’llgivethemanestimatedvelocityrangesothattheycan
makeareleaseplan.”
Wendiwentontoexplainindetailhowthewait-and-seeapproachwouldwork.The
teamagreedtogiveitatry.Theteamwasevenabletoconvinceaveryreluctantgroupof
managerstowaitforrealdata.
Beginningwithsprintone,theteamposteditstrackingchartsinpublicplacesand
updatedthemateachdailyscrum.Attheendofthreesprints,theteamgavemanagement
anestimatedvelocityrangeandexplainedwhatthatmeantintermsoflikelyfunctionality.
Theproductownerplannedareleasebasedonthepessimisticandoptimisticranges
providedbytheteam.
Thoughtheteamultimatelydeliveredtowardthelowendofitspredictedvelocity
range,itsestimatesprovedaccurate.Asaresult,noonewassurprisedbythereleased
product.Theproductownerhadallthemust-havefeaturesheexpected.Thecustomers
hadallthefeaturestheyhadbeenpromised(andhadnotbeenteasedwithanyofthe
featuresthatwerequestionablefortheinitialrelease).Intheend,everyoneagreedthat
combiningrealdatawithpatience,education,andtransparencyhadbeenthebestsolution.
TheModel
Wendiandtheteamstruggledwithhowtogivetheirproductowneranideaofhowmuch
theycouldaccomplisheachsprint.Afterabriefbrainstorm,theycameupwiththree
solutions:historicaldata,blindestimation,andwaitandsee.Wendi’steamultimately
electedtowaitandsee.Whydidtheyeliminatehistoricaldataandblindestimation?
TheProblemwithHistoricalData
Historicaldataisaviableoptionwhenyourteamhasworkedtogetherbefore.Inthis
chapter’sstory,Wendi’steamdismissedthisoptionbecauseitsmemberswerenewto
workingtogether,sotheonlyhistoricaldataavailabletothemwouldhavebeendatafrom
anotherteam,likelyoneatadifferentcompany.Whileit’spossibletoderiveanumber
basedonanotherteam’sperformance,it’snotanoptionIwouldrecommendbecauseofall
thedifferentvariablesinvolved:
Thenewnessoftheteamanditscomposition
Thepoliticalenvironment
Projectsizeandcomplexity
Theproductownerandcustomer
Thefirstvariabletoconsideristherelativenewnessoftheteam.Thenewnessofateam
willaffectitsvelocityascomparedtoanexistingteam,ifonlybecauseanewteamtakes
timetogel.Teammembersmustlearneachother’sstrengthsandweaknessesand
determinehowbesttoworktogether.Further,anewteammaynotbeideallycomposed.It
mightstillbeclingingtosometraditionalapproaches,suchashavingtwoseparateteams
—onemadeuponlyofdevelopersandonededicatedtotesting.Thismono-functional
teamapproachwillnotseethebigperformanceimprovementsoutofthegatethatatruly
cross-functionalteamwill.
Thenextvariablethatcanaffectvelocityisthepoliticalenvironment.Overtime,
companiesrestructureorchangedirection.Whatmayhavebeenthecompanygoalayear
agomaybeobsoletenow.Keymanagersmayhaveshiftedroles,changingthedynamics
oftheenvironment.Sometimespoliticalchangesareobvious;othertimes,thechangesare
pervasivebutmuchmoresubtle.Teamsusinghistoricaldatamustallowforthesenuances,
whichcanbeverydifficulttodo.
Thethirdvariabletoconsiderisprojectsizeandcomplexity.Teams,neworwell
established,thattakeonaprojectthathasdifferenttechnologyorachangeincomplexity
cannotrelyonhistoricaldata.WhenJeffSutherlandwasCTOatPatientkeeper,heoften
toldmethathisteamscouldsnuffoutthecompetitionbecausetheywereabletorespond
tochangequickly.Oneofthereasonstheywereabletodothiswasthattheworkwas
similarandtheteamsremainedconsistent;theywerenotswitchingfromalegacyC++
applicationtoaJavawebplatform.Theyweren’tsloweddownbyalearningcurve.
Thelastvariablehastodowiththeproductownerandcustomer.Whilethesemaynot
seemlikelikelyvariablestoconsider,theycanbehugefactors.Saytheteamremainsthe
samebuttheproductownerisnew.Thevelocitythattheteamsawinthepastissureto
changebecausenotallproductownersarecreatedequal.Therelationshipbetweenthe
teamandtheproductowner,muchliketheoneamongnewteammembers,needsto
developovertime.Bythesametoken,workingwithanewcustomerrequiresadjustments
thatcouldaffectvelocityaswell.
Ifyoufindyourselfinthesituationwhereyouareconsideringusinghistoricaldata,take
thesefactorsintoaccountandbecareful.Usingotherteams’velocitiesasadatapointis
fine.Usingitasyourteam’spredictedvelocity,however,isfraughtwithperil.Proceedat
yourownrisk.
SheddingLightonBlindEstimation
Ifyouareforcedtoprovidesomesortofvelocityestimatebeforeyou’vedoneanywork
together,blindestimationisprobablyyourbestbet.Ratherthanjustactuallyguessing
blindlyatanumber,however,youcandosomebackgroundworktoshedasmuchlighton
yourguessasyoupossiblycanbeforepredictingyourvelocity.
Thestepsinthis“not-quite-blind”estimationtechniqueareasfollows:
1.Estimatetheproductbacklog.
2.Decomposeareferencestory.
3.Determineapoint-to-hoursapproximation.
4.Identifytheteamcapacity.
5.Estimatetheteamvelocity.
6.Communicatethevelocityasarange.
BeforeIdiscussthesestepsinmoredetail,letmeaddafewdisclaimers.
First,pleasedon’tthinkI’madvocatingthisasageneralpractice.Ionlysuggestit
forthosesituationswhenyouabsolutelyhavetotakeaguess.
Second,onceyouhaverealnumberstohelpyoumakeamoreeducatedguess,throw
awaytheestimatesthatyouderivedinthisexercise.They’reartificial,acrutchto
helpyougetstarted.And,worstofall,they’rebasedonabigagileno-no,mapping
storypointstohours.
Third,everynumberyouprovideinablindestimateshouldbereplacedwithreal
dataassoonasyou’vehadtimetodoasprintortwo.Andeverypoint-to-hours
approximationyoumaketoarriveatthatnumbershouldbediscarded,no
exceptions.
EstimatetheProductBacklog
Thefirststepinthisadmittedlyless-than-idealtechniqueistoestimatetheproduct
backlog.Iftheproductbackloghasalreadybeenestimatedinpoints,skiptothenextstep.
Ifnot,youneedtoprovidepointestimatesfortheproductbacklogitems.Todothis,
browsetheproductbacklogwiththeteam.Youarelookingforatwo-pointreferencestory.
(Iuseatwo-pointstoryasareferencesothatiftheteamcomesacrossastorythatis
smaller,ithasroomtoassignitasmallerpointvalue.)Oncethetwo-pointreferencestory
hasbeenidentified,workwiththeteamtocomparethatstorywitheveryotherstoryinthe
productbacklog,assigningpointvaluesuntileachstoryinthebackloghasapointvalue
associatedwithit.IliketousePlanningPoker[GRENNING]forthis,butyoucanuseany
methodyouchooseaslongastheentirebacklogisestimatedinpointswhenyouaredone.
(Forhelpwithestimatingparticularlylarge,rawproductbacklogs,seeChapter29,
“PrioritizingandEstimatingLargeBacklogs—TheBigWall.”)
DecomposetheReferenceStory
Oncetheproductbacklogisestimatedinpoints,it’stimetoidentifyaroughhour-to-point
mapping.Todothis,youmustfirstdecomposeareferencestory.Withtheteam,choosea
two-pointstory(thismayormaynotbethesamestoryasyouusedinstepone).Forthat
story,identifythetasksnecessarytocompleteit.IuseOsborn’sbrainstormingtechnique,
asidentifiedinChapter7,toidentifythetasks.Thisisahighlyinteractiveprocessandis
mosteffectiveifeveryoneparticipates.
Oncetheteamhasidentifiedthetasksforthereferencestory,itneedstoestimatehow
manyhoursthosetaskswilltaketocomplete.YoucanusePlanningPokertoarriveat
theseestimates;justhavethenumbersonthecardsrepresenthoursinsteadofpoints.
BecauseIadvocatethatteamslimitestimatesofanytasktonomorethan13hours(ifa
taskwilltakelongerthan13hours,breakitintomultipletasks),removeallthehigher
cardsforyourdeck—youmightevenconsiderremovingthe13.Further,sticktothe
numbersonthecards.IfyouareusingtheFibonaccisequence,thatmeansyouwillhave
sixvaluestochoosefrom:1,2,3,5,8,13.Whatifapersonestimatesthetaskatfour
hours?Thepersonmustchoose—threeorfive.Thisisdonetomaintainalevelof
accuracythatisgoodenoughwithoutbeingtooprecise.
Points-to-HoursApproximation
Supposethatwhenthehoursforeachtaskareaddedtogether,thetotalforthereference
storyis14hours.Thenextstepistoextrapolate:Alltwo-pointstoriesintheproduct
backlogwilltaketheteamroughly14hourstocomplete.Ofcourse,inactualitytheteam
mightcompletethis,oranyothertwo-pointstory,in2hours,8hours,or16hours.It’sa
roughestimate—andforthepurposesofblindestimation,that’sokay.Fromthis,youcan
determineapoints-to-hoursapproximation.
14hours/2points=7hoursperpoint
Letmeemphasizeonemoretime:Youdon’tknownearlyenoughaboutyourteam,your
project,ortherealitiesofthedevelopmenttohaveanumberthisspecificinmind.Usethe
points-to-hoursapproximationtohelpbuildyourinitialestimate,andthenthrowitaway
assoonasyouhaverealdata.
TeamCapacity
Next,youneedtodeterminetheteam’scapacity.Tocalculatecapacity,decidehowmany
hourspersprinttheteam(notindividuals)willcontributetotheproject.Togetthis
number,youhavetocollectdataattheindividuallevel;however,remembertoreportback
whattheteamcommitmentis,notwhattheindividualcommitmentsare.
Haveeachteammemberwritethenumberofhoursheorshehasavailableperweekto
contributetotheproject.Remember,thisnumbershouldnotbe40.Everyonehas
unrelatedprojectactivitiesthatofteninterferewithdedicatedprojecttime.Thisoverhead
ultimatelytakesawayfromthetimeeachpersonhasavailabletodoactualtask-related
work.Askpeopletocommunicatetheiravailabilityastwonumbers:bestcase(high)and
worstcase(low).
Addtheworst-caseestimatestogetherandmultiplybythenumberofweeksinasprint.
Thisisyourlow-rangeteamcapacity.Addthebest-caseestimatestogetherandmultiply
thatbythenumberofweeksinasprint.Thisisyourhigh-rangeteamcapacity.Table4-1
presentsanexample.
TABLE4-1TeamCapacityEstimates
Inthistable,Itookthesumofthelow-rangeweeklycapacityandmultiplieditbythe
numberofweeksinasprint(2)togetatotalof176hours.Ithentookthesumoftheteam
members’high-rangecapacitiesperweekandmultiplieditbysprintlengthinweeks(2)to
getatotalof236availablehours.Itislikelythattheactualteamcapacitywillnotbeeither
theabsolutehighortheabsolutelow;theteamshouldfeelconfident,though,thatits
capacitywillfallsomewhereinbetweenthetwoextremes.
EstimateTeamVelocity
Thefollowinghasbeendeterminedforthisexample:
Theproductbackloghas200points.
Atwo-pointstoryisestimatedtotakeroughly14hours,or7hoursperpoint.
Theteambelievesitcandeliverbetween176and236hoursofworkpersprint.
Younowhaveenoughdatatoestimatethevelocityfortheteam.Determinealow
velocityandahighvelocitybydividingthelow-andhigh-capacityrangesfortheteamby
theestimatedhoursperpointandroundingdown,asshowninTable4-2.
TABLE4-2EstimatingVelocity
Inthisexample,theestimatedvelocitycouldbecommunicatedasbeingbetween25and
33pointspersprint.Tolearnhowtousethatinformationtodeviseaninitialreleaseplan,
seeChapter11,“ReleasePlanning.”
IncreasingConfidenceintheTechnique
Extrapolatinganumberofhoursperpointfromonereferencestoryisveryroughandvery
quick.Ifyou’dlike,youcanrefineyourestimatefurtherbyestimatingafewdifferent
referencestoriesofthesamepointvalue.Supposeyouestimatefivetwo-pointreference
stories:
1.Yourinitialstory:14hours
2.Story2:22hours
3.Story3:30hours
4.Story4:25hours
5.Story5:20hours
Fromthis,youcanseethattheinitialreferencestoryof14hoursmightbeabitlow.A
moreaccurateestimateseemstobeapproximately22hours,with14and30hoursbeing
outliers.Addingmorereferencepointsmayhelpincreaseyourconfidenceinyourblind
estimates,butitdoestakesignificantlymoreworkandreliesonarelativelywell-
estimatedproductbacklog.Itmightalsogiveyouafalsesenseofsecurityinyourvelocity
estimate.
Rememberthatnomatterhowmanydatapointsyouuse,atthetimeinwhichyouare
estimatingthesestories,yourteamhasyettodoanywork.Youreallydon’tknowfor
certainatthispointhowlongitwilltakeyoutoaccomplishanytask.Whetheryouchoose
toextrapolatefromonedatapointormore,remindyourteamandyourstakeholdersthat
theestimatedrangeisveryroughandwillcertainlychangeastheteambeginstogeta
betterpictureofhowmuchworktheycanaccomplishpersprint.
WaitandSee(UseRealData)
Wendiandtheteamultimatelydecidedtowaitandseewhattheiraveragevelocityisfor
thefirstfewsprintsbeforeprovidinganinitialestimate.Theteamplannedtorunthree
sprints,capturingtheobservedvelocitydataattheendofeachone.Ipreferthisoption
becausetheteamismakingitscalculationsbasedonrealdata,datathatdoesnotexist
untiltheprojectisrunning.
Evenifyouhavetocommunicateanestimateusinghistoricaldataorblindestimation,
youshouldstillcollectyourowndataandreviseyourestimatebasedonthatdataassoon
asyoucan.Infact,youshouldtrackyourvelocityandrefineyourestimatesthroughout
thelifeofyourproject,nomatterwhichinitialmethodyouchoose.(Fordetailsonhowto
maintainareleaseplan,seeChapter11).
Thebenefitofwaitingandseeing,ofusingrealdata,isthattheteamisusingitsown
dataonitscurrentproject.Basinganestimateonrealityenablesareleaseplantoreflect
fact,notfiction.Themainhurdletothisapproachisthatmanagementisgoingtowanta
numberbeforeyouarereadytogivethemone.Toovercomethisobstacle,communicate
thebenefitsofwaitingtothemanagementteamandlayoutaplanastowhentheteam
willprovidenumbersandwhyrealdataisworthwaitingfor.
Whenusingrealdata,youshouldfollowthesesteps:
1.Collectandchartyouractualvelocityforatleastthreesprints.
2.Calculateaveragevelocity,butcommunicatearange.
3.Mapyourvelocityrangetotheproductbacklog.
4.Updateyourvelocityandlikelyvelocityrangeaftereachsprint.
CollectandChartRealData
Thefirststepistocollectyourteam’sdata.Trackyourvelocityforthreesprints,mapping
thedataonapublicchartpostedintheteamarea.Attheendofthreesprints,yourchart
willlooksomethingliketheoneinFigure4-1.
FIGURE4-1Samplevelocitiescapturedduringthreesprints
CalculateAverageVelocitybutCommunicateaRange
Onceyouhavethreesprintsworthofdata,calculateyouraveragevelocity.Donot,
however,communicatethisaverage.Why?First,becauseyoursisanewteam,thisinitial
datamightnotbestableenoughtouseforanextendedperiodoftime.Second,youwant
toavoidmisunderstandingsthatarisefromfixatingonasetnumber.Forexample,asIwas
writingthischapter,mywifeaskedmewhattimeIwouldbedoneandreadyfordinner.I
toldher,inabout30minutes.WhatImeantwhenIsaidaboutwassomewherebetween20
and40minutes.Whatsheheardwas30minutes—not29,not31—30.Shefixatedona
numberandheldmetoit.
Whenthe30-minutemarkcame,therefore,shesaid,“Okay,it’sbeen30minutes.You
shouldbedone—comeandeatorthefoodwillbecold.”Myresponse,whichwasless
thanstellar,was,“ButI’mnotdone.Isaidabout30minutes;Ineedanother10minutes.”
Thisfrustratedherbeyondbelief.Iwas95percentconfidentthatIwouldbedonein20to
40minutes,butIdidnotcommunicateitthatway;Icommunicated30.Mywifeheardthis
asacommitment.Imeantitasanapproximation.
IfIhadgivenmywifearange,“Iam95percentsureI’llbedonein20to40minutes,
shewouldhaveknowntohavedinnerreadyin40minutes(myworst-caseestimate)but
wouldhavehadsomehopethatIwouldbetheretohelpsetthetableandopenthewineif
Ifinishedearlier.Bysaying30straightup,Icommunicatedahighdegreeofconfidencein
anumberthatwasmerelythemiddleofalikelyrange.
Youdon’twantfrustratedmanagementanddisappointedcustomers.Ifyougivethema
setvelocity,theyarelikelytomakeplansaroundthatonenumber.Ifyougivethema
range,ontheotherhand,theyaremorelikelytoplanforabest-caseandworst-case
scenario.Ifyoutellthemyouhavearangebutarenotveryconfidentinityet,theyare
likelytomakethoseplansalittlemoreflexible.It’sallaboutcommunication.Avoiding
misunderstandingsgoesalongwaytowardhavingasatisfiedcustomer.
So,eventhoughyou’llcaptureyouraveragevelocity,youwillnotcommunicateyour
averagevelocity.Instead,talkintermsofalikelyvelocityrangeandyourconfidencein
thatrange.ImaginethatthenumbersinthechartinFigure4-1reflectyouractualvelocity
duringyourteam’sfirstthreesprints.Youraveragevelocitywouldbe12((10+14+
12)/3).Yourlikelyvelocityrange,however,wouldbe10to14(lowvelocitytohigh
velocity).Andbecauseyouareanewteamthatisstillformingandstorming,yourdegree
ofconfidencethatfuturevelocitieswillfallinthisrangemightbe75percent.(Formore
onstagesofteamdevelopment,seeChapter20,“AddingNewTeamMembers.”)
Itisimportanttorememberthatthedegreeofconfidenceissubjectiveandqualitative.
Whiletherearewaystomeasuresubjectiveconfidence,Idon’trecommendthatyoutryto
measureitinanyabsolutesense.Ifindtheefforttoohighcomparedtothesmallamount
ofvaluegained.
Todeterminetheteam’sdegreeofconfidence,haveeachteammemberwritedownhis
orherconfidenceratingonapieceofpaper—wheneveryonehasanumberwrittendown,
flipoverallthewrittenpagesatonce.Youcanskipthisifyourteamiscomfortabledoing
thisverballyandjusthaveeveryoneansweratthesametime.Thepointistoprevent
peoplefrombeinginfluencedbyothers’answers.Next,reviewthenumbersthattheteam
memberschose.Ifthenumbersaresimilar,you’vefoundyourteam’sconfidencerating.
If,ontheotherhand,youhavealargerangeofconfidencenumbers,youneedtohave
morediscussions.Askiftherearespecificreasonswhysomeoneislessconfidentthan
others(e.g.,Ithinkwe’llgofasterinfuturesprintsbecauseitwilltakesometimeto
understandthetechnology;Ithinkwemightslowdownforafewsprintsbecausewe’re
goingtolearnhowtoworkinpairs).Encouragetheteamtotalkitout.Then,voteagain.
Dothisuntilalloftheindividualconfidenceratingsconverge.
TruncatedDataCollection
Toooftenteamswhoinitiallygetpermissiontocollectdataforthreesprintsbefore
communicatingarangearecoercedintoprovidingdataafteronlyoneortwosprints
togethersothattheproductownercancommunicatethereleaseplantomanagement,
customers,andstakeholders.Thelimitedamountofdatamakesitdifficulttocommunicate
arealisticrange.
Ifyoufindyourselfinthissituation,considertwoapproaches.ThefirstistouseMike
Cohn’smultipliers.Thistableofmultipliersgivesteamsawaytocommunicateaprobable
velocityrangeuntiltheyhaveenoughmeaningfuldatatocalculatetheirown.InAgile
EstimationandPlanning,Cohnexplainshowtousemultiplierstoestimatefuturevelocity
basedonobservedteamvelocity,asseeninTable4-3.YoucanseethatCohnrecommends
thatifyouhavecompletedonlyoneiteration,youshouldmultiplyyourvelocitybythe
highandlowmultiplierstoarriveatarangeyoucancommunicate[COHN].
TABLE4-3MikeCohn’sObservedTeamVelocityMatrix
InmyearlyprojectsIstartedwithCohn’svalues,aslistedinTable4-3.Theywere
extremelyhelpful.AsIencountereddifferentteamswithdifferentbackgroundsworking
ondifferenttechnologies,however,IfoundthatIstartedmodifyingCohn’smatrixinto
somethingthatwasnotdrivenbythenumberofiterationsateamhasunderitsbelt.Why?
Well,Ididn’talwayshaveexperiencedteams,andoftenafterthefourthsprint,we’dfind
thatwestillneededthesamemultiplierthatweusedinsprintone.Inotherwords,Ifound
thatthenumberofiterationsnecessarytoreachastablevelocitywasdifferentfor
experiencedteamsthanfornewlyformedteams,andIwantedamatrixthatwouldreflect
thatreality.ThismodifiedmodelisillustratedinTable4-4.
TABLE4-4ModifiedVelocityMatrix
Toseehowthemodifiedmatrixworks,supposeTeamSuperBunnyisanewlyformed
teamforaprojectanditsteammembersareworkingtogetherforthefirsttime.Theteam
runsitsfirstsprintand,liketheteaminFigure4-1,achievesavelocityof10.Ifafterthat
sprint,theteamispressedtocommunicatearange,itshouldusethelowandhigh
multipliersforCategory2,newteam.
AsIprogressedwiththismodifiedmatrix,Irealizedanotherfactorthatcontinually
surfaced:thetechnologyorprojecttypetheteamswereworkingon.Istartedasking
myself,“Istheprojecttechnologycompletelynewtothepeople(unknowntechnology),or
isitsomethingthatthemajorityoralloftheteamhasexperiencewith(known
technology)?”Thisquestionisimportanttoconsiderbecauseunknowntechnologies
createalevelofuncertaintyandriskthatwouldnototherwisebepresentforateamthatis
justdoing“thesameoldtypeofwork.”
Becauseofthis,thenextfilterIpassateamandaprojectthroughexpandsonmy
originalmodifiedvelocitymatrixinTable4-4andincludesproject/technologyfamiliarity,
illustratedinTable4-5.Ofcourse,howfamiliarateamiswithatechnologyissubjective,
butitisanimportantfactorwhenateamisintheearlystagesofaprojectandworkingto
communicateitsvelocity.
TABLE4-5ModifiedVelocityMatrixwithProject/TechnologyFamiliarity
Thevolatilityoftheteam’svelocityvariesaccordingtoitsmakeupandprojecttype.I
findthatmorestableteamscanstopusingthistableafteroneortwomonths(twotofour
sprintsdependingonsprintlength).Newerteamshaveahardertimeandshouldcontinue
tofactorinhighandlowmultipliersforaboutthreemonths(threetosixsprintsdepending
onsprintlength).Oncethreemonthshavepassed,teamsshouldnotusethistableand
shouldinsteadrelysolelyontheirrealizedvelocityranges.
KeystoSuccess
Determiningyourteam’svelocitydoesnothavetobeachallenge.Thethreetechniques
listedinthischaptercanhelpyougetstartedandkeepyourvelocityontrack.
IamoftenaskedwhichapproachIrecommend.Eachapproachhasatimeandaplace
anditsownmeritsandissues.Table4-6summarizesmyrecommendations.
TABLE4-6WhentoUseEachApproachtoEstimatingVelocity
Waitingfortheobservedvelocityproducesthebestresult,asitisbasedonactualteam
data.Itcan,however,makemanagementorcustomers(especiallythosenewtoScrum)
feeluncomfortable,astherewillbenoinitialreleaseplangeneratedbytheteam.Thismay
negativelyaffectthetrusttheteamhaswithitsstakeholders,resultinginincreased
pressurefortheteamtodeliver.Compoundthiswithaninitialvelocitythatisnotwhat
managementorthecustomerswereexpecting,andtheteamwillfinditselfinaholethatit
mustdigitselfoutofbeforeitevenstarts.Tomakewaitingaviableoption,the
ScrumMastermustworkespeciallycloselywiththeproductownertohelpmanage
expectationsandcommunicateprogress.Somesuggestionsforwaystokeepstakeholders
inthelooparetoinvitethemtothesprintreviewmeetings,toallowthemtowatchthe
dailyscrum,and,ifneeded,tohostsmall,informalmidsprintreviewsforpeople
concernedabouthowmuchprogressisbeingmade.
Usinghistoricaldatacanbeanexcellentoptionwithanexistingteamworkinginan
environmentsimilartothoseithasencounteredinthepast.Asthelistofvarying
environmentalconditionsincreases,however,theviabilityofthisoptiondecreases.
Factorssuchaschangestoteamcompositionandtechnologyoranewcustomeror
productownerwillimpactateam’sabilitytorelyonhistoricaldataelements.
Blindlyestimatingtheteam’svelocityusingreferencestorieswill,atleastatfirst,give
theteam,customers,andmanagementthecomforttheyarelookingfor.Doingso,
however,setsatrap.Toooftentheteam’sinitialreleaseplan,whichwasbasedsolelyon
itscalculatedvelocity,isturnedintoacommitment.TheScrumMastermustworkwiththe
productownerandcustomertocorrectlymanagetheirexpectations.TheScrumMaster
mustbesurethateveryoneinvolvedunderstandsthatthevelocitycalculationtheteam
derivedisacalculatedguessintendedsolelytohelpbuildareleaseplan.Allstakeholders
mustunderstandthatonceobserveddataisinplace,thecalculatedestimateswill
disappearandbereplacedwithrealdata.
Whatevertechniqueyouusetoarriveatyourvelocity,rememberthatthisnumberisthe
epitomeofaball-parkestimate.Assoonasyouhaverunasprint,anyestimatesderived
fromhistoricaldataorblindestimationshouldbethrownoutandreplacedwitha
predictiverangebasedonwhatevervelocitywasattainedinyourfirstsprint.Asthe
projectprogresses,yourconfidenceinyourvelocityrangewillincrease.Continueto
throwoutpreviousestimatesandreplacethemwithnewonesasyougathermoredata.
Mostimportantly,communicatewithandbetransparenttoyourcustomersand
stakeholderssothattheytoounderstandhowmuchconfidencetheyshouldplaceinyour
initial,interim,andlaterestimates.Inallcases,remembertouserangesandtoexpress
youranswerwithadegreeofconfidencethatwillhelppeopleunderstandhowmuchtrust
toputintoeachestimate.
References
[COHN]Cohn,Mike.2006.AgileEstimatingandPlanning.UpperSaddleRiver,NJ:
PrenticeHall.
[GRENNING]Grenning,James.
http://renaissancesoftware.net/files/articles/PlanningPoker-v1.1.pdf(accessedon11June
2009).
Chapter5.ImplementingtheScrumRoles
TraditionalteamsthatshifttoScrumencounternewrolesanddifferentwaysofthinking.
TheyfirstneedtofindaScrumMaster,whosejobistokeeptheprocessrunningsmoothly
withoutdictatingsolutions.Theyalsoneedtohaveaproductowner,whoistaskedwith
steeringtheteaminthebestdirectionbasedonbusinessvalue.Then,theymustlearntobe
goodteammembers,whichinvolvescultivatingacross-functionalattitude,learningnew
skills,andworkingontasksthatmightbenewtothem.
FittingexistingteammembersintotheScrumrolescanbechallenging.Afterall,
sometimespeopledon’tfitneatlyintojustonecategory—ajack-of-all-tradesdeveloper
mightalsohavegreatleadershipskills.Aformerteamleadermightalsohavecrucial
codingskillsandadeepknowledgeoftheproductmarket.Whenthathappens,should
thesemultifacetedindividualstakeonmorethanoneroleforthegoodoftheteam?What
arethepotentialbenefitsandtheanticipated,andunanticipated,negativeimpactsthatcan
comefromallowingonepersontowearmultiplehatsonaScrumteam?
TheStory
MarcuswasaskedtobuildanewScrumteamtoreleaseamission-criticalapplicationfor
thecompany.MarcusandhisnewlyformedteamdidnothaveanyexperiencewithScrum,
buttheydidhavethesponsorshipandsupportofmanagementtolearnhowtousethe
Scrumframework.
Thefirstthingtheteammembersdidwasgettogethertofigureouthowtheywould
structurethemselves.Onformerteams,MiguelandJosehadbeendevelopersandHugo
andDominichadbeentesters.TheirrolesontheScrumteamwereclear:Theywouldbe
theteammembersresponsibleforexecutingthevisionofthebusinessascommunicated
andmanagedbytheproductownerthroughtheproductbacklog.
Marcus,ontheotherhand,wasmoredifficulttoposition.Onhisformerteam,hehad
beentheprojectmanager.Healsohadadevelopmentbackground.Theteamwentback
andforthonhowtobestuseMarcus.Shouldhebeamemberofthecoreteam,doingwork
onthesprintbacklog?OrshouldheserveastheScrumMaster,formalizinghiscurrent
leadershiprole?Or,shouldheperhapsbetheproductowner,helpingsteertheprojectin
therightdirection?AndifMarcusdidn’tfilltheseroles,whowould?Nooneelseonthe
teamhadhisunderstandingofScrumandthemarket.
Becausetheteammembersdidnothavetheauthoritytobringinconsultantstohelpfill
anygapsintheirskills,1theycameupwithwhattheythoughtwasthebestpossible
solution:Marcusshouldsplithistimebetweenallthreeroles.Mostlikely,theyimagined,
thetasksoftheScrumMasterandproductownerwouldn’trequiresomeonefull-time.
Marcuscoulddoboth.Then,theyreasoned,whenMarcuswasn’tbusydoing
“managerial”things,hecouldfillhistimewithteamtasks.Afterall,itwasn’tthatunusual
intheircompanytohavemanagersalsotakeonhands-onwork.Sotheyforgedahead.
1.SeeChapter3,“UsingTeamConsultantstoOptimizeTeamPerformance.”
Thingswentwellatfirst.Eachdayduringthedailyscrum,Marcusreportedhisstatus
liketherestoftheteam,highlightingthathedidnothaveanyimpedimentsandthatthe
taskshesignedupforwereontrackacrosstheboard.
Bytheendofthefirstweekoftheone-monthsprint,hehadmetwiththeknown
stakeholderstofurtherbuildtheproductbacklog,facilitatedallthenecessarymeetings,
andsignedupforseveraltasks(40hours’worth).Hisplanwastoaccomplishtenhoursof
workonthesprintbacklogperweek.Thisfirstweek,however,hehadnotspentevenone
houronthosetasks.Heandtheteamweren’tworried,though.HisScrumMasterand
productownerdutieswouldprobablyscalebacknow,theyreasoned,andtherewerestill
threeweeksleftinthesprintforhimtocompletehistasks.
ThesecondweekyieldedboththefirstScrum-relatedimprovementandthefirsthintof
troublebrewing.Marcusspentthefirstfewdaysoftheweekreviewingthesprintbacklog
lookingfortrends,oneoftheScrumMasterdutieshehadn’tgottentoduringthefirst
week.Whilereviewingthebacklog,henoticedthateveryworkitemthatHugoworkedon
wastakingabout50percentlongerthantheteamhadestimated.Thepatternwasuniqueto
Hugo;everyotherteammemberwaswithin20percentoftheestimates.Marcusraised
thispatterninthedailyscrummeetingthatFriday.Hugoconfirmedthathewasstruggling
withthework.TheteamdecidedthatHugowouldpairwithJose,aseniordeveloper,and
thattheywouldworktasksdowntogether.Marcuswasecstatic.Afteronlytwoweeks
usingScrum,theyalreadyhadidentifiedanissuethatwouldhavegoneunaddressed
otherwise.
Theteam,whilehappyabouttheimprovement,wasabitconcernedaboutthefactthat
Marcushadn’tfoundtimetoworkonanyofhistasks.Hisfellowteammemberspointedly
askedMarcusabouttheworkthathesignedupfor—couldhedoit,ordidtheyneedto
findanothersolution?Marcusagreedthatitwastimeforhimtoputonhisteammember
hat.Nextweek,hepromisedtheteam,hewouldbeginhisteammembertasks.
Weekthreequicklypassed.Despitehisbestefforts,Marcuscouldnotfocusonthe
sprinttaskshehadonhisplate.Histimewasconsumedbyacrisiswiththestakeholders
(15showedupoutoftheblue,claimingtheyhadbeenleftoutoftheoriginalmeetingsand
demandedbetterplacementoftheirstoriesonthebacklog),theresultingreprioritizationof
thebacklog,negotiationswithstakeholders,andall-aroundfirefighting.Marcusandthe
teamhadnoideahowtimeconsumingthejobofproductownercouldbe,nottomention
thetasksrequiredofaScrumMaster.
Marcuswasstuck.Hewasacommittedcoreteammemberandknewhisteammates
wererelyingonhimtodowork,hewastheproductownerandhadtoensurethe
stakeholderinterestswerebeingaddressedbytheteam,andhewastheScrumMasterand
wassupposedtobefocusingonteamhealth.Somethinghadtogive.
Atthestartofthefourthweek,Marcusdedicatedhimselftohissprintbacklogitemsin
solitaryconfinement.Heshutoffhisemail,retreatedtohisoffice,lockedhisdoor,and
shuttheblinds;however,theworkthathesigneduptodowasinfactmuchlargerthanhe
oranyoneelsehadanticipated.
ByWednesdayeveningofthefourthweek,Marcusacceptedthathewasnotgoingtobe
abletodeliverhistasksthissprint.Heburiedhisfaceinhishandsandbegantosearchfor
waystotelltheteamthathehadfailed.Suddenly,Marcushadwhatheconsidereda
brilliantidea.Hewasn’tjustateammember,hewastheScrumMasterandtheproduct
owner,too,wasn’the?
Delusionalfromburninghimselfout,hedecideditwastimeforMarcustheproduct
ownertomeetwithMarcustheScrumMasterandMarcustheteammember.How
convenientthattheycouldallgettogetherinoneperson!Togethertheydecideditwould
bebesttopostponethetasksthatMarcushadsignedupforandputthembackonthe
productbacklogtobereprioritizedandaddressedinafuturesprint.Marcustheteam
memberupdatedthesprintbacklog,revieweditwithMarcustheproductowner(allwith
MarcustheScrumMasterwatching),anditwasdone.Noharm.Nofoul.Marcustookoff
hismanyhats,wipedhisbrow,andsighedwithrelief.Problemsolved.
Thenextmorningatthedailyscrummeeting,Marcushandedcopiesoftheburndowns
toeveryonetoreview,justashehaddoneeveryday.Today,though,somethingsmelleda
bitrottenonthesesuddenly-on-targetburndowns.Miguelspokeup.
“Marcus,whatisthisdropintheworkremaining?”heasked.
“Ohthat,”Marcusreplied.“IremovedsometasksthatIhadsignedupfor.Nobigdeal.
Asproductowner,Idecidedwecandothatstuffnextsprint.”
Theroomwentsilent,andalleyesturnedtoMarcus.
“Youdidwhat?”exclaimedDominic.“Youcutyourownwork?”
“Marcus,youcan’tdothis.Whatwereyouthinking?”demandedHugo,without
allowingMarcustimetorespondtoDominic’squestion.
TheteamcontinuedtodigintoMarcus,demandinganexplanation.Marcusrealizedhe
wasinadeephole.
“Guys,I’msorry.I’msorry!”pleadedMarcus,tryingtogetawordin.“Ididn’tthinkit
wouldmatter!”
“Well,itdoes!”saidJose.
Marcusfelthisheartsinkintohisshoes.Hefeltawfulforlettingtheteamdownand
understood,toolate,thatwhathedidwaswrong.
“Marcus,ifthisisgoingtowork,youcan’tdothatagain,”saidHugo.
“Youneedtodecidewhatyou’regoingtodoonthisprojectbecauseitisclearthat
splittingyourtimebetweenScrumMasterandproductownerandsigningupforworkon
thesprintbacklogdoesnotwork,”addedDominic.“We’vegotteamissuesthataregoing
unaddressed.Weneedyoutoseewhatwearemissingontheteamdynamics!Allofyour
timehasbeenspentwiththecustomers,andyou’relettingusdown.”
Marcusacknowledgedhismistake.
“Guys,I’msorry.IdidnotthinktherolesofScrumMasterorproductownerwouldbe
sodemanding.Myjobasproductownerhastakenalmostallofmytimethissprint,andI
don’tseeitgettinganybetter.Idon’twanttoletyouguysdownasateam,butIknowmy
jobasScrumMasterisnotworkingouteither.Andmyteammemberrole!Man!Iblew
thatone,”admittedMarcus.
“Look,Marcus,”saidHugo.“Iknowthisistoughonyoumostofall,butweneedyou
tofocusononethinganddoitwell.Whatdoyouwanttodo?”
Marcuscontemplatedhisoptions.He’ddoneadecentjobasScrumMaster.Hereally
likedtheproductownerroleandcouldseehimselfdoingthatjob.Still,hehatedtogiveup
developmentworkaltogether.
“C’monMarcus,it’sbeenfiveminutes,”saidHugo.“Timetodecide.”
“I’llbetheproductowner—I’vegotthemostinvestedinthatnow.Whatarewegoing
todoforaScrumMaster?”
“Wewereallprettysureyou’dpickthat,”saidHugo,withtheteamnodding.“Whatif
wetalktoFilipaaboutbecomingourScrumMaster?She’sbetweenprojects.Shewould
probablyneedtoturntoyouforsomeanswersatfirst,buttheworkofScrumMaster
wouldbeherpriority,notyours,”saidHugo.
Marcusfelttheweightoftheworldbeingliftedoffhisshoulders.
“Thanks,guys,Ireallyappreciatethis.Icouldn’thavekeptupthecircusactforanother
sprint!”exclaimedMarcus.“IhadnoideawhatIwasgettingmyselfinto.Nooneona
Scrumteamshouldwearmultiplehats;Idon’tcarehowsmalltheteamis!”
TheModel
Inthischapter’sstory,MarcusandtheteamstruggledwithwhatroleMarcusshouldplay.
Hehadexperienceindevelopmentandhadmovedontobecomeaprojectmanager,where
hehadbeensuccessfulforseveralyears.Hewrestledwithhowtonarrowhisdiverseskill
settooneoftherolesinScrum.TeamsnewtoScrumoftenstrugglewiththesameissues.
TheytrytomapcurrentcompanyrolestotheonesinScrum,attempttomixroles,or
misinterpretwhatthetimeandskillrequirementsareforthevariousScrumroles.
Considereachoftheroles,startingwiththeproductowner.Theprimaryjobofthe
productowneristobethecustomerrepresentative,turningthevisionintoaworkable
productbacklogfortheteamtoexecuteandmanagingthereturnoninvestment/business
valueoftheproductorserviceunderdevelopment.Theproductownerhelpsbringclarity
tothevisionoftheproducttobedeliveredandthenworkswiththecustomersand
stakeholderstobuildalistofrequirementscalledaproductbacklog.Theproductbacklog
containsanythingthatneedstobedonetodeliverthevision—everythingthatwillcontend
fortheteam’stimeoverthelifeoftheprojectandthatbringsvaluetotheenddeliverable.
Typicalitemsinaproductbacklogincludestories(orfeatures),taxes,spikes,and
preconditions(formoreinformationonthis,seeChapter25,“OptimizingandMeasuring
Value”).Inservicetothestakeholders,theproductownercreatesasenseofurgencyand
pushestheteamtomaximizevelocityandquality.
TheroleoftheScrumMasterismultifaceted.First,theScrumMastershouldworkto
buildandmaintainahigh-performingteam.Onthesurface,thisroleappearstobesimple:
Managethedailyscrum,collectstatusfromtheteam,andcoordinatethevariousmeetings
andtasks.Thetruthis,though,thatnothingaboutaScrumMaster’sjobissimple.The
ScrumMasterisnottheteammanagerorleader;onthecontrary,theScrumMasteristhe
fluidthatensurestheteam’sengineisfiringonallcylinders.ScrumMasterskeeptheteam
focusedandontrack.Todothiseffectively,ScrumMastersneedtobeabletoseetheforest
forthetrees—toseethebigpictureandlookforplaceswheretheteamcanmake
improvements.
Second,theScrumMasterisexpectedtoworkwiththeproductowner,helpingas
neededbutalsoprovidingakeycheckandbalanceagainstthewilloftheproductowner.
Theproductownerwillwantteamstohaveahighvelocity;it’stheScrumMaster’sjobto
makethathappenwhileprotectingtheteam’shealth.GoodScrumMastersincrease
velocitybyremovingimpediments,monitoringteamhealth,andhelpingtheteamknow
howitistrackingtowarditsdeliveryofthesprintgoal.
ThefinaljoboftheScrumMasteristobeachangeagentintheorganization.Good
ScrumMastersreachbeyondtransformingtheteamtohelptransformthecompanyaswell.
ScrumMasterslookforopportunitiestocreatealearningorganization,offeringlunch-and-
learnandotherinformaltrainingonScrum,advisingnewScrumteams,andgenerally
beingaresourceandadvocateforScrumthroughouttheorganization.AScrumMaster’s
subtle,servantleadershipcanhaveatremendousimpactontherestofthecompany.
Coreteammembersdeliverthevisionandbuildtheworkneededtoimplement
featuresfromtheprioritizedproductbacklog,asmanagedbytheproductowner.Todo
thismosteffectively,teammembersdependontheScrumMasterandproductownerto
dealwiththenoise,enablingthemtofocussolelyonthetaskathand.
WhatmakesScrumuniqueisthateachrolehasminimaloverlapandisinintentional
conflictwiththeotherroles.Theproductownerisexpectedtodrivetheteamtoexecute
thevisionofthebusinessanddeliverbusinessvaluetothestakeholdersandcustomers.
TheScrumMaster,ontheotherhand,focuseshisorhertimeonensuringtheteamis
healthyandabletoexecutethevisionoftheproductowneraseffectivelyandefficiently
aspossible.Theteamconcentratesondeliveringthework.Withoutaproductowner,the
teamrisksmissingthemark.WithoutaScrumMaster,theteamrisksburnout.Without
committedteammembers,theprojectitselfisatrisk.
ChoosingRoles
Onceyouunderstandwhattherolesareandwhytheyexist,youneedtolookathowto
matchpeopletothoseroles.Thefirstthingtorememberwhenpickingyourrolesisto
disregardanyassignedtitlesthatmightexistinyourcompany.Titlesdonotmaptothe
rolesinScrum.Skillsdo.Considerthecompetenciesyouwanttohaveineachroleand
matchpeopletothem.
Table5-1summarizessomekeycompetenciestolookforwhenfillingeachofthethree
rolesinScrum.
TABLE5-1ScrumRolesandCompetencies
Youcanflipthisaroundtheotherway,too,ifit’smorehelpful.Listthekeydutiesofa
typicalprojectmanagerinyourorganization,forinstance.Thencomparethosetothe
responsibilitiesofeachScrumrole.InTable5-2,youcanseehowsomeoftheworkofa
traditionalprojectmanagermapstotherolesinScrum.
TABLE5-2MappingProjectManagerDutiestoScrumRoles
Thoughthesechartscanhelpyoustopthinkingintermsoftitlesandstartthinkingin
termsofskillsets,choosingpeopleforataskisneverthatcutanddried.Ifaformer
projectmanagertakesontheroleoftheScrumMaster,shemightfindherselfwishingshe
couldbemorehands-onwiththetasksinthesprintbacklog.Atesterwhoisparticularly
giftedatfacilitation,ontheotherhand,mightdesiretotakeontheScrumMasterrole.
Givepeoplefreedomtochoosetherolesthatbestsuitthem.Whatpeoplearecalledisso
muchlessimportantinScrumthanwhatpeopledo.
MixingRoles
MixingrolesinScrumiswhenasinglepersontakesonmorethanoneScrumrole.For
example,apersonisateammemberandtheScrumMaster,orapersonistheScrumMaster
andtheproductowner,orapersonistheproductownerandateammember.Theother
option,ofcourse,isMarcus—apersonwhotakesoneachrolesimultaneously.
Itistemptingtothinkthatasinglepersoncansplittimeacrossroleseffectively;after
all,thatiswhatpeoplehavebeendoingforyearsontheirprojects.I,too,havemixedall
therolesinScrum.LikeMarcus,IthoughtthatIcoulddoeverything,and,likeMarcus,I
failedmiserably.Ihavealsotriedeachofthepossiblecombinationsatleastthreetimes.
Eachtimetheyendedinfailure.Icanandwilltellyounottodoit.However,Ialso
understandthatyou’llwanttotryitanyway,soconsiderthisadisclaimer.
Peopleoftenaskwhat’swrongwithmixingroles.Thisquestioncomesfroma
fundamentalmisunderstandingofwhateachroleentailsandhowthatroleinteractswith
theothermembersoftheScrumteam.
Acombinationproductowner/ScrumMasterisakintoatwo-headeddragon.Eventually
oneheadwilleattheother.TheproductownerandScrumMasterrolesareinconflictwith
eachotherbydesign.Oneexiststodrivetheteam,theothertoprotecttheteam.Suppose
youaretheproductownerforateamandthecustomerisridingyouhardtopushtheteam.
Ifyoudonothaveamechanisminplace(theScrumMaster)tomonitorthedemandsbeing
placedontheteam,itislikelytheteamwillbepushedbeyondthelevelofitsendurance.
Productivitywillberealizedintheshorttermduetotheextrapush,butthecustomerwill
cometoexpectthisasalong-termthing.Theissuehereisthattheteamisnotworkingata
sustainablepace.Eventuallyproductivitywilldecreaseandmoralewillsufferasteam
membersburnout.Theproductownermightnotcareiftheteamiskeepingitsmoraleup
andmightnotevencareifqualitysuffers,aslongasthedatesarehitandthecustomer’s
needsareperceivedtohavebeenmet.
It’sequallyineffectivetohaveaScrumMasterbutnoproductowner.Withoutthesense
ofurgencyaproductownerbrings,itislikelythattheteamwillmovetooslowly.Without
aproductownertokeeptheteamfocusedondeliveringvalue,theteammightinclude
functionalitythatitthinkswouldbefuntobuildorwouldbesomethingnicetohave.The
team,notthecustomer,soonendsupdrivingtheproductvision.Inthesecases,the
customerwillsoonbegintoquestionwhethertheproductwilleverbedelivered,will
wonderwhethertheteamcouldbemovingfaster,andwillwonderwhyitispayingfor
featuresitdidnotaskfor.Eventually,thecustomerwilllosefaithintheteamaltogether.
Ifonepersonisplayingbothroles,andthereisapushycustomer,thepersonplayingthe
roleswillneedtodecidewhichjobwinsout:theScrumMaster,wherebyhewillnotsatisfy
thecustomerbutwillprotecttheteam,ortheproductowner,wherebyhewillsatisfythe
customerandwilllikelyhurttheteam.Theseconflictscomeupineveryproject,whichis
whyitisimportanttonotmixroles.
Incontrasttothetwo-headeddragonthatacombinationproductowner/ScrumMaster
becomes,acombinationteammember/ScrumMasterseemsinnocuous.Still,thereare
issuesthatmakethiscombinedrolelessthanideal.Earlierinthischapter,Isaidthatthe
ScrumMaster’sjobistobuildandmaintainahigh-performingteam.Considertheteam
member’sprimaryfunction:tobuildandexecuteonthevisionascommunicatedbythe
productowner.ThisjobdirectlyconflictswiththejobofaScrumMaster.Ifthe
ScrumMaster’sjobistoseetheforestforthetrees,thejobofateammemberistobein
theforest,staringatindividualtrees(e.g.,userstories)forthedurationofthesprint.The
problemwithmixingthesetworolesisthatiftheScrumMasterisdoingtask-specific
work,shewon’tbeaslikelytoseethebiggerissueswiththeteamandthereforecannot
elevatetheteamtogreatness.
Acombinationproductowner/teammemberroleisnearlyimpossibletodowell.A
productownerjusthastoomuchnoisetocontendwith.Earlyinanyproject,thejobofthe
productownerwillbeintense.Theproductownerwillbeworkingwithcustomersand
stakeholders,workingtobuildaproductbacklogandtounderstandthevisionindepth.
Thefastertheteamgoes,theharderitwillbefortheproductownertokeepupwiththe
team’sdemandforstories,clarification,andverification.
Theonlybenefittothiscombinationisthattheteamwillhavetheproductownerinthe
roommostofthetime.Eventhen,itisdifficultforproductownerstohaveaclearvision
ofwhattheteamisworkingtoexecutewhentheyaresplittingtimebetweendevelopment
andproductownerduties.
WhataboutExtremeProgramming’scallforhavingthe“customerintheroom,”which
Iamallforbutrarely,ifever,see?Couldhavingacombinationclient/productownerbe
thenextbestthing?Sometimes.
First,considerthatiftheproductowneristheactualclient,heholdsthemoney.This
giveshimthepowertopullthe“Ihavethemoneycard,sodowhatIsay”routine—nota
funsituationfortheteamortheScrumMasterwhoistryingtoprotectthatteam.
Second,considerthatacustomerhasotherworktodo.Forexample,acolleagueof
minewasrunningaScrumprojectatanorganizationI’llcallZDev.Theclient,I’llcallit
WiCo,hadarepresentativeactingastheproductowner,butnotbychoice.Shewas
electedproductownerbyhermanageratWiCobecauseshehadthemostknowledgeof
thesystemthatwasunderdevelopment.Sheservedastheactualproductownerfor
ZDev’sScrumteambecauseWiCodidn’twanttopayZDevforaproductowner’stime.
Thingsweregreatinthebeginning,andtheteamgotwhatitneeded.Astimewenton,
though,theclient/productownerbecameincreasinglydistant,eventuallyonlyshowingup
forthedemomeetingsandeventhennotbeingengaged.ZDevnoticedthistrendand
stoppedtheproject.ThebigwigsatWiCoflippedoutanddemandedanexplanation.
ZDevexplainedthattheproductownerwasnotengaged;underthosecircumstances,
ZDevwouldratherbuildnothingthanbuildthewrongthing.
WiCodidaninvestigationastowhytheproductownerhadcheckedoutoftheproject.
Theclient/productownerexplainedthatshehadtokeepdoingherprimaryjobinaddition
totheextraworkofproductowner,whichhadbeenhandedtohersimplybecauseshewas
asubjectmatterexpert.Whenshelookedatwhatshewouldbejudgedonforher
performancereview,sheelectedtosacrificetheZDevprojectbecauseitwastiny
comparedtoherothergoalsfortheyear.WiCorealizedthatitwouldhavebeencheaper
andmoreeffectivetohavehiredaproductownerfromZDevinthefirstplace,soit
restartedtheprojectwithaproductownerfromZDev.
Combiningallthreeroles,asituationIaffectionatelyrefertoasthetrifectaofdeath,is
anexceedinglybadidea.Ican’tthinkofanupside.Still,I’vebeenwhereyouareand
knowmanyofyouwilltryitanyway.Ifyouhaveaone-ortwo-personteam,thismight
work—eventhen,don’tbetonit.Thisdysfunctionisasymptomoflargerissues,
includingthepossibilitythatyoushouldnothavetakenontheprojectinthefirstplace.
When,NotIf,YouDecidetoMixRolesAnyway
HopefullybythispointIhavemadeaclearcasefornotmixingtheroles.Irealize,though,
thatyou’reprobablygoingtotryitanyway.Whenyoudo,Isuggestyoumixonlythe
teammemberandScrumMasterroles.Why?It’stheleastdamagingcombination,putsthe
ScrumMasterintheroomwiththeteam,andwilllikelybetheeasiesttosellto
management,especiallyiftheyaresayingyou“cannotdoScrum”becauseofthe
perceivedadditionaloverheadintheroles.Becarefulwhenyoudoit,andmakesure
everyoneontheteamisonthesamepagefromthebeginning.Readmoreabouthowto
justifyhavingafull-timeScrumMasterinChapter8,“TheCaseforaFull-Time
ScrumMaster.”
KeystoSuccess
Marcus’storyillustratesanextremecaseofmixingrolesinScrum,butthemoreIwork
withteams,themoreIfindmixingtobeanoptionthatteamspropose.Itisalltoo
commontoseeaScrumMasteralsoactasateammemberandisnotunusualtoseeone
personactingasbothScrumMasterandproductowner.Neithercombinationresultsinthe
bestsolutionforyourcustomersoryourteam.
Evenonasmallteam,resistthetemptationtodeviatefromtheScrum“recipe,”
especiallyatfirst.Youwillfindthatyouultimatelyperformmuchbetterifyoutakethe
extraefforttobuildtheintentionaltensionthathavingthreedistinctrolesonyourteam
provides.Ifyou’restrappedforcashorpersonnel,shareaScrumMasterorproductowner
withanotherteam.Havingapart-timeScrumMasterismuchbetter,andsomuchmorein
keepingwiththechecksandbalancesinherentinhavingthreedistinctroles,thanhaving
oneteammembertryingtofulfillmultipleroles.Ifyoudothis,keepinmindthatit’sfar
easiertoshareaScrumMasterthanaproductownerbecauseonceateamgetscranking,a
part-timeproductownerlikelywillhavetroublekeepingup.
Ifyouareinacoreteammemberroleandmissyourleadershipduties,lookforwaysto
leadfromwithintheteam.Headupanimprovementcommunity.Startorgrowalocal
user’sgroup.BecomeaScrumadvocateinsideyourorganizationbybeingamentoror
offeringlunch-and-learnopportunities.Youmightevenconsidervolunteeringasthepart-
timeScrumMasterforanotherteam,whereyourdutieswon’tdirectlyconflictwithyour
teammemberresponsibilities.
Becreative.Bepersistent.Andbepreparedtojustifytheimportanceofhavingthree
separateindividualsservingasScrumMaster,productowner,andteammember.
SometeamshaveclaimedsuccessinrotatingtheroleofScrumMasteramongtheteam
members.I’veneverattemptedthismethodandhavenotseenthisresultinagood
outcomeforanyonewhohas.Withoutlong-termfocus,theScrumMasterbecomesa
glorifiedadministratorfortheteam.TheskillsrequiredofaScrumMasterarespecialized
andtheroletooimportanttobehandedaroundthecampfirelikeabottleofmoonshine.
IprefertohaveoneScrumMasterforaslongastheteamistogether,andseenothing
wrongwithhavingtheteamchooseanewScrumMasteriftheteamfeelsthatperson
mighthavetherightstufftoshepherdtheteamandstanduptotheproductowner.
Irecognizethatrealityoftengetsinthewayofwhatwouldbeideal.Butinallyour
decisions,keepinmindthatwhenpushcomestoshove,someonemixingrolesultimately
willbeforcedtochooseasinglerole.Someonesplittingtheteammemberand
ScrumMasterroleswillbeforcedtochoosewhichonetodoandwhichonetosacrifice.
Thisisinevitable.Recognizethiscostifyouestablishateaminwhichpeoplemustactin
multipleroles.
Chapter6.DeterminingSprintLength
Thereisnoone-size-fits-allmagicbulletfordeterminingasprintlengththatworkswell
foreveryteam.Originally,Scrumcalledforone-monthsprints,butnowadaysmanyteams
havebeensuccessfulwithtwo-weekorevenone-weeksprints.
Withalltheseoptions,howdoyouchoosethesprintlengththatworksbestforyour
teamandproject?Isthereeveratimewhenit’sokaytohaveasprintlengththatislonger
thanonemonth?Whywouldateamwanttoimplementone-weeksprints?Andwhatare
thefactorsthatinfluencewhattherightsprintlengthshouldbe,especiallyonanew
project?
Thischapterexploresthesequestionsandtheconceptofsprintlengththroughastory.
Afterthat,itdescribeswaysyoucandeterminewhichsprintlengthmakesthemostsense
foryourteam.
TheStory
JayandRachelweretappedtodeliveranorder-trackingsystemfortheirclient,Tegla,an
executiveaircraftmanufacturer.Astheypreparedtomeetwiththeclient,JayandRachel
struggledwithhowbesttodealwiththeproject’srequirementsandconstraints.
“Rachel,I’vebeengoingovertheserequirementsdocumentsfordaysandI’mstuck.
We’vegotsomedifficultconstraintstoworkwith.Teglaneedsustobuildanorder-
trackingsystem.TheproblemisTegla’sbudgetcycleendsattheendoftheyear,which
meanswehavetogetthesystembuiltquicklyandcorrectly,becauseafterthefirstofthe
yearwewon’thaveabudgettofixit.FromwhatIcantell,thestakeholdersatTeglaare
notyetveryclearontheirrequirements.Alltheyknowistheydon’twantthenextversion
ofwhattheyhavetoday;theywantsomethingcompletelynew.”
“Wow,”saidRachel.“That’svague.”
“What’sworseisthattherequirementsareabitdated.Thereisavisualdesign
documentthatsomeoneputtogetherforthem,completewithUIelements,behaviors,and
systeminteractions.Wecanusethisasastartingpoint.Wealsohavethelegacysystemas
areference,alongwiththeusecasesthatwerebuilttosupportit.Therestofitwillneedto
bedevelopedovertime,”repliedJay.
“Hmmm,”saidRachel.“Let’sdivideandconquerandseewhatwecanfigureout.If
youcanidentifytheessentialpeopleforthecoreoftheteambasedonwhatyouknow,I’ll
talkwiththecustomerandcomeupwitharecommendationforhowtostructurethis
project.”
Afterafewdays,RachelandJaymetagain.
Jaydescribedhiseffortstobuildacoreteam.Rachelrespondedwithherown
conclusionsandquestions.
“IthinkusingScrumtomanagethisprojectwouldreallyhelp,”saidRachel.“How
muchexperiencedoesthecoreteamhavewithScrum?”
“Actually,theteammembershavebeenonafewScrumprojectstogetherandare
experiencedwithtechnicalskillsliketest-drivendevelopmentandcontinuousintegration.
They’reprettystrong,”Jayreplied.
“Good,”saidRachel.“IspokewiththestakeholdersatTeglaandgavethemahigh-level
overviewofourvariousprojectengagementstyles,andtheylikedScrumthebest.Igave
themabunchofinformationonhowScrumworks,andtheywerereceptive.WhatIdid
nottellthemishowlongthesprintswillbe,soweneedtofigurethatout.Here’swhatwe
knowsofar.”
Rachelwroteabulletedlistonthewhiteboard.
“Greatsummary,Rachel,”saidJay.“ThebiggestissuesIseearethe31December
deadlineandthelooserequirements.Wedon’tknowexactlywhatitistheywant,and
neitherdothey.Withourteam,theirdeadlines,andthefuzzynatureoftherequirements,
I’mthinkingshortfeedbackcycleswillworkbest.I’mthinkingweeklysprints.”
Rachelsmiled.“Metoo.TherealvariableisTegla.Idon’tthinkthestakeholdershave
thetimeneededforone-weeksprints.Theywouldhavetobeverycommittedand
availabletotheprojectforone-weeksprintstowork,andIjustdon’tknowifthey’llbe
willingorabletodothat.Logistically,itwouldbepossible.Allthecustomersarelocated
intheonelocaloffice.Still,they’reprettybusywithotherduties.Ican’tthinkofany
projectwithTeglawherewe’veaskedforthatlevelofcommitment,”saidRachel.
“Well,alloftheotherprojectswehavedonewithTeglahavebeenbetterdefined,”
counteredJay.“Wedidn’tneedthatlevelofcommitment.”Hepausedforamoment,then
continued.“Let’ssetupameeting.We’llaskthemafewquestionsandgetanideaofhow
muchtimetheyhavetocommittotheproject.Then,we’llmakeourcaseforweekly
sprintsandseewhattheyhavetosay.Neverhurtstoask,right?”
ThecustomermeetingcamebeforeJayandRachelknewit.Theyhadagoodturnout:
AllfivecustomerrepresentativesfromTeglawerepresent.Jaydecidedtostartby
recappingwhyeveryonewasthereandwhatthepurposeofthemeetingwas.“Hifolks,”
hesaid.“Weareheretodaytotalkabouthowwearegoingtoworktogethertobuildthe
systeminthetimeframeyouwant.”
Jaycontinued.“Wehaveasolutionthatwethinkwillworkforourteam,buttomakeit
workweneedyourcommitment.Ourapproachwillprovideyoufrequentopportunities,
wehopeweekly,toreviewthesystem.Wewillworkwithyoueachweektodetermine
whatwebuildthefollowingweek.Wewilldelivercodeinastatethatispotentially
shippablesowecanputitonyourproductionsystemsandnothaveissuesdealingwith
operations.”
Thecustomers’eyesgotverywideuponhearingthis.“Weeklyreleasesandfeedback?
Soundsgreat.What’sthecatch?”oneofthemasked.
“Well,inordertoshipeveryFriday,wearegoingtoneedyoutospendalotofyour
timeguidingusintherightdirection.We’retalking20to30hoursperweekforone
person,lessfortherestofyou.Thattimewillbespentrefiningtheproductbacklog,
runningthroughallthestoriestoensuretheyworktoyoursatisfaction,helpingprioritize
theupcomingsprints,reviewingandhelpingtriagebugs—andwritingbugsforthat
matter.Also,we’llneedyouto—”
Jaystoppedinmidsentenceashesawtheeyesofthecustomersglossoverindisbelief.
Oneofthemspoke:“Wehavejobstodo.Howarewegoingtobeabletodothis?”
“Well,”saidJay.“Itisalot,Iagree.”
Rachelnoddedheragreement.
Jaycontinued,“However,doingthiseachweekwillgiveusabout22opportunities—
oneperweekforsixmonths—toadjustthedirectionoftheproject.It’sthebestwayto
makesurethatwhatyouhaveattheendoftheyeariswhatyouwant.”
“Itmightbethebestway,”repliedoneofthecustomers,“butIknowIcan’tcommitto
theamountoftimeandfeedbackyouwillneed.Isthereawaywecanstretchitoutabit?
Whatifwedidthisonceamonthinsteadofeveryweek?”
Rachelchimedinthistime,“JayandIdiscussedthisoptionbeforethemeeting.
Monthlyfeedbackcyclesdoworkformanyprojects.Theissuewebothidentifiedis
aroundtherequirements;theyareverysoft.Combinethatwiththefixedend-of-year
deadlineandwerunahighriskofbuildingthewrongthing,whichnoneofuswant.
Monthlysprintsonlygiveusfiveopportunitiesforfeedback,becausebythesixthsprint,
we’redeliveringthesystem.”
“It’sallaboutstimulus-to-responsetime,”agreedJay.“Ifyouweretocutyourfinger,
howquicklywouldyouwantyourbraintorespondtothecut?Immediately,right?Well,
whenwebuildapieceoffunctionality,it’sbetterfortheoverallhealthofthesystemif
yourespondtothesechangesalmostasimmediately.So,wewanttobuildinseveral
stimulus-to-responsecycles.Doingthatrequiresaninvestmentonyourpart.However,
makingthatinvestmentmeanswehaveahigherlikelihoodofgettingitright.”
Rachelnoddedandsaid,“Iknowyou’dratherhavemonthlysprints,butgiventhetime
durationoftheproject,Istronglyfeelwewon’tgetenoughresponsethatway.Whatifwe
compromiseandhavetwo-weeksprints?That’snotasideal(orintense)astheone-week
sprintsweproposedoriginally,butitstillgivesusanadequatestimulus-to-responsetime.”
Thecustomerteamknewthemath.Asix-monthreleasewindowwithtwo-weeksprints
meanttheywouldhavebetweentenandtwelveopportunitiestoimpactthedirectionofthe
project.Theyalsoknewthatgettingthesysteminplacebytheendoftheyear—and
havingitbetherightsystem—wascritical.Otherwise,thebudgetwouldbegoneandthey
wouldbestuckwitheitherhalfasystemoracompletesystemthatnoonecoulduse.
“We’llfindawaytomaketwoweekswork.We’lljusthavetojuggleourschedulesa
bitforthenextsixmonths,”saidtheirdecisionmaker.“We’regoingtoneedyourhelp,
though.WearenewtothisScrumthing,soyouwillhavetohelpguideusthrough.Weare
alittleskeptical,butyoumakeagoodargument.Let’sgowithatwo-weeksprintlength
andgetthisthingoutthedoor.”
TheModel
Choosingtherightsprintlengthisaboutdetermininganappropriatestimulus-to-response
cycle.Inthischapter’sstory,theteamandthecustomerwerepressedwithatightdeadline,
somethingnotuncommoninsoftwareprojects.RachelandJayexplainedtothecustomers
that,withtheirfirmenddateandfuzzyrequirements,thestimulus-to-responsetimeforthe
projectneededtobeasshortastheycouldhandle.
Inasprint,theinitialstimulusisthecustomersettingthepriorityofthestories.The
responseistheteambuildingworkingsoftware.Thebuildingoftheworkingsoftwareisin
turnastimulustothecustomer.Thecustomer’sresponseisfeedback.Themore
instantaneousthatfeedback,thehigherthelikelihoodthatwhatisdeliveredintheendis
whatthecustomerreallywanted,whichisnotnecessarilywhatthecustomeroriginally
askedfor.Thisstimulus-responsecyclecontinuesuntiltheprojectends.
RachelandJayrecommendedone-weeksprints.Theyarrivedatthisconclusionby
takingintoaccountthefollowingcriteria:
Theexpecteddurationoftheproject
Customers/stakeholders
–Availabilityforfeedbackandguidance
–CulturalbarriersandScrumfamiliarity
–Environmentalfactors
TheScrumteam
–Scrumexperience
–Technicalcapabilities(suchasautomatedacceptancetesting,test-driven
development[TDD],automatedreleases,etc.)
–Abilitytodecomposework
Theseelementsplayacriticalpartwhenanyteamistryingtodeterminehowlongits
sprintsshouldbe:fromoneweektoonemonth.Thenextsectionlooksateachelementin
moredetail.
ProjectDuration
Sprintlengthsshouldbechoseninrelationtoprojectduration;however,theyshouldnever
belongerthanonemonth.Considerathree-monthproject.Ifithasone-monthsprints,the
stakeholderswillgettoparticipateinonlytwodemosbeforetheprojectisreleased.This
isnotenoughfeedbacktomitigatetherisks.Shortersprintlengthsareanecessity.
Whataboutayear-longproject?Assumingone-monthorfour-weeksprintsthatbreak
approximatelymonthly,thatprojectwouldoffer11opportunities(plusthefinalrelease)
forstakeholderstoseethedevelopingproduct.One-monthsprintsarearealisticoption,
dependingontheotherfactorsinvolved.
Theareabetweenthreemonthsandoneyearislessclear.Theprojectinthischapter’s
storywasasix-monthproject.Ifasix-monthprojectisrunonaone-monthsprintmodel,
theteamhasascantfiveopportunitiesforfeedbackbecausethelastsprintwillendthe
project.Isthatenoughfeedbacktoproducetherightproduct?Itdepends.Myadvicefor
projectsofgreaterthanthreemonthsandlessthanoneyearwouldbetochooseasprint
lengththatmirrorstheprojectlength—meaningaprojectthatisshortshouldhaveshort
(one-totwo-week)sprints.Aprojectthatislongcanhavelongersprints,betweentwo
weekstoonemonth.Again,andIcan’tstressthisenough,sprintlengthshouldneverbe
longerthanonemonth.Infact,Irecommendtwoweeksfornearlyallofmyprojects,asI
finditgivestheidealamountofstimulustoresponse.
Ifaprojectisexpectedtolastmorethanoneyear,rethinktheproject.Multiyearprojects
aretoobig.Findawaytodivideitintoshorterreleases.Justasyouwouldn’twantatwo-
monthsprint,youdon’twantatwo-yearproject—divideitintofour,six-monthprojects
instead.Ifyoufindyoucan’tcomeupwithawaytobreaktheprojectintosmaller
components,thenyoulikelyhavelargerculturalissuesinyourcompanythatshouldbe
addressedbeforeyoutrytouseScrum.
Figure6-1showsanotherwaytoviewit.Thelighterthesprintlengthbaris,themore
likelyyouaretobuildtherightthingbecauseyouwillhavemoreclarity.Asyoumoveto
theright,toadarkershade,youhavelessclarity,lesscustomerinvolvement,andyou
increasetheriskofbuildingthewrongthing.
FIGURE6-1Amountoffeedbackandcustomerinvolvement
ProductOwnerandStakeholders
Whenworkingtodeterminesprintlengthwithyourproductownerorcustomer,itis
importanttofactorinthreethings:
1.Productownerandstakeholderavailability
2.CompanycultureandfamiliaritywithScrum
3.Productenvironment
Lookattheavailabilityfirst.It’scrucialtobalancetheteam’sneedswiththeneedsof
yourstakeholders.Havingaone-weeksprintmightsolvesometeamproblemsbutcould
puttoomuchstressontheteam’srelationshipwithitsstakeholders.Discussthesituation,
justasRachelandJaydid.Explainwhystakeholderfeedbackissoimportanttothe
project’ssuccessandtellthemwhattheywillgetoutoftheaddedexpenseoftheirtime.
Remember,inmostcases,thosewhoarenewtoScrumarenotaccustomedtoScrum’s
demandforinteraction.Theyareusedtodroppingtherequirementsoffatthebeginningof
theprojectandpickinguptheirproductattheend.Regardlessofthesprintlengthyou
choose,youwillwanttospendtimeeducatingeveryoneaboutthevalueoffrequent
feedbackandthenecessityofstakeholderinvolvementinthedevelopmentprocess.
Next,examinethecompanyculture.IsworkingwithaScrumteamnewtothecompany
anditsculture?Isitperhapsbeingtestedinapilotgroup?Ifso,considerthatshortsprints
maygiveafalseimpressionthatScrumteamsputunduepressureonstakeholders.Find
waystoalignthevaluesofthecompanywiththoseofthestakeholdersandScrumteam.If
thestakeholdersandScrumteamvaluetheideaofhavingworkingsoftwareeverytwo
weeksbutthecompanyvaluesanall-encompassingprojectplan,theprojectwillstruggle
regardlessofsprintlength.Again,educationiskey.Meanwhile,themoreresultstheteam
produces,themorewillingareluctantcompanywillbetoacceptthevalueoffrequent
iterationsoveracomprehensiveplan.
Theenvironmentinwhichthefinalproductwilllivecanalsoaffecttheproject.
Environmentalfactors(suchasregulatoryissues)areoftenoverlookedbut,frommy
experience,affectwhattheteamcandeliverandwhen.Ifyourteamhasalargenumberof
regulatoryissuestocontendwith,youwillprobablywanttochoosesprintlengthsofat
leasttwoweeks(threemaybeideal)becauseoftheamountofcomplianceassessmentand
auditingthatwillneedtooccurbeforeanysoftwarecantrulybeconsideredshippable.
ScrumTeam
TheScrumteamitselfisafactorindeterminingsprintlength.Considerationsincludethe
following:
TeamexperiencewithScrum
Teamengineeringcapabilities(TDD,continuousintegration,pairprogramming,
etc.)
Teamabilitytodecomposework
Inthischapter’sstory,theteammembershadsomeexperiencewithScrumandXP.
Theykneweachotherwell,wouldnothavetostartfromscratch,andfeltcomfortable
workingwithintheScrumframework.Noneofthemhadthe“Iownthispartofthecode”
mindset.JayandRachelknewthis,sotheyfeltconfidentwithone-weeksprints.Asarule,
aninexperiencedScrumteamshouldstickwithtwo-weektoone-monthsprintsbecausea
one-weeksprintismuchmoreadvancedandbettersuitedforestablishedteamsthathavea
goodworkingrelationship.
Ifyoudochooseaone-weeksprintlength,youneedtohavetheengineeringcapabilities
inplace(orlearnthemawfullyquickly)toexecuteworkinthatcompressedtimeframe.
Goodengineeringpracticesarejustasimportantwithlongersprintlengths,buttheir
absenceisnotaspainfulorglaringlyobviousonaone-monthsprintasitisonaone-week
sprint.Toefficientlyexecuteaone-weeksprint,theteammustbeabletoestablishand
maintainacontinuousintegrationserverandshouldfollowengineeringpracticessuchas
TDDandpairprogramming.Utilizinganautomatedacceptancetestingframeworkisalso
arequirement,asthetestingdebtgrowthwillbefastandfurious.Iflearningthese
mechanismsinahurryisyourgoal,thenaone-weeksprintmightmakesense.Otherwise,
don’teventhinkaboutaone-weeksprintuntilyouhavethesecapabilitiesinplace.
Theabilityoftheteamtodecomposeitstasksiscriticalonshortsprintsandstill
importantonlongersprints.Decomposingtasks,inandofitself,isanartform,onethat
requirescriticalthinking,aclearmind,andtheabilitytoseebeyondthesurfaceofa
problem.Ifyourteamisgoodatdecomposingtasks(orwantstogetgoodinahurry),
shortersprints(onetotwoweeks)canwork.Ifyourteamisstilllearninghowto
decomposetasks,youmightwanttostartwithlongersprints.Readmoreabouttask
decompositionandwaystoimproveinChapter12,“DecomposingStoriesandTasks.”
DeterminingYourSprintLength
“Sothisisallfineanddandy,”yousay,“butattheendoftheday,Ineedtofigureouthow
longmyteam’ssprintsshouldbe!”
IfindmyselfaskingaseriesofbasequestionseachtimeIsetupaproject,justasJay
andRacheldidinthestory.Thesequestionshelpdeterminesprintlengthbutarenot
conclusiveandshouldbeconsideredguidelinesratherthanhardandfastrules.WhatI’ve
endedupwithisaquizsimilartooneyou’dseeinamagazine,whereyouansweraseries
ofmultiplechoicequestions,giveacertainvaluetoeachanswer,andthentabulatethose
answerstoarriveatyouridealsprintlength.Aswithanytool,useitwisely—iftheanswer
itgivesyoudoesnotfeelright,thenchooseadifferentsprintlength.
First,answerthemultiplechoicequizshowninTable6-1.Choosetheanswerthatmost
closelymatchesyoursituation.
TABLE6-1QuestionstoAsk
Onceyou’veansweredthemultiple-choicequestionsinTable6-1,usethescoringkey
inTable6-2toassignanumberforeachquestionbasedonyouranswer.You’llnoticethat
therearethreenumbersinthekey:1,3,and5,where1=BadChoice;3=Workable;5=
Ideal.You’llcarryallfournumbersthatmatchyouranswerintothesprintlengthmodel.
TABLE6-2ScoringKey
Asanexample,recallthatinthestoryJayandRachelhadatimelineofsixmonths.I’ll
usethisasthedatapointtoanswerthefirstquestion:Whatismyprojectduration?Their
answerwouldbeC.6–8months.IfyouhadaprojectdurationthatmatchedJayand
Rachel’s,youtoowouldgotothe6–8monthsrowinthescoringkeyandcopythe
numbersintothesprintlengthmodel,asshowninTable6-3.
TABLE6-3AnswerstotheProjectDurationPortionoftheSprintLengthModel
Youwouldthenanswerthenextquestion,takingthenumericalvaluesfromTable6-2
andpopulatingthesprintlengthmodel.Youshouldproceeduntilyouhaveansweredall
questions.
Forexample,assumethatyouranswersforquestions2through5(aboutthecustomer)
wereallB,andyouranswerstoquestions6through8(abouttheteam)wereB,A,andA.
YourcompletedmodelwouldlookliketheoneshowninTable6-4.
TABLE6-4ExampleSprintLengthModel
Inthelastrow,youwillseesomenumbers(3,4.75,4,and3).Thesearecalculated
averages(Iaddedthenumbersineachcolumnanddividedbyeight,thenumberof
questions).Basedontheaveragerating,yourhypotheticalprojectshouldworkbestwitha
two-weeksprint,withthreeweeksbeingaclosecontenderandone-weekandone-month
sprintscominginlast.
BeWarned
Youshouldhavenoticedthatsomeoftheoptionsinthemodelaboveruncontrarytothe
recommendationsIhavemade.Youcanselect,forinstance,aprojectdurationgreaterthan
12months.Theseoptionshavebeenhighlightedaswarningsignanswers(denotedbyan
asteriskinthescoringkey).Ifyoufindyourselfwithoneormorewarningsignanswers,
youlikelyhavelargerissuesthansprintlengthtodealwithinyourcompany.Theseissues
shouldbeaddressedbeforeyoutryScrum.TeamsthatstartusingScrumwiththeseissues
inplaceusuallyendupfailingandthenblametheframework.UsingScrumforthefirst
timeisstressfulenough.Tryingtodoitwithtoomanypreexistingwarningsignsisnearly
impossible.
BeyondtheQuiz
Afteryougainmoreexperiencewithchoosingsprintlength,youwillbeabletogaugea
goodsprintlengthforyourteaminmuchthesamewaythatJayandRacheldid,by
analysis,discussion,andfeel.
Whenteamsdonothaveexperiencerunningsprints,theytendtoexperimentwith
varioussprintlengths,whichcanbeconfusingforthecustomeranddebilitatingtothe
team.OnecompanyIworkedwithwantedtoexperimentwithallthesprintlengthstosee
whichonefitthembest,eventhoughtwoweekswastheobviouschoicebasedonthe
sprintlengthmodelandjustsimplegutfeel.Theystartedoutwiththree-weeksprints,then
two-weeksprints,thenaone-weeksprint,thenaone-monthsprint,thenbacktoatwo-
weeksprint,andfinallyathree-weeksprint.Theyendedupsettlingontwo-weeksprints,
butbythetimetheteamfinallylandedonanoptimalsprintlength,theteammembersfelt
liketheyhadjustwalkedawayfromacarcrashwithnoseatbelts.
IfyourteamisnewtoScrum,youshouldstartwithoneofthetwoacceptedstandards
(twoweeksoronemonth)anddothatconsistentlyforseveralsprints.Ifyoufindthatyou
arehavingdifficultyatthatpoint,itismuchmorelikelythatScrumhasuncovereda
problemyourteamishavingthanthatthereissomethingwrongwithyoursprintlengthor
withScrum.
KeystoSuccess
Determiningsprintlengthdoesnotneedtobeadauntingtask,butitalsoisnotsomething
thatshouldbetakenlightly.
Whendeterminingsprintlength,itisessentialtorememberthefollowingthings:
Stabilityofrequirements
Projectduration
Amountofriskyouarewillingtotakeon
Theabilityoftheteam
Thetoleranceofthestakeholders
Foreachofthesepoints,rememberthatthestimulus-to-responsetimeshouldbethe
maindriverinyourdecision.Themorefeedbackyouhave,thebetteryouwillbeableto
toleraterisk.Shortersprintsenableyoutoidentifypotentialriskssoonerbutcomewitha
costofincreasedcustomerinteractionandmoreinterruptionsfortheteam.Longersprints
meanthatitwilltakelongertofindtherisksbuthavethebenefitoffewerinteraction
pointsandinterruptions.Balancingfeedbackwithinteractionandinterruptionsisa
challengethatyourteamandstakeholderswillperfectovertime.
ThequestionsandcorrespondingscoringkeyIprovidedinthischaptermaynotwork
foreveryone.Whenyou’readjustingsprintlength,consideritastartingpoint.Don’tstress
aboutfollowingthemodel’sexactquestionsandanswerweightings.Focusinsteadon
whetheryouareconsideringbothyourcustomerandyourteamwhendeterminingyour
sprintlength.Finally,don’tforgetthereasonsyouareworkinginshortiterations:togather
stakeholderfeedback,toshowprojectprogressthroughworkingsoftware,andto
frequentlydeliveraprojectthatispotentiallyshippable.Therightsprintlengthbalances
theteam’scravingforfeedbackandinputwiththeteam’sabilitytodeliverandthe
stakeholder’sabilitytorespond.
SprintsLongerthanOneMonth
InAgileProjectManagementwithScrum,KenSchwaberwrites,“TheSprintistime-
boxedto30consecutivecalendardays….Thisisalsothemaximumtimethatcanbe
allocatedwithouttheTeamdoingsomuchworkthatitrequiresartifactsand
documentationtosupportitsthoughtprocesses.Itisalsothemaximumtimethatmost
stakeholderswillwaitwithoutlosinginterestintheTeam’sprogressandwithoutlosing
theirbeliefthattheTeamisdoingsomethingmeaningfulforthem”[SCHWABER].
Thisadviceisimportantbecausethewholepurposebehindtheone-month-or-lesssprint
lengthisthefeedbackloop—validatingthatthepaththeteamisonisthecorrectone.
Anotherwaytoputitisthatinthebeginningofthesprint,theteamheardwhatthe
productowner(andthestakeholders)askedfor.Attheendofthesprint,theteampresents
whatitthinksitheard—essentiallytryingtopreventwhatIcallthe“whatImeant
syndrome,”whencustomersandstakeholderssayafterasprintorrelease,“Youbuiltwhat
IaskedforbutnotwhatImeant.”Thissituationplaguessomanyteams,whywouldthey
notwantshorterfeedbackloops?
ExtendingSprintLength
“Ifwecan’tfinishourwork,shouldweextendthelengthofoursprinttofinishthework
wehaveinthebacklog?”Igetthisquestionallthetime.Extendingthesprintlengthis
dangerous.RonJeffriestoldmeonce,“Ifyoucan’tgetyourworkdoneinthreeorfour
weeks,whatmakesyouthinkyoucangetitdoneinfive,six,seven,oreight?”His
suggestiontomeatthetime?Switchtoone-weeksprints.Why?Simplybecauseyoucan
comeinonaMondayandseewhatyouneedtodobytheendoftheweek.It’sclearand
simple.Further,itforcesstoriesandtaskstobesmallerandprovidestheteammore
opportunitiestoreflectandcorrectissues.Inotherwords,itgreatlyimprovesthestimulus-
to-responsetime.
Reference
[SCHWABER]Schwaber,Ken.2004.AgileProjectManagementwithScrum.Redmond,
WA:MicrosoftPress.
Chapter7.HowDoYouKnowYou’reDone?
“Areyoudoneyet?”Thisseeminglyharmlessquestionisaskedcountlesstimesonalmost
everysoftwareproject.Thewayyouanswer,however,isanythingbutinnocuous.Ifyou
respondwitha“yes,”youmightbeforcedtotakeonadditionalwork.Ifyousay,“no,”
youcouldbebrandedassomeonewhocan’tgetthingsdone.Ifteammembers’answers
vary,youcouldloseyourstakeholders’trust.
Establishinganup-front,commonunderstandingofdonecansaveteamsandbusinesses
countlesshoursofprocessthrash,ambiguity,andhiddenwork.Inthischapter,youlearn
whatadefinitionofdoneis,howitaddsvalue,andthevalueitcommunicatesto
stakeholders.Youalsolearnanexercisethathelpsyoubuildyourowndefinitionofdone
andmanageitovertime.
Considerthestoryofoneproductowner,MaryAnne,andwhathappenedwhenshe
hostedastoryworkshopwithhercustomersandstakeholders.
TheStory
MaryAnnewashostingastoryworkshoptobuildtheproductbacklogforalargeback-
officeinfrastructureproject.Shehadgatheredroughly20stakeholdersintheroom,
includingsystemsengineers,generalmanagers,developers,marketingpeople,product
managers,andsalespeople.Eachhadvaryingdegreesofinterestinthisproject,andeach
thoughtthathisthingwasthemostimportantandshouldbebuiltfirst.MaryAnneknew
fromthebeginningthatsuchadiversegroupwouldbeachallengeandthatshewould
needtofocusthemeetingtofilteroutthenoiseandgettherealstoriesneededtostartthe
work.
MaryAnnehadherteam’sdefinitionofdoneinherbackpocket.Evenso,shedecided
nottodistributeitatthebeginningofthemeetingfortworeasons:Shewantedtoensure
peoplefeltopenandunconstrainedbyapredeterminedlist,andshedidnotwantthemto
gettheimpressionthatthedefinitionofdonewasupfornegotiation.
Twohoursintothemeeting,thegrouphadauthoredapproximately200stories.This
achievementwasahugesuccess.However,peoplewerestartingtogethunguponthe
sequenceinwhichthestorieswouldbeaccomplished—theirordering.Naturally,the
systemsengineerwantedtoseemonitoring,security,andallhardwareandnetwork
infrastructurebuiltfirst.Themarketingpeoplejustwantedareleasedate.Thedevelopers
andtesterswantedtodrillintotheweedstofigureoutwhateachstorymeant.
ItwasatthispointthatMaryAnnerevealedtheinitialdefinitionofdonethattheteam
hadidentified.Thelistcoveredwhatitmeanttobedonewithastory,asprint,areleaseto
theintegrationenvironments,and,finally,toreleasetoproductionenvironment.
Thestakeholderswereshocked.Neverbeforehadtheyseensuchdetailand
commitmenttoaqualityrelease.Billwasthefirsttospeakup.
“MaryAnne,thislooksprettycomprehensive.Areyousayingthatwitheachsprint,you
willdoperformancetesting?”Billasked.
“Yes,”saidMaryAnne,“andmore.Weplantodoeverythingneededtoshipoursystem
asit’sbeingbuiltinsteadofwaitinguntiltheend.”
Billcontinued,“Whatabouttheenvironmentbuildouts?Whyaren’ttheyonhere?”
“Simple,”statedMaryAnne.“Ourdefinitionofdonecoversthethingsthatwillbe
presentthroughoutthelifeoftheproject.Forexample,thetroubleshootingguidesthatwe
willwriteforyourteamwillbeexpandedwitheachstory,meaningwewon’thavea
documentationphaseattheendoftheprojecttodoallthedocumentation.Wewilldo
everythingneededtodeliveroursystemeverysprint.”
“IthinkIgetit,”saidBill.“Yourdefinitionofdoneisallaboutthingsthatareconstant
throughouttheproject—notone-timethings.Thosewouldbemoreonthetasklevel,or
maybestorylevel?”
“That’sright,”saidMaryAnne.“Ourdefinitionofdoneisaboutkeepingour
professionalismhigh.LikeaChristmaspresentisdonewhenit’sunderthetreeand
wrappedwithabow,ourdefinitionofdoneiswhenatask,astory,asprint,orarelease
meetsthiscriteria.We’lltestforiteachsprintanddemoitinthereviewmeetingaswell.”
Afterseeingtheteam’sdefinitionofdone,thesystemsengineerswerecomfortablewith
thelevelofworkMaryAnne’steamwasdoingtoaddresstheirneedsoneachstory,sprint,
andrelease.Thesecurityteamwasequallypleasedtoseethelevelofarchitecture,
documentation,andthreatmodelingthattheteamwasdoing.Equally(ifnotmore)
satisfiedwerethesales,marketing,andgeneralmanagementstakeholders.Knowingthat
theteamwouldcompletetheworktosuchadetailedleveleachandeverysprinthelped
assureeveryonethathisownpieceofthepiewouldbedonecorrectly.
TheModel
Havingadefinitionofdoneforyourteamhelpsallthestakeholdersvisualizetheteam’s
commitmenttodothebestpossibleworktomeettheneedsofthebusinessandits
customers.Toteamsthatuseit,thedefinitionofdonebecomesawayoflife—nota
checklisttofollowbutacommitmenttoexcellence.
Thedefinitionofdonealsogivesteamsawaytocommunicate.Wheneveryonecansee
whatitmeansforthisteamtobedonewithitswork,theteamnolongerhears,“Areyou
donewiththisorareyoudonewiththat?”Insteadtheyareasked,“Isthisstorycomplete?
Isthesprintcomplete?Haveyoureleasedtoyourenvironments?”Andeveryone
understandswhatanaffirmativeanswerreallymeans.
Eachteamshouldcreateadefinitionofdonethatisuniquetoitsownparticular
company,product,orsituation.Thefollowingexercisehelpsyourteamexplorewhatit
meanswhenastoryisdone.Figure7-1liststheitemsyouneedforthisexercise.
FIGURE7-1ItemsneededfortheDefinitionofDoneExercise
TheDefinitionofDoneExerciseisconstructedfromthesecomponents:
1.Brainstorming
2.Categorization
3.Sortingandconsolidation
4.Creationandpublishing
Thisexercisehelpsyoucreateyourinitialdefinitionofdone.NoticeIsaid“initial.”
Definitionsarenotcreatedonce,buriedinadrawer,andneverseenagain.Teamsshould
periodicallyscheduletimeduringretrospectivestoreviewtheirdefinitionsofdoneand
determinewhetherthereareopportunitiesforimprovementormodification.Feelfreeto
reviseyourdefinitiontomeettheneedsoftheteamanditsstakeholders,butbecareful.
Toomanyrevisionstoofrequentlymaycreatedoubtaboutthevalidityofthedefinitionof
doneintheeyesofthestakeholders.Investthetimetocreateasolidbaselineandmodify
thelistbasedontheexperienceandfindingsoftheteam.
Introduction
Peopleoftenaskmewhoshouldparticipateinthisexercise.Thisquestionalwayscatches
meoffguard.Theansweristhateveryoneontheteam,nomatterhisorherexpertiseor
background,canaddvalueandshouldbeinvolved.
Iftheteam’sprojectisabackenddatabasesystemorathree-tierwebapplication,some
peopleontheteamwillhavelessexperienceandknowledgeofthetechnologythanothers.
Thatisjustafact.Keepingteammembersoutofthisexercise,though,onlydeprivesthe
teamofthevaluablecontributionsofotherteammembers.Italsokeepsthemfromhaving
avaluableteambuildingexercise.
Therefore,Irecommendhavingtheentireteamdothisexercise,regardlessofskills,
background,orroleontheteam.Considerincludingyourproductowner,whohasinsight
andinformationthattheteammayfindvaluable.Oftentimestheproductownerwillshare
companywideorprojectmandatesthattheteammaybeunawareof.
Brainstorming
AlexF.Osborniswidelyknownasthefatherofbrainstorming.Inhisbooks,Your
CreativePower[OSBORN01]andAppliedImagination[OSBORN02],Osbornoutlines
thebrainstormingtechnique,asystemthatusesthebrainto“storm”creativesolutionstoa
problem.Thiscreativeapproachisdesignedtogenerateafreeflowofideas.Inthe
brainstormingsession,therearenorightorwronganswers.Thereisnocriticism,thereare
onlyideas.Therulesofbrainstorming,outlinedbyOsborn,aresimple:
Nocriticismofideas.
Goforlargequantitiesofideas.
Buildoneachother’sideas.
Encouragewildandexaggeratedideas.
Atthebeginningofthebrainstormingsession,setthepropertone.Thereasontheteam
istogetherintheroomistoidentifyeverythingthatitneedstodotoshipsoftware,the
truestmeasureofprojectprogressIhaveencounteredtodate.
Itisimportanttosetthedirectionofthebrainstormingsessioninthebeginning.Startby
writingonawhiteboardthequestiontheteam’sdefinitionofdonewillanswer.What
questionisthat?Itdepends.Themostcommonquestionis,“Whatdoweneedtodo,asa
team,toshipsoftwaretoourcustomers/stakeholders?”
Yourquestionmayvary,dependingonyourownteam’suniquecircumstances;
however,thepurposeisthesame.Makesurethequestionisonawhiteboard,oranyother
viewableplaceintheroom,beforeyoucontinue.
Handeveryoneapenandastickypad.Ihavefoundthatindividualslikehaving
differentcolors,butthisultimatelydependsonyourteammembers’preferences.
Kickoffbrainstormingbyrepeatingthequestionyouareanswering(seetheprevious
oneforanexample).Haveeachteammemberwriteanansweronasticky,calloutthe
answer,putthatstickyinapileinthemiddleofthetable,andrepeat.Teammemberscan
writeonlyoneanswerperstickyandmustcalloutaftereachansweriswritten.Asteam
membersstarttosharetheiranswers,inevitablysomewillsay,“I’vealreadywrittenthat.”
Donotworryaboutcreatingduplicateitemsatthemoment.Thiswillbeaddressedlater.
Whythecallout?Ihavefoundthatthecalloutputsthestorminbrainstorm.Ihave
personallyidentifiedmultipleadditionalitemswhenfellowteammembershavecalledout
adefinitionofdoneitem.Ihavealsowitnessedteammembersbuildingoneachother’s
ideaswhencoachingteamsthroughthisexercise.Ihaverunthisexercisewithouttheteam
memberscallingoutwhattheywroteonthestickies,butitresultedinastale,
unproductivemeeting.Callingoutmightfeelweirdanduncomfortableatfirst,butthe
feelingsubsidesveryquicklyoncetheteambecomesengagedandthemeetingbeginsto
flow.
Runthebrainstormingpartofthesessionuntilpeoplerunoutofideas.Itisgenerally
easytoidentifywhenpeoplearedonebecausetheflowofstickynotesbeginstodwindle.
Peoplebeginlookingaroundtheroomandateachother,searchingforsomethingnew.
Categorization
Thereshouldnowbealargepileofstickynotesonthetablewaitingtobesorted.Startby
askingtheteammembershowtheyeachthinkthestickynotesshouldbecategorized.The
mostcommonresponsesIseelooklikethis:
Development
Test
Projectmanagement
Other
Allofthesetypicalresponsesdecomposetheworkbyfunctionalarea,whichmakes
sense.Afterall,peoplehavebeenconditionedtoworkinfunctionalsilos,notincross-
functionalmultidisciplineteams.Unfortunately,categorizingthiswayreinforcesthe
functionalroleswithinateamanddoesnotpromotecross-functionalinteraction.
Additionally,customersandstakeholdersgenerallydonotcareaboutwhatitmeanstobe
donewithdevelopmentortest.Theycareaboutwhatitmeanstobedonewithareleaseto
productionorareleasetomanufacturing.
Ifyougetthoseresponses,challengetheteamtochangeitsfocustocustomervalue.
Remindtheteamthatdeliveringstoriesattheendofasprintandreleasingtoan
integrationenvironmentwherecustomerscanuseandinteractwiththesoftwareprovides
value.Mentionthatthecustomerreceivesevenmorevaluewhenasystemisreleasedtoa
productionenvironmentandofficiallydone.
Afterintroducingthislineofthinking,theresponsesItypicallyseechange
dramatically:
Donewithastory
Donewithasprint
Releasedtointegration
Releasedtoproduction
Oncethecategoriesareidentified,clear,andunderstoodbytheteam,createareasonthe
wallsintheroomtopoststickynotesundereachcategory.
Thisisateamexercise.Eachpersongrabssomestickynotes(preferablyothers’notes)
fromthetableandbeginsplacingeachnoteundertheappropriatecategory.Team
membersshouldnotspendagreatdealoftimewrestlingwithwhethercategoriesareright
orwrong—atthispoint,theyshouldjustgetthenotesonthewall.Sortingthenotes
shouldtakebetween10and20minutes,dependingonthequantityofstickynotes
generated.
Onceallstickynotesareonthewall,haveeveryonetakeafewstepsbackandlookat
thewall.
Thatisalotofstuff.
SortingandConsolidation
Onceeverythingisonthewallandcategorized,it’stimetobeginconsolidation,usingan
approachsimilartothatadvocatedbyMikeCohn[COHN].
Thisisalsoateamexercise.Instructtheteamasfollows.Eachpersonreviewsthe
stickynotesonthewallandlooksforduplicateitems.Duplicatesincludeexactmatches
andsimilarmatches.Forexactmatches,simplytakethetwomatchingstickynotesand
stickthemtoeachothersothatthetoponefullycoverstheoneunderneath,asillustrated
inFigure7-2.Dothisonlywhenitisclearthattheitemsareexactmatchesandthatthe
decisiontomarrythemwillrequirelittlediscussionlaterinthesession.
FIGURE7-2Exactmatches
Forsimilarmatches,havethetopstickynoteoffsettheoneunderneath,asillustratedin
Figure7-3.Thishelpseveryoneeasilyidentifyitemsthataresimilarbutnotexact
matches.Theseitemsrequirefurtherteamdiscussiontodetermineaclearmeaningforthe
definitionofdonecandidateitem.
FIGURE7-3Similarmatches
Forstickiesthatseemtomakenosense,asillustratedinFigure7-4,createanew
categoryofftothesideandnameit“other”orsomethingsimilar.Theteamwillrevisit
theseitemslast.
FIGURE7-4Initiallymakesnosense
Again,havetheteamstepbackandlookatthewall.Theymaystillbeoverwhelmedbut
theyshouldstarttofeelthatthingsareabitmoremanageable.
Onceallnotesareconsolidated,theteamisreadytodiscusswhatthecommentswritten
oneachstickynotemean.Asfacilitator,startwithasetofstickies(Iliketostartwiththe
exactmatches,astheyaretheeasiest),takethemoffthewall,readthemtotheteamand
ask,“Whatdoesthismean?”Oncetheteamanswers,moveon.Ifcertainteammembers
consistentlydonotanswer,askthemdirectlyforconfirmationthattheyunderstandwhata
specificstickysetmeansbeforemovingon.Itisimportanttotimeboxthisactivity.Ifthe
teamcannotcometoconsensus,orevenagreement,shelvetheitemforlater.Ifitis
happeningrepeatedly,invitetheteamtotakeamentalbreak,perhapsbygoingforawalk.
Aftertheexactmatchesarefinished,moveontothesimilarmatchsets.Theseare
managedslightlydifferently.Takethesetoffthewall,readthemtotheteam,andask,
“Whatdoesthismean?”Whentheteamanswers,pickthecardfromthesetthatmost
closelymatchestheteam’sanswerorcreateanewstickytoreflectwhatwasdiscussed.
Afterasetofsimilaritemshasbeendiscussed,theteamshouldbeleftwithonlyonecard
thatrepresentswhatwillbecommittedtoandcommunicatedbytheteam.Throwout
duplicatesorconsolidatethemliketheir“exactmatch”counterparts.Throughoutthe
process,newstickiesmaybecreatedandoldonesdiscarded.Thisiskey.
Lastly,moveontothestickiesthatareanomaliesormakenosense.Usethesame
processofpullingastickyoffthewall,readingit,andaskingwhatitmeans.Followthis
byasking,“Doestheteamneedtoincludethisiteminitsdefinitionofdone?”Iftheteam
respondswitharesounding“yes,”movethestickytoitspropercategory.Iftheteam
respondswithaplainno,tossit.Iftheteamisunsureorhesitant,facilitateadiscussion
thatresultsineitheraresoundingyesortheitembeingtossed.
Ifyourproductownerisintheroomatthistime,shemayhavequestionsastowhy
itemsarebeingtossed.Helptheproductownerunderstandthatwhilethingscanalwaysbe
addedtothedefinitionofdoneatalaterdatewithoutbreakingordamagingtrust,
removingitemsafteritispublishedisdetrimentaltobuildingandmaintainingtrust.
Oncetheteamhasreviewedallstickynotes,understandswhateachonemeans,and
agreesthatallstickiesareintherightcategory,thisportionoftheexerciseiscomplete.
CreationandPublishing
Teamsaregenerallytiredatthispoint.Theentireexerciseisadrainingexperiencebut
veryrewarding.Creatingandpublishingthedefinitionofdoneistheeasiestpart.
Takeadigitalpictureofwhatisonthewallandputitinelectronicformat.Publishthe
definitionofdoneontheinternalteamwebsiteandcreateaposterthatcanbehungforall
tosee.Itservesasareminderofwhatitmeansfortheteamtobedoneandactsasa
powerfulcommunicationtool.Whenstakeholdersask,“Areyoudoneyet?”theteamcan
pointtothedefinitionofdoneandanswer,“Wearedonewithourstoriesandworkingon
ourrelease”—andthepersonaskingwillunderstandexactlywhattheteammeans.
Whatabout“Undone”Work?
Supposetheteamgetstotheendofasprintandsomeitemsdon’tmeetthedefinitionof
donecriteria.Whathappens?Shouldtheygetpartialcredit(whateverthatmeans)for
workcompleted,orshouldtheworknotcounttowardtheteam’svelocity?
Formeit’ssimple.Isuggestyoufollowthismodelaswell:Ifastorymeetsthe
definitionofdone,it’sdone.Ifitdoesnot,thenitisnotdone.It’sprettyblackandwhite.
Discussionsaboutcomplexity,almostornearlydone,oranyotherexcusetosayacertain
storyshouldbeincludedinthevelocityshouldbediscounted.Undoneworkgoesbackto
theproductbacklogforprioritizationbytheproductownerandmaybere-addedtothe
sprintbackloginthefollowingsprint.
KeystoSuccess
Creatingandpublishingadefinitionofdoneaccomplishesthreethings.
First,ithelpsbuildabondbetweenteammembers.Itgivesthemthefeelingthatthey
areallinthistogetherandthatitismoreimportanttodeliverontheirjointcommitments
thantofocusonspecificindividualtasks.
Second,itprovidesclearcommunicationtostakeholders,andbyimplication,drives
downtheriskoftechnicalorotherdebtbeingdeferredtolaterinthecycle.Ifa
stakeholderaskstheteamifthesoftwareworks,andtheresponsebytheteamis,“Sureit
works,butwedon’thaveaninstaller,”theteamhasrackeduptechnicaldebtbecausethe
componentisnottrulydone.Throughthedefinitionofdone,andthroughits
communication,stakeholdersgainaclearunderstandingthatiftheteamsaysitwillbe
releasingtoproduction,theteamwillhavemetaseriesofrequirementsthatareneededfor
sucharelease.Nomorewilldonesimplymeanthatsomethingiscoded.Onthecontrary,
donewillmeanexactlywhatispublishedinthedefinitionofdone.Stakeholderswillcome
toexpectthislevelofqualityandcommitment.
Third,itkeepstheteamontrackandfocused.Whenplanninganiteration,theteamhas
areferencepointastowhatitmeanstobedone.Theguessworkisremoved,enablingthe
teamtofocusondeliveryinsteadofspeculatingaboutwhatcouldbe“ifonlythisorthat
happened.”Thecommitmentismade,published,andadheredto.Thedefinitionofdone
shouldberevisedasneededandrepublished.Whenusedcorrectly,itprovidestheteam
withacommonvisiononhowtoachievetheiteration,release,andprojectgoals.
References
[COHN]Cohn,Mike.2004.UserStoriesApplied.UpperSaddleRiver,NJ:Addison-
Wesley.
[OSBORN01]Osborn,AlexF.1948.YourCreativePower.NewYork:C.Scribner’s
Sons.
[OSBORN02]Osborn,AlexF.1957.AppliedImagination.NewYork:Scribner.
Chapter8.TheCaseforaFull-TimeScrumMaster
OrganizationsoftenhavedifficultytransitioningtoanagileprocesssuchasScrumor
ExtremeProgramming(XP).Evenaftertheygetpastthe“it’safad”or“wecan’twork
likethat”excuses,someintheorganizationstillhavetroublelettinggoofoldideasand
habits.Oneofthemostcommonargumentsteamsfaceisthattheorganizationcan’tjustify
oraffordafull-timeScrumMaster.Organizationsareusedtohavingpeoplewearmultiple
hats,andhavingateamleaderwhodoesn’talsocontributetotheworkoftheprojectis
hardformanytounderstand.
Chapter5,“ImplementingtheScrumRoles,”talkedabouthowScrumMastersdotheir
jobsmosteffectivelywhentheyarefullydedicatedtothatroleanddonotalsoserveas
teammembersorproductowners.Butdidyouknowhavingafull-timeScrumMastercan
actuallysavetheorganizationmoneyinthelongrun?Formore,followSteinashe
explainsthisnotiontohisboss,David.
TheStory
“Ijustdon’tgetit,”saidDavid.“YouareoneofthebestdevelopersIhave.WhyshouldI
sacrificeallyourcodingabilityandmakeyouadedicatedScrumMaster?Can’tyoudothe
coachingbitontheside?FromwhatIcansee,it’sjustaglorifiedreportingjob.”
Steinsighedquietly.Davidwashisimmediatemanager,andwhiletheyhadbeendoing
Scruminthecompanyforsometime,Davidhadalwaysfoughtit.
“I’vetakenthetraining,talkedtoteamswhohavedoneseveraldifferentagile
processes,andreadseveralbooksandarticlesonthis,David,”explainedStein.“Theyall
saythesamething—thatafull-timeScrumMasterisessentialtomakingthemostof
Scrum.”
Davidsneered.“Yoursisn’ttheonlyteamaroundheredoingScrum,youknow.Take
theCollettiproject,forinstance.I’veseenwhatthatScrumMaster,Teresa,does—shejust
schedulesthemeetingsandcollectsdata.Bigdeal.Ineedyouinthetrenches,Stein.You
candoyourreportinginfourhoursaweek.”
Steinwasreadyforthis.“IhearyouandIagree.Icoulddoreportinginfourhoursa
week.Let’stalkabouttheCollettiprojectforasecond,though.Whatkindsof
improvementshaveyouseenfromthatteam?”
“Well,atfirsttheteamgotbetter—about20to30percentfasterthanitwasbefore—but
thenitstalledout.Ihaven’tseenanyofthosehugegainstheconsultantspromisedme.”
“Iagreewithyou,David.TheCollettiprojecthasn’tmademuchprogressafterthe
initialimprovements.ButoneofthereasonsisthatTeresahasn’tbeenallowedtoletgoof
herteammemberrole.Shedoesn’thavetimetodoanyofthethingsScrumMastersare
supposedtodo.”
Steinwenttothewhiteboardandbeganwritingsomeofthekeyfunctionsofa
ScrumMaster.
“ReportingisatoolgoodScrumMastersuse,butit’sonlyacontributingactivitytothe
realthingsScrumMastersaresupposedtodoeveryday,”explainedStein,ashecontinued
towrite.“Thesearejustafewoftheothers,”heexplained.
“Iagreethosethingsareimportant,butI’mnotconvincedthey’reworthremovingyou
fromyourhands-onwork.I’vegottobeabletojustifythecost,”saidDavid.
“I’mworkinguptothat.IndulgemeforaminutewhileIwalkthroughthiswithyou.
Yourememberwhatvelocityisright?”askedStein.
“Yes,theamountofwork,orstorypoints,orsomething,thatateamgetsdoneina
sprint,”statedDavid.
“Closeenough.SomyjobastheScrumMasteristohelptheteamgetitsvelocityup,
whichessentiallydrivesthecostperpointdown,”saidStein.
“Costperpoint?”askedDavid.
“Yes,yousee,thevelocityiscalculatedonthenumberofstorypointsateamdeliversin
eachsprint.Everyteamhasacost,theloadedteamcost.Let’ssaythatmyteamcoststen
grandpersprint,justtokeepourmathsimple.Everysprint,myteamdeliverstenstory
points.Whatisthecostperpoint?”askedStein.
“Athousandbucksperpoint.Sowhat?”askedDavid.
“Well,ifyougivemetimetodoallthethingsonthiswhiteboard,”saidStein,pointing
backtohislist,“Icangetthatcostperpointdowntosomethingmuchclosertofive
hundredtosixhundredfiftydollars.”
“Youmustbejoking.Cutitinhalf?Whatmagicareyougoingtoperformtodothat?”
askedDavid.
“That’swhatmakesthisafull-timejob.Ilookupfromthenoiseofeverydayworkto
seetrends,patterns,problemsthatareblockingtheteamfromgoingfaster.Ifyouhavea
kinkinahoseandthewaterison—howmuchwatercomesout?Notmuch.Ifyouunkink
thehose,thewaterflowsagain.Oneofmyjobsistolookforkinksandthenworkwith
youorotherteamsorHRorseniormanagementtogetthosekinksoutasquicklyas
possible.”
“I’mwithyousofar,”saidDavid.“Keepgoing.”
“Somyotherbigresponsibilityistobuildacohesiveteam.Likeacoachonafootball
team,right?He’snotouttherepassingtheballorrunningplays,ishe?Butdoesthatmean
hehasnothingtodobutrecordthestats?No.He’sleadingtheteam,watchingfilm,
workingoutstrategies,helpingthemembersoftheteamseewheretheyneedtoimprove.
Heliftsteammembersupwhenthey’redownandkeepsthemgroundedwhentheystartto
strayfromtheirgoals.Hefocusestheteamonfundamentalsanddisciplineandjustplain
functioningasaunit.
“TheScrumMasterjobislikethat.I’mstillhands-on.Myhandsarejustbusydoing
otherthings.Ileadandguidetheteamthroughtheprocessallthewhilekeepingahigher-
levelviewofwhatneedstobedone.Iremoveimpediments,keepeveryonefocused,find
waystoimprove,allinthenameofincreasingvelocity.Now,unlikeafootballcoach,I
canhelpoutandwritecodefromtimetotime.ButIwouldonlydothataspartofhelping
theteamwhenitisstuck.It’snotreallypartofmyjob.Makingsense?”askedStein.
“Goodpoints,”saidDavid.“Butyouhaven’texplainedhowthiscostsmeless,yet.”
“Well,ifIdomyjobwell,theteamwillgofasterandgetmoredone,right?Therewill
befewerimpediments,moreteamwork,improvedperformance.Withmeasapart-time
ScrumMaster,we’dperformaboutlikeTeresa’steam.Somyteamwouldgofroma
velocityoftenstorypointstotwelveorthirteen,andyourcostwouldgofromathousand
perpointtosomethingmorealongthelinesofsevenhundredfiftytoeighthundreddollars
perpoint.Withmeonfull-time,we’llstartoutwiththoseimprovementsbutthenquickly
movetoseventeentotwentystorypointspersprint.Sopointsthatusedtocostagrandare
nowcostingyou—”
“Aroundfivehundredtosixhundreddollarspersprint.Isthatrealistic,though?Can
havingadedicatedScrumMasterreallydoubleteamperformance?”askedDavid.
“TheresearchI’vedoneandthepeopleI’vetalkedtosuggestthatit’spossible,butI’m
hesitanttomakethosekindsofpromiseswithouthavingdoneitbefore,”saidStein.
“However,theresearchdoesleadmetobelievethatwecan’treachourtruepotential
unlessI’mworkingonthisfull-time.You’repayingmetoworkontheteamanyway.So,if
havingmeworkasaScrumMasterinsteadof‘inthetrenches’meanswedoevenalittle
betterthanTeresa’steam,wouldn’tyougrantmethatthecostsavingsmakehavingafull-
timeScrumMasterworthwhile?”
“Yes.Ifyoucanshowmethatthisworksandgetsomegoodgainsandimprovements,
andhelpdrivethechangebackintotheorganization,I’llgrantyouanythingyouwant,”
Davidsaid,smiling.“Showmewhatyoucando.”
TheModel
OneofthejobsoftheScrumMasteristoremoveimpediments.Anotheriseducatingthe
Scrumteam(includingtheproductowner),stakeholders,andothersintheorganizationso
thattheteamiseventuallyabletoresolveitsownimpediments.Thisultimatelyresultsin
moreefficientteamsthatcangetmoredone.GoodScrumMastersfunctionastheoilinthe
engine—theyreduceorremovefriction,enablingteamstomovetoahigh-performing
state.
IamoftenaskedhowImeasuretheimpactaScrumMasterhasonateamor
organization,andhowIconvincemanagementto“allow”apersontotakeonthe
ScrumMasterrolefull-time.Myanswer:Startwithratesofimprovementandrelatethatto
dollarsandcents.
Thougheachteamwillhavedifferentvelocities,runadifferentnumberofiterations,
andhavedifferentoverallresults,therateofimprovementonanagileteamfollowsthe
samegeneralpattern.Iwanttowalkyouthroughanexamplewithfictitiousnumbersso
thatyoucanseehowthatpatternforms.Idon’twantyoutogetcaughtupintheexact
numbers,though.It’sthepatternthatmatters—anditisevidentinthetrendlineineach
chartthatfollows.OnceI’veestablishedthepattern,Ishowyousomerealchartsfromreal
teamssothatyoucanseethepatternrepeatitselfagainandagain.
Supposeafictitiousteamisrunningtwo-weeksprintsandisabouttwomonthsintothe
project.Afterfoursprints,theteamisaveragingaround10pointspersprint(seeFigure8-
1).Duringsprint3,theteamcomestotherealizationthatitwantstoaddafull-time
ScrumMaster.
FIGURE8-1Outputpatternforthefirstfoursprints
Afteraddingafull-timeScrumMaster,theteambeginstoseesomeinitial
improvements.Theteamlearnstoworktogetherinanagileway,andtheScrumMaster
startstoremoveimpediments.Theaverageinsprints5through8goesuptoabout12
pointspersprint(seeFigure8-2).
FIGURE8-2Outputpatternforsprints1through8
Torecap,sprintblock1(sprints1–4)hadanaverageof10points.Theteamdidnot
haveafull-timeScrumMasteruntilsprint4.Sprintblock2(sprints5–8)hadanaverageof
about12,a20percentimprovementwithafull-timeScrumMaster.
Whathappensifyourollbacktheclockandworkonthepremisethattheteamnever
addsafull-timeScrumMaster(assumetheteamhasone,butthatpersonissplittingtime
betweenteamtasksandScrumMasterresponsibilities)?Afterthefirstsixsprints,theteam
wouldbegintoplateau,asillustratedinFigure8-3,whichshowsaflatteningtrend.The
teamhaslikelyachieveditsmaximumoutput.
FIGURE8-3OutputwithoutaScrumMaster
Noticethattheteamsputteredout,forwhateverreason.Itcouldbethatthe
ScrumMasterwasdoingtoomuchsprinttask-relatedwork,itcouldbethatthe
impedimentsfacingtheteamcouldnotberesolved,itcouldbethattherewasnocohesive
coreteambutinsteadonlyavarietyofpeopleworkingonvariouspartsofthecode.
Regardless,thisiswheremostteamsendup.TheystartScrum,getsomesmallgains(10
to25percent),andthensputterout.Theybecomefrustratedbecausetheyknowtheycan
domore.Someevenenduprevertingtotraditionaldevelopmentapproaches.
Returntotheoriginalpremise,wheretheteamaddedtheScrumMasteraftersprint3.
Theteamiswellcoachedandisinareasonablysupportiveorganization.Thisteamalso
stumblesabitbutthenrightstheship—afterthefirstsixsprints,theteamtakesonestep
backandthenbeginstoshowsteadyimprovementinsprints11through14,asidentified
bythetrendlineshowninFigure8-4.
FIGURE8-4OutputwithaScrumMaster
Nowwhatyouseeisthattheaveragefromsprint8tosprint14ismorelike15to17
pointspersprint.Sofar,thisteamhasmovedfroma10-pointaverage,througha12-point
average,toa16-pointaverage,a60percentimprovementsofar.
Thisteamisn’tdoneyet,though.Theteamcontinuestoimprove,clearingblocking
issuesandobtainingtraininginmanyengineeringpractices.Theteambeginstoperformat
highlevels,findingtheirsweetspotandhoveringatasustainablepace(seeChapter23,
“SustainablePace,”formoreonthistopic).
ThefinalchartinFigure8-5showsanaverageof24pointspersprintfromsprints15to
23.Theteamthentopsoutatanaveragevelocityof28forthenextfivesprints,sprints24
through28.Thetrendlineflattenseventually,butatamuchhigherlevelthantheteam
withoutafull-timeScrumMaster.
FIGURE8-5Outputpeak
Table8-1summarizestheimprovements.
TABLE8-1SummaryoftheImprovements
Improvedvelocityisgreat.Buttohelpbringthepointhometoyourmanager,youneed
toputitintermsofcostperstorypoint.
Teamsgenerallyhaveaflatcostthatcanbeeasilyidentifiedbytheemployeeloaded
cost.Consideramodelwherethenumbersareroundedbutarebasedonanactual
company(seeTable8-2).
TABLE8-2EmployeeLoadedCosts
Whatyouseehereisthatanemployee’scostaddsuptoaboutdoublehisorhersalary.
Soifyouhaveanemployeeloadedcostof$200,000(rounding),andthatemployeehas
about47workingweeksavailable(52weeksavailableminus3forPTO,minus2for
governmentholidays),thenthatemployeehas1,880hoursavailabletowork.At$200,000,
thatcomesouttojustunder$107/hrincosttothecompany.Nowcompaniesdoneedto
makemoney,especiallyconsultingcompanies,sotheyoftenmarkupwhattheycharge
theirclientsfortheiremployees’time,sayto$150or$225anhourdependingontheskill
setoftheemployee.Butjustusetheactualcostforthisanalysis.
Now,youhavethecostperpointandknowwhatitcostsperhourforoneemployee.Ifa
teamcanbustout10pointspersprintandtheloadedteamcostisabout$440anhour
(teamoffourpeopleatarounded$110anhourloadedcost),theteamcosts,persprint,
$35,200($440times80hoursforatwo-weeksprint).Thisassumesadedicatedteam
workingononeprojectandthatthetimespentonnon-project-relatedtasksisstillaloaded
costontheproject.
Ateamcosts$35,200persprint.Table8-3showshowallthismapstoTable8-1.What
youseehereisthatthisfictional,well-coachedteam,withafull-timeScrumMaster,over
thecourseofeightsprints(4months),reduceditscostperpointby17percent.
TABLE8-3SummaryoftheImprovements
Overthecourseof14sprints(7months),thesameteamwasabletoreducethecostper
pointby38percentfromthestartandincreasevelocityby60percent.Overthecourseof
28sprints(14months),theteam’scostperpointwasreducedby64percent,aresultofthe
team’s180percentimprovementfromthebeginning.
Now,Ijustmadethesenumbersuptoshowyoutheoverallpattern.Canrealteams
expectthissamekindofimprovement?Yes.Realteams,withfull-timeScrumMasters
whoareengagedincoachingtheteam,management,andthestakeholders,andwhoare
activelyremovingimpediments,cananddoexhibitastrikinglysimilarpattern.I’llnow
showyousomerealteamsandrealdata.Itypicallycapturetensprints’worthofdata
becauseafterabouttensprints,thereisenoughtoseeiftheteamisgoingtosucceedordo
poorly.
ThefirstexampleisshowninFigure8-6.Noticethatthoughthenumbersaredifferent,
thepatternisthesameasinthechartspresentedearlierinthechapter.Figure8-7presents
anotherexample.Noticethatthoughthischartreflectstheperformanceofadifferent
team,thepatternisalmostidenticaltothepreviouschart.Notethatinbothexamples,the
teamfailedasprint.Figure8-8showsachartfortheteamthatfailedasprintbydelivering
zeropoints,andstruggledallthewaythroughsixsprintsbeforeimprovingdramatically.
FIGURE8-6Outputfromarealteam
FIGURE8-7Outputfromanotherrealteam
FIGURE8-8Actualnumbersfromateamthatstruggled
Therealteamsyouseeintheseexamplesallhadvaryingstartingpointsandsustained
velocities,butineachcaseyoucanseeaquickrise,followedbyashallowdip(wherethe
teamwasindangerofstagnating),followedagainbyasharpriseandagradualincreaseto
highsustainablepace.
Sowhydidtheseteamsachievetheseimprovementswhileotherssputterout?The
reasonsfortheinabilitytogrowareasplentifulasthenumberofagileteams.Somefailto
implementengineeringpractices,othersdon’tlearntoworkasatruecross-functional
team,stillothersstrugglewithorganizationalblockers.WhileIcannotsayforcertainwhat
causedtheteamsshowninthesechartstoachievetheirresults,Idoknowthatallthese
teamshadfull-timeScrumMasterswhowerefocusedonclearingimpedimentsand
maximizingteamefficiency.
Don’thandicapyourScrumMasters.Staffingfull-timeScrumMastersgivesthemthe
abilitytorealizetheirtrueroleofteamcoachandadvocate.Anditallowsteamsthe
opportunitytoreachtheirfullpotentialaswell.
KeystoSuccess
ThejobofScrumMasterisreal.Itcanhaveabigimpactoncosts,asillustratedpreviously,
savingthecompanymoney—period.ButwhatdoesaScrumMasterdoalldaytojustifya
full-timerole?Thefollowinglistencompassesmost,butnotall,oftheday-to-daytasks.
Removeimpediments/resolveproblems
Breakupfights/quarrels
Actasateammom
Reportteamdata
Facilitate
Helpoutwhereneeded
Educatetheorganization
Driveorganizationalchange
In2007,MichaelJamespublishedagreatchecklist(seetheURLinthereferences
section[JAMES]).Ifthelinkisbrokenbythetimeyoureadthis,searchfor“ScrumMaster
ChecklistMichaelJames”inyourfavoritesearchengine.
RemoveImpediments/ResolveProblems
Removingimpedimentsanddealingwithproblemsarealwaysatthetopofthelist
because,especiallyearlyinaproject,therearemanyissuestodealwith,bothexternaland
internal.Manychaptersinthisbookareaboutremovingimpediments—runningagood
dailyscrum,havingpeopleshowupontime,andsoon—buttherearemorethanjust
team-centricimpediments.Impedimentsorproblemsfoundbyteamsinlargecompanies
couldbethatanotherteamdoesnotrespondinatimelyfashion,yourteamneedsto
understandhowaninterfaceworksandthepeoplewhowroteitarenotavailable,oryou
havesomeoneoutsickforafewdays.Youalsohavetointerfacewithmanagement,doing
everythingfromconvincingthemthatanagileprocessisworthwhiletoremindingthemof
thefactthatyourjobisfull-timetopavingthewayforbonusesbasedonteam,ratherthan
individual,performance.Finally,ScrumMastersworktohelptheteamsolveitsown
problemsandremoveitsownimpediments.AgoodScrumMastershouldstrivetowork
himselfoutofajobbycreatingahigh-performingteam.Forsome,thismaybeimpossible
toachieve,butitshouldalwaysbethegoal.
BreakUpFights/ActasTeamMom
Peoplearehuman.Sometimestheygetinweirdmoods,don’tgetenoughcoffee,havea
badnight’ssleep,orjustseesomethingthatsetsthemoffinabadway.Oneday,my
youngestdaughter,Emma,whowassixatthetime,couldnotdecidewhatshirttowear
andthoughteverythingshepickedoutlookedterrible—shelostit.MywifeandIhadto
bringherbacktoreality.
I’vehadperfectlyrationalteammemberswhocometoworkonedayandloseitovera
challengetotheirideas.Talkingthemofftheledgetakespatienceandpeopleskills.Good
ScrumMastersneedtohavethesoftskillstointervenewhennecessaryandtheemotional
intelligencetoknowwhenit’sbettertostayoutofitaltogether.
ReportTeamData
TheScrumMasterneedstohelptheteamimprove.Oneofthewaysthisisdoneisthrough
reporting.AgoodScrumMastertrackstheteam’shistoricvelocities,burndownrates,and
manyotherproject-relatedmetrics.ThisisdonenotonlytoassisttheScrumMaster’sown
teaminimprovingitsestimationandsmoothingtherateatwhichstoriesarecompleted
duringasprintbutalsotohelpotherteams(especiallynewones)intheorganizationas
well.
Theimportantthingtorememberisthatreportingteamdata,eitherinternallytothe
teamorexternallytomanagement,doesnotmeanyouarecollectingdatatoholdagainst
theteam.Instead,youareprovidingvisibilityintodatathattheteamneedstoseeand
understandtoimprove.
FacilitateandHelpOutWhereNeeded
FacilitatehasitsrootsintheFrenchwordfacile,whichistomakethingseasy.Thejobof
theScrumMasteristomakeiteasierfortheteamtoachieveitssprintgoal.
ManyScrumMasterswronglyassumethatthebestwaytofacilitateistojustdothe
work,whichisnotthecase.Facilitationisnotsolvingotherpeople’sproblemsordoingit
forthem:it’smakingthingsrunmoresmoothly.Thegoaloffacilitationshouldbetohelp
theteamanalyzeandgetabetterunderstandingofthesystemandtheproblemsathand.
ThejoboftheScrumMasterisnottoprovideanswersortodoeverythingfortheteam,
butratheritistohelppeoplegettotheanswersthattheymayalreadyknow.It’slike
teachingamantofishsohecansustainhimselfversusjustgivinghimafisheverytime
he’shungry.GoodScrumMasterspracticeafacilitativeformofcoachingcalledservant
leadership.JeanTabaka,inCollaborationExplained,sumsthisupnicely.
Simplystated,theservantleaderasleaderdrivestoservethegroupfirst.Leadership
evidencesitselfinservantleadersthroughtheiruseofpower;thatis,inapositionof
strength,theydeterminethatthegreatestpowertheycanwieldisinservicetotheir
teamsasleader.Fortechnicalleads,teamleads,andprojectmanagerswhohaveused
powertocontrolversusserveteams,thisparadigmshiftcanseemantitheticaltotheir
role.Ultimately,however,asevidencedintheagilesoftwaredevelopmentcontext,a
shifttothisalteredstyleofleadershipreapsthegreatestrewards.[TABAKA,p.18]
Now,theremaycomeatimewhenyouneedtohelpouttocompleteatask.Helpingout
isokayonoccasionaslongasyoudon’tforgetthatitcanhinderyouinyourroleas
ScrumMaster.AsI’vementionedinpreviouschapters,astheScrumMasteryouneedto
seetheforest,notjustthetrees.Teammembersarelookingatthetreeinfrontofthem.If
youstarttohelpoutwithtasksandstories,youtoowillbegintoseeonlythetrees,
potentiallycausingyoutomisslargerthingsorimpedimentsthatarenegativelyimpacting
theteam.
Whenyoudohelpout(andIknowyouwill),rememberthatdoingsomustbethe
exceptionratherthantherule.Whiletheteammembersmightkickandscreamwhenyou
refusetohelpmoreoften,remindthem(andyourself)thatyouarenothelpingfortheir
owngood.Iftheystartrelyingonyoufortoomuchandyouimpacttheirvelocity,you
havejustdonethemadisservicebyartificiallyinflatingit.Thelinebetweenhelpinga
teamandhurtingitisafineone.Oneexampleofthiscouldbethattheteamhas
overcommitted.IftheScrumMaster“helpsout”bydoingsomeworkitemsonthesprint
backlog,therebyenablingtheteamtomeetallormostofitscommitmentstotheproduct
owner,thedamageisdone.Theteamwillnotlearnwhyitendedupinthissituationand
willlikelyrepeatitagain.Thegoalhereisfortheteamtolearn,notfortheScrumMaster
tobailouttheteam.
EducatetheOrganizationandDriveOrganizationalChange
“Ifyouwanttomakeenemies,trytochangesomething.”Thatquote,commonlyattributed
toformerUnitedStatesPresidentWoodrowWilson,definitelyappliestothesetwo
ScrumMastertasks:educatetheorganizationanddriveorganizationalchange.It’stoughto
convincepeopletotrysomethingnewbecausechangeisfrightening.Ifindthatwhennew
concepts,policies,ormethodologiesarebeingintroducedinsideanorganization,people’s
firstresponseistowonderhowitwillaffectthem.Willtheirjobsbesafe?Willtheybe
abletoprovidefortheirfamilies?Othersopenlyrebel,refusingtoconsiderthenewway
becausethestatusquo,asbadasitmightbe,isbetterthantheunknown.
AgoodScrumMasterfindsawaytointroducenewideasandinitiatechangewithout
shockingpeople’ssystems.Forexample,afriendofmineisrunningforthelocalschool
board.He’snotgoingtojustwaitandshowuponelectiondayandaskpeopletovotefor
him.Instead,he’srecruitingpeople,likeme,tohelpwithhiscampaign.Thatcampaign
willintroducethevoterstohimandtohisideas,andeventuallyattempttoconvince
peoplethatheisthebestchoiceforthepost.
AsaScrumMaster,youshouldhaveastrategiccampaignforeducatingthe
organization.Consideryourtimelineforrollingitout,whereyouwanttobeandwhen,
andwhomyoucanrecruittohelpbringinyourmessage.Oncethemessageissent,
hopefullyinmultipleformats,youwillbeabletobringpeopleovertoyourwayof
thinking.Asaresult,theywillbeopentohowyouareworkingandwhatyouarebuilding.
Intheend,thejoboftheScrumMasteristwofold:tohelpbuildteamsandtohelpthe
organizationbecomesuccessful.Toreadmoreaboutgettingpeople,andorganizations,on
boardwithScrum,seeChapter2,“GettingPeopleonBoard.”
InSummary
It’stoughtojustifyyourjob,especiallywhenitmightnotbeajobyou’vedonebefore.
WhenapproachingyourmanageraboutmakingtheScrumMasterroleafull-timeone,be
suretohaveacoolheadandagooddealofammunition.
Comearmedwithasmanyfactsasyoucangetyourhandson.Bringdatashowingthe
improvementrealizedbyteamsinyourorganization.Ifyoudon’thaveanyfromyourown
company,bringdatashowingimprovementsinotherorganizations.Relatethose
improvementstothecostperpoint.
Bepreparedtoremindmanagementofthereasonsteamsuseagileprocessesinthefirst
place:lowercosts,reducedrisk,happieremployees,moresatisfiedcustomers,andan
increasedabilitytorespondtochange.
References
[JAMES]James,Michael.2007.“ScrumMasterChecklist.”
http://www.scrummasterchecklist.org/pdf/scrummaster_checklist09.pdf(pulledon15
November2009).
[TABAKA]Tabaka,Jean.2006.CollaborationExplained.UpperSaddleRiver,NJ:
Addison-Wesley,p.18.
WorkConsulted
Benefield,Gabrielle.2008.“RollingOutAgileinaLargeEnterprise.”Proceedingsofthe
AnnualHawaiiInternationalConferenceonSystemSciences(1530–1605),p.461.
PartII:FieldBasics
Chapter9.WhyEngineeringPracticesareImportantin
Scrum
Scrumdoesnottalkaboutengineeringpractices.Assuch,youmightassumethat,likea
Jedimaster,youareexpectedtousetheforcetostartthesprint,committothework,and
thenmagicallymakeitappear.You’renot.ButScrumdoesexpectthatastheteambegins
toworkinanagilemanner,itwillsoonrealizethatgoodengineeringpracticesare
essentialtobecomingahigh-performingScrumteam.Engineeringpracticesarethemagic,
theforce,thepixiedustneededtogetthingsoutthedoor.Idonotattempttodefineallthe
engineeringpracticesthatyouneedtoimplementonyourScrumprojects—thereareother
booksforthat.WhatIdotrytoconveyinthischapter,however,iswhyyouneed
engineeringpracticesandhowtheycanhelpdevelopmentteamsovercomesometypical
obstacles.
IstartwithastoryaboutPatrick,adeveloperbytradewhomovedintoproject
managementatanewcompanyandgotalittlelessthanhebargainedfor.
TheStory
WhenPatrickacceptedhisnewjobatHelix,hewasexcited.Helixhadagreatreputation,
afabulousproducthistory,andhadlongbeenPatrick’sdreamcompany.Patrickcould
hardlywaittogetstartedinhisnewroleasScrumMaster.True,he’dbeinheritingateam,
buthehadbeentoldtheteamwasusingScrum,aframeworkPatrickwasveryfamiliar
with,havingusedScrumandExtremeProgramming(XP)fromboththemanagerand
developerperspectives.Andhewassure,sincetheteammembershadbeentogetherfor
sometime,thatthey’devolvedintoafairlyhigh-performingteam.
Patrickwasinforarudeawakening.
Onhisfirstfulldayonthejob,Patricksatdownwiththeteammembersindividuallyto
gettoknowthemandseehowtheirprocessesworked.HisfirststopwasNiall,reputedto
beoneofthebestcodersontheteam.
“So,Niall,”saidPatrick.“I’vebeenwatchingyouworkforawhile.You’reasgooda
coderaseveryonesays.I’msurprised,though,thatyoudon’tseemtousetest-driven
development.Whynot?”
“Toomuchup-frontworkandtoomuchthrowawaycode,”Niallresponded
offhandedly.“It’skindofawasteoftime.We’veonlygotsixmonthstogetallthisoutthe
door,”hesaidpointingtothewallfullofindexcards.
“Sure.TDDmighttakesomeextratime.Andthereisusuallyextracode,butthedesign
thatemergesismuchcleaner,andwhenyouencounterissuesanderrors—”saidPatrick
beforeNiallinterruptedhim.
“We’vegotadebuggerandacross-functionalteamthatincludestesters,”saidNiall.
“Weaddanybugswefindtothenextsprintandsolvethemaswefindthem.Asfor
design,wehadalotofdesignmeetingsupfront,sothat’sanonissueforus.”
“SoyourissuewithTDDisextratime,correct,Niall?”askedPatrick.
“Yes,Idon’thaveenoughtimetodoit.Weneedtocrankonthecode!”exclaimed
Niall.
“Whataboutwhenyouaredealingwithbugs?Howdoyouknowwheretolook?”asked
Patrick.
“Iusethedebuggertofindtheissues,”statedNiall.
“Whatdoyouconsiderallthetimeyouspendinthedebugger?Isn’tthatextratime,
too?”challengedPatrick.
“Well,Iguess.I’veneverthoughtaboutitthatway,”Nialladmitted.
Patricknodded.“Onemorequestion.Inoticedthathereyouwrotesomecodetocreate
drop-downboxes.What’stheacceptancecriteriayou’retryingtofulfillwiththat?”
Niallsmiledconspiratorially.“Noonehasaskedforityet.Butoncetheyseehowit
works,Ithinkthey’rereallygoingtolikeit.Iliketogivethecustomersalittlesomething
extra.Itgetsthemexcitedaboutthefinalproduct.”
“Yourheart’sintherightplace,butit’sthewrongthingtodo,”saidPatrick.“Wehave
towritejustenoughcodetosatisfythestory.Nothingmore.Nothingless.We’vegottoo
muchtodotospendtimegold-platingthesoftware.Andjustthinkofthetimeitwould
freeuptodothatTDDweweretalkingabout.Thinkaboutit.We’lltalkmorelater.”
PatrickmovedontoClaire,anexperiencedteammemberwithareputationforfinishing
herworkquickly.
“Thismorninginthedailyscrum,youmentionedthatStory1822wasdone,right?”
“Right,”Clairereplied.
“SowhatifIaskedyoutoproveittome?Whatwouldyoudo?”askedPatrick.
“Well,itworks.See?”answeredClaire,givingPatrickaquickdemo.
“Right.ButwhataboutallthethingsIcan’tsee?Isitcheckedinyet?”
“No.I’mplanningtodothatlater.I’mdemoingitonmyownmachine.”
“Howoftendoyoucheckin,Claire?”
“Itrytocheckinonceaday,butit’smoreliketwotothreetimesperweek.”
“Ifyoucheckinthatinfrequently,howdoyouknowthatyourcodeisn’tbreaking
anythingelse?”
“Ummm.IguessI’llfindoutatcheck-inanddealwithitthen.Itusuallyisn’ta
problem.”
“Dootherteammemberscheckintwoorthreetimesperweekalso?”
“Yes,”saidClaire.
“Sorightnow,andmaybeevenwhenwecheckin(ifalltheothers’codeisn’tchecked
inyet),wedon’treallyknowifastoryisdone,dowe?”askedPatrick.
“Weknowit’swritten.Weknowit’sworking,”respondedClaire.
“I’vealsonoticedthereisnocontinuousintegrationserver,anditappearsthattheteam
membersdon’tsharealotofcodewitheachother,”saidPatrick.
“WekeepmeaningtosetupaCIserver,butwedon’thavetimeorfundingforit,”said
Claire.“Iknowweneedtosetitup,andweneedtofollowmorerigorousengineering
practices,butwe’reundersomuchtimepressure.Withmanagementbreathingdownour
necks,doingtherightthingisalmostimpossible!”
Abitworriedbytheattitudeoftheseniorteammemberstowardwhatheconsidered
basicengineeringpracticeslikeTDD,staticcodeanalysis,sharedcodeownership,and
continuousintegration,Patrickspentthenextthreedayslookingatthebacklogand
burndownchartsanddiggingaroundinthecodetoseewhatothertrendshecouldfind.He
lookedattheacceptancetestframeworkthattheteamhadimplementedandsawthatit
hadnotbeenupdatedatallduringthesprint.Helookedatthesizeofthecheck-ins,often
1,000linesofcodeormore.Henoticedthattheteamtendedtoleaveoneortwostories
unfinishedalmosteverysprint.Healsosawfromtheburndownthattheyoftenhadalotof
workinprogressuntilquitelateinthesprint.What’sworse,hefoundnothingrelatedto
goodcodingpractices—variouspartsofthecodewerewrittenindifferentstyles,there
werefewcommentsinthecode,andoverallitjusthadapoorstructure.
Patrickwasshockedwithhowfewbasicengineeringpracticestheteamhad
implementedandthenumberofbugsthatcontinuedtopopupineverysprint.Hisbig
concernwashowtheywouldreleaseaqualityproduct,ontime,withoutmakingsome
seriouschanges.Hethoughtbacktoacoupleofprojectssomecolleagueshadworkedon
athislastcompany—projectsthathadslippedbymonthsbecauseoflargearchitectural
issuesthatwereidentifiedlate.Heneededtheteamtounderstandwhythesepractices
wereworththeeffort.Afterthinkingforsometime,hecameupwithaplan.
Thesprintended,andafterthedemotheteamgatheredintheconferenceroomforthe
sprintretrospective.Whiletheywerefilteringin,Patrickquietlytapedtwobigsheetsof
paperuparoundtheroom.Theretrospectivestarted.Patricknoticedthattheaccepted
formatseemedtobeforeveryonetotaketurnstalkingaboutwhathadgonewellandwhat
couldgobetter.Perfect,hethought.Hegotupastheybegantalkingandbeganwriting
downafewofthethingstheteamwassayingononeofthesheets,whichhelabeled
“StuffGoingWrong.”
Whentheteamhadfinishedtalking,Patrickmovedtothenextsheetofpaper.He
labeledit“StuffWeCanDotoFixIt.”
“So,team.I’vetalkedtoeachofyouindividually.I’velookedthroughthecode.I
imagineyouhaveproblemsliketheseeverysprint,right?”
“Sure,”saidNiall.“Butwe’regoingtohaveproblems.There’snoblamehere.”
“Notblaming,”saidPatrick.“Justwondering.Isthereanythingwecandoaboutit?”
“Notifwewanttokeepupthispace,”saidClaire.
“That’sinteresting.Let’stalkaboutpace.Tome,alltheseproblems,”Patrickpausedto
pointtothesheetfullofissues,thencontinued,“theseissuesthatkeeprecurring—theyare
slowingyoudownthemost:bugs,gold-plating,findingpoorcodebutleavingittofester
until‘someone,someday’fixesit.Wehaveadeadlinetomeetandrightnow,letmetell
you,we’renotgoingtogetthere.We’vegottoinvestsometimeingettingbettersothat
wecangofaster.”
Helettheteamsoakthatinforafewseconds.
“Sowhatarewegoingtoinvestin?”askedPatrick.Thenhesatdownandletthesilence
growuncomfortable,waitingfortheteamtorespond.
“IseestuffeverydayIwishIhadtimetofix,”saidoneoftheteammembers.“Soone
thingwecoulddowouldbetoincludesomerefactoringtasksineachsprint.”
“Good,”saidPatrick,standinguptowriteitonthepaper.“Whatelse?”
“Test-drivendevelopment,”saidNiall,inalow,resignedvoice.“Irealizeit’stheright
thingtodo,butit’llslowmedown.”
Patricknoddedbutwroteitanyway.“Maybeatfirst.Butwe’llmorethanmakeupfor
thatbycreatingfewerbugsandspendinglesstimefindingandtrackingtheonesthatdo
slipthrough.”
“Continuousintegration,”saidClaire.“Andmorefrequentcheck-inswillbothreally
help.Thatdoesn’ttakethatmuchtimeonceit’ssetup.It’sjustnotahabit,andwedon’t
havethetoolsrightnow.”
Patrickwroteit.“Yep.Toolsareeasywhencomparedtohabits.It’schangingthehabit
that’llbehard.”
Theteamwentontomentionpairprogramming.
“AndpairprogrammingandTDDwillhelpensurethatwebuildonlywhatisnecessary,
whichwillultimatelyincreaseourvelocityaswell,”saidPatrick.“Clearly,youknowwhat
youshouldbedoing.Sowhyaren’tyoudoingit?”
Theteamwentontolistseveralobstaclestoadoptingengineeringpractices,including
lackofknowledgeandfearoftheunknown.Theydiscussedeachobjection,withPatrick
providingtherationaleforatleastgivingitatry.Intheend,theyagreedtoanew
definitionofdone,whichwouldnecessitateimmediatelyimplementingcertain
engineeringpractices.
Theteamdecidedtoimplementothernewpracticesoneatatimeinsubsequentsprints.
Inturn,Patrickagreedtodiscusstheirplanwithmanagementandexplaintothemwhythe
team’svelocitymightdropforasprintortwowhiletheyinvestinpracticesthatwould
ultimatelyhelpthemgofaster.
“We’vegotalongwaytogotogetfromheretothere,”finishedPatrick.“ButIpromise
oncewestartworkingthisway,youwon’twanttogoback.ThefirstthingIwantusallto
doisstartchallengingeachother,everyday,toproveit.Proveit’stested,proveit’snot
goingtobreakthebuild,proveit’sdone.That’llgetusalongwaytowardwherewewant
togo.”
ThePractices
Thechallengesinthischapter’sstoryarenotuncommonwithteamsnewtoScrum.It’s
easytogetfocusedondoingScrumrightandnevergetaroundtoimplementingthe
engineeringpracticesnecessarytobecomeatrulyhigh-performingScrumteam.Patrick’s
solutionwastoworkwiththeteammemberstomakevisiblewheretheywerefalling
short,toshowthemhowitwasslowingthemdown,andtoconvincethemthatinvesting
inbetterpracticeswouldbothsolvetheirissuesandincreasetheirvelocityaswell.Patrick
andtheteamagreedtoimplementthefollowingkeyengineeringpractices:
Implementingtest-drivendevelopment(TDD)
Refactoring
Continuousintegrationandmorefrequentcheck-ins
Pairprogramming
Automatedintegrationandacceptancetests
Alltheseengineeringpracticesmeanmoreworkupfront.Youwillrecoverthetime
laterthroughefficiency,improvedquality,andcodestability.Whatisoftenforgotten,
however,isthatthesearenotsilverbullets.Doingoneordoingthemallwillnotguarantee
yousuccess,buttheywillgetyoufartherawayfromfailureiftheyaredonewithdiligence
anddiscipline.
ImplementingTest-DrivenDevelopment
Contrarytowhatthenameimplies,TDDisnotatestingprocess.It’sasoftwaredesign
techniquewherecodeisdevelopedinshortcycles.YoucanthinkoftheTDDprocessas
Red,Green,Refactor.Thefirststepistowriteanewunittestwithoutwritingthe
necessarycodetomakeitpass.Obviously,whenyourunthattest,itwillfail:Thisisthe
redstate.Thenextstepistowritenewcode,justenoughtomaketheunittestpass,orturn
green.Whenthetestpasses,thecodeisworking.Thefinalstep,whichcanoccurmultiple
times,istorefactor.Thenextsectiontalksmoreaboutrefactoring,butinTDDit
specificallyreferstoremovingany“codesmells,”elementsthatmightnotbeoptimized,
well-designed,oreasytomaintain.Designchangescanoccurduringthisphaseaswell
(alsocalledtest-drivendesign).Duringrefactoring,thecodeischanged,butallthewhile,
thetestsmuststaygreen,orcontinuetopass.Theprocessthenbeginsagainfromthe
beginning,writingnewtestsandnewcodeuntilthestoryiscomplete.
I’mnotgoingtoattempttoteachyouTDD;otherbooksdoagoodjobofthat,andI’ve
listedtheminthereferenceswiththischapter.WhatIwantyoutoknowiswhyTDD
helpsaScrumteamandhowyoucanhelpyourownteamreachtheconclusionthatTDD
isthebestchoiceformeetingsprintgoalsatahighvelocity.
ImplementingTDDisakintoweavinganet.Eachautomatedtestbuildsonthenext,
formingasafetynetofautomatedunittests.Considerthehigh-wireactatthecircus.The
personhundredsoffeetabovetheground,tryingtobalanceonathinwire,cantakethe
riskshedoesbecauseheknowsthatifhefalls,thenetwillsavehim.Ateammightchoose
towalkthehighwirewithoutasafetynet,butiftheteamfalls,itcouldtakedays,maybe
weeks,torecover.Yes,morecodeiswrittenwithautomatedunittests,and,yes,
sometimesthereismorethrowawaycode.Inmanycases,though,thereduceddebugging
timemeansthatthetotalimplementationtimeisactuallyshorter.
In2008,researchersfromMicrosoft,IBM,andNorthCarolinaStateUniversity
publishedapaper[NAGGAPAN],lookingatthebenefitofTDDinteams.Theylookedat
threeteamsfromMicrosoft(oneofwhichIhappenedtobeworkingonatthetime)and
oneteamfromIBM.Theyfoundthatallteamsshowed“significant”dropsindefect
density1:40percentfortheIBMteamandbetween60percentand90percentforthe
Microsoftteams.TheincreasedtimetakentofollowTDDontheseprojectsrangedfrom
15percentto35percent.Thisinitialdecreaseinvelocityiscompensatedforbythe
increasedstabilityandmaintainabilityofthecode.
1.Definedinthepublished[NAGGAPAN]paperas“Whensoftwareisbeingdeveloped,apersonmakesanerrorthat
resultsinaphysicalfault(ordefect)inasoftwareelement.Whenthiselementisexecuted,traversalofthefaultor
defectmayputtheelement(orsystem)intoanerroneousstate.Whenthiserroneousstateresultsinanexternally
visibleanomaly,wesaythatafailurehasoccurred.”
ByfollowingTDD,youwillbelessreluctanttoimplementachangethatisrequiredby
akeystakeholderorcustomer,evenlateintheproject.WithoutTDD,thesekindsof
eleventh-hourchangesareconsideredtooriskybecausetheteamcan’tpredictthe
ramificationsofthechange(coupledcode),andthereisn’ttimetotestallthepossibleside
effects.WithaTDDsafetynetinplace,theriskissubstantiallylessened,soteamsfeel
confidentthattheycanmakerequiredchanges,nomatterwhentheyoccurintheproject.
Goodversioncontrolpracticesalsohelptoincreaseconfidence.Iftherefactoringgoes
horriblyawryorbreakstoomanytests,youcaneasilyrollbacktoaknowngoodpoint.
Anecdotally,intheprojectsthatI’verunandonteamsthatI’vecoached,Ihavefound
thatTDDsignificantlystreamlinesthewritingofcode,helpspreventanalysisparalysis,
minimizesimpediments(likespendinglotsoftimeinthedebugger),andgivesteamsthe
pricelessgiftofpeaceofmind.Forthesereasons,TDDisessentialforteamsthatwantto
growtobehigh-performingScrumteams.And,remember,thecompileronlytellsyouif
yourcodeobeystherulesofthelanguage;yourtestsprovethatyourcodeexecutes
correctly.
Refactoring
Refactoringistheactofenhancingorimprovingthedesignofexistingcodewithout
changingitsintentorbehavior.Theexternalbehaviorremainsthesame,whileunderthe
hood,thingsaremorestreamlined.
Howdoyouknowwhenyouneedtorefactoryourcode?Whenitbeginstosmell.It’s
besttofollowtherulesofGrandmaBeck:“Ifitstinks,changeit”[FOWLER01].Her
commentwasinrelationtodiapers,butitappliesequallyaswelltocode.Acodesmellis
aseeminglysmallproblemthatissymptomaticofadeeperissue.JeffAtwoodlistssome
commonsmellsina2006blogpost,breakingthemoutintosmellswithinclassesand
smellsbetweenclasses[ATWOOD].
Table9-1presentsafewofmyfavoritesfromAtwood’slist.
TABLE9-1SampleofCodeSmellsfromJeffAtwood
Thesemightseemlikebasiccodingstandards,yettheyaretoooftenleftundone.
Whenshouldyourefactor?Refactoringcanbeperformedatanytimeduringasystem’s
lifecycle,wheneversomeoneisworkingonlegacycode,seesapieceofcodethatcould
beimproved,noticesthesystemactingsluggish,orifbugsstartpoppingupinaparticular
sectionofthecode—basicallywheneversomeonenoticesasmell.
Rememberthatrefactoringcodedoesnotmeanrewritingthecode.Instead,torefactor
istooptimizethecodewithoutchangingtheintentofthecode.RobertMartin[MARTIN]
hasasetofprincipleswiththeacronymSOLID,inwhichhetalksaboutthefirstfive
principlesofclassdesign(seeTable9-2).
TABLE9-2RobertMartin’sSOLIDOODClassDesignPrinciples
Whyshouldyourefactor?Simplyput,it’saninvestmentinyourcode.Ananalogy
mightbedoinglittlethingseverydaytohelpkeepyourhouseclean.Ifyouwalkinthe
housewithdirtyshoesonyourfeetonaregularbasis,yourhousewillgetdirtierfaster.If
youthrowyourdiscardedwrappersonthecounterratherthaninthegarbagecanbecause
you’reinahurry,inafewdaysyou’regoingtolookupandfindacounterfullofwrappers
thatyou’llhavetoclean.Refactoringislikecleaningupafteryourselfalittlebiteachday
—sometimesyou’llcleanmore,sometimesless,buttheimportantthingisthatyou’re
cleaningonaregularbasisandnotwaitingforittopileuporforsomeoneelsetodoit.
ContinuousIntegrationandMoreFrequentCheck-Ins
Withdiscipline,continuousintegrationcantakeaScrumteamfromsomewhatgoodtojust
plainawesome.MartinFowlerdefinescontinuousintegrationbest[FOWLER02]:
ContinuousIntegrationisasoftwaredevelopmentpracticewheremembersofateam
integratetheirworkfrequently,usuallyeachpersonintegratesatleastdaily—leading
tomultipleintegrationsperday.Eachintegrationisverifiedbyanautomatedbuild
(includingtest)todetectintegrationerrorsasquicklyaspossible.
Continuousintegrationenablesustohavequickerfeedbackloopsonthecode.Instead
ofcheckinginathousandormorelinesatonetime,eachpersonorpairmightcheckina
hundredorsolinesofcodeatonce.Theyalsocreatealocalbuildbeforecommitting
changestothelargerbranch.Boththelocalbuildsandthemainbuildshouldbefully
automated,meaningthepersonsubmittingthecodecan“clickandforget,”whilethebuild
machinerunsasuiteofautomatedacceptanceandunittests.Thesebuildsmightalsorun
staticanalysischecks,verifydesiredunittestcoverage,andsoforth.Oncethecodeisrun
andinthebranch,acontinuousintegrationserverpullseverythingintothebranchand
runsalargersuiteoftests.Becausethecheck-inissmall,thefeedbackloopisshort.
Therefore,itismucheasiertoseewhatbrokethebuildandwhy.
Whyshouldyoucareaboutanyofthis?Letmeassociateitwithreallife.
Headlights—Whydodriversuseheadlightsatnight?Becausetheycanseewhere
they’regoingandotherscanseethem,itreducestheriskofanaccident.Continuous
integrationprovidesthesamebenefit—itshinesalightdownthepathofcode
developmentsothatyoucandeterminewhetheryou’regoingtohitsomething.A
continuousintegrationbuildwarnsyoualmostinstantlyifsomethingisblocking
yourway.
Washingmachines—Whatmakesthemgreat?Theyeliminatetherepetitive,labor-
intensiveprocessofwashingclothesbyhand.Similarly,continuousintegration
reducesall(ormost)manualprocesses.Notonlydoesthisfreeuptimeand
resources,italsohelpsimprovequality.Whenpeopledothingsmanually,thereisan
increasedchanceforvariation.Automatinghelpsensurethatthesametestrunsthe
exactsameway,everytime.Havingautomationinplaceallowsteammembersto
focusonwhatmatters,thesystemunderdevelopment.
GPS—HavingaGPSinyourvehicleallowsyoutoknowwhereyouareandwhere
youaregoingatalltimes.Followingcontinuousintegrationdoesthesamething.It
providesvisibilityintowherethesystemunderdevelopmentiscurrentlyandwhere
itisgoingnext.Thisenablestheteamandtheproductownertoreactinrealtime.
Further,ifteamsaddqualitygatesorcheckssuchasunittestcoverage,teamsare
abletoidentifypatternsortrendswiththecodeandmakenecessarycorrectionsin
timetoalterthesystem’scourse.
Scrummaintainsthatattheendofthesprint,allcodeshouldbepotentiallyshippable,
yetmanyteamsstrugglewithwhatthatreallymeans.Continuousintegrationenables
teamstobuildandreleaseatanytime.Whatdoesthislooklikeintherealworld?Itmeans
goingintoacustomerreviewmeetingandpullingthelatestbitsoffthebuildserverand
installingthem.Nomanualtesting,nohumanvalidation,justpuretrustinthesystem.And
wheredoesthistrustcomefrom?Asteamsdevelopgoodengineeringpractices,their
confidenceintheirsystemsgrows.TDDgivesthemasafetynetandshowsthemwhere
theirerrorsarewhiletheircontinuousintegrationstrategyallowsthemtofullyunderstand
theimpactsofcodechanges.Itisimportanttokeepthebuildinagreenstateandnotgo
homewithabrokenbuild.Whenbreaksdooccur,itisbettertorollbackanychangesat
theendoftheday,gohomeinagreenstate,andfixanyissuesinthemorning.
OneofmyfavoritebooksoncontinuousintegrationisContinuousIntegration:
ImprovingSoftwareQualityandReducingRiskbyPaulM.Duvall.Asstatedbefore,this
isnotabookonsoftwaredesignorhardcoreengineering;buyPaul’sbookandlearnjust
abouteverythingyouneedtoknowabouthowtostartdoingcontinuousintegration.
Alternatively,apaperpresentedatAgile2008byAdeMiller,“OneHundredDaysof
ContinuousIntegration,”isalsoagreatread.
PairProgramming
Toooftenaspeopleshiftinandoutofprojects,theinformationshiftsinandoutwiththem
—informationrangingfromhowstorieswereimplementedtowhyonedesignpathwas
chosenoveranother.Thereisonetechniquethatensuresthateveryoneontheteamknows
whatthecodeactuallydoes:pairprogramming.
Pairprogrammingdrivescollectivecodeownershipandincreasesknowledgesharing
amongteammembers.Atitscore,pairprogrammingistwoteammembersworking
togethertoaccomplishasingletask.Onepersonisthedriverandtheotheristhenavigator
(formoreadvancedpairingtechniques,seeChapter19,“KeepingPeopleEngagedwith
PairProgramming”).Whenthepairsswitchonafrequentbasis,informationpassesamong
allteammembers,ensuringthateveryoneisfamiliarwiththeentirecodebase.Pair
programmingusesconversationtoproducecode,causingeachpersoninthepairto
questioncontinuallywhattheotherisdoing.Pairsreinforcetheteam’scodingstandards
onhowtogetthesystembuilt.
Itisimportanttounderstandthatpracticingpairprogrammingdoesnotnecessarily
meanworktakestwiceaslong.WhileattheUniversityofUtah,LaurieWilliamsfound
thatpairprogrammingwas15percentslowerthantwoindependentdevelopersbut
produced15percentfewerbugs[WILLIAMS].Thisfindingreemphasizesthatteams
makingtheinvestmentinmoredisciplinedengineeringpracticesupfrontgoalittleslower
atfirstbuthaveabetterpaybackdowntheroadthroughreducedrisk,bettervisibility,and
lowerdefectrates.
Anotherbenefitofpairprogrammingisthereductionofnoise.Idefinenoisetomean
phonecalls,email,instantmessages,unnecessarymeetings,andotherdistractions.It
wouldberudetobeworkingonaprojectwithsomeoneandcheckyouremailorbeona
phonecall,right?Whenyou’repairingwithsomeone,youfocusonwhatyouareworking
onwiththeotherperson,justbecauseit’sgoodmanners.Thesidebenefitofthatfocusis
thatyougetmoredone.
Iliketothinkofpairprogrammingas“real-timecodereviews,”meaningthecodeis
beingreviewedandupdatedandmanagedinrealtime.Agoodpairisconstantlyreflecting
ontheworkathand.Asaresult,thereisnoneedforseparatepre-check-inmeetingsto
reviewwhatindividualteammembersdid.Andbecausefewerbugsarecreated,fewerbug
triagemeetingsareneeded.Allthisnoisereductionoptimizestheteam’stime.
Pairinginadistributedenvironmentischallengingbutpossible.Sharingascreen,using
webcams,andhavingadirectlinetothepersonyouarepairingwithareallrequirements.
Itisobviouslyeasiertopairwhenyouareinthesametimezone,soifyoudistribute,go
north/southinsteadofeast/west.
Todopairprogrammingwell,you’vegottostartwitheducationandgroundrules.Let
peopleknowthat,aswithtryinganythingnew,pairingwillbeuncomfortableandabit
awkwardatfirst—remindthemtokeepanopenmindandtohavecourage,knowingthat
theentireteamisgoingthroughthistogether.Encouragethemtohaveawillingnesstotry
—toreallygiveitagoodeffortandbeabletomoveforwardobjectively.Ifneeded,invest
intrainingandalwaysbesuretoimplementcodingstandards.
Therearesomegreatpapersandbooksonpairprogramming.Startwithanolderbook,
PairProgrammingIlluminated,byLaurieWilliamsandRobertKessler.Anothergood
referenceisTheArtofAgileDevelopmentbyJamesShore.Tounderstandthecostbenefits
ofpairing,read“TheEconomicsofSoftwareDevelopmentbyPairProgrammers”by
HakanErdogmusandLaurieWilliams.
AutomatedIntegrationandAcceptanceTests
Likeunittests,integrationandacceptancetestsareaformofautomatedtests.Theyfillout
thetopofthetestautomationpyramid,withunittestsmakingupthebaselayer(seeFigure
9-1).
FIGURE9-1Testautomationpyramid
Integrationtestsaredesignedtotestvariousintegrationpointsinyoursystem.Theyare
designedtoensurethatAPIs,dataformats,andinterfacesareallworkingasexpected
whiletheteamisdevelopingthecode.Whileitisgenerallyeasiertobuildautomated
integrationteststhanitwouldbetobuildautomatedacceptancetests,automated
integrationtestsstillhaveahighmaintenancecost,especiallyinsystemswithhundredsor
thousandsofpeopleworkingtogetherinasinglecodebase.
Acceptancetestsaredesignedtoemulatethebehavioroftheuser.Ifyouarebuildinga
systemwithaUI,thesetestswillbescriptedtoclickandemulatethebehavioroftheuser.
Thesetestsareofteneasilybroken,asinterfacesandcustomers’mindschange.Theyare
alsopronetoprovidingfalse-positivereadings,whichmakestheirmaintenancecosthigh
—butnotashighasadedicatedteammember.Avarietyoftoolsaredesignedtolet
stakeholdersandcustomerswritetestsandeasilyinterfacewiththesystemtoensurethe
systemisworkingfromanend-userperspective.
Evenwiththeaddedworkandmaintenancecost,though,automatedintegrationand
acceptancetestsareoneofthebestwaystoensurethataScrumteamisbuildingtheright
thing.First,eachallowstheteamtohaveacontinuousfeedbackloop,similartothat
providedbycontinuousintegrationandunittests,whichhelpspreventlongtestingphases
attheendofeachsprintandeachrelease.
Second,it’scheapertofixissueswhentheyoccurthanitistofixthemafewsprints
downtheroad.Plus,fixingbugsearlyhelpskeeptheoverallcodebaseclean.Whileit
mightinitiallyslowtheteamdownwithwhatcanseemtobeconstantbugfixing,theend
resultwillbeacleaner,morerefinedcodebaseovertime.
Third,havingautomatedintegrationtestshelpsensurethattheexternalintegration
pointsthesysteminterfaceswithareworking,andcontinueworking,asexpected.Ioften
askteams,“Whatareyourcomputersdoingafteryougohomeatnight?”Theyshouldbe
workingfortheteam,identifyingweaknessesinthesystemsothattheycanbecorrected
thenextmorning.
Finally,havingautomatedacceptancetestshelpsensurethattheUI/interfacebehaviors
ofthesystemworkasexpected.Acceptancetestsshouldbefiredoffateverybuildif
possibleandshouldbecontinuallybuiltoutovertime,justastheunittest“safetynet”is
builtoutovertime.Thismeansthatacceptancetestsneedtobewrittenbeforethesprint
startssowhentheproductownergoesintothesprintplanningmeeting,theteamknows
whatitneedsto“prove”tothecustomersandstakeholdersfromaninterfacepointof
view.Whenyoudon’tinvestinautomation,eachsprintthetestdebtbuildsandtheamount
oftimespenteachsprintonmanualtestingrisestoanearlyunmanageablestate,as
illustratedinFigure9-2.
FIGURE9-2Timespentwithautomationversusmanualtesting
KeystoSuccess
Yes,youcandoScrumwithoutimplementingevenoneoftheengineeringpracticeslisted
inthischapter.Andyes,youwillfindsomesuccess.Eventually,however,youwillfind
yourselfasking,“Whyarewenotgettingsignificantlybetter?”Theanswerliesinhowyou
actuallybuildyoursoftware.IfyoudoScrumbutrefusetochangeyourengineering
practicesandyourmindset,you’llstruggle.
AllIamaskingyoutodoisthinkaboutthingsdifferently;understandthewhywhen
youdothings.Intheend,however,youmayhavetodosomeadditionalinitialwork
(TDD,pairing,settingupcontinuousintegration,creatinganautomatedtesting
framework)andmakeacalculatedbetthatitwillpayoffintheendbyhavinglessrework
todo—simplybecauseyouwillhavefewerbugsandintegrationissues.
NotaSilverBullet
Icannotstressthisenough:UsingScrumoranyofthepracticeslistedinthischapterwill
notmagicallygrantyoutheabilitytomorefrequentlydeliversoftwareofhigherqualityor
guaranteethatwhatyoubuildwillbewhatyourcustomersreallywant.Ican,however,
predictthatchoosingnottousethesetechniqueswilllikelyresultinfailure.Inthesame
way,ridingahigh-qualityroadbikewon’tguaranteeawinattheTourdeFrance;ridinga
tricycle,however,willvirtuallyguaranteealoss.
Thepracticesinthischapterarenotsilverbullets.Insteadtheyaremeantassignposts
thathelpguideyoutoyourfinaldestination,whichisahigh-performingteamabletoreact
quicklyandwithhighqualitytocustomerrequests.
StartingOut
Gettingstartedwithengineeringpracticescanseemoverwhelming,butitdoesn’thaveto
be.Ihaveseenteamstaketwodifferentapproaches:Theystartslowly,graduallyadding
onenewpracticeatatime,ortheygofullthrottleandembraceeverything.Eitherwayyou
doit,it’simportanttorealizethatyouaregoingtostumbleasyouworkyourwaythrough
revampingyourengineeringpractices.Thisisokay—justremembertostaydisciplined
andfocused,andseekhelpwhenyouneedit.
GettheTeamtoBuyIn
Inthestory,Patricktooktheteamtothewhyrightoutofthegate.Sure,hecouldhave
said,“Doitmywayorthehighway,”butinmyexperienceitworksmuchbetterifthe
teammembersunderstandwhytheyarebeingaskedtochange.Patrickhelpedthe
membersoftheteamseehispointofview,helpedthemunderstandwhathewanted,and
gavethemguidanceonhowtogetthereandwhyitwasimportant.Thisisessential,
becauseifyouhavesomeoneontheteamwhoisnotsupportingtheeffort,therewillbe
sabotage.Avoidthecommand-and-controlaspectandfocusonthewhy.
DefinitionofDone
InChapter7,“HowDoYouKnowYou’reDone?”youfindanexerciseonhowtobuilda
donelistandhowtoworkwithyourteamtodefinewhatdoneis.Attheendofthis
chapter’sstory,Patrickdidasimilarexercisewithhisteam.Knowingwhatitmeanstobe
doneislikehavingasetofblueprints.Thepracticesinthischapterarethetoolsthatallow
teamstobuild.Withoutunderstandingwhatdoneis,youwillbelost.
BuildEngineeringintoProductBacklog
Fortheteamtogettractiononthepracticeslistedinthischapter,theproductownermust
understandthepayoffandbewillingtomakethelong-terminvestmentintheteam.One
waytodothisandtohighlightthevisibilityoftheteam’sefforttoimproveitsengineering
practicesistoincludeengineeringrelatedstoriesortasksintheproductbacklog.
GetTraining/Coaching
Don’texpectpeopleto“justdoit.”Teachthemhow.Youcandothisbyhiringacoach,
findingsomeoneinyourcompanywhohasthisexperience,buyingbooks,orsending
themtotrainingclasses.Ifmoneyisanissue,realizethatthebenefitthatmodern
engineeringpracticeswillbringbacktothecompanywillbefargreaterthanthefinancial
investmentbeingmadeupfront.
PuttingItTogether
Goodengineeringpracticesbreakthingsintosmallpieces,fromunitteststocheck-ins,so
thatthetimebetweenissueandfixisminimal.Further,theabilitytochangedirectionwith
smallbatchesincreasescustomerconfidence.Asaresult,riskisreducedacrosstheboard.
Theautomatedunittestframeworkthattheteamhasbuiltoverthelengthoftheproject
alertsteammemberswhenthereisaproblemandtellsthemwheretheproblemis—no
morespending“daysinthedebugger.”Theyalsohavebettervisibilityintothecode
throughpairing,asthe“Iownthecode”mentalityisreplacedwithcollectivecode
ownership.Thismixtureresultsinlowerdefectratesandoverallincreasesinquality—
givingcustomersmoretrustintheteamandtheteammoreconfidenceinitsabilityto
deliverworkingsoftwareeverysprint.
Pleaseunderstandthatthesepracticesarenoteasytodo.Theyrequirededication,focus,
andperseverance.Youwillbepulledawayfromthem,naturally,duetothetime
investmentneededtodoeachofthem.WithTDD,youwilllikelyneedtogettrainedon
howtodoitandinitiallyyouwillslowdown.Thesamewithsettingupacontinuous
integrationserver—thoughyoucanhaveitupandrunninginunderanhour,gettingitset
uptosuityourneedsandtheneedsofthebusinesstakestime.Refactoringandpair
programmingalsotaketimetobecomeaccustomedto.Overall,Ifindittakesfromthree
tosixmonthstogofromatraditionalwayofbuildingsoftwaretofullyunderstandingand
implementingtheengineeringpracticeslistedinthischapter.
Attheendoftheday,theinvestmentyoumakenowwillpayhugedividendsinthe
future,givingyoustability,quality,andtheconsistentreleasesyourcustomersdesire.
References
[ATWOOD]Atwood,Jeff.“CodeSmells.”CodingHorror.
http://www.codinghorror.com/blog/2006/05/code-smells.html(accessed01/11/2009).
[FOWLER01]Fowler,Martin,KentBeck,etal.1999.Refactoring:Improvingthe
DesignofExistingCode.Reading,MA:Addison-WesleyProfessional,p.75.
[FOWLER02]Fowler,Martin.“ContinuousIntegration.”martinfowler.com.
http://martinfowler.com/articles/continuousIntegration.html(accessed03/15/2010).
[MARTIN]Martin,RobertC.“ThePrinciplesofOOD.”
http://butunclebob.com/ArticleS.UncleBob.PrinciplesOfOod(accessed07/11/2011).
[NAGGAPAN]Nagappan,Nachiappan,E.MichaelMaximilien,ThirumaleshBhat,and
LaurieWilliams.“RealizingQualityImprovementthroughTestDrivenDevelopment:
ResultsandExperiencesofFourIndustrialTeams.”MicrosoftResearch.
http://research.microsoft.com/en-us/groups/ese/nagappan_tdd.pdf(accessed07/11/2011).
[WILLIAMS]Williams,Laurie.“TheCollaborativeSoftwareProcess.”TheUniversity
ofUtah.http://www.cs.utah.edu/~lwilliam/Papers/dissertation.pdf(accessed01/11/2011).
WorksConsulted
Astels,David.2004.Test-DrivenDevelopment:APracticalGuide.UpperSaddleRiver,
NJ:PrenticeHall.
Erdogmus,Hakan,andLaurieWilliams.“TheEconomicsofSoftwareDevelopmentby
PairProgrammers.”TheEngineeringEconomist48(4):283–319.
Shore,James,andShaneWarden.2008.TheArtofAgileDevelopment.Sebastopol,CA:
O’ReillyMedia.
Williams,Laurie,andRobertR.Kessler.2003.PairProgrammingIlluminated.Boston:
Addison-Wesley.
Chapter10.CoreHours
Scrumteamsshouldworktogetherasmuchaspossible,ideallyinacolocatedspace.
Workinginthiswaydrivescommunicationefficiencies,collaboration,andcollectivecode
ownership.Butwhatdoyoudowhensometeammembersliketoarriveat7a.m.and
othersliketoarriveat11a.m.?Whataboutend-of-dayscenariosandlunchtime?
Whilethesemayseemliketrivialthings,formanyteamstheyarenot.Infact,a
seeminglyharmlessandsmallimpedimentlikecorehourscansinkanentireteam.That’s
whatnearlyhappenedinthestorythatfollows—whenamoreseniordeveloperhadan
issuewithayoungdeveloper’sperceivedlackofdedicationtotheteamandthecompany,
itnearlytoretheteamapart.
TheStory
Russwasayoungdeveloper,fiveyearsoutofuniversity.Helovedgaming,ridinghis
bicycle,andworkingonhisclassiccar.Hehadspentthelastfewyearsworkingina
startupinSeattlewheretherewaslittleprocess;careersweremadenotbyteamworkbut
bypurerawtalent.Thoughnewtohiscurrentcompany,Russwasalreadyknownasone
ofthego-toguys,someonewhogotthingsdone.Russpridedhimselfonkeepingagood
work–lifebalance,whichdrovehimtowardhavingsomeofthebestcodeinthecompany.
Hefrequentlyrodehisbikebackandforthtowork,especiallyinthesummer,oftentaking
thelongroutefromtheeastsidetoSeattlebygoingaroundLakeWashington.Heloved
spendinghisnightsplayingonlinegameswithfriendsfromwork.
Jacobwasthecompany’sauthorityoncustomerrelationshipmanagement(CRM).He
hadbeenworkingonCRMsolutionsformorethan20yearsandwasfamiliarwithevery
systembothinsideandoutsidethecompany.Jacobwasnotmuchofasocialcreature—he
preferredtofocusonhiswork.Inhissparetime,hebuiltLegorobotswithhisyoungson.
ThesolutionsthatJacobdeliveredwerealwayshighinquality—customersravedabout
howstableandeasytomaintainhissolutionswere.Jacobwashappyinhisroutineand
enjoyedtherhythmofhisownregularschedule.
BothRussandJacobwereabouttoexperiencethestressthatoftenaccompanies
change.TheyweretappedtoworkonanewCRMtoolforaclient.Itwasacomplex
projectthatwouldrequireallofJacob’sexperienceandallofRuss’screativecoding.
Atfirst,thingswentwell.RussandJacobassembledacoreteamandhiredtheirteam
consultants(seeChapter3,“UsingTeamConsultantstoOptimizeTeamPerformance,”for
moreonteamconsultants).Theysetastartdate,July15,ameretwoweeksaway.The
productowner,Sally,workedfeverishlytogetthebacklogestablishedwhileRussand
Jacobworkedwiththeteamtoestablishengineeringpractices,getthelabandmachines
setup,andcreategeneralteamrulesofengagement.Letitia,theScrumMaster,workedto
scheduletherecurringmeetings:thesprintplanningmeetings,dailyscrums,andcustomer
reviewmeetings.Everyonefeltconfidentgoingintothefirstsprint.
OnJuly15,theteammembersassembledforitssprintplanningmeeting.Therewasthe
normalambiguitygoingonintheplanningmeeting,butnothingalarming.Thiswas,after
all,ateamoffairlyexperiencedpeople.
Thenextday,theteamassembledintheteamareatobeginworkontheirchosentasks;
everyoneexceptRuss,thatis.Jacob,whohadarrivedearlysohewouldhavetimetogo
oversomethingswithRuss,wasvisiblyfrustrated.Helookedattheclock—8:30—thenat
everyoneelsebusyatwork.“Russissupposedtobeoneoftheleaders,”hethought.“How
canhesetsuchapoorexample?”Jacobstarteddoingsomeprojectworkwithoneother
teammemberandlosttrackoftime.Justbefore10:00,Jacobheardthestirringoftheother
teammembers.“Timeforthescheduleddailyscrum,”hethought.AsJacoblookedup,
Russpoppedin,freshoutoftheshowerafteralongbikeride.
“Heyguys!It’sdailyscrumtime!”exclaimedRuss.
“Gladyoucouldjoinus.Ithoughtyouweretakingthedayoff,”saidJacob,withahint
ofsarcasminhisvoice.
“Nope,juststartingthedaywithamorningbikeride,”respondedRuss,notpickingup
onJacob’ssarcasm.
Letitiadidpickuponthesarcasm,butchosetoignoreit.Themeetingwentoffwithout
ahitch.Itranjustahairover15minutes,notbadforthefirsttime.
Overthenextcoupleofweeks,thesamepatternoccurred.Jacobgottoworkat7:30and
workedwithanotherearly-risingteammember.Russ,ontheotherhand,poppedinright
beforethedailyscrum.Onemorning,asRussbouncedin,Jacobfinallysnapped.
“ThedayisalmosthalfoverRuss,”quippedJacob.
“Orhalfbegun,myfriend!IfeellikeIcouldclimbmountains.Igotupatfivethis
morning,wenttothegym,andthenloggedsixty-fivemilesonthebike.Whatagreat
day!”saidRuss.Heturnedhisattentiontotheotherteammemberswhohadgatheredfor
thedailyscrum.“I’mreadytorock.I’llgofirst!”andoffwentRuss,sharinghis
accomplishmentsandplans.
Jacobwentnext.“YesterdayIcompletedtaskeighteen,‘Createthestoredprocedure.’
TodayI’dliketoworkwithRussontheassembler,providedheisactuallyworkingtoday
andnotleavingearlyorsomething.”
“Sure,”saidRuss,laughingatwhatheconsideredJacob’squirkysenseofhumor.“We
canworkonthattogether;that’dbegreat!”hesaidenergetically.
LetitiabegantoobserveRussandJacobastheyworkedtogether.Itwasobvioustoher
thatRusscomingintoworkjustbeforethedailyscrumbotheredJacob.Infact,itseemed
tobotherhimthattheentireteamwasnotthereat7:30a.m.whenhearrived.Asthemost
seniormemberontheteam,bothfromanagestandpointandfromanexperience
standpoint,heseemedtofeelthathewastheonewhoshouldleadtheteamandset
boundaries,suchaswhoarriveswhen.Letitiaknewtherewasboundtobeablow-up
soon.Sureenough,asRussandJacobwerecoding,Jacobbegantogetirritated.
“Russ!Whyareyoucallingtheinterfacelikethis?You’renotanidiot;whyareyou
codinglikeone?Nowweneedtorewriteeverything!”saidJacob,throwinghispencilat
thescreenindisgust.
Russwasastoundedatthesuddenoutburst.“Jacob,calmdown.What’stheproblem?
Yellingatmewon’thelp.”
“Thisisjustlaziness,pureandsimple.Though,whatelsecouldIexpectfromsomeone
likeyou.Howhaveyoulastedheresolong?”Jacobroared,asherolledhischairbackand
threwuphisarmsindisgust.
Letitiaquicklysteppedin.
“Jacob,I’mhearingsomeverypersonalattacks.Whatisgoingon?”
“Russislazyandsloppy.Justlookatthiscode!”saidJacob.“Ican’tworklikethis.”
“Okay,Jacob.Let’sgoforawalkandtalkaboutit.”LetitiatookJacobdowntothe
localcoffeeshop.ShehadanideaofwhatwascausingJacob’sdiscontentbutneededto
hearitfromhim,one-on-one.
“Jacob,whatisreallybotheringyou?”Letitiaaskedhimdirectly.“Whenwestarted,
youwereexcitedaboutthisproject,butnowyouseemsoagitatedanddistressed,”she
added,gesturingtoJacob’sshakinghands.
“It’sRuss.Ican’tworkwithhim.He’sclearlynotcommittedtothisproject.Ican’tdo
alltheworkonmyown,andIdon’tknowwhattodoaboutit,”admittedJacob.
“Whatmakesyouthinkthatheisnotcommitted?”askedLetitia.
“Firstofall,hecomesinatteno’clock!Igethereatseven-thirtyandhegetshereat
ten?What’sthatabout?It’snotlongafterhesauntersinthatheleavesforlunch—‘Gotta
fillup,man’—suchabadexample,”quippedJacob.“Thatcodetodayshowsthathe’s
obviouslyrushingthrough,readytoleaveearlysohecangettothenextparty.Whendoes
hework?”
LetitiawashappytohearJacobspellouttheproblemshehadbeennoticing.
“Ithinkyou’vegotthewrongidea,Jacob.True,Russcomesinatteno’clock.Buthe
doesn’tleaveuntilafterseven,wellafteryouandtheotherteammembershavecalledit
quits.He’sonadifferentschedulethanyou,butheisworking.Still,Icanseethatthisisa
problem,notonlyforyoubutfortherestoftheteamaswell.”
“Soyou’llhavehimcomeinearlier,then?”askedJacob,stillhopingtokeepthingson
histerms.
“Ihaveadifferentidea,”Letitiareplied.“Onmylastproject,wehadasimilarissue,
andwesolveditbysettingcoreteamhours.”
“Coreteamhours?”askedJacob.
“Yes,coreteamhours—thesearethehourswheneveryonecancommittobeinginthe
teamspace.Thatway,ifweneedtoschedulepairingorifweneedtotalktosomeone,we
can.”explainedLetitia.“Ifyou’rewilling,wecangoproposeittotheteamnow.Ithinkit
willhelpeaseanyconcernsthatmightbelingeringaboutpeople’savailabilityand
dedication.”
Jacobthoughtaboutit.“Wecantryit.Butdomeafavor.KeepyoureyeonRuss.I’m
stillnotconvincedhe’sfullycommittedtothisproject,”saidJacob.
Letitiasmiled.“I’lldothat,Jacob.Butdomeafavor.Givehimachance.Andgivethe
ideathatpeoplecanhavedifferentschedulesandstillworktogetherachance,too.Ithink
you’llbepleasantlysurprised.”
TheModel
Letitiawasaveryin-tuneScrumMaster.ShequicklynoticedthefrustrationthatJacobwas
havingbutdidtherightthingbykeepingquietuntilitwasobviousthattheteamwas
readyforhertodosomethingaboutit.
ColocatedTeams
Ifyourteamisstrugglingwithdifferentschedules,orifyouareatalldistributed,acore
schedulehelpstorestoreasenseofbalanceandcohesiveness.Togetstarted,write
everyone’snameacrossthetopofthewhiteboard(I’llusethenamesfromthischapter’s
storyasanexample).Downtheleftside,write“startofday,”“lunchstart,”“lunchend,”
and“endofday.”Intheend,you’llhaveachartthatlookssomethingliketheoneinTable
10-1.
TABLE10-1ABlankCoreHoursChart
Fromthere,theteammembersneedtoindicatewhentheyliketostart,takelunch,and
leavefortheday.Youcancollectthisinformationinoneoftwoways.Optiononeisto
havetheteammembersfillinthechartdirectlyonthewhiteboard.Theonlydrawbackto
thatoptionisthatpeoplemightbeinfluencedbyothers’answersandbetemptedtoadjust
theirschedulestomorecloselymatchtheotherteammembers’schedules.Optiontwoisto
haveeachteammemberwritedownhisorherscheduleprivatelyandgiveittoyou.Then,
onceallthescheduleshavebeensubmitted,youcanpopulatethechartwiththeiranswers.
TheresultingchartwilllookmuchliketheoneinTable10-2.
TABLE10-2ThePopulatedChart
Oncethechartisfilledin,youcanextrapolatethecoreteamhours,thehoursthatall
teammembersareavailable,ontheboard.Ifyouweretodothatusingtheinformationin
Table10-2,you’dcomeupwiththefollowingcorehours:
Morning:10:00–11:30(1.5hoursofsharedtime)
Lunch:11:30–1:00
Afternoon:1:00–4:30(3.5hoursofsharedtime)
Thechartshowsthattheteamhasfivehourswhentheyareallavailableforwork
together.Thenextthingtodoisasktheteamwhethertherearesmallwaysinwhichthis
chartcouldbeoptimized.Forinstance,thelunchhouriscurrentlyanhourandahalf.If
RussandSam,however,weretodelaytheirlunchstarttimesby30minutes,itwouldgive
theteammorecorehours.ThenewmodelwouldlooklikewhatisshowninTable10-3.
TABLE10-3TheRevisedCoreHours
Theresultingcorehourswouldbeasfollows:
Morning:10:00–12:00(2hoursofsharedtime)
Lunch:12:00–1:00
Afternoon:1:00–4:30(3.5hoursofsharedtime)
Thissmalltweakwouldgivetheteam5.5hoursofworkingtimetogether.
Onceyouhaveestablishedyourcoreteamhours,itisbesttopublishthembyposting
themonawall,makingsmallindexcardsforteammemberstopostontheirmonitors,or
byputtingthemintothegroupcalendar.Anysoloworkcanbedoneoutsidecoreteam
hours;theworkthatrequirespairingorcollaborationshouldbedoneduringcoreteam
hours.
DistributedandPart-TimeTeams
Notallteamshavetheluxuryofhavingafullydedicatedteamworkingtogetherinateam
workspace.Whileteamsstriveforthis,sometimesitjustdoesnotworkoutthatway.The
approachlistedabovecanbeappliedtoteamsthatareeitherdistributedorhavepart-time
people—orboth.
DistributedTeams
Ifyouhaveadistributedteam,dothesameexerciseasabove,butdoitremotely.As
ScrumMaster,hostavirtualonlineconferencingsession.Havepeopleemailtheir
preferredstart,lunch,andendtimesbeforethescheduledmeetingstarttime.Theninsert
thedataintoaspreadsheettosharewiththeteamduringtheonlineconference.Picka
commontimezonereference.InTable10-4,IchoseCentral.Thenmapoutthetimes,
listingstartandendtimesinboththecommontimezoneandtheprimarytimezonefor
thatindividual.
TABLE10-4TimeMap
Basedonthisinformation,thesampledistributedtotheteamshowninTable10-4has
somebigtimegapstodiscuss.Rockylikestostarthismorningat11:30a.m.Centraltime,
thoughit’sonly9:30inhistimezone.SamalsogetsalatestartinCentraltime(10:30
a.m.),thoughagain,forhimit’sonly8:30.Therestoftheteamstartswithinanhouranda
halfofeachother.Theendofthedayisnobetter.Herearetheircorehours:
Morning:Thereareactuallynomorninghourswhenthewholeteamisavailable.
Russ,Jacob,andMikaelahavesharedmorninghoursbetween9:00and10:30a.m.,
Centraltime.SamandRockyarebothavailablewhilemostoftheteamisatlunch,
between11:30and1:30Central.
Lunch:10:30–3:00
Afternoon:3:00–4:30
Thisgivestheteamonly1.5totalhoursofsharedtime.Notgood.Someoftheproblems
couldbemitigatedbyhavingSamandRockypairandcollaboratetogetherinthemorning
separatelyfromtherestoftheteam,whointurnareavailableforeachotherinthe
morninghours.Theafternoonhourscouldbedesignatedforwholeteamactivities.
Imagine,though,iftheteamhadteammembersaroundtheglobe!Thismodelmakes
thechallengesfacingdistributedteamsveryvisible,especiallywhentheteamhasastrong
desiretoworktogether.Ininstanceslikethis,wherepeoplearemanytimezonesapart,
oneviableoptionistocreatesmallworkingteamslocally,eveniftheyarestilldistributed.
Thisensuresthatatleastsometeammembersaresharingcommonworkhours.
Part-TimeTeamMembers
Inthecaseofpart-timeteammembers,ratherthanlistthestartandendtimesdownthe
leftside,listthedaysoftheweek(seeTable10-5).Unlessateamchangesitshoursfrom
sprinttosprint,you’llonlyhavetodothisonceatthebeginningoftheproject,updatingit
onlyifsomeone’shourschange.Oncethetableispopulated,youcandeterminecore
hoursforeachdayoftheweek.Inthisexample,onMonday,theteamhasnotruecore
teamhours,asJacobcanonlyworkintheafternoonandSamcanonlyworkinthe
morning.
TABLE10-5TimeMapforPart-TimeTeamMembers
KeystoSuccess
Yousawinthischapter’sstorythatJacobwasveryunhappywithwhatheperceivedas
Russ’sslackerattitude.IfLetitiahadletthisproblemcontinue,theteamcouldhavesplit
apart,justoversomethingassimpleaswhentostarttheday.Payingattentiontolittle
detailslikethiscanhelppreventbigteamissuesdowntheroad.Whendeterminingthe
team’scorehours,it’simportanttokeepafewthingsinmind.
First,makesuretheteamunderstandswhyhavingcorehoursisimportantandthatthe
teamhasacollectiveagreementaboutpreferredworkhoursandworkhabits.Whetherthe
teamiscolocatedordistributed,workingwitheachotherdailyhelpskeepthingsincheck
andcommunicationflowing.Peoplewhoworktogetherhavetheabilitytolearneach
other’sbehaviors,skills,andpatterns.Astheygettoknoweachotherpersonallyand
professionally,theybegintogelasateam.Workingtogetheralsoprovidespeoplethe
opportunitytolearnmoreabouttheproject,astheyspendmoretimeworkingonpartsof
thesystemtheymaynotordinarilytouch.
Second,establishcorehoursatthebeginningofeachprojectandregularlythroughout
theprojectasneeded.JacobwouldnothavebeenasupsetifhehadknownRuss’swork
patterns;thoughRusstendstoarrivelate,healsostayslate.Thesepatternsmaychange
throughoutthecourseoftheproject—school,work,weather,andholidayscanallimpact
availability—soitisimportantfortheteammemberstocommunicatewhentheirhours
changedramatically.Iftheteamisdistributedorispart-time,andcorehoursvary
dramaticallyfromsprinttosprint,thisactivitymayneedtobedoneasoftenasatthe
beginningofeverysprint.
Last,encouragetheteamtoremainopenandwillingtobend.Everyonemaynotbeable
toworktheexactscheduleheorshewouldprefer.Remindtheteamthatthebenefitsof
workingtogetherfaroutweighthepersonalhabitsofoneperson.
Establishingcorehoursisasmalldetailthatcanhaveabigimpactonteam
performance.Calculatehowmanyhoursyouarespendingasacolocatedteamandwork
togethertomaximizethosecorehourswhilemaintainingindividualteammembers’
flexibility.Chartingcorehoursisasmallbutpowerfultoolthatcanhelpyoubuildahigh-
performingteam.
Chapter11.ReleasePlanning
Themostfrequentlyaskedquestiononanyprojectmightbe,“Whenwillyoubedone?”or
“Willyouhaveitalldonebyacertaindate?”Allteamshaveexperiencedtimeswhentheir
projectionshavebeenwayoff,andthey’vesufferedasaresult.Scrumpromisesamore
adaptableplanningscenario,onethatallowsforandindeedexpectschange.Butdoesthat
promisemeanthatteamsdon’thaveto—orcan’t—doreleaseplanninganymore?
Fortheanswer,followBrianandStephen,bothrelativelynewtoScrumand
approachingitfromdifferentpointsofview,astheyexplorereleaseplanningwithScrum.
TheStory
BrianwasanexperiencedprojectmanagerandnewScrumMasterabouttostarthisfirst
Scrumproject.HeandhisteammembershadbeentoanintroductoryworkshoponScrum
andwerewellversedintheby-the-bookanswersonhowtogetstartedandwhattodo.
Brianwasexcitedaboutthepossibilities.
Until,thatis,hegotavisitfromhiscorporatevicepresident,Stephen.
“Bytheway,Brian,”Stephensaidashepassedbyhisworkspace.“I’mgoingtoneeda
releaseplanassoonaspossible.”
Brianfrowned.Releaseplan?hethoughttohimself.Isn’tthatarelicfromthewaterfall
days?IthoughtthatwhenIswitchedtoScrumIcouldleaveGanttchartsbacktherewith
brokenpromisesandhalf-formedhandoffs.
Outloud,however,hesaidonly,“ButStephen,we’reusingScrum.Ican’ttellyou
exactlywhenwe’llbedone.That’sthewholepoint.”
ThatanswerstoppedStepheninhistracks.Hecameinandplantedhimselfinthechair
byBrian’sdesk.“Buddy,weneedtotalk.WhatmakesyouthinkthatwithScrumyou
don’tneedareleaseplan?”
Brianleanedforwardexcitedly.Hewasreadyforthisone.“Simple.Theproblemwith
alltheprojectsI’vebeenoninthepastisyouthinkyou’reontrackandthenrealizeatthe
veryendthatyou’renotgoingtomakeit.Thatmakeseveryoneanxiousandcausesusto
rushtomeetdeadlines,whichaswebothknowjustmeanspoorqualityandburnedout
teams.WithScrum,ifyoudon’tgettosomething,itjustdoesn’tgetdone.That’sthe
beautyofpriorityorder.”
“I’mnotsurethat’sexactlyright,”saidStephen.
“Well,rememberthatlastdisastrousprojectIwason?”counteredBrian.“Theteamand
Iworkedsohardandsowell,andthenattheendeverythingjustkindofcaughtuptous.
Suddenlywerealizedweweretoofarbehindtocatchup.Ifinallygotitafterthatproject.
Wecan’ttrustourplans.Theyarejustmilestones,dates,andsoftthingsthatwemakeup
togiveeveryoneasenseofsecurity,butintheendtheyarejustabunchoflies.”
“Ican’tdisagreewithyouaboutthat,Brian,”saidStephen.“ButScrumdoesn’tgive
youafreepassonreleaseplanning.Istillhavetoknowandbeabletocommunicatetomy
bosseswhenwe’llbedone.Ineedareleaseplan.”
“Well,Stephen,Icouldtellyouadatewhentheentireproductbacklogwillbedone,but
webothknowitwon’tbeaccurate,”Briansaid.“I’mtryingtothinkhowtoexplainitto
you.HowmuchdoyouknowaboutScrum?”
“Iwentthroughthesamecourseyoudid—justonadifferentdate,”answeredStephen.
“Infact,I’mtheonewhosuggestedweuseScrumandwhobroughtthetrainerintohelp
usgetstarted,”repliedStephen.
“Great,”saidBrianwithasighofrelief.“Thenyouknowthatwedeliverworking
softwareeverysprintandhowtheframeworkisstructured.Soyouunderstandwhyit
wouldbeirresponsibleofmetogiveyouanexactdatewhenwe’llbedone.”
“No,IthinkitwouldbeirresponsibleofyoutothinkthatScrumgivesyoufreereignto
justdowhatyouwanttodowhenyouwanttodoit,”challengedStephen.“I’vegotto
answertoourexecutiveVPandthentheCEO.Wehaveplansthatweneedtomakein
advanceastowhenthingswillbeinthemarket.”
“Stephen’sright,youknow,”cameavoicefromtheirright.ItwasCarol,aScrum
consultanttheyhadhiredtohelpthemintheirtransition.“Sorry.Ididn’tmeanto
eavesdrop,butIwasfascinatedbytheconversation,”sheaddedasshewalkedcloserto
them.
“Youcan—andshould—stilldoreleaseplanningwithScrum.ButBrian’sright,too,
Stephen.Scrumdoestreatreleaseplanningdifferently.”
CarolwalkedovertothewhiteboardthathungnearBrian’sdesk.
“You’vebothheardofvelocity,right?”shesaid,asshebeganwritingontheboard.
Theybothnodded.
“It’sthemeasureofhowmuchworkateamcangetdoneovertime,usuallyexpressed
instorypoints,”saidBrianconfidently.
“Exactly,”saidCarol.“So,let’slikenittotheodometerinyourcar.Velocityislikea
sprintodometer:Ittrackshowmuchgroundyoucoverintermsofstorypoints,perdayor
persprint.So,ifinthreesprintsyoutraveled1,000miles,900miles,and1,100miles,
you’dhaveaveraged1,000milesasprint.Putadifferentway,youcantravelabout100
milesperdayandcoverabouta1,000milesinasprint,sincewe’redoingtwo-week
sprints.Withmesofar?”
“Yep,”saidBrianandStepheninunison.
“Buthowdoesthatgetusclosertoknowingwhatdayyou’llbedone?”askedStephen.
“Bearwithme.There’samethodtomymadness,”Carolsaidsmiling.“Sowehavea
velocitymeasure.Howelsecanwetellwhereweareonourlittletrip?Well,backtomy
carmetaphor,we’dlookoutthewindowandseewhatwesee.Inasprint,wedothatevery
dayinourdailyscrum.Wepullourselvesupfromourwork—ourdriving,ifyouwill—
andlookaroundtoseehowwe’redoing.”
BrianandStephennodded.
“So.Wecantravel1,000mileseverytwoweeks,andwemakesurethosemilesare
takingusintherightdirectionwithourdailyscrums—andatahigher-levelview,withour
reviewsandretrospectives.Butourfinalmeasureisaburndownchart.Thespeedometeris
theslopeoftheburndown,”saidCarolasshecontinuedtosketchontheboard.
“Theburndowncharttellsusifouraveragespeedissuchthatwecanexpecttocover
our1,000miles.Wemightnotgoexactly100mileseveryday,right?Butaslongasour
speedissuchthatweaverage100milesaday,we’regoingtoreachour1,000-milegoal,”
Carolfinished.
TheteamlookedconfusedasCarolcontinued.“Sowhere’sthereleaseplan,right?The
releaseplanisourroadmap,markedwithstoppingpoints—waypoints—alongtheway.”
Figure11-1summarizesthemetaphor.
FIGURE11-1Indicatorsusingthetravelingcarmetaphor
“So,Brian,”saidCarol.“Stickingwithourcartripmetaphor,let’ssaywewanttogo
fromSanDiego,California,toJacksonville,Florida.You’resittingatacomputer.Pullup
amap.Howfaristhat?”
“Well,itdependsonwhichwaywego.Buttheshortestis2,339miles.Thatincludesa
stopinJuarez,though,andIreallywantedtostopinAlbuquerque,”Brianmusedoutloud.
“So,I’lljustdragtherouteandgothroughAlbuquerqueinstead,givingus2,413miles,
butputtingusinmuchmorepopulatedareasalongtheway:Phoenix,Albuquerque,
Amarillo,Dallas,thenHouston.”
Stephennodded,“So,ifwetravel1,000mileseverytwoweeks,we’relookingatabout
2.5sprintsorfiveweekstoJacksonville.That’sallareleaseplanhastobe.Itanswersthe
question,‘Atourcurrentrateandroute,howlongwillittaketogettheresafely?’Ifit’sthe
reverse—IneedyoutobeinJacksonvillebyacertaindate—you’dneedtodothemathto
seeifit’spossible.Youcoulddetermineifitisfeasibletodrivemorehoursinadayor
discussotherwaystomakethatarrivaldatepossible.Itoldyoureleaseplanningwas
possiblewithScrum.”
Brianfrowned.“Butwait.Thatsoundsgreatifyou’reactuallydrivingfromPointAto
PointB,butonourproject,wedon’talwaysknowexactlywherewe’regoingorhowfarit
isfromheretothere.Andevenifwedid,thisperfectlittleroadmapdoesn’ttakeinto
accountroadwork,whichmightslowusdown.Orallowforfoulweatherthatcausesusto
deviatefromourroute.Orfactorinagreatopportunitythatrequiresweincludeanout-of-
the-waystop,St.Louis,forinstance,toouroriginalplan.”
Carolintervened.“Well,that’swhereyou’rebothright.Releaseplansaregoodtohave
andworthwhilebecausetheygiveusapreliminaryroadmap—wecan’tjustsetoutwithno
directionandnoideaofwhenwe’lllikelyarrive.ButBrian’srightthatit’simportantfor
thestakeholderstounderstandthataplanisjustthat,aplan.Andtheplanwillchange.
Despiteourbesteffortstoestimatetheproductbacklog,wemightfindoutthattheactual
distancewehavetotraveltocompletetheprojectismuchmoreorlessthanwe
anticipated.Anddespiteourbesteffortstotrackandestimateourexpectedvelocity,we
mightrunintoproblemsthatslowusdown.Andmaybeweincludeaworst-casescenario
forissueslikethat.AndiftheydiscovergoldinSt.Louis,whichistheequivalentof
addingsomenew,undiscoveredstoriestotheproductbacklog,wecanincludeitinour
trip,butit’sgoingtocostintermsofeithertimeormoney,right?”
“Rememberourvisualindicators?”Carolcontinued.“Ourdailystandups,reviews,and
retrospectives?That’showwedetermineiftheteamisontherighttrack.Wedon’twantto
goallthewaytoJacksonvilleonlytorealizethatit’sthewrongplacetobe.Ifwewantto
gotoSt.LouisorLouisville,sobeit,butweneedtocommunicatetothecustomerswhat
therisksareandthat,shouldtheydecidetoincludeadetour,wejustcan’tmagically
reappearinJacksonvilleasoriginallyscheduledasifwenevertookaleftturn.Thetime
willhavebeenlostandthescheduleadjustedaccordingly.”
“Understood,”saidStephen.“Ithinkwe’vegotanideaofwhattodo.Thanksfor
stoppingbyandendingourstalemate.”
BrianandStephencontinuedtheirconversationwellpastlunch.Brianacknowledged
thatStephenneededaplan.HethenexplainedtoStephenhowheshouldexpectvarying
levelsoffunctionalityalongtheway,andwhygivingaharddateandacompletelistof
functionalitythisearlyintheprojectwouldbeirresponsiblebecausetherewasn’tenough
knownabouttheprojecttobeabletonaildownboththings.Intheend,Brianwasableto
giveStephenenoughinformationforhimtodohisstrategicworkwhiletheteamexecuted
theday-to-daytacticalwork.
ThismodeloffersadetailedlookattheprocessthatBrianproposed.
TheModel
Inthestory,bothStephenandBriansatthroughthesameScrumtrainingbutheardtwo
verydifferentthings.Stephendidn’tbelievethatusingScrumtoruntheprojectshould
affecthisreleaseplanning,whileBrianfeltthatScrum’sflexibilitymeanthewouldno
longerbetiedtoanunrealisticfixeddeadline.Inaway,theywerebothright.
Historically,thereleaseplanisoneofthedeliverablesthatcomesoutofatraditional
projectinitiationphase[PMI]andoftenmanifestsitselfinaGanttchart.Itslevelof
precisionisveryhigh.InaScrumproject,however,releaseplanningisquitedifferent.
ThereisnoGanttchartthattellsyouexactlywhenacertaintaskwillbedone.Instead,an
agilereleaseplanshows,onahigher,morestrategiclevel,whatrangeoffunctionalitywill
likelybedonebyacertaindate.Yourreleaseplanshouldbebuiltbeforethefirstsprint
andshouldberevisedcontinuallythroughouttheprojecttoreflectchangingprioritiesand
projectrealities.
Anagilereleaseplanonlyhastoincludethebest-andworst-casefinalreleasedates,but
manyplansalsoincludesmallerreleasesalongtheway,includingend-of-sprintreleases
andquarterlyreleases.Ihaveevenscheduleddailyreleaseswhenworkingonasustained
engineeringteamsupportingaproductionsystem.Whatmakessenseforyourreleaseplan
dependsonthespecificsofyourprojectandtherealityofyoursituation.
Iapproachreleaseplanninginaverystraightforwardway:
1.Sketchapreliminaryreleaseplan.
2.Establishdegreeofconfidence.
3.Revisetheplaneverysprint.
SketchaPreliminaryReleasePlan
Tobuildtheinitialreleaseplan,youstartwiththreeinputs:
1.Anestimated,ordered,andprioritizedproductbacklog
2.Theteamvelocity
3.End-of-sprintdates
Anestimated,ordered,andprioritizedproductbacklogisthefoundationoftherelease
planatthebeginningofandthroughouttheproject.Withoutawell-groomedbacklog,I
finditnearlyimpossibletobuildareleaseplan,letalonegettheprojectoutthedoor.This
doesnotmeantheproductbacklogshouldcontainonlysmall,manageablestoriesorthat
yourestimatesorprioritieswon’tchangeasyoulearnmoreabouttheproject.Butstory
pointestimatesandpriorityorderingareessentialforplanning,evenatthisearlystage.
Onceyouhaveanestimatedandprioritizedproductbacklog,youneedtoknowthe
team’svelocity.Itcaneitherbepulledfromhistoricaldataorestimated.Velocityshould
alwaysbeexpressedasarangeratherthanonesetnumber.(SeeChapter4,“Predicting
TeamVelocity,”formoreinformation.)
Next,youneedtoknowanappropriatesprintlengthforyourproject.(SeeChapter6,
“DeterminingSprintLength,”tohelpfigureoutyoursprintlength.)Fromyoursprint
length,youcaneasilycalculateallyourend-of-sprintdates.
Onceyouhavetheseinputs,youcansketchthepreliminaryreleaseplan.Forthis
example,imaginethefollowingistrue:
Theproductbackloghasbeenprioritizedandestimatedat200points.
Theteamhasdetermineditsvelocitytobe15to20pointspersprint.
Theteamisusingbimonthlysprints.Thefirstsprintendson30June.
First,usingthelowervelocityof15points,startatthetopoftheproductbacklog,count
down15points,anddrawaline.Thisrepresentsthefirstsprint,whichendson30June.
Repeatthisuntilyoureachtheendofeitheryourfunding(discussedbelow)oryour
projectduration,oruntilyoureachapointintheproductbacklogwherethestory
estimatesaretoobigtofitintoasprint(theyarelargerthantheplanned15-pointvelocity).
TheresultingreleaseplanatthisvelocityisshownontheleftsideofFigure11-2.
FIGURE11-2Mapestimatedvelocitytotheestimatedproductbacklog
Next,dothesameusingthehigher,20-point,velocity.Startatthetopoftheproduct
backlog,countdown20points,anddrawaline.Thisisyourfirstplannedsprintifthe
teammanagesahighervelocity;theenddateforthatsprintisstill30June.Repeatas
describedabove.Theresultingreleaseplanatthisvelocityisshownontherightsideof
Figure11-2.Noticethatwithavelocityof20,theteamwillcompletemoreproduct
backlogitemspersprint,andasaresult,delivermorefeaturesoverallinthesamenumber
ofsprints,whencomparedtothelowervelocityof15points.
Nowthatyouhavearoughpictureoftheinitialreleaseplan,youcanextrapolatesome
preliminaryreleasedatesforthehighestprioritystories.Keepinmind,though,thatthese
dateswilllikelychangeastheteamobtainsactualvelocitydataandlearnsmoreaboutthe
productitisbuilding.Thepreliminaryreleaseplanisaneducatedguess,onethattheteam
willrefineasitlearnsmoreabouttheproduct,theproject,anditself.
ConsiderDegreeofConfidence
Atthisstage,thereleaseplanisnotyetreadytobeshared.Youstillneedtoaddtheteam’s
degreeofconfidence.Inthebeginningofanyproject,theteamwillhavevaryingdegrees
ofconfidencewhenitcomestoitsconfidenceintheestimatesitputforthintheproduct
backlogaswellasitsownvelocityestimates.Theteamhasahigherdegreeofconfidence
inthestorypointestimatesforthestoriesthatarenearthetop—thosethatwillhappenin
thefirstfewsprints.That’sbecausethesestoriestendtobesmallerandtheteamhasa
deeperunderstandingofitsdirectionanditsownapproachtothosestories.Becausethese
storiesarebetterunderstood,theteam’sconfidenceincompletingthemshouldbehigher
aswell(80to100percent).
However,asthestoriesdropinpriority,theytendtogetbiggerandmoreambiguous.As
aresult,theteam’sconfidenceintheestimatesforthosestoriesalsodrops.Bythetimethe
teamreachesthebottomoftheproductbacklog,itreallyhasnoideahowitplansto
implementthestoriesorhowmuchworkisinvolved,sotheteamcan’tsaywithahigh
degreeofconfidence,oranyconfidenceforthatmatter,howmucheffortthestorieswill
take.Theteam’sconfidenceincompletingthem,therefore,isverylow,10to25percent.
Asmentionedpreviously,thedegreeofconfidenceisalsoaffectedbyhowcertainthe
teamisinitsownvelocityestimates.SometeamshaveworkedtogetheronsimilarScrum
projectsbefore,somembersarefairlyconfidentinthevelocityrangestheyhaveputforth.
Equallycommon,though,areteamsthatarerelativelynewtoScrumandhavenotyet
workedtogether.Theirinitialvelocityestimatesmaybewayoff,resultinginhugedegrees
ofvariance.Thesenewteamsalsowouldhaveamuchlowerdegreeofconfidencein
completingallofthestoriesatthelowerestimatedvelocity,andanevenlowerdegreeof
confidenceincompletingallofthestoriesatthehigherestimatedvelocity.
Figure11-3showsthecompletepreliminaryreleaseplan,withtheteam’sconfidence
rangeadded.Theteamisfairlyconfident(80percent)thatitcancompletethemajorityof
storiesthatarecontainedinsidethetenplannedsprintsata15-point-per-sprintvelocity
(abovetheminimumvelocityline).Theteamisnotnearlyasconfidentthatitcan
completemostofthestoriesthatarecontainedinsidethetenplannedsprintsata20-point-
per-sprintvelocity(abovethemaximumvelocityline).Notonlyistheteamunsureabout
completingtheselower-prioritystoriesatitsestimatedvelocity,thesestoriesarealsobig
andambiguousrelativetothestoriesatthetopoftheproductbacklog.Theteamisabout
50percentconfidentinthestoriesthatfallbetweentheminimumandmaximumlines—
thosemightgetdoneifthingsgobetterthanplanned,butnooneshouldcountonhaving
them.
FIGURE11-3Addthedegreeofconfidencetothereleaseplan
Teamsneedtostartworkingasquicklyaspossible,sotheycangatherrealvelocity
data,createworkingsoftware,andbegintobetterunderstandthemoreambiguousitemsin
theproductbacklog.Doingsogiveseveryonerealinformationwithwhichtoadjustthe
releaseplan.Nomatterhowconfidenttheteamis,theinitialreleaseplanisonlyan
educatedguess.Itshouldbeusedtoinformthebacklogordering(ifastoryorsetof
storiesabsolutelyneedstobedonebyaspecificdate)ortohelpstarttheconversation
aboutwhatrangeoffunctionalitycanbeexpectedatasetbudget.Itshouldnotbeusedto
holdtheteamtoanunrealisticsetofdatesoranunattainablerangeoffunctionality.The
preliminaryreleaseplanisroughandfluid—andpurposefullydesignedtoadapttoproject
realitiesthroughouttheprojectlifecycle,asisdiscussedinthenextsection.
MaintainingtheReleasePlanthroughouttheProject
Scrumteamsknowthatmostprojectsdon’tmovefromPointAtoPointBataconstant
rateofspeedwithnodetoursormishaps.That’swhytheyupdatetheirreleaseplansto
reflectrealityattheendofeverysprint.Theteamsadjusttheirprojectionstoincludethe
actualaverageteamvelocitiestodateandtheknowledgetheyhavegainedfromcreating
workingsoftwareandreceivingfeedbackfromtheircustomersandproductowners.
Intheexampleinitialreleaseroadmap,theassumed/guessed/calculatedvelocitywas15
to20storypointspersprint(refertoFigure11-3).Suppose,though,thatinthefirstsprint,
theteamaccomplished11storypointsratherthanthe15to20ithadestimated.Theteam
updatedthereleaseplantoreflectthisactualvelocity,thentheteamranfivemoresprints,
eachtimeupdatingthereleaseplan.Bytheendofsprint6,theteamhadanobserved
velocityrangeof8to14points,withanaverageof11.Thoughnooneissurprisedbythe
lowervelocity,theteamandproductownerdoneedtoupdatethereleaseplantoreflect
thedropinvelocity.
TherevisedreleaseroadmapfortheremainderoftheprojectisshowninFigure11-4.
FIGURE11-4Updatereleaseplantoreflectcurrentreality
Noticethatintherevisedplan,theteamisusingitsnewvelocityrangeof8and14
pointspersprintwithanaverageof11.Asbefore,everythingbetweentheobserved
minimumandmaximumobservedvelocitiesisaquestionableareafortheteam.Theteam
shouldgeteverythingdoneaboveitsobservedminimumvelocityof8,willbe50/50
betweentheminandmaxobservedvelocities,andhasverylowconfidenceforanything
belowthemaximumobservedvelocityof14.
ProjectCost
Manyteamshaveprojectsthatareboundbymoney—fixedcost.Assumethattheproject
intheexampleisalsoafixed-costproject,withatotalbudgetof$200,000.Whenthe
moneyisgone,theprojectwillend.Assume,too,thattheteamburnsabout$20,000per
sprint,whichmeansthattheteamcanrunapproximatelytentwo-weeksprintsbeforethe
projectisoutofmoney.
Intheexample,theteam’sinitialplanassumedavelocityof15to20pointspersprint.
Atthisrate,theteamcouldhaveaccomplishedtheentireproductbacklog(10sprints×20
pointspersprint=200points,theestimatedtotaloftheproductbacklog).However,the
actualvelocitiesatabitmorethanhalfwaythroughtheprojectwererunningbetween8
and14,withanaverageof11.Thismeansthatiffuturesprintsfollowthissamepattern,
theteamwillaccomplish110pointsworthofstoriesintheproductbacklog,leavingthe
restunaccomplished(11points×10sprints).
Theteamcaneasilycommunicatethepointinthebacklogatwhichthemoneywill
likelyrunoutbyaddingthatitemtothereleaseplan,asshowninFigure11-5.
FIGURE11-5Releaseplanupdatedtoshowwhentheprojectbudgetmightrunout
DeterminingtheEndGame
Allagilepracticesfocusondeliveringvaluetocustomersearly.Sometimesthismeans
thattheprojectendsearly.Sometimesitmeansdeliveringsomewhereintheoriginal
estimatedvelocityrange.Othertimes,it’sclearafterthefirstfewsprintsthattheproject
willrunoutofmoneybeforeitrunsoutofstories.Eachofthesescenariosismanageable
withintheScrumframeworkandwithanagilereleaseplan.
Releasingaprojectthatmeetseveryone’sexpectationsunderbudgetandearlierthan
expectedisagoodgoal.Thekeytodoingsoistobuildonlywhatisneededtoaccomplish
thebusinessobjectives,withoutalltheextragoldplatingandunnecessaryfunctionality
thatisrarelyorneverused.(Chapter30,“WritingContracts,”offerstwomethodsthatyou
canusetoreleaseyoursystemearlyandunderbudget.)
Mostoften,however,evenhigh-performingteamsreleasesomewherebetweentheir
minimumandmaximumvelocityestimates.Theydeliverenoughfunctionalitytomeetthe
needsofthebusiness,asplanned,bythetimetheprojectrunsoutofmoney.Theyhitthe
arrowdead-centeronthetarget.Everyoneishappy.
Thethirdoutcome,andtheoneIseemostfrequently,isareleaseplanthatpredictsthat
thebudgetwillrunoutlongbeforethestoriesdo.Thestakeholderslookattheplanand
decidethatthereisalargegapbetweenthebest-casescenarioandthetruebusinessvalue
thecustomersareaskingfor.Whenthathappens,youhaveseveraloptionsforsolvingthe
problem:
Increasevelocitybybuildingahigher-performingteam.Perhapsthereare
impedimentsinthewayortheteamneedstoimplementbettertechnicalpractices.
Increasethebudget,resultinginmoreoftheproductbacklogbeingexecuted.
Dropdesiredfunctionality.Reevaluatewhatistrulyessentialandwhatisdesirable.
Someitemsarelowerinprioritythanothersforareason.Thoughacertainfeature
mightnotmakeitintothisrelease,itmightbeacceptabletoincludeitinfuture
releases.
Regardlessofwhetheryoudeliverclosetoyourinitialbest-casescenarioorevenfalla
littleshortofyourinitialworst-casescenario,thegoodnewsis,withScrum,thecustomers
areintheloopfromdayone.Astheprojectprogresses,andthelargerstoriesbecome
clearerandreceivemoreaccurateestimates,theproductownerandstakeholdersmight
needtoconsiderwhethercertainthingscanbeleftundone,whattradeoffsneedtobe
made,orwhetherthescheduleshouldbepushedouttoaccommodatemorefunctionality.
TheadvantagetomakingthesetradeoffsinScrumisthattheyarebeingmadebasedon
softwarethatisworkingand,ideally,hascustomerandstakeholderfeedback.
Additionally,havingworkingsoftwaremeansthatlast-minuteglitchesandintegration
issuesaremuchlesslikelytoimpactthereleaselateintheschedule.
Theinitialtradeoffofhavinglessclarityinthereleaseplanactuallyprovidesan
opportunitytogivepeopleabetterproductorservice.Stakeholdersseeworkingsoftware
attheendofeachsprintandcanmakeadjustmentsbasedonwhattheysee.Theyaregiven
anupdatedreleaseplanaftereverysprint,sotheyknowwheretheprojectstandsinreal
time.Theyhaveunderstoodandbeenapartofthetradeoffsandhavehelpedguidethe
teamalongthepathofoptimaldelivery.Whentheproductisultimatelyreleased,there
shouldbenosurprises.
KeystoSuccess
Releaseplanningismeanttoprovideanswersastowhenspecificfunctionalityandthe
entireprojectwillbedone.Thoughtheseareoftenmovingtargets,withagilerelease
planning,teamsacknowledgethisfactupfrontandrevisetheirplansaccordingly.To
makeagilereleaseplansworkmosteffectively,youneedtodofivethings:communicate
upfrontandoften,updatethereleaseplanaftereverysprint,trytodothehighest-priority
itemsfirst,refineestimatesonbiggeritems,andcreateworkingsoftwareeverysprint.
CommunicateUpFrontandOften
Ilearnedthehardwaythatnomanagerorleaderinanycompanylikessurprises,nomatter
howwellintentioned.Yourreleaseplanshouldbeatopicfordiscussiononaregularbasis.
Peopleshouldbeinvitedtoaskquestionsaboutit,andyoushouldnotgetdefensive.Itisa
goodideatopostyourreleaseplanonthewallforeveryonetoseeandtorepostitevery
timethereisachange.
UpdatetheReleasePlanafterEverySprint
Scrumisallaboutchange.Yourplanwillchangefrequently,perhapsasoftenasevery
sprint,asbacklogitemsshiftinpriority,estimatesgetmorerefined,andvelocitybecomes
morereliableandpredictable.Itisimportanttoupdatethereleaseplaneverysprint.This
helpsstakeholdersknowwhattoexpectandwhentoexpectitbasedonthefeedbackthey
provideateachsprintreview.
TrytoDotheHighest-PriorityItemsFirst
Whiledoingthehighest-priorityitemsfirstmaysoundlikeano-brainer,it’soften
forgotten.Wheneverpossible,teamsshouldbeworkingonthehighest-priorityitemsfrom
theproductbacklogineachsprint.Attimes,thehighest-priorityitemmightnotbethe
itemtheteamisdyingtosinkitsteethinto,butunlessthereisacompellingreason
(dependencies,size,lackofunderstanding),it’stheonethatshouldbecompletedfirst.
RefineEstimatesonBiggerItems
Asitemsmoveuptheproductbackloginpriority,theyrequiremoreattentionsothatthey
canberefinedandsizedappropriatelyforasprint.Assuch,remembertohaveagrooming
meetingwiththeproductownerandtheteamtoreviewtheproductbacklogeverysprint.
Thisisaforward-lookingmeeting,wheretheteamandtheproductownerworktogetherto
refineandbreakdownstoriesintosmallchunksandestimatethem,andtore-estimate
otherstoriesthatmighthavegainedmoreclarity.
DeliverWorkingSoftware
OnScrumprojects,releaseplanningdoesnotdisappear;it’sjustapproacheddifferently.
Bycouchingestimatesinrangesandconfidencelevels,agilereleaseplansactually
providemoreinformationthanmoretraditionalreleaseplans.And,bycreatingan
adaptablereleaseplanthatnotonlyisdesignedforbutactuallyexpectschanges,your
stakeholdersandmanagershaveatransparentlookintohowtheprojectisdoing,where
it’sgoing,andwhattheycanexpecteverystepoftheway.
NoreleaseplanorGanttchartgivesthetruestatuslikeworkingsoftware.Working
softwareisyourbestinsuranceagainstlast-minuteproblemsderailingarelease.Nomatter
whatthereleaseplan,Ganttchart,ortaskboardsays,workingsoftwareistheonlytrue
measureofprojectprogress.Neverforgetthat.
Reference
[PMI]ProjectManagementInstitute.2008.GuidetotheProjectManagementBodyof
Knowledge,FourthEdition.UpperDarby,PA:ProjectManagementInstitute.
Chapter12.DecomposingStoriesandTasks
Writingstoriesandtasksoftherightsizecanmeanthedifferencebetweensucceeding
withScrumandfailingmiserably.It’sthatimportant.What’sworseisthatthesymptoms
ofwrong-sizedworkelements(notfinishing,poorestimation,painfuldailyscrums)don’t
atfirstseemrelatedtostoryandtasksizeatall.
Duringoneofmyfirstprojects,wefoundwewerehavingproblemscompletingour
workbytheendofthesprint.Ourfirstinclinationwastoblamethesprintlength—we
assumedalongersprintwouldsolveourproblems.Throughaserendipitousencounter,we
discoveredthatitwasactuallyourtasksandstories,ratherthanoursprints,thatneededto
beresized.
TheStory
OurswasateamthatwasnewtoeachotherandnewtoScrum.Thingsweregoingokay,
buttheteamandIwerebaffledastowhy,forthepastthreesprints,wehadneededto
returnahandfulofunfinisheditemsbacktotheproductbacklog.Wehaddiscussedthe
problematlengthbutwerenoclosertoananswer.Becauseitalwaysfeltlikeweneeded
onemoreweektotieupallthelooseends,extendingthelengthofoursprintstofive
weeksseemedtoustobetheperfectsolution.Atthesametime,though,we’dneverheard
ofsprintslongerthanonemonthsowewereunsureaboutwhattodo.
Luckilyforus,theannualAgileAllianceAgileconferencewasjustaroundthecorner.
Wehadproblemsweneededtosolveand,withenoughbegging,wereprettysurewecould
getmanagementtosendourteam.Afewweekslater,withplaneticketsandsprint
backlogsinhand,wewereofftotheconference.Ofalltheanswerswehadsetouttofind,
determiningthebestlengthforoursprintswasatthetopofourlist.
OnthefirstdayoftheconferenceIfoundmyselfinlineforcoffeewithRonJeffries.I
introducedmyselfandaskedhimifhe’dbewillingtogetlunchwiththeteam,ourtreat.
Heagreed.Itrackeddowntheteamandtoldthemthegoodnews.Everyonewasvery
excited.LaterthatdaywefoundourselvessittingwithRondiscussing,amongother
things,howlongoursprintsshouldbe.
Afterourinitialintroductionandagoodservingoffood,IstartedoffbytellingRon
aboutourteam,ourproject,andthehurdleswewerefacing.Ithenpulledoutaprioritized
listofitemswewantedtodiscussandstartedwithnumberone:thelengthofoursprints,
thetroublewewerehavingtryingtoaccomplishourworkinthirtydays,andhowwe
thoughtweshouldextendoursprintlength.
“Ron,wecannotgetourworkdoneinthirtydays.Therearealwaysstoriesleftundone.
We’vedecidedweneedtoextendthelengthofoursprintstofinishthesestories,butthe
bookssaynottogolongerthanonemonth.Whatdoyouthink?”Iasked.
“Ithinkyouneedtoshortenthelengthofyoursprints,”saidRon,bluntly.
I’msureRoncouldseetheteam’scollectivejawdropping.Ofalltheanswerswewere
expecting,thiswasnotoneofthem.
Mikechimedin,“Uh,Ron,Idon’tthinkIheardyoucorrectly—didyousaytoshorten
oursprints?”
“Yup,”saidRon.
“Imusthaveexplaineditwrong,”Isaid.“Yousee,theproblemis,oursprintsaretoo
shortalready.Wecan’tgetourworkdoneinthetimewehave.”
“No,Igetit.Butwhatmakesyouthinkthatyou’llbeabletogetanythingdoneinforty-
fivedaysorsixtydays?”askedRon.
“Simple.Wejustneedmoretime,”answeredJohn.
AtthispointRonaskedtoseeourmostrecentsprintbacklog.Weshowedhimour
backlog,aportionofwhichisillustratedhere.
RonpointedtotheestimateforCreateaConfigurationSettingsSingletonclass.
“Whyisthistaskeighthours?”askedRon.
“That’showlongitwilltake,”statedJohn,sincehewasfamiliarwiththework.
“Icandothatinfiveminutes.Whydoesittakeyoueighthours?”inquiredRon.
Johnsatthereforaminute,thinkingabouthisanswerandofalltheworkneededto
accomplishthetask.Therewasjusttoomuchtolistout.
“Ittakeseighthoursbecausethat’showlongittakes.Whyareyoupepperingmeon
this?”statedJohn,abruptly.
“Ijustwonderwhyafive-minutetaskformewilltakeyourteameighthours.How
manyotherfive-minutetasksareintherethatyouareallmaskinginhugenumbersof
hours?”askedRon.“It’snowonderyouguysthinkyouneedtoextendyoursprints;you
haveeasytasksthattakeforever!”
ItwaseasytoseethefrustrationquicklybuildinginJohn’sface.
“Ron,I’llhaveyouknowthatI’vebeenworkinginsoftwareformorethantwenty
years.IfIsaythat’showlongitwilltake,that’showlongitwilltake!”exclaimedJohn.
Ronsmiled.HehadJohnjustwherehewantedhim.
“Okay,John.You’reright.Tellme,whatdoesthistaskentail?Whatwillyoudoto
completeit?”askedRon.
Atthispoint,Johnstartedlistingnotonlyworkrelatedtothistask,likewritingtheclass
andmeetingthedonelistcriteria,butalsomanyadditionalworkitemsthatwerenot
capturedonthesprintbacklogatall.Bythetimehehadfinished,Johnhadlistedabouta
dozentasks,onlyoneofwhichwascapturedonthesprintbacklog.
“Okay,John,doyouseetheproblem?Youhavejustlistedadozentasks,eachoneof
whichshouldbelistedseparatelyonthesprintbacklogandshouldhaveanestimate.
Instead,you’veburiedthemallintothissingle,seeminglysimpleeight-hourtask,”stated
Ron.
Thelightbulbwentonforallofus.
“Iamthinkingofallthestuffwe’dhavetolistforthisDLLtask,”Isaid.“Howmanyof
ourtaskshaveotherworkitemsburiedinthem?”
“Lots,”whisperedMikeunderhisbreath.“Lots.”
“Yousee,guys,”explainedRon.“Yourproblemisn’tyoursprintlength.Youneedto
improveyourtaskdecomposition.”
Asweweresettlingintothistruth,Ronshockedusevenmore.
“So,likeIsuspected,tofixyourproblems,youneedtoshortenyoursprints.Doingthat
willforceyoutothinkcriticallyaboutyourworkanddecomposeyourtasksintosmall
units.Isuggestyoutryone-weeksprints,”saidRon.
“Aone…What?”saidJohn,verbalizingouramazementthatthiswouldevenbe
proposed.
Ronsmiledandnodded,“Oneweek.Trustme.”
Bytheendofthelunch,wewereconvinced.WelefttheconferenceandtookRon’s
advice,usingone-weeksprintstoforceustolearntodecomposework.Inthebeginning,
wefoundone-weeksprintstobeachallenge.Afterawhile,though,workinginshort
sprintsnotonlyforcedustodecomposeourtasksandourstories,italsomadeuslookat
theminanewlight.Astimewenton,wefoundourselvesaccomplishingmore,inless
time,thanweeverimaginedpossible.
TheModel
Ithappensallthetime.Scrumteamsfeelpainandimmediatelystartsecond-guessing
Scrum.Inthestoryabove,wecouldn’tfinishourstories,sowethoughtweneededlonger
sprints.Inreality,wecouldn’tfinishourstoriesbecausewewerehidingallsortsofwork
insideseeminglystraightforwardtasks.Thiscausedourestimatestobeoff,ourdaily
scrumstobepainful,oursprintstofeeltooshort,andallofustoexperienceanimmense
amountofpressure.Oursprintlengthwasn’ttheproblem.Ourabilitytodecomposework
was.
SettingtheStage
Tobetterunderstanddecomposition,firstexaminethedifferentelementsofworkthat
mightneedtobebrokendown.Theproductbacklogcontainsuserstories.Thesestories
typicallycomeinthreeflavors:epics,themes,andstories.Ateachsprintplanning
meeting,storiesarebrokendownintotasks,whicharestoredinthesprintbacklog.
LiketheLordoftheRings,anepicisvastinscope.“Redesigntheuserinterfacefor
thedocumenteditor”wouldbeanepic.
Athemeisacentralideaofthelargerepic.“Redesignthemenubarforthe
documenteditor,”mightbeathemeinsidethelargerepic,“Redesigntheuser
interface.”
Astoryisoneeventinatheme.“Redesigntheeditmenu”mightbeastoryinthe
“Redesignthemenubar”themeoftheepic“Redesigntheuserinterface”
[DICTIONARY.COM].
Tasksarethespecificactionsneededtoaccomplishastory.Theyaretheelements
thattogethermakeastorycompleteandpotentiallyshippable.
Figure12-1showstherelationshipbetweenstoriesandtasksandhowtheyfitinto
sprintsandreleases.
FIGURE12-1Task,story,sprint,andreleasemodel
Astorywillhavemultipletasks.Asprintwillhavemultiplestories.Areleasemight
havemultiplesprints.Isaymightbecausewhilemanyteamsdon’treleaseeverysprint,
someteamsactuallydo.That’swhyScrumusesthewordspotentiallyshippableto
describecompletedstories;thestoriestheoreticallycouldbereleasedattheendofthe
sprint,butitmightnotbeeconomicallypractical.
Atheme,ontheotherhand,istypicallytoobigtobecompletedinonesprintandmust
insteadbedoneoveracoupleofsprints.Anepicisevenbiggerandisoftenthesubjectof
severalsprintsorasinglerelease.
Inthestory,ourteamhadtasksthathadnotbeenbrokendownenoughandwerehiding
smallerassociatedtasks.Inmanycases,though,theproblemisthatthestoriesaretoobig
andhidelargeamountsofwork.Problemswithdecomposingworktypicallybeginatthe
storylevelandcancontinuethroughtothetasklevel.It’stoughtofindtherightbalance
betweentoomuchandnotenoughdecomposition.Thereisnoone-size-fits-allmodelout
there;everyteammustfinditsownbalance.
StoryDecomposition
I’llwalkthroughsomeworkdecompositionwithyou.Pretendthatyouarewritingthe
storiestocreateanaccountmanagementpageforanonlineretailer,Big’nHuge(see
Figure12-2).Whatarethethingsyoucandoonthispage?
FIGURE12-2Big‘nHuge’s“YourAccount”page
First,youcanmanageseveralfactorsofanaccount.So,turnthatintoastory:“Asa
BnHcustomer,Icanmanagemyaccount.”Thisisagoodstory,butbysimplylookingat
thescreenshowninFigure12-2,youcanseethatitistoohighleveltomeananything;it’s
anepic.Infact,itwilllikelybringmorequestionstoyourmindthananswers:
WhatcanIdotomanagemyaccount?
CanIupdatecreditcards?
CanImanagemyaddressbook?
Giventhatyoustillhaveallthesequestions(andmore),youneedtobreakthestory
downtosomethingmoremeaningful(seeFigure12-3),basedagainonthescreenshot
showninFigure12-2.
FIGURE12-3Initialstorybreakdown
Younowhavethreelargestories(Icanmanagemyorders,Icanmanagemypayment
options,andIcanmanagemysettings).Thesecanbeconsideredthemesthatrollintothe
largeepic,“Icanmanagemyaccount.”Still,youhavemorequestionsthananswers.So
youdecompose“Icanmanagemypaymentoptions”somemore(seeFigure12-4).
FIGURE12-4Furtherstorybreakdown
Whathashappened?Youhavedecomposed“managepaymentoptions”intoitsfour
primarystories,allofwhichrollupintothe“managepaymentoptions”theme,whichin
turnrollsintothe“managemyaccount”epic.
Icanmanageexistingpaymentoptions
Icaneditordeleteanexistingcreditcard
Icanaddacreditcard
Icanmanagegiftcards
Lookcloselyatthesestories.You’vecomealongway,butyoucouldgofurther.The
story“Icaneditordeleteanexistingcreditcard”shouldbedecomposed.Why?Becauseit
isreallytwodifferentstories,albeitcloselyrelated,inone(seeFigure12-5).
FIGURE12-5Decomposingthestory“Icaneditordeleteanexistingcreditcard”
Formorepractice,youmightalsodecomposethetheme“managemysettings.”(see
Figure12-6).
FIGURE12-6Decomposing“managemysettings”
Whenyoulookatthelistofstories,youprobablywanttofurtherdecomposeboth
“changemyname”and“managemyaddressbook.”Forthesakeofbrevity,just
decompose“managemyaddressbook.”OnthefictionalwebsiteBig’nHuge,Icanedit
existingaddressesorIcandeleteexistingaddresses.So,writeastory:“AsaBnH
customer,Icaneditexistingaddressessavedinmyaddressbook.”Thisisagoodlevelof
decompositionbecauseitistestableandsmall.Couldyoudecomposethatfurther?Sure.
Thequestionis,doyouwantto?LookatthedecompositionshowninFigure12-7.
FIGURE12-7Decomposingcangotoofar.
Figure12-7probablygoestoofarforastoryattheproductbackloglevel.Usersdonot
oftenupdateonlytheaddressfieldorthestate;mostofthetime,theyupdatetheentire
address.Asaresult,therightlevelofdecompositionforthisstorywouldbeattheoverall
addresslevel,notatthespecificsthatcomprisetheaddressitself.
Thisbegsthequestion—howdoyouknowwhenthestoryistherightsizeonthe
productbacklog?Iusethisgeneralruleofthumb:Astoryisattherightsizewhenit
describesthesmallestactionthatauserwouldtypicallywanttodo,oritisthesmallest
pieceoffunctionalitywithbusinessvalue.Now,thisisnotasteadfastrule,butit’sagood
guidelinethathelpsyougaugethesizeofyourstory.
TaskDecomposition
Inthischapter’sstory,theteamstrugglednotwithstorydecompositionbutwithtask
decomposition.Foramoment,I’llpretendthatmystrugglingteamisworkingonthe
interfaceforBig’nHuge’saccountmanagementscreen.Theydecideduringsprint
planningtoestimatethestory“AsaBnHcustomer,Icaneditexistingaddressesinmy
savedaddressbook”(seeFigure12-8).
FIGURE12-8Theteamdecomposesthestory“Icaneditexistingaddresses.”
AsshowninFigure12-8,theteammembersfirstbreakthestoryintotwotasks:update
configurationfileandupdateextensionDLL.Theteamthenestimatesthenumberof
hoursforeachtask.UpdatingtheextensionDLLwilltake50hours.Updatingthe
configurationfilewilltake80hours(seeFigure12-9)
FIGURE12-9Theteamestimatesthetwotasksthatmakeupthestory“Icanedit
existingaddresses.”
Atthispoint,theteamrealizesthesetasksarefartoobigtoestimateandtrack
accurately,sotheteamcontinuestodecomposethesetwotasks(seeFigure12-10).
FIGURE12-10Sampletaskbreakdownfromauserstory
Oncethesetasksareamoremanageablesize,theteamestimateseachofthesmaller
tasks(seeFigure12-11).Noticethatthetotalofthesetasks(42and96)doesnotmatchthe
team’soriginalestimatesforthelargertasks(50and80).Thatistobeexpected.Because
thelevelofdetailismuchmorefinegrained,theestimatesaremoreprecise.
FIGURE12-11Estimatingtasksizes
Arethesetasksstilltoolarge?Thatdependsontheteam.First,followtherulethattasks
shouldbecompletedinnomorethantwodays.Soimaginethatateamhasfourpeople.
Onthissprint,theteamcommittedtoanaverageof16hoursofworkburneddownper
day(4people×4hoursofburned-downwork).Theteam’smaximumtasksizeshouldbe
about8hours(2days×4maxhoursperday).Iftheteam’smaximumhoursperdaywere
6,themaximumtasksizewouldincreaseaccordingly,forexample,12hours.Iftheteam’s
maximumhoursperdaywere3,themaximumtasksizewoulddecreaseaccordingly,for
example,6hours.
Sometimesataskcanappeartobetherightsizeatfirstbutlaterturnouttobetoobig.
Forinstance,supposethatduringsprintplanning,theteamdecidesthat“Identifycustomer
acceptancetests”issizedappropriatelyat12hours.However,oncetheteamstartsdoing
thetask,itsmembersquicklyrealizesomehiddenitemsneedtobeaddressed.Theystop
anddecomposethetaskfurthertobettertrackthework:
Brainstormanddocumentacceptancetests:2h
Reviewacceptancetestswithstakeholder:3h
Brainstormadditionaltestswithstakeholder:2h
Reviewacceptancetestswithteam:1h
Addacceptanceteststoautomatedtestframework:3h
Updateacceptancetestdataelementsfornewtests:2h
Setuptheprofilertohandlenewtests:2h
Addnewteststobuildmachine:1h
Thesetasksmayfeeltoosmallatfirst;however,formanyteams,especiallythosejust
startingwithScrumorhavingtroublefinishingtheirsprints,thesetasksaretherightsize
becausetheyaresmall,easilymanageableandunderstood,andprecise.
ThebestadviceIcangiveistoquestioneverything.Lookbackatsimilarpasttasksto
seehowcloseyourcurrentestimateistotheactualtimeyouspentonthetask.Thegoalis
nottohave100percentprecision,becausethatisnotachievable,butratheradegreeof
accuracythattheteamiscomfortablewithwhileprovidingtheproductownerwithenough
informationtomaintainareleaseroadmap.
KeystoSuccess
Decomposingstoriesandtasksischallengingbutisessentialtocreatingrealisticestimates
andbecomingahigh-performingteam.Thebestwaytofigureoutifastoryortaskistoo
largeorsmallistoaskaseriesofquestionstohelpdeterminethelevelofdecomposition.
Whilethisisnotanexactscienceandnotwoteamsarealike,understandingthesebasic
elementshelpsindeterminingwhatlevelofdecompositionisrightforyou.
Cantheteamestimatetheproductbackloginstorypoints?(Ifthisisansweredwith
aquestionlike“Whatarestorypoints?”thenyouranswerwilllikelybeno.)Putting
thataside,ifyoucanestimateyourproductbackloginstorypoints,youareofftoa
goodstart.
Isthereclarityonthestoriesintheproductbacklog?Someteamshavetheirproduct
backlogsestimatedbuthavefactoreddozens,ifnothundreds,ofassumptionsinto
theirestimates.Theydothisbecausetheyhavemorequestionsthananswersbut
have“mitigated”thatbyaddingtheassumptions.Alackofunderstandingisabig
redflagthatthestoriesarelikelytoolargeandneedfurtherdecomposition.
Howpreciseareyourstories?Andwhatistherightlevelofprecision?WhileI
cannotrecommendananswerthatwouldapplytoallstories,youcanaskyourself
thefollowingsimplequestions:
–Isittestable?
–Isitsmall?
DoIunderstandthisstorywellenoughtodoitmyself?Thisrelatestobothclarity
andprecision.Ifyoudon’tunderstanditpersonally,it’slikelythatthestorylacks
eitherclarityorprecision.
Sometimesyou’llfindyourstoriesortaskshavebecometoogranular.Thishappened
earlierinthechapter,whenItooktheBnHstorydowntothemanagingaddresslinefields.
Itisimportanttoremember,though,thatsometeamsmaynotfindthislevelofdetailtoo
granular.Myadviceis,atfirst,togopastthepointofwhereyoufeelcomfortable,to
experimentwithwhatlevelofgranularityworksforyourteam.Whatmightatfirstseem
liketoomuchdetailcansometimesbetherightgranularitylevelforyou.Youcanalways
gobackalevelduringthenextsprint.
Oneindicatorthatagroupofstoriesmightbetoogranularisiftheestimateforthe
storiesiszeroorhalfapoint.Iflumpingallthestoriestogethermakesa1-pointstory,but
individuallytheyareallahalfpoint,thestoriesarelikelytoogranular.Similarly,ifeach
tasktakes15minutesbutthegroupoftaskswouldtake2hours,you’veperhapsgonea
steptoofarindecomposingthework.
Anotherindicatorthatstoriesortasksarethecorrectsizeisthatindependentestimates
duringplanningareroughlyequal.Whenstories/tasksarehidingadditionalwork,
estimateswillvarywildly.ThisdisparitycanbeuncoveredusingPlanningPoker
[GRENNING].
Intheintroductorystory,everyonewasgoodatmaskingwork.Myteamhidtasksinside
taskswithoutevenrealizingit.Topreventthisfromhappeningonyourteam,challenge
estimates.Donotchallengetheminanegativeway;strivetoseekknowledgeonhowthe
teamcameupwiththetaskestimates.Estimatecollectively.Lookforassumptions.Ensure
thatyouunderstandthestoryortaskwellenoughtodoityourself.Aboveall,don’tbe
afraidtogoabitoutsideyourcomfortzoneindecomposingwork.Youmightbesurprised
withtheresults.
References
[DICTIONARY.COM]“epic,”“story,”and“theme.”Dictionary.com.TheAmerican
Heritage®NewDictionaryofCulturalLiteracy,ThirdEdition.HoughtonMifflin
Company,2005.http://dictionary.reference.com(accessed23July2011).
[GRENNING]Grenning,James.
http://renaissancesoftware.net/files/articles/PlanningPoker-v1.1.pdf(accessedon13
March2015).
WorksConsulted
Beck,Kent.2005.ExtremeProgrammingExplained,SecondEdition.UpperSaddle
River,NJ:Addison-Wesley.
Cohn,Mike.2004.UserStoriesApplied.UpperSaddleRiver,NJ:Addison-Wesley.
Chapter13.KeepingDefectsinCheck
Potentiallyshippable.You’veheardthisphrasebeforeandyou’llhearitagain.Sowhat
aboutdefectsthen?Isastorypotentiallyshippableifattheendofthesprintithassome
defectsthatwereleftunaddressed?Thisthenleadsteamstowonderwhethereachproject
shouldhaveasprint(ortwo)dedicatedtofixingdefects.Howaboutastabilizationphase?
Afterall,youneedtoreachthemythicalcodecompletephase,right?
Defectmanagementisoneoftheeasiestthingstoforegoonanyagileproject.Afterall,
teamshavebeenconditionedtoaddressdefectsattheendofprojects.Failingtodealwith
thesedefectsonanongoingbasis,though,resultsonlyinlow-qualitysoftwareandthe
needto“buildqualityintothesystem”duringadefect-fixingphaseorsprint.
Sohowshouldateammanageitsdefects?Dozensofbooksandtechniquesexplainhow
tomanagedefects,butIhavefoundastrategythatIparticularlylikeandhavesuccessfully
appliedtoalargemajorityofmyprojects,bothbigandsmall,newcodeandlegacy
systems.I’llstartbytellingyouthestorythatledtothispattern.
TheStory
MiguelandhisteamwerenewtoScrum.Miguelwasexcitedabouthisnewroleas
productownerandhadbeenreadingbooksandpracticing.Unfortunately,becausethe
projectstartedaboutoneweekaftertheteamwasformed,hehadn’tlearnedasmuchashe
wantedyet.
Abouthalfwaythroughitsfirstsprint,theteamaskedMiguelhowhewantedtohandle
defects.
“Miguel,rightnowwe’retwoweeksintothisone-monthsprintandwe’vegotabout
twentydefects,”saidateammember.“Shouldweholdoffandhaveadedicateddefect-
fixingsprintorplansometimebetweensprintstofixdefects?”
Miguelwasabitperplexed.Everythinghehadreadledhimtobelievethatadedicated
defect-fixingsprintwasthewrongwaytogo,buthereallywasn’tsurewhattosuggestas
analternative.Asfarasheknew,theonlywaydefectswereeverfixedatthecompany
wasinadedicateddefect-fixingphase.
“Justholdoffandlet’sseehowitgoes,”saidMiguel.Bytheendofthesprint,theteam
hadfoundevenmoredefects—somany,infact,thatthesprintreviewmeetingwas
cancelled.
TheteamwentintotheretrospectiveandaskedMigueltojoin.Astheyallstarted
writingthingsonthewall,theonethatemergedwasthemostobvious:wehavetoomany
defects.
RaulwastheScrumMaster,andhesawthisasabigimpediment,soheasked,“How
shouldwehandlethedefects?”
Theteamsatthereforawhile,andthentheideasstartedcoming.
“Haveadefect-fixingsprint,”saidoneteammember.Raulwroteitonthewhiteboard.
“Howaboutwejusthaveafixedamountoftimeeachsprinttofixdefectsattheend,
sayadayortwo?”saidanother.Raulagainaddeditonthewhiteboard.
“Ireadsomewherethatteamsputdefectsontheproductbacklog.Iguesswecanjustdo
that,”addedanotherteammember.
Thentheroomwassilent,untilfinallyMiguelspokeup.
“Whydon’twejustfixtheminrealtime?”heasked.
Theteamlookedathimlikehewascrazy.
“Realtime?Asin,wejustfixthemaswefindthem?”
“Yes,”saidMiguel.“Thatwaywekeepourcostslow.Theotherideaswe’veidentified
sofarallhavethesamepattern—pushingoffthedefectsuntilsomelaterdate.The
problemis,ifwedothat,thenwe’llbewritingcodeonbuggysoftware.Howwillwe
knowifthecodewewritedayslater,ontopofabuggyAPI,let’ssay,won’thavedefects
init,too?”
Theteamsatthereabitstunned.
“Whatifwejustdothesecondoption,settingsometimeasideattheend?”someone
asked.
“Iwasthinkingaboutthat,”saidMiguel.“Sinceweneverknowhowmanydefects
we’llhaveduringasprint,it’llbenearlyimpossibletosettimeaside.Whatifwehavetoo
manydefectsandnotenoughtime?Thenwe’rebacktowherewearetoday.I’mwillingto
tryitasalastresort,butI’drathertrythereal-timethingfirst.”
Theteamsatthere,thinking.
“Yes,Ithinkyou’reright,”saidateammember.“I’mwillingtotryreal-timebugfixing,
butarewegoingtofixeverybuginrealtime?Eventhelow-prioritystuff?”
“Ithinkthemostcriticaldefectsneedtobefixedinrealtime.Anyitemsthatarelower
prioritycangobackontheproductbacklog,andI’llprioritizethem,”saidMiguel.
“That’safaircompromise.I’mwillingtotrythis,too,”saidanotherteammember.
Raultooksomenotes,“Great!Thenit’ssettled.We’lldoreal-timebugfixing.Now,
let’sfigureoutwhatthatreallymeans.”
TheModel
Iamastrongadvocateforprioritizingqualityaboveallelse,whenitmakessense.Ialso
understandthatevencodethatlivesuptothehighestqualitystandardswillneverbe
perfect—willneverbeforeverfreefromdefectsandmaintenance.Assuch,Iacceptthat
defectsareapartoflife.Thatdoesn’tmean,however,thatteamsdon’tneedatechnique
formanagingthem,preferablyinrealtime.
Thefirststepindefectmanagementistorealizewhatisimportanttothecustomers,the
productowner,andtheteam.Customerswanttheteamtodeliverfeaturesthat,atleastin
theory,canbereleasedtoproduction;theydon’twanttopayfordefects.Productowners
wanttomeettherequestsoftheircustomers;theyknowthatcustomerswantdefect-free
code.Teamswantthefreedomtomitigateagainstdefectswhentheyrealizeapieceof
theircodeisinbadshape.
Thesecondstepindefectmanagementistounderstandthatfrequenttestingreduces
overallprojectcostsanddefects[BECK]andthatit’sjustplaincheapertofixdefects
soonerratherthanlater.Howmuchcheaper?Industrystandardnumberstofixadefect
rangethatfollowsfamiliar1:10:100rule,wherethingsidentifiedontheteammember’s
desktophaveacostofone,andastheymovethroughthesoftwarelifecycle,theyget
exponentiallymoreexpensive.InBarryBoehm’sdatafromSoftwareEngineering
Economics,heobserveda4:1cost/fixratio[BOEHM].Myfavoritestudy,however,was
donein2002.JohannaRothmanpublishedanarticleontheStickyMindswebsite
[ROTHMAN].Inherarticle,basedoncustomerssheworkswith,sheconcludedthata
teamthatwaitsuntiltheveryendtofixdefectswillhaveadefectcostprice400percent
largerthanateamthataddressesdefectsinrealtimeornearrealtime.
Allthisdatawasenoughtoconvincemethatitisjustbettertofixdefectsinrealtime.
Fromthis,Icreatedasimplemodelthatgivesteamsthefreedomtofixcriticaldefectsin
realtime,whileallowingtheproductownertoprioritizelesscriticaldefectsusingthe
productbacklog.Inessence,alldefectsareratedonapriorityscaleof0to3,where
priority0andpriority1arecriticaldefectsandpriority2andpriority3arelesscritical.
Theteamhasfullauthoritytofixp0andp1defectsastheyseefit.Thep2andp3defects
areputontotheproductbacklogtobereviewedandprioritized.
Atthispointyoumaybewonderinghowtodifferentiatebetweenp0,p1,p2,andp3.
Determiningthesevaluesisessentialinguidingtheteammembersastowhentheyshould
usetheirfullauthoritytofixabug.Table13-1illustrateshowIratethesevalues.Usemy
definitionsasconsiderationsandnotthedefaultstandardforyourprojects.Inotherwords,
talkaboutthiswithyourteamanddecideonamodelthatworksforyouandyour
company.
TABLE13-1DefectRatingSystem
Onceyouandtheteamhavedeterminedastandardizedwaytoprioritizedefects,the
teamcanbeginrankingthemduringthesprints.Whenanewdefectsurfaces,thepersonor
pairwhodiscoveredthedefectmustquicklytriagetheissueinrealtimeanddetermineits
priorityusingthescaleinTable13-1(ortheoneyoucreate).Ifthedefectisap2orp3,the
defectisimmediatelyloggedwithasmuchdetailandasmanysupportingfilesorlogsas
possible.Thedefectnumberisthenputontheproductbacklogforreviewand
prioritizationbytheproductowner.
If,ontheotherhand,thedefectisap0orp1,thepersonorpairhasonehourto
accomplishthefollowingtasks:
Stoptheworktheyaredoing.
Identifytherootcauseofthedefect.
Fixtherootcause.
Updatealltests(unit,integration,acceptance).
Buildorupdateanybuildverificationtests(BVTs).
Ensurethatalltests(acceptance,unit,etc.)arepassing.
Checkthecodein.
Releasethesoftwareto,atleast,anintegrationenvironment.
Ifthepersonorpaircandothiswithinonehour,theyarenotrequiredtoopenthedefect
trackingsoftwaretologthedefect.However,ifthiscannotbedonewithinonehour,they
musttakethefollowingsteps:
1.Stopattheendofonehour.
2.Logthedefectinthetrackingsystem.
3.Continuedrivingthedefecttocompletionwiththecriterialistedabove.
4.Oncethisiscomplete,openthedefectinthetrackingsystem,writingallstepstaken
togetittocompletion,andthenclosethedefect.
5.Createanewiteminthesprintbacklogforthedefectthatwasfixed,alongwiththe
hoursspentondrivingthedefecttocompletion.
Thisisasimpleyeteffectiveapproach.Withtheteamfocusedonqualitythroughout
eachsprint(andautomatedteststofixissueswhendefectssurface),youcanreleaseona
regularbasiswithextremeconfidence.Formoreonautomatedtests,seeChapter9,“Why
EngineeringPracticesAreImportantinScrum.”
KeystoSuccess
Managingdefectsinanysoftwareprojectisachallenge.Teamshaveyearsoflearned
brainmusclememorythattellsthemtofixdefectsattheend,afterdevelopment.Asa
result,asteamstransitiontoagile,theynaturallyfeeltheyshouldfixdefectsattheendof
thesprintorhaveabug-fixingsprint.Teamsneedtoretraintheirbrains.
Thekeytoasuccessfulagileprojectistodeliverapotentiallyshippableproduct
increment,wherethecodeistestedandthetechnicaldebtislow,attheendofeachsprint.
Doingthisrequiresashiftinmindsetthatisdifficultforpeople,teams,andcompanies
alike.Changingsuchingrainedhabitsrequiresdisciplineandeffort.
Thoseofyouworkingonalegacysystemmightbelievethatthisapproachcouldnever
workforyou.Reconsider.You’reprobablyconcernedaboutthesheervolumeofissues
youwillfind.Ihearyou.Afriendofmineatalargesoftwarecompanyspenttwoyears
fixingdefectsbeforehewasabletowritemainlinecode.Thisapproach,however,allows
youtomoveforwardwithnewcodewhileloggingnoncriticaldefectsandfixingcritical
onesinrealtime.Irealizethatitcouldbedays,weeks,orevenmonthsbeforeyoufinda
goodrhythmandgetoutofdefect-fixingmode.That’sokay.Theimportantthingisthat
youstartpayingdownallthetechnicaldebtthatthelegacysystemhasaccruedwhilealso
demonstratingnewfunctionality.
Whetheryouaredoingthisonanewprojectorwithlegacycode,thisapproachexposes
technicaldebtanddefectsquickly.Beforeyoubegin,youneedtoexplaintothecustomer,
management,andstakeholderswhatyouaredoingandwhyitisimportant.Startby
buildingacommonunderstandingonwhatadefectmeansfortheteamandideallythe
grouporcompany.Next,educateyourcustomersandstakeholders.Givethemexamples
ofwhatmakesagooddefectreportandteachthemhowtowritedefectreportsthatgive
theteamtheinformationitneedstoduplicateandfixdefects.Onceyouhaveasystemin
place,communicatethepriorityofthedefectsinyourbacklogandmakeitvisibleand
available.
Byestablishingawell-understoodprocessandmakingdefectsvisible,youandyour
companycanbreakfreeofthewaterfallattitudetowarddefectsonceandforall.
AdditionalInformation
Asthesecondeditionwasbeingwritten,Iwaslookingatdifferentbugmanagement
techniques.OneofthembymyfriendBillHanlonofMicrosoftissimplybrilliant.Please
readmoreathttp://www.mitchlacey.com/blog/managing-bugs-in-scrum-and-agile-
projects.
References
[BECK]Beck,Kent.2005.ExtremeProgrammingExplained,SecondEdition.Upper
SaddleRiver,NJ:Addison-Wesley.
[BOEHM]Boehm,Barry.1982.SoftwareEngineeringEconomics.EnglewoodCliffs,NJ:
Prentice-Hall.
[ROTHMAN]Rothman,Johanna.StickyMindsWebsite.
http://www.stickyminds.com/article/what-does-it-cost-fix-defect(accessed30June2015).
WorkConsulted
Ward,WilliamT.1991.TheCBSInteractiveBusinessWebsite.“CalculatingtheReal
CostofSoftwareDefects.”
http://findarticles.com/p/articles/mi_m0HPJ/is_n4_v42/ai_11400873/(accessed30June
2011).
Chapter14.SustainedEngineeringandScrum
“Quick,wehaveaserviceoutageandweneedyoutofixit,ASAP!”Thisis,
unfortunately,quitecommoninsoftwaredevelopmenttoday.Afterdecadesof
developmentandbillionsoflinesofcodehavingbeenwritten,oftenwithoutastrategyto
managethem,teamsfindthemselvesinasituationwherelegacysystemsmustbe
maintainedevenasnewsystemsarebeingdevelopedtoreplacethem.Howcanthisissue
betackled?DoessustainedengineeringhaveaplaceinScrum?Oneteamstruggledtofind
awaytomakeitwork.
TheStory
Rohitwastheproductownerforaflightprice-trackingsystem.Theprojectgoalwasto
replaceanexistingsystemusedonlybytravelagentswithanewonethatwouldbeused
bybothagentsandconsumers.Thelegacysystemwasquiteold,withintegration
componentsthattiedbacktoanIBMS/390mainframe.Veryfewpeoplewereleftwho
knewhowtomanagethemainframetie-ins,andmaintainingthecodebasehadbecomean
ever-growingchallengefortheentirebusiness.Thenewsystemwouldbeawinfor
everyone.Inthemeantime,though,Rohit’scompanystillhadtokeeptheexistingsystem
goinguntilitsreplacementwasonline.
Thingsransmoothlyatfirst.Rohitenjoyedattendingdailyscrumsasaflyonthewall
andhearingfirsthandabouttheworktheteamwasdoing.Thenewsystemwasworking
beautifully,andhecouldtellthatheandtheteamweregoingtobeverypleasedwiththe
producttheyultimatelydelivered.Hewasconcerned,though,abouthowoftenworkon
thelegacysystemseemedtointerferewithprogressontheproject.Justyesterday,hehad
createdanalarminggraphthatshowedthereleaseplannearlyspiralingoutofcontrol,not
becauseofbadestimates,newstories,orchangestothesystemtheywerebuildingbut
becauseoftheconstant,yetstillunpredictable,demandsofthelegacysystemonthe
team’stime.Hehademailedthegraphtotheteam’sScrumMaster,Kent,withanotethat
said,“Let’sdiscuss.”
Withtheseconcernsfreshonhismind,Rohitjoinedthenextmorning’sdailyscrumjust
asTimbegantalking.
“Yesterdaythelegacysystemhadanoutage,andtheoperationsteamcalledmeforan
immediatefix.Iinitiallythoughtitwouldbeaboutanhour,butitendeduptakingmeall
dayandhalfoflastnight,soIaccomplishednothingonthestoryIhadsaidIwouldwork
on.”
“Yourstorywas‘allowanexternalcustomertoseepricesforthethreedaysbeforeand
aftertheirplannedtraveldates,’right?”askedGeorge.
“Yes,”saidTim,“andIthinkit’sathugeriskrightnow.Iwashereuntiljustafter
midnightandmybrainisfried.Idon’tknowhowusefulI’llbetoday,guys.Sorry.”
“I’llhelpyouout,”saidGeorge,“butIdon’treallyunderstandthatstory.Ifwework
together,though,wecangetthroughit.”
“Doyouhavetime,George?”askedKent.
“Tim’sstoryisahigherprioritythanmine,soI’llmaketime.Butit’sgoingtomeanthe
storyIhadplannedtofinishtodayandtheoneIhadplannedtopickupwillhavetogo
backtothesprintbacklog.Ifwedon’tbringinsomeoneelsetohelpus,we’relikelygoing
tohavetodropatleastonestoryfromthesprint,”explainedGeorge.
RohitcaughtKent’seye.KentunderstoodthatRohitwantedtotalkaboutthesystem
outageissue.Afterthemeeting,thetwowenttogetacoffee.RohitandKenthadagood
relationship.Theyhadworkedtogetheratacompanyinthepastandkneweachother’s
styleswell.Sowhentheysat,Rohitgotrighttothepoint.
“Kent,yousawthegraphIsentyouyesterday.Theteamisconsistentlyfailingto
completeallthestoriesforthesprint.And,justlikeTimthismorning,peoplealwaysgive
thesamereason:‘Wegotcalledawaysuddenlyforanurgentfixinthelegacymainframe
code.’”
“Yes,”saidKent.“Unfortunately,weseemtobetheonlyonesleftwhoknowhowtofix
it.Wehaveinheritedthetribalknowledgeofthatsystem;wecan’tjusthandtheproblem
toanothergroupbecauseexplainingitwouldtaketoolong,andthereisnoreal
documentationtohelpgetthemuptospeed.Ihavetobelievethatkeepingthelegacy
systemgoingisworththelossinvelocityonthenewproject.Wehavetobalancekeeping
ourcurrentcustomershappywiththedemandsofthenewproduct.”
Rohitwasalltoofamiliarwiththerealityoftribalknowledge.Still,hehadcustomer
expectationstomanage.
“Kent,Iunderstandourteamhasuniqueexpertiseonthelegacysystem,butweneedto
findasolutiontotheseinterruptions.Outoffourteensprints,elevenhavebeeninterrupted.
That’sunacceptable,”saidRohit.
“Idon’tseeawayaroundit.Ifthesystemgoesdown,wehavetofixitrightthen,”said
Kent.
“Agreed—livelegacysystemoutagesmusttakeprecedenceoverthenewversion.Still,
youandIbothknowthatwecannotaffordtobedistractedeverytimethereisasmall
systemhiccup,right?”askedRohit.
“Yes,butlikeTimdescribedthismorning,someofthesehiccupsaren’tsosmall.The
systemisdown,anditcantakeanywherefromonehourtoonedaytobringbackonline.
Weneverknowuntilwestartfixingithowlongit’sgoingtotake,”saidKent.
“Ihaveanidea.InthelastcompanyIwaswith,wehadadedicatedengineeringteam.
Theirprimaryjobwastokeeptheexistingsystemrunning.Ithinkitwouldworkhere,
too,”saidRohit.
“Whowouldwegettorunit?”askedKent.
“That’stheproblem.Wewouldneedtotakeoneortwopeoplefromeachteamor
projecttostafftheteam,”explainedRohit.
“That’snotlikelytohappen,”saidKent.
“You’reprobablyright,”saidRohit.“Doyouhaveanotheridea?”
“Well,whenIwasateammemberononeofourpastprojectshere,wehadthissame
issue.Whatwedid,whileitwasnotideal,wassetasideacertainnumberofhourseach
sprinttoworkontheissuesthatcameup,”saidKent.
“Whatiftherewerenoissues?”askedRohit.
“Thenwewouldtakeanotheritemofftheproductbacklogthatwethoughtwecouldget
doneforthesprintorwewouldcleanupthenewcode.Therewasalwayssomethingto
workon,”saidKent.
“I’mnotsureaboutthis—yousaidyouactuallyblockouttimeinthesprint?Doesn’t
thattakeawayfromthetotalworktheteamcando?Rightnowtheydelivertwentypoints
persprintwhentherearenolegacysystemissues.Won’tthismodeltakethatnumber
down?”askedRohit.
“Yes.That’softenwhatitcomesdownto,”admittedKent.
“Idon’twanttoloseteamvelocity.Dowehaveanyotheralternatives?”askedRohit.
“Well,wecouldtalkaboutthelossintermsofstorypoints,insteadofhours,for
example,buteverysolutionIcanthinkofwillultimatelylowerteamvelocity.Rohit,the
thingis,teamvelocityisdownanyway.Likeyousaid,forelevenoutofthepastfourteen
sprints,we’vefailedtodeliveronourcommitmentandalwaysforthesamereason:the
legacysystem.Whentherearenoproductionissues,ourvelocityisgreat.Theteamis
workinghardandwantstogetintoadependablerhythm.Eachtimetheyhavetocometo
asprintreviewwiththeirheadshunglowbecausetheyhaven’tfinished,moraledropsa
littlefurther,”saidKent.
“Yes.Icanseehowthatwouldhappen.Wedon’twantthat.Loweringvelocityisn’t
ideal,butitiscertainlybetterthanloweringmorale.And,partoftheproblemnowisthatI
can’tmakeanyplansbecauseIhavenoideawhatthevelocityisgoingtobefromsprintto
sprint.I’matthemercyofthelegacycode.Ifwecouldfindawaytostabilizethevelocity,
Icouldpredictreleases,”Rohitmused.“Let’ssitdownandseeifwecanworksomething
out.”
KentandRohitspentthenextcoupleofhourslookingatthehistoricalsprintdatathe
teamhadaccumulated.Theyfoundthattheteammissedabout50hours’worthofworkon
averagefromsprinttosprint.Somesprintswerehigher,somewerelower,butthisgave
themthenumbertheywerelookingfor.Theyusedthisdatatobuildasustained
engineeringplanthatwasadequateforthebusinessandtheteam.
TheModel
Therearemanyapproachestosustainedengineering(SE),twoofwhichwerediscussed
byKentandRohit.ThemodelsIdiscussnext,allocatingacertainamountoftimefor
unforeseenissues(dedicatedtime)orhavingadedicatedteam,areonesIhaveusedwith
greatsuccess.
DedicatedTimeModel
Inthededicatedtimemodel,asingleScrumteamisresponsiblebothforworkingthrough
theproductbacklogonthedevelopmentprojectandalsofordedicatingtimetoworkon
issuesthatmightsurfacefromanexistingproductorservice.Theseteamstreatlegacy
systemworkjustastheywouldvacationtimeorothercommitments:Theysubtractitfrom
thetimeavailableforsprintbacklogitems.Oncethat’sdone,theteamcontinueswith
sprintplanningandbreaksdownandestimatestheworkaccordingly.Anexampleofthis
isshowninFigure14-1.
FIGURE14-1Workonlegacysystemissubtractedfromtimeavailableforsprintbacklog
items.
Determininghowmanyhourstodeductfromtheteam’ssprintisachallengeifyou
havenotbeencapturingthedata.Iftheteamdoesnotknowhowmanyhoursithasspent
inthepastonoutagesandsustainedengineering,takeaguess.Afterall,yourteamwill
haveexperience,sotheguessshouldbesomewhataccurate.Then,aseachsprint
progresses,gatherrealdataeachtimeaneventrequirestheteam’sattention.Thisdata
allowsfutureguessestobemoreaccurate,asshowninFigure14-2.
FIGURE14-2Collectdatatomakeestimatesbetter.
DataGatheredoverTime
InFigure14-2,noticethattheteamhasfivesprints’worthofdata.Ithasspentbetween50
and90hourspersprintonsustainedengineeringtasksand210to250hoursonstories
fromthesprintbacklog.Theproductownernowhasenoughinformationtoseethathecan
havehighconfidenceintheteambeingabletocompleteupto210hoursofstory-centered
workandsomepotentialfortheteamtocompleteupto250hoursofstorytaskspersprint.
Theproductowneralsoknowsthatanyhoursabove250willlikelybededicatedto
sustainedengineeringwork.
Theadvantagesmayincludethefollowing:
Thepeopleworkingonthenewversionfeelthepainofthelegacysystem,helping
toensurethesameshortcomingsarenotrepeated.
Itmaybecheaper.
Fixestothelegacysystemcanbe(nearly)immediatelyincorporatedintothenew
system.
NoonefeelsexiledtoadedicatedSEteam.
Thedisadvantagesmayincludethefollowing:
Personnel-andteam-switchingcostsarehigher,especiallyifissuessurfaceinsmall
pieceseachday.Multitaskingandcontextswitchingcandramaticallyhurt
productivity.
Theconstantshiftfromthenewsystemtoaddressinglivesystemproblemscan
causefrustrationovertime.
Someissuesmaytakelongerthantheallocatedtime.
Fluctuationsinoutageissuesmaycauseteamvelocitytovarywidely,resultingin
poorpredictability.
DedicatedTeamModel
Inthestory,Rohitsuggestedcreatingateamdedicatedsolelytomaintainingthelegacy
system.IusedasimilarapproachmanytimesatMicrosoftandothercompanies.Itisa
pull-basedsystemthatborrowssomeelementsfromScrumandKanban/Lean.
JustlikeaScrumteam,adedicatedsustainedengineeringteamshouldcontainallthe
skillsnecessarytodothework.Also,justlikeaScrumteam,theteamshouldworkfroma
backlogthathasbeenprioritizedbythestakeholders;onlyinthiscase,thebacklogis
madeupofissuesratherthannewfeatures.Theissuescanbeinaproductbacklog,orthey
canbeonaboardfortheteamtopullfrom.And,justlikeaScrumteam,thesustained
engineeringteamshouldhaveaproductownertoensurethatissuesarebeingtriagedsuch
thatthehighest-priorityitemsarebeingworkedonatalltimes.
Thesprintlengthmustbeshorterthanby-the-bookScrum,allowingtheteamtoreact
morequicklytochangingconditionsandproblems,astheyariseinrealtime.Iadvocate
andpreferdailysprints,butsomeissuesmaynotbeabletobefixedandreleaseddaily.In
theseinstances,theissuesarereportedondaily,bothtostakeholdersandtotherestofthe
team.
Whenateamisreleasingdaily,teamactivities(planning,meetings,releases,and
commitments)happenfast.Inthemorning,thereisabriefplanningsession,addressing
anynewissuesthatsurfacedovernightandanylegacyissuesstillbeingaddressed.As
teammembersaddresstheissuesandfixthem,theyreleaseeitherthroughoutthedayorat
theendofthedayinonebatch—thetimingisalwaysdependentontheorganization.At
thebeginningofthenextday,theteambrieflytalksaboutitsprioraccomplishments
beforemovingontoplanningthecurrentday.
Forlargerissues,however,dedicatedSEteamsshouldhaveweeklygoalplanning,daily
scrums,andaweeklystakeholdermeeting.
GoalPlanning
Onthefirstdayofeachweek,theteammeetstoreviewthequeueoflargeworkissuesand
toplanaweeklygoal,suchas“move50,000ordersthroughthesystem”or“reduce
customersupportresponsetimesbytenseconds.”Theteammembersreviewthecurrent
issuesbacklog,associatethemwiththegoal,anddeterminetheworktheythinktheycan
accomplishthatweek,basedontheinformationtheyhavethatday.Theyunderstandthat
theplanmaychangeifthereisasuddenoutageorunforeseenissue.
DailyReleasesandScrums
Attheendofthegoalplanningmeeting,andthenagainatthestartofeachday,theteam
shouldperformitsdailyscrum,wherethemembersdeterminewhichissuestheybelieve
theycanaddressthatdayandthencommittocompletingthem.Theyshouldalsoprovide
statusonthesame-dayissuestheyareintheprocessoffixingandplantorelease.Nearthe
endofeachday,theteamshouldalsomeettoshareprogressandimpediments,checkfor
newissuesthatneedtobeaddedtothebacklog,anddemoandreleasethework,if
possible.Theteamshoulddrivetoreleasefixesasquicklyascanbemanagedbythe
business.Myteamsreleaseddaily.Otherteamsinourorganizationreleasedweekly.The
urgencyofaparticularfixdependsonmanyfactors,includingthepainthecustomersare
experiencing.Ifanewissuesurfacesduringtheweek,itisthejobofthededicatedteamto
reacttoit.
StakeholderMeetings
Atleastonceaweek(moreoftenifanurgentissuearises),theteamshouldmeetwiththe
stakeholderstodemofixesandprioritizenewissues.Stakeholdersshouldunderstandthat
somefixesbeingdemoedmayhavealreadybeenreleased.Havingthisformalreview
helpstheteamfeelasenseofaccomplishmentandallowsthestakeholdersanopportunity
bothtocommendtheteamforitseffortsandprovideanyfeedbackintowaystheprocess
couldbeimproved.
Workingonadedicatedteamcanberewarding,butitcanalsobeachore.Still,itisone
waytomanagetheproblemoflegacysystems.
Theadvantagesmayincludethefollowing:
Dedicatedteamscanbeatraininggroundfornewemployees,whocancycletonew
productteamsafterthey’velearnedtheexistingsystems.
Dedicatedteamscanrespondquickly.
Teammembersare(orwillbecome)expertsonthelegacysystem.
Theyaregenerallyabletoaddressissuesmorequicklythannewproductteams.
Theyareabletoreleasemoreoften.
Thedisadvantagesmayincludethefollowing:
Theworkisoftenboringandrepetitive.
Teammembersmaynothavepassionfortheirwork.
Moralesuffersifquickwinsarenotshown,recognized,andacknowledged.
Dedicatedteamscostmorebecausepeoplewithspecializedskillsetswillneedtobe
replicatedonthededicatedteamandinthelargerorganization.
Otherteamsmaybelessconscientiousknowingsomeoneelsewilldothecleanup
workforthem.
KeystoSuccess
JeffSutherlandoncetoldmethebestcodeisnocodebecausenocodeissustainable,
maintainable,performswell,andcostsnothing.Unfortunately,teamsmustwritesome
codetobuildsystems.Andallcode,oncewritten,islegacycode.Legacycodethathas
supportingelements,suchasunittestsandautomatedacceptancetests,andiswell
documentedandcleanischeapertomaintain,butmaintenancecostsandresponsibilities
canneverbeeliminated.
Regardlessofthemodelyouchoosetosupportlegacycode,remembertotreatyour
sustainedengineeringprocessasyoudotheprocessforteamsbuildingnewfeaturesand
functionality:Evolveitovertime,makingchangesthatallowtheteamtobemore
successful.
CycleDedicatedMaintenanceTeamMembers
Ifyourorganizationusesadedicatedmaintenanceteam,itsmemberscan(andprobably
should)becycledbetweenateamthatisdevelopingnewfeaturesandfunctionalityand
themaintenanceteam.Thisisimportantforseveralreasons.First,ithelpsprevent
specialistskillproblems.Specialistskillsarethosebuiltupovertimewhenpeopledothe
samerepetitivetasksoverandoveragain,especiallyiftheyareworkingonthesame
technologyorfunctionalareaofasystem.Switchingpeoplekeepsthemfresh.Italso
helpsmitigatefrustrationandboredom.Ioftenseepeoplebecomeenviousofotherswho
areonnewfeatureteamsbecausethatiswherethefunis—theyarecreatingsomething
newascomparedtocleaningupsomeoneelse’smessthathasbeenlongsinceforgotten.
Cyclingpeopleamongteamshelpsalleviatetheseissues.
RetrofitLegacyCodewithGoodEngineeringPractices
Ioftenseeteamsthatfollowthededicatedtimemodelbutonlyapplytheirengineering
practices,suchastest-drivendevelopment,tothenewproductsorservicesthattheyare
building.Whenalegacysystemissuesurfaces,theythrowthosepracticesaside.Inmost
cases,thisisthewrongapproach.Thededicatedtimemodelneedstofactorintheeffortto
buildintothelegacycodeunittests,acceptancetests,andanyothermodernengineering
practicetheteamhasimplemented.AgreatresourceformanaginglegacycodeisMichael
Feathers’WorkingEffectivelywithLegacyCode[FEATHERS].
IntheEnd
Supportingalegacysystemwhiletryingtobuilditsreplacementisinherentlydifficult.
Whateverapproachyouchoose,themostimportantthingistoacknowledgethatmakingit
workrequiresacommitmentofeithertimeorpersonnel.MartinFowlerintroducedthe
worldtothetermstranglerapplicationin2004[FOWLER].Hisblogpostillustratesatrip
hetooktoAustraliawherehesawfigtreeswithstranglervinesdroppingdownfromthe
highbranches.Thevineswraparoundahosttree;asthevinestakeroot,thehostdies.The
samethinghastohappentoyourlegacysystem:Thenewstranglerapplicationwill
eventuallychokeitout.Meanwhile,youneedaviableapproachthatallowsyoutoreduce
riskandmaintainfunctionalityuntilthestranglerapplicationisliveandthehostisretired.
References
[FEATHERS]Feathers,Michael.2005.WorkingEffectivelywithLegacyCode.Upper
SaddleRiver,NJ:PrenticeHall.
[FOWLER]Fowler,Martin.Personalwebsite.
http://martinfowler.com/bliki/StranglerApplication.html(accessed3March2011).
Chapter15.TheSprintReview
Ideally,thesprintreviewshouldbeplug-and-play,meaningthingsshouldjust“work.”The
teamissupposedtobeabletopreparefornomorethananhourand,withouta
presentationorunduefuss,demotheworkinanenvironmentasclosetoproductionas
possible.Inreality,though,asprintreviewcanbeenoughtomakeyourteethrattle.
Sometimesthecustomerasksaquestionyouaren’tpreparedtoanswer;occasionallya
featuredoesn’tworkasexpected;andeverynowandthen,somethingcrashes
unexpectedly.Insomecases,oneortwooftheseembarrassingmomentsoccurinevery
reviewmeeting,degradingthetrustbetweentheteamanditsstakeholdersandmaking
everysprintreviewanunpleasantanduncomfortableexperienceforallconcerned.
Thatwasthecasefortheteaminthefollowingstory.Readontofindouthowaseries
ofunfortunatemeetingsandabreakdownintrustcausedtheteamtoreconsiderthewayit
preppedforandranitssprintreviews.
TheStory
Onebyone,theteammembersslunkoutofthesprint-threereviewwithheadshunglow.
“Ican’ttakemanymoremeetingslikethat,”commentedMike,lookingsidewaysathis
teammates.“Welookedlikecompleteidiots,again!”
Mikeandtheteammadetheirwaytotheteamroomfortheretrospective.Assoonas
thedoorclosedbehindthem,thefloodgatesopened.Groans,moans,andgrumblesfilled
theroom.
ScrumMasterJimstoodupandaskedforquiet.“Writeitdown,”hesaid,motioningto
thewhiteboard.
Theteammembersgrabbedmarkersandmadetheirwaytothe“whatwentwrong”
whiteboard.Inamatterofminutes,itwasfilledwithproblems,mostofwhichcentered
aroundsprintreviews.“Everymeetingisworsethanthelast,”onesaid.Anotherread,
“Reviewsarepainfulandembarrassing.”Mike’swaswritteninlargecapitalletters,
“COULDWEPLEASESTOPSHOWINGOURBUTTSTOTHECUSTOMER!”
Oncethewhiteboardwasfilledtocapacity,theteammemberssatdown,tossedtheir
markersonthetable,andlookedatJimexpectantly.
Jimchuckled.“Okay,then.Let’sclarify.Mike,wanttostart?”
Mikedovein,“I’vereadalltheScrumbooks,andtheysaythatthesethingsare
supposedtobeplug-and-play.Noprep.Justshowup,showthenewfeatures,andthe
customersacceptit.ButeverytimeI’minasprintreviewmeeting,IfeellikeI’mina
nightmarewhereI’mbackinuniversity,thereisatestineverycourse,andIhaven’t
studiedforanyofthem!”
Michelefollowedup.“Therehasgottobeabetterway.Irefusetobelievethereare
teamsouttherethatjustshowupandeverythingworksforthem.”
“Exactly,”saidMaria.“TheyaskedsomanyquestionsthatIjustwasn’tpreparedto
answer!”
Jiminterjectedwithanideahe’dgottenfromoneoftheotherteamsinthecompany.
“YouallknowSam’steam,right?Italkedtohimlastweek,andhesharedhisteam’s
reviewprocedurewithme.”Jimuncoveredapre-preparedflipchartandbegantopoint
outkeyelements.“Theteamusestheseslidetemplateswithspotstoplugininformation
abouteachsprint.Theslidesdetailwhattheteamcommittedto,whatitgotdone,project
metrics,andmore.”
“Doesn’tthatgoagainstthe‘noPowerPointrule,’though?”askedMike.
“Sureitdoes,”saidJim.“Butitdoesnotgoagainstthe‘usecommonsense’rule.”
Micheleandtherestoftheteamlaughed.
“Reviewsshouldbeeasy,”Jimcontinued.“Ithinkifwedoafewthingsdifferently
duringthesprint,involvetheproductownermore,anddoabetterjobofcommunicating,
we’llfeelmuchmorepreparedandmuchmorecomfortable.”
Jimandtheteamdecidedtohaveaselectgroupworkonthesprintreviewproblem
immediatelyfollowingtheretrospective.Thegroupspenttwohourscomingupwitha
planforthenextsprintreview.Theplanincludedmidsprintactivitiesmeanttoinform
stakeholdersandengagetheproductowner.
Sprintfourpassedquickly.Theteamcouldhardlysitstillasthesprintreviewgot
underway.Eachofthemfeltexcitedandready.“Nicechange,huh?”whisperedMiketo
Jim.
Jimnodded.
Dan,theproductowner,kickedoffthemeeting.“Thanks,everyone,forcoming.
Normally,I’dstartthedemorightnow,butI’vebeenworkingwiththeteamonanewway
torunthemeeting.Ithinkwecanallagreethepastthreemeetingshavebeen,well,
special.”
Everyoneburstoutlaughing,eventheteam.
“Withoutfurtherdelay,letmehanditovertoMichele,oneoftheteammembers.
Michele?”Danmotionedforhertotakethefloor.
Aperplexedlookcameoverthecustomersandstakeholders.WhywasMichele,ateam
member,doingthedemo?Theproductownerhadalwaysdonethedemo.
“Thanks,Dan,”saidMichele.“Letmestartbyshowingyouapresentationofthework
wecommittedto.I’mgoingtofollowthatwithalookatwhatweactuallygotdoneand
whatwedidn’tgetdone;thenI’llshowyousomeprojectmetricsandrecapthedecisions
weallmade.WhenI’mthrough,Jimwillwalkusthroughthedemo.Ithinkthe
paperwork,allthestuffbeforethedemo,shouldtakeabout10to15minutes,leavingus
plentyofhands-ondemotime.Howdoesthatsound?”
Theyallnoddedandsatforwardintheirseatsexpectantly.Michelewentthroughher
slides,showingthecustomersandstakeholderswhattheteamhadcommittedtodoing,
whatithadgottendone,whatithadfailedtogetdoneandwhy,andthedecisionsmadeby
boththeteamandDanwiththecustomer.
“Anyquestions?”askedMichele.“No?Then,I’llturnitovertoJim.”
“Thanks,Michele,excellentrecap.Now,today,aswedemoourfeatures,wearegoing
todothingsalittledifferently.Danisonboardwiththischangeaswell,right,Dan?”
Dannoddedeagerly.
“Okay.”Jimcontinued,lookingatthestakeholdersintheroom.“Whowouldliketo
validatethefirststoryforus?”
“I’msorry.I’mnotsurewhatyou’reasking,”saidoneofthestakeholders.
“Well,”saidJim.“Whatwewouldlikeyoutodoispickanitemthatislistedas‘done’
intheslideswehandedout.Anyonewilldo.Thenwewantyoutoworkit,validateit,
playwithit—seeifitdoeswhatyouexpectordoessomethingyoudon’texpect.”
Akeystakeholder,Meher,walkedhesitantlytothelaptopandgrabbedthemouse.
“Areyousureyouwantmetodothis?”sheaskedwithgreathesitationinhervoice.“I
knowhowitshouldwork,becauseit’smystory,butIdon’twanttomessanythingup.”
“Don’tworry,”saidMike.“Youwon’tmessanythingup.Andifyoudofindaproblem
now,thatjustmeanswecanfixitsooner,whichisagoodthing.Pleasegiveitatrylike
younormallywould.”
Theteamtookacollectivedeepbreathasshemovedthemouseandstartedvalidating
theuserstory.“Asanaccountant,Icanpulluplastmonth’ssalesreportbyregion.”She
startedwiththesouthwestregion.Sheclickedit.Atfirst,nothinghappened.Threelong
secondspassed,andthenthereportappearedonscreen.
“Huh,”saidMikeunderhisbreath,butloudenoughforeveryone’seyestoshiftupin
hisdirection.Mikewasexpectingamuchfasterloadtime.
Mehercontinuedclicking,selectingvariousdrop-downs,shiftingthroughdifferent
views,andclickingthroughthedatatogetexpandedreportsontheregion.
“Thisreportlooksreallygood!”shesaid.“Iwouldexpectsomedifferentdatahere,but
it’smoreorlessmyjobtovalidateit.Ihavetimenextsprinttoworkwithyouonthat.
Hey!I’vegotaquestion:WhathappenswhenIclickthroughhere?”
Theteamhadagoodideaofwhatwouldhappen,especiallyMichele.
“Clickitandfindout,”shesaidfirmly.
Thecustomerclickedthroughoneofthereportsthathadnotbeenfullydeveloped.The
teamwatchedthepathshetooktogettoit,onethattheteamdidnotpredictorthinkthata
userwoulduse.Sheclickeditandtheapplicationsatthere.After30secondsorso,it
crashedwithanalertbox.
“IknewI’dmessitup,”saidthestakeholderapologetically.
“I’mgladyoudid,”saidMichele.“That’saworkflowwehadn’tconsidered.Mike,did
youwatchthatflowandtakenotes?”
“Gotit,”saidMike.“I’llhaveitfixedbeforewegohometoday.”
“Waitaminute,guys,”saidDan.“Ididn’tspecificallyaskforthatpartoftheworkflow
tobedonethissprint.Shouldwereallyfixit?”
“Yeah,Icanfixit.Iknowwheretheproblemis,soitwon’ttakemelong,”saidMike.
Jimnodded.“Okay.Whowouldliketotrythenextstory?”
“So,I’mabitconfused,”anotherstakeholderadmitted.“Whyareyouhavingusdo
this?Itseemslikeitwouldbemuchquickerforyoutojustshowusthefeature.This
seemslikeawasteoftime.We’renotevensurehowyou’vedesignedthesetowork.”
Jim,theScrumMaster,spokeup.“We’rehavingyoudoitbecauseyouareultimately
theoneswhowillbeusingthesoftware.Ifitdoesn’tworkthewayyouintuitivelythinkit
should,thenweneedtochangeit.Watchingyouexperimentwiththefeatureteachesus
quiteabitabouthowyouthinkandwhatatypicalworkflowlookslike,soIdon’tthink
it’sawasteoftime.Andyou’reright.Youdon’tknowhowwe’vedesignedthesetowork.
So,it’simportantthatyouputyourhandsonitandmakesureitworksforyoubeforeyou
acceptthework.
“Weknowthatthingshaven’tgonewell.Andweknowthatyou’relosingfaithinour
abilitytodeliverwhatyouneed.Bycommunicatingmoreandusingthistimetoletyou
experiencethesystem,wewillgetbetterandcanadjustwhatwe’redoingbasedonwhat
wesee.Alreadyinthisdemowe’velearnedthatthereportsloadtooslowlyandour
understandingoftheworkflowwasinaccurate.Andwe’vegottenbuy-infromMeherto
workwithusonsomedatanextsprint.Itisn’tperfect.Butit’sprogress.”
Thestakeholdersagreed,andthemeetingcontinued.Afterafewsprints,theteamand
thecustomerweremuchbetteraligned.ThePowerPointslidesreplacedformersprint
statusreports.Danworkedwiththecustomerstohelpthemunderstandthatcertain
featuresweren’tquiteperfectyetdidmeetthestatedacceptancecriteria,andusedtheir
feedbacktohelprefinetheproductbacklog.Asfeaturesbegantomatchexpectations,
everyone—theteam,thestakeholders,andDan—begantofeelmorecomfortablewiththe
process,moreateasewitheachother,andmorehopefulforthefinalproduct.
TheModel
Itiscommonfornewteamstostrugglewithsprintreviews.Afterall,formany,it’sthe
firsttimethey’vehadtostandinfrontofcustomersandstakeholdersandshowthemwhat
theyhavecreated,letalonedemoitemsthataredeliberatelylackingfullfunctionalityand
haveonlybeenaliveforafewweeks.Iftheteamfailedtocompleteitsstories,orifwhat
hasbeendoneworkspoorly,thesprintreviewcanbeawfullyembarrassing.
Thepurposeofthesprintreviewisfortheteamtoshowthecustomersandstakeholders
theworkithasaccomplishedoverthesprint.Whileideallythismeetingshouldbe
informal,manybusinessestreatitveryformally.Myadvice:Treatitinawaythatworks
foryourcompanycultureanddoesnotcausetheteamunduestress.
Thesprintreviewoccursonthelastdayofthesprint.Thedurationvariesdependingon
sprintlength.Foraone-weeksprint,themeetingshouldlastabout30minutestoanhour;
atwo-weeksprint,anhourtoanhourandahalf.Teamsrunningfour-weeksprintsshould
allowtwotofourhoursforthismeeting.Besidesacceptanceofthesprintwork,other
intendedoutcomesofthemeetingincludeadditionalcustomerinteraction,increased
feedback,andadditionsorchangestothebacklogbasedontheinteractionandfeedback.
Inthestory,Jimandhisteamdiscoveredthatitismuchmoredifficulttorunasprint
reviewthanitsounds.Thinkingthereviewwouldjusttakecareofitself,Jim’steamtried
toraceallthewaythroughthesprintwithnothoughttowardthedemo.Thesnafus,
failures,andmisunderstandingstheteamexperiencedasaresultarealltoocommonfor
newteams.Inthestory,theteam’ssolutionwastopreparemoreduringthesprint,
formalizethepresentationtoensurealltherightinformationwascommunicated,and
encouragethestakeholderstoparticipateinahands-onway.Theamountofpreparation
youdowillvarydependingonyourcompany’scultureandteam’sexperiencelevel.I’ll
walkthroughhowIprepareformysprintreviews.
PreparingfortheMeeting
First,Iamafanofusingslidesinthesprintreviewmeeting.Theyhelptheteammembers
organizetheirthoughtsandprovidesomethingforthestakeholderstotakewiththem
whentheyleave.Foreverysprintreviewmeeting,Iusethefollowingtemplate;each
bulletisanindividualslide:
Thesprintgoal
Thestoriestheteamforecasted/committedtodeliver
Thestoriestheteamcompleted
Thestoriestheteamfailedtocomplete
Keydecisionsmadeduringthesprint,whichmayincludetechnical,market-driven,
requirements,etc.(Thesecouldbedecisionsthatweremadebytheteam,the
productowner,thecustomers,oranyoneelse.)
Projectmetrics(codecoverage,etc.)
Demonstrationofthecompletedwork
Priorityreviewforthenextsprint
Dependingonthetypeofsystemyouarebuilding,youmayalsofindyouneedto
preparetechnically—forinstance,haveabuildreadyordeployedtoapreproductionor
stagingenvironment.Inthesecases,consideramockrun-through.Haveateammember
spendsometimegoingthroughandreviewingsomeofthestoriesthatwillbeshownto
thecustomers.Isthedatacorrect?Aretheconnectionstringsworking?Diddataget
repopulatedfromoneenvironmenttothenext?Thesesmalldetailscancomebacktobite
youwhenyou’reinthemeeting.
Atfirst,allthatpreparationseemstobeindirectviolationoftherules,andiftakento
theextreme,itis.Preparationshouldnotgetinthewayofthework.Spendingsometime
gettinginformationtogetherandplanningthereviewmeetinggoesalongwaytoward
makingateamappearmoreinformedandhelpingthestakeholdersunderstandwhatthey
aresupposedtobelookingfor.Idon’tlivebyahardandfastrulethattheteamcanspend
onlyonehourpreparingforthemeeting,asthisisextremelydifficultfornewScrum
teamstodo.Thatbeingsaid,there’salimittohowmuchpreptimeistoomuch.I’veseen
teamstaketwodaystogetthingsready,which,especiallyinashortsprint,isoverkilland
pointstolargerissues.Therefore,althoughnewteamswilllikelyneedmoretimetoprep
thananexperiencedteam,allteamsshouldbeworkingtowardthegoalofpreppingforan
hourorlesseachsprint.
RunningtheMeeting
Inthischapter’sstory,thecustomersendedupexecutingthefunctionalitytheteam
deliveredduringthemeeting.Thisisdifferentthanthestandard“theteamdemonstrates
thework”approachinScrum.Onmyteams,Iasktheproductownertofacilitatethe
meeting(justasaScrumMasterwouldfacilitatetheretrospective),theteamtodoahigh-
levelreviewofthestories,andthecustomersandstakeholderstoexecute/demonstratethe
storieswhiletheteamobservesthebehavior.
Wheneverpossible,Ialsoliketoincludetheprocessofmovingtheinstallationoffthe
buildpathandintotheenvironmentaspartofeveryreview.Ialwaysstartwithafresh
install,doingtheactualinstallduringthereview,withthecustomerintheroom.For
example,ononeWindowsapplicationprojectwhereIwasaproductowner,astoryinthe
firstsprintreviewwastoinstalltheapplicationfromthelatestbuildoffthebuildbox,an
acceptancetestthatwouldrecurinfuturesprints.Thebuildwasnevermorethanfiveor
tenminutesold.Ononeoccasion,tosavetime,wehadtheapplicationpreinstalledona
computerwewereusing.Thecustomershadgrownsoaccustomedtoseeingthe
installationatthedemothattheycomplained.Weuninstalledtheapplicationonthe
computer,wenttothebuildmachine,andinstalledfreshbits.Sincethen,Itrytoinclude
installs(whenappropriate)aspartofeverysprintreview.Ithelpsbuildcustomer
confidenceandexhibitsaleveloftransparencythatcustomersmighthaveneverwitnessed
before.
Oncethedemoiscomplete,teamsshouldaskthecustomerstoacceptthework.After
all,thecustomershavehadthechancetodemothefeaturesthemselves,andtheteam
shouldbeconfidentthattheworkispotentiallyshippableatthispoint.
KeystoSuccess
Thesprintreviewisatimefortheteamtogoonrecordandexplainwhatithas
accomplished.Likeachildeagerlyawaitingherparents’approvalfortakingoutthetrash
onherownforthefirsttime,newteamswantandneedpositivefeedbackfromthe
customersandstakeholders.
Don’tcancelthesemeetingsduetopressureorotherexcuses.Theyarecriticalandhelp
theteamdothefollowing:
Buildandmaintaintrustwiththecustomersandstakeholders
Makeprojectcoursecorrectionsinnearrealtime
Identifyrisksandissues
Gatherfeedback
Positivesprintreviewsincreasecustomerconfidenceintheteamandtheteam’s
confidenceinitself.Ifthereviewgoesbadly,thetrustwilldegrade,justashappenedinthe
story.
Atthesametime,don’tforgetthatthemainpurposeofthesprintreviewisnotaround
ofapplauseforajobwelldone.Therealgoalofasprintreviewistostopandascertain
whethertheprojectisontherighttrack.Theteamandproductownerneedtoknowwhat
theydidn’tknowthelasttimetheygroomedtheproductbacklog:howthisincarnationof
theproductworksinacustomer’shandsandwhatfeaturesneedtobechangedoraddedto
makeitmorevaluable.Thebestsprintreviewshaveanimpactonwhatoccursinthe
followingsprint.Theyincreaseboththeteam’stoleranceforchangeandthecustomer’s
expectationthattheyhavepermissiontobreakthings,tofindflaws,allinthenameof
makingtheendproductbetter.Reviews,whendoneright,fosteraspiritofcontinuous
improvementthatistrulycontagious.Tomakesureyourreviewsaredonetherightway,
taketimetoplan,documentdecisions,askforacceptance,andbebrave.
TakeTimetoPlan
Becauseapositivereviewexperienceissoimportanttotheproject’ssuccess,takealittle
timetoplanthereviewsothattheteamfeelspreparedandconfident,especiallyifyouare
anewteamorateamnewtoScrum.It’sperfectlyacceptabletouseascriptorfollowthe
slideoutline.Inaddition,theteamshouldtakethetimetopreparethedemoenvironment.
Setitupsothatthecustomerscanputtheirhandsonthenewfeaturesandexperienceit
firsthand.You’llfindthattheteamalsolearnsmoreabouthowtheendusermightinteract
withtheproduct.Allowtimenotonlyforthemeetingbutforsomecasualchat-time
betweenthecustomerandtheteamafterthemeeting.Iallow30minutesbetweenthe
sprintreviewandtheretrospectivesothattheteamandthecustomerfeelfreetointeract
informally.
DocumentDecisions
Peopleallforgetthings,butitseemspeopleareespeciallygoodatforgettingthethings
thattheyneedtoremember,suchasaworkflow,acolorchoice,abuttonlocation.Whena
decisionismadeduringareview,writeitdownandkeepitsafe.I’vebeenintoomany
reviewswhereacustomersays,“Idon’trememberthatdecision.I’mnotpayingforthis
andthat.”
Havingawrittenrecordofverbalagreementsmeansthatevenifsomethinghappens
fourmonthsaftertheoriginalconversation,youareabletopointtothedocumentationand
say,“Yes,wediscussedthisandagreedtomoveinthisdirectiononSeptember1oflast
year.”
AskforAcceptance
Askingthecustomertoacceptthestoriesduringthemeetingallowsyoutoreleasesoon
afterthereview.Often,however,customerswanttimetousetheapplicationbefore
pushingitout,sosometeamsopttogivethecustomersuptoaweektoformallyaccept
thework.Regardlessofwhenyougetacceptance,makesureyougetit.
Acceptanceisimportantbecauseitputspressureontheteamtoensurethefeaturesmeet
thepublisheddefinitionofdone.Onewaytohelpthecustomersfeelcomfortable
acceptingonthespotistofollowyourdonelistfaithfully,highlightitduringeachreview,
andupdateitasneeded.Ifcustomersarestillhesitanttogiveyouthegreenlight,besure
toaskwhatyoucandotomakethemfeelmorecomfortableinacceptingworkthatyouall
agreemeetsthedefinitionofdone.
BeBrave
Evenwithplanning,sprintreviewscanbeintimidating.Helpteammembersfightthrough
thefearbyremindingthemthattheyarereadyforthisreview.Theyhaveimplemented
eachfeatureasdescribedbytheproductowner.Whentheyhadquestionsaboutafeature,
theyreachedouttotheproductownerforhelp—andreceivedit.Theyhavefollowedtheir
definitionofdone,whicheveryonehasagreedshouldresultinpotentiallyshippable
workingsoftware.Thedemoenvironmenthasbeenprepared.Slidesaretheretoguide
themthroughallthattalkingandallthosepeskyquestions.
Encouragethemtorememberthatanychangesandanynewstoriesthatarisefromthe
reviewarenotonlynaturalbutdesired.Withoutchanges,theproductwouldbeexactlyas
imaginedonthefirstdaytheproductbacklogwaswritten,whicheveryoneknowsisnever
preciselywhatthecustomerwants.Byembracingreviewsandthemyriadchangesthat
comefromeachencounterwiththecustomer,theteammovesonestepclosertothegoal:
afinalproductthatdelightstheenduser.Nowthat’sachangeeveryonecanlivewith!
WorksConsulted
Schwaber,Ken,andJeffSutherland.2013.“TheScrumGuide.”
http://www.scrumguides.org(accessed28October2015).Alsoavailableat
http://www.mitchlacey.com/resources/official-scrum-guide-current-and-past-versions.
Schwaber,Ken.2004.AgileProjectManagementwithScrum.Redmond,WA:Microsoft
Press,p.137.
Chapter16.Retrospectives
RetrospectivesareoftenoneofthefirstScrumelementstogobythewayside.Asschedule
pressuresmount,teamsfeeltheydon’thavetimeforwhattheyconsidertheluxuryofa
retrospective.Onceteamsstartskippingretrospectives,thedownwardspiralbegins.
Inthestorythatfollows,visitoneteamthatiscomingoutofitsthirdsprintand
struggling.Frustrationisrunninghighanddisciplineisrunninglow.Peoplearereadyto
abandontheretrospective—notatypicalforateamgettingstartedwithScrum.
TheStory
Jamieknewtheteamwasindangeroffallingapart.Storieswereslipping,buildswere
breaking,unittestswerefailing.Itfeltliketheprojectwascrashingaroundthem.
Everyonewasfrustrated.SeveralteammembershadcometoJamieindividuallyand
expresseddisappointmentwithotherteammembersandwithScrum.Tomakeitworse,
justabouteveryonewantedtoskipthisretrospective,citingthattheyhad“realworkto
do”and“besides,theseretrospectivesareacompletewasteoftime.”Theteamneededan
intervention.
Jamieknewthattheupcomingretrospectivewouldbeaperfectopportunitytoaddress
alltheseissuesbutwasunsurewhethertheteamwouldstepuptotheplatewithout
facilitation.Thepasttworetrospectiveshadnotbeenproductive.Looselyrunbytheteam,
theretrospectiveshadmainlyconsistedoflisteningtooneortwopeoplesoapbox,either
theircomplaintsortheirideasforimprovement.Attheendofthediatribes,everyoneleft
exhaustedanddeflated.Intheend,nothinghadchanged.
Jamiedecideditwastimetotakemattersintohisownhands.Hespentacouplehours
readingaboutwell-runretrospectivesandcameupwithaplan.Afewminutesbeforethe
scheduledretrospective,Jamiewentintothemeetingroomandremovedeverychair.He
alsohungthreewhiteboardsandlabeledthem“WhatWentWell,”“WhatNeedsto
Improve,”and“Rants.”
Astheteamfilteredintotheroom,thesechangescaughttheteammembersoffguard,
especiallyBenandTodd.
“Whatgives,Jamie?”askedBen.
“Yeah,whereareourchairs?”askedTodd.
“Iremovedthem.Everytimewemeetandtherearechairs,wegooffintola-lalandon
ourdiscussionsandnothingseemstogetdone.Itookthemoutsowecanstand,likeinour
dailyscrums,”statedJamiematter-of-factly.
Theteamletoutacollectivemoan.
“What’sallthis?”askedTodd,motioningtothewhiteboards.
“Simple,”saidJamie.“Forthissprint’sretrospective,wearegoingtotalkaboutwhat
wedidwell,whatweneedtofix,andrants.”
“Rants?”askedBen.
“Yeah,rants.Thesearethingsthatwejustplaindon’tlike,andtheydon’treallyfitin
thethingstoimprovecategory.Onmylastteam,ourteamspacewasreallyhotandone
teammemberkeptwriting,inbigletters,fixtheairconditioning.Mostofusjustthought
hewasmakingnoise,butafterseeinghimwritethatforthreesprints,werealized,‘Hey!It
isprettydarnhotinhere!’andwefixedit.”
Theteamnodded.
“Sowhatdowedonow,Jamie?”askedTodd,shiftingfromonefoottoanother,anxious
togetonwiththings.“Dowetaketurnsgoingaroundtheroomtellingyouwhatwelike
anddon’tlike,andyouwriteitdown?”
“Nope.Youaregoingtowriteitdown,”saidJamie,lookingatTodd.
ToddlookedoverhisshouldertoseeifJamiewasspeakingtohim.
“Uh,me?”saidTodd.
“Everyone.”AndwiththatJamiestartedthemeeting.“WhatIwanteachofyoutodois
grabamarkerandjustwriteonthewhiteboardwhatevercomestomind,thegoodandthe
bad.Don’tworryaboutthingsthatoverlaporaresimilar;thekeyhereistogatherdataso
wecanfigureoutwhattodoaboutit,ormaybewhatnottodoaboutitinsomecases.So,
upandat’em!”
Theteammemberslookedateachotherandbegantomoveslowlytowardtheboards.
Theyallstoppedthereandstared.
Toddsteppedupandwroteunderthingstoimprove,“Testcoveragesucks!!!”
TheteamlookedatToddasifhewereasheepamongwolves.
“See,that’sallthereistoit!”saidJamie,whointurnwrotesomethingunderthethings
thataregoingwellsection,“Teamworkandtrustareimproving.”
Onebyoneeachoftheseventeammemberswrotevariousitemsonthewhiteboard.
Afterabout20minutes,theteamhadidentifiedacoupledozen.Nowcamethehardpart,
sortingthedata.
Theteammembersspentthenext15minutesclarifyingwhattheyperceivedtobe
duplicatesandclearingthemfromthelist.Oncetheduplicateswereremoved,Jamie
introducedatechniquehehadlearnedathislastcompany.
“Team,whatwearegoingtodonowisprioritizethislist.Todothat,we’reallgoingto
imaginethatyoueachhaveonehundreddollarsinyourwallets.Youronehundreddollars
canonlybespentontheitemsontheboard.Ifsomethingisreallyimportanttoyou,you
mayspendallyourmoneyonit—it’sahundred-dollarissue.Otherissuesmightonlybe
worthfivedollarsorfiftydollarstoyou.Someitemsyoumightnotwanttopayforatall.
Divvyupyourmoneyanywayyouwantamongtheissuesthatmattertoyouorthatyou
feelaremostimpactingtheteam.Oncewehaveallspentourmoney,wewilladdupthe
paymentsforeachitem.Theitemsthatreceivedthemostmoneywillbeourhigher
priorityitems;thosewiththeleastmoneyareourlowerpriorityitems,”explainedJamie.
Toddheldhishandout.“I’mbroke.I’lltakethathundreddollarsnow,”hesaidwitha
smile.
Theroomchuckledandeveryonestartedspendingthevirtualmoney.After15minutes,
theteamhadaprioritizedlistofissuesthatitneededtodiscuss.Jamiecopiedeachof
themontoanewwhiteboardandputaplussign,aminussign,oranRnexttoeachoneto
symbolizewhetheritwasanissuethatwentwell,asuggestionforimprovement,orarant.
Jamielookedattheclockandsawtheyhadabout45minutesleftintheir90-minute
retrospective,nowherenearenoughtimetoworkthroughthehigh-priorityissues.Jamie
realizedthat,giventheitemsontheboard,theretrospectivewouldrunover.Sincehe
knewaddressingtheseissueswasvitaltothisteam,heaskedtheteamtheirthoughtson
extendingthemeetingby30minutes.“Wehavetoomanyitemsinourhigh-prioritygroup
tocoverinthetimeremaining,”Jamiebegan.“Basicmathsaysthatwecanonlyspend
fourminutesoneachitem;clearlythatwon’twork.Weneedtoconsidercullingthelistor
addingmoretime.Let’sdoagutcheck.Isthislistcorrect?Isthiswherewewanttostart?”
askedJamiepointingtothetopitem.
Theteamnoddedyes.
“Great,here’swhatIproposewedo.Ratherthancutthelist,let’saddthirtyminutesto
themeeting.Thatgivesusseventy-fiveminutes—enoughtimetogetthroughmostof
theseitemsifweaddresstheminseven-toten-minuteblocks.Soundreasonable?”
Theteameagerlynoddeditsagreement.
“Perfect,”Jamiecontinued.“Let’sspendthefirstcoupleminutesofeachtimebox
makingsureweallunderstandwhattheissueis.Oncewehaveclarifiedtheissue,we’ll
makeadecisiontoactonitortoskipitforthissprint—basicallytomitigateornot.Ifwe
choosetomitigate,thenwewilldecideonactionsthatwe,asateam,willworktowardto
eitherensurewecontinuedoingit(ifit’sagooditem)oralterourbehavior(ifit’sanitem
wewanttoimprove).Then,foreachitemwechoosetoacton,we’lldeterminewhowill
volunteertodriveit,”saidJamie.
“Driveit?”askedBen.
“Yes.Eachitemwechoosetoworkonormitigateneedstohavesomeoneontheteam
lookingafterit,soifabehaviororapatterncropsup,itcanbecalledout.It’sawayto
ensurenoneofusforgetwhatwedecidedtodointhismeeting—itkeepsthediscipline
andthefocus.”
Theteambeganworkingthroughtheissues,whichrangedfrom“pairprogramming
sucksandisworthless”to“retrospectivesareawasteoftime—justmoremeetings!”The
teamclosedwithaplanonhowtohandlesomeoftheirbiggestissuesforthenextsprint.
Everyonefeltgoodcomingoutofthemeeting.Theteamwalkedoutfeelinghopefulfor
thefirsttimeinalongtime.Theyhadnotjustcomplained;theyhadactuallycomeupwith
aplantofixthings.Asdrivers,eachpersonhadsomeresponsibilityforensuringthatthe
actionstheteamhadcommittedtotakewouldcontinuethroughoutthesprint.Itwasn’tall
rightyet,butitfeltlikeitmightgetbetter.
ThePractice
Retrospectivesarenoteasy.Theytaketime,theytakecommitment,andtheytakecourage.
Whenthepressure’son,they’reoftenthefirstthingtogo.Intherushtodeliveraproduct,
havingameetingtoventyourfrustrationcanseemlikeawasteoftime.Andifthat’sall
yourretrospectiveshavebecome,Iagree.Youarewastingyourtime.
Thatdoesn’tmean,however,thatyoushouldaxeretrospectives.Onthecontrary,ifyour
retrospectiveshavebecomemeaningless,it’sevenmoreimportantthatyoudothem,and
dothemright.Whyareretrospectivessoimportant,youask?Andwhataresomekey
componentsofaneffectiveretrospective?Thenextsectionsanswerbothofthose
questions.
GiveRetrospectivesTheirDueDiligence
Retrospectivesareakeypartofateam’sinspect-and-adaptcycle.Everyotherpartofthe
sprint(sprintplanning,dailyscrums,thesprintreview)focusesondeliveringworking
software.Theretrospectiveisthesoleopportunityforteammemberstoexaminehowthey
workandhowtheyareworkingtogether.It’satimeforthemtolearnhowtoimprove,
howtoworkmoreefficiently,andhowtodeliveratahighervelocityandwithsuperior
quality.Retrospectivesareaconstantreminderoftheneedtocontinuouslyimprove,to
alwaysbemovingforward.Becausewhenateamstandsstill,itfallsbehind.
Retrospectivesalsoremindteamsoftheimportanceofcontinuingnewtechnical
practices.Ononeofmyfirstprojects,wecuttheretrospectivebecausewehadtoomuch
worktodo.(Orsowethought.Itturnsoutourrealproblemwasinproperlysizingour
storiesandtasks.)Afterafewskippedretrospectives,wehadasprintwhereateam
memberwasoutoftheofficesickforfourdays.Nooneontheteamknewwhathewas
doingwiththecode.Sure,weknewthetasks,butwedidnotknowwhathe’ddone,where
hewasatwiththestory,orthedesignapproachhewastaking.Welostfourdaysofwork.
Howdidthishappen?Ithappenedbecausewefailedtoidentifyinaretrospectivethatwe
werefallingintooldhabitsandworkingonwhatweeachspecializedin.Weweren’t
distributinginformationandknowledgeinthecodethroughpairing.Itwasn’tuntilthis
eventthatwerealized,“Wow,Iguesstheretrospectivedoeshelp!”
Retrospectiveshaveamoresubtlebenefitaswell.Theygiveteamstheopportunityto
changebehaviorandteamculture.Theyareasafeplaceforteammemberstoshare,
constructively,feedbackwithothersonhowtheiractions(orinactions)impactthequality,
morale,andattitudeoftheteam.Leftalone,thesesmallproblemsfester,likean
unattendedcut.Retrospectives,ontheotherhand,allowteamstoattendtotheirsmall
woundsquickly,reducingthechanceofinfection.
PlananEffectiveRetrospective
Everyonecanagreethatretrospectivesarevaluable.Buthowdoyourunthemeffectively,
sothattheydon’tdegenerateintoanunproductivecomplain-and-moansession?Likeany
goodmeeting,retrospectivesrunmuchmoresmoothlyiftheyarewellplannedandwell
executed.Theteamshouldnotconsideraretrospectivedoneuntilithasasolidplanof
actionbasedonwhatwasdiscussedandawaytoholdeachotheraccountabletothe
commitments.
Planninganeffectiveretrospectiveisatrickything.Someteamsareabletoaccomplish
successfulretrospectiveswithnoplanning;someteamsneedadaytoplan.Whetheryou
aretheScrumMasterornot,ifyouaretheonefacilitatingthemeeting,youareresponsible
formakingsureafewcomponentsareinplacebeforethemeetingbegins:communication,
physicalsetup,andgroundrules.Remember,doingScrumdoesnotmeanyoudon’thave
toplanandthatyougettodowhateveryouwant,wheneveryouwant.
Communication
Goodcommunicationstartswithsettingexpectationsupfront.Atthebeginningofany
Scrumproject,publishallscheduledmeetingsandsendoutnotificationstoallinvited
parties.Retrospectivesshouldbeoneveryteammember’scalendar.Inaddition,priorto
eachretrospective,sendameetingreminderthatincludesanagenda.Thisagendashould
remindeveryonewhatthemeetingisabout,whyitisimportant,thebenefitstotheteam,
andhowtoprepare.Itshouldalsoincludearoughoutlineofhowthetimewillbespent
andasentencesummarizingtheendgoal.
PhysicalSetup
Newteamsoftenstrugglewithstayingonscheduleduringthemeetings.Assuch,Ioften
removethechairsfromtheroom,asJamiedidinthischapter’sstory.Iwanttoprevent
peoplefromsittingsothattheyfocusonmovingthroughthepartsofthemeetingwithout
endlessdiscussion.Ihavebeenintoomanyretrospectiveswhereeveryoneisleaningback
inachair,debatingwithnorealoutcome.Ihavefoundthatremovingthesittingaspectof
themeetinghelpskeeppeopleontrack.Ifyouwanttokeepyourseatandhaveamore
subtlereminderoftimemarchingon,youcouldalsodisplayatimerthatcountsdowneach
timeblock,perhapsflashingyelloworredaseachsectionofyourmeetingapproachesits
close.
Hangonthewallacopyofthesprintburndown,theteamdonelist,theteam’svision
fortheprojectandthesprintitself,andacopyofanynotesfromthelastretrospective.
Thenclearnewspacetocollectdata.Pre-labelwhiteboardsorflipchartswiththeitems
fordiscussion.Inthestory,Jamieused“whatwentwell,”“whatdoweneedtoimprove,”
and“rants.”DerbyandLarsen’sAgileRetrospectivesdescribesa“mad,sad,glad”
approachtodatacollection[DERBY].Whateverapproachyouuse,makesureit’sclear
whatdatatheteamneedstocollect.
GroundRules
Manymeetingsarederailedsimplybecausetherearenogroundrules.Basicgroundrules
thatshouldbepostedintheroomandverballysharedatthebeginningofeachmeeting
includethefollowing:
Berespectful.
Donotinterrupt.
Putawaylaptopsandphones.
Parklongdiscussionsinthedesignatedlot.
Recapwhatyouheartoensureunderstanding.
Whatissaidherestayshere.
Egos,Emotions,andFocus
Everyoneontheteamshouldattendtheretrospective:BythatImeantheScrumMaster,
theproductowner,andthecoreteammembers.Teamconsultantsareoptional.
AstheScrumMasterfacilitatingthemeeting,youshouldbeonthelookoutfor
distractions,disrespectfulbehaviororcomments,andthefocus(orlackthereof)that
peoplemighthave.Ihaveseentoomanytimeswhereoneteammemberhijacksthe
retrospectivetotelltheotherteammemberswheretheywentwrong,tookabadapproach,
didn’tlistencorrectly—younameit.It’snotthatthisisbad;onthecontrary,opennessisa
corevalueofScrum.However,there’sarightwayandawrongwaytohavethese
discussions.Ifpeoplefeeldisrespected,theyshutdownandtheretrospectivecomestoa
grindinghaltofsilence,headnodding,andforcedagreementtowhatevertheaccusing
personissaying.
TheScrumMaster’sjob,then,istonoticethisaccusatorybehaviorandkeepitatbayas
muchaspossible.Ifitisarecurringpattern,theScrumMastershouldspeaktotheperson
whoiscausingtheissueoutsideofthemeeting,explaininghowthebehavioriscausinga
distractionandactuallyblockingtheteamfrombecominghighperforming.Ifthepattern
continues,moredrasticmeasuresmayneedtobetaken,suchasremovingthepersonfrom
theteamoraskinghimorhernottoattendtheretrospective.
RuntheRetrospective
AstheScrumMaster,arriveinthespaceabout15minutesearlytosetuptheroom.And
beginthemeetingassoonaseveryonearrives.Howeveryouchoosetorunthemeeting,
thefirst15minutesorsoshouldbededicatedtodatacollection.Thiscouldbeeveryone
writingonwhiteboards(asinthestory),open-space-stylestickynotes(withpeoplecalling
outtheirissues),orascribebeingresponsibleforwritingdownideasonwhiteboards.I’ve
usedallthesetechniques;therightonedependsonthegroupdynamics.Thefacilitator
shouldwatchforinteractionsbetweenpeople.Ifpeoplearewritingontheboardsor
flipcharts,theroomisusuallyveryquietatthispoint.Watchcarefullyandlistenforthe
soundsofsilentagreementordisagreement,perhapsloudhighlightingofsomething
someoneelsewrote,andmakeamentalnote.Noticeifsomeoneisnotparticipating,
hangingbackandnotengaging.Don’tsayanythinginthemeeting!Justmakeamental
notetoseehowthatpersoninteractsonceitcomestimetodiscuss.Ifsomeoneissilent
throughouttheretrospective,youshouldhaveaone-on-onediscussionwiththatindividual
afterthemeetingtogettotherootoftheissue.Onceallthedatahasbeencollected,takea
pictureofanyartifactsintheirrawstate.
Aftertheideashavebeencollected,theyneedtobeprioritized.Therearemany
prioritizationtechniques,butIfindthe“buyanissue”techniqueusedinthestorytobe
veryeffective.Itworkslikethis.Everyonehassomevirtualcurrency.Itcanbeanything.I
haveseenpeopleusePing-Pongballs,playmoney,evenjustplain“units.”Theimportant
thingistomakesurethateveryonehasthesameamountandeveryonehastheright
amount.
Iftherearemanyhigh-priorityitemsonthewhiteboardandeachpersononlyhasthree
unitstodistribute,importantthingswillgetlost.Atthesametime,ifeveryonehas1,000
unitstodistribute,toomanyitemswillbechosen.Ifyouhaveafive-personteam,start
with$25or$50.Ifyouhavealargerteam,startwith$75or$100.Youcanalways
increasetheamountifitlookslikethehigh-prioritylististoosmall.Youcanfurtherlimit
theactionbyhavingpeoplespendin$5increments.(LukeHohmanndescribesasimilar
gamecalled“BuyaFeature”inhisbookInnovationGames[HOHMANN].Itcanalsobe
usedinthissituation.)
Aftereverythinghasbeenallocated,sumupthetotals.Younowhaveyourprioritized
list.Again,takeapictureatthispointanddoagutcheck—doeseveryonethinkthisisa
goodlist?Issomethingsoblatantlyobviousthatitshouldbereconsidered?Usecommon
sense.
Nowthatthingsareprioritized,theteamneedstimetodiscussthehigh-priorityissues
andmakedecisions.Timeboxthispartofthemeetingbasedontheamountoftimeyou
haveleftandthenumberofitemsyouneedtodiscuss.Bringaneggtimerifpeoplehavea
hardtimegettingtothepoint.Onceeveryoneunderstandstheissue,ask,“Isthis
somethingwearegoingtomitigate/fixornot?”Ifno,discusswhynot.Then,moveonto
thenextitem.WhenIfacilitateretrospectivesandIencounteranissuethattheteam
decidesnottofix,Iwriteitdowninmynotestocaptureontheteamwiki.Thatway,when
theteamcomesbackandsays,“Whydidn’tweeverfixthis?”Icantellthemexactlywhy
wedecidednottodoit.Ifinditjusthelpswithdealingwithnoiselateron.
Forthethingsyoudowanttoworkonormitigate,comeupwithaplan.Forcomplex
issueswheresolutionsarenotimmediatelyapparent,nameoneortwopeopleresponsible
forcomingupwithideasandoptionsthattheteamcanactupon.Keeptheactivity
timeboxedtoacoupledays,haveafollow-upmeeting,anddecideonacourseofaction.
YouwillanswerquestionssuchasWhatarethethingspreventingXfromhappening?How
canweworkaroundit?Thisisateamexercise,somakesureeveryoneparticipates.
LikeJamiedidinthischapter’sstory,onceyouhaveaplaninplace,writeitdownand
asksomeonetohelpdriveit.Thedriveristhepersonontheteamwholooksforthat
specificbehaviororpatternandalertstheteamwhenitoccurs.Rememberthattheteam
shouldn’tjustworkonitemsthatneedimprovement;itshouldalsocontinuedoingthings
thatareworkingwell.Ononeofmyprojects,wewerenotswappingpairsenough,so
someonechosetodrivethat.Thenextsprint,thedriverspentalotoftimekeepinguson
track.Wechosetocontinueemphasizingswitchingduringournextsprint,eventhoughit
hadimproved.Duringthefollowingsprint,ourswitchingwasbetter,butwestillheard
fromourdriver,“Hey!Youswitched!Iguessyoudon’tneedmetotellyou.”Itwasa
formofpositivereinforcementthatworkedwell.
Onceyouhavealistofactionableitems,orthetimehasexpired,it’stimetoclosedown
theretrospective.Iliketoenditwithaone-to-fivequickratingof“Howwasthissprint?”
and“Whatdoyouthinknextsprintwillbe?”Oneisbad;fiveisgood.Peoplecananswer
anywaytheywant,butItrytoweedoutanyillwillintheteamthatmayrequiresome
discussion.Icareabouttheanswer,sure,butI’mmoreinterestedinhowpeopleanswer.
WhatIlookforarenonverbalcluesthatmayindicatethatwhatpeoplearesayingdoesnot
matchwhattheyarefeeling.Idon’tcallitoutinthemeeting;insteadIapproachthe
personone-on-one,tellhimwhatIobserved,andjustaskifthereisanythingwrong.Ifthe
personsaysthateverythingisfineagain,I’llletitgoandjustobservetheteamoverthe
sprint.
Afterthemeeting,theScrumMasterorfacilitatorneedstostickaroundtodocument
things.Itakedigitalpicturesandnotesandputthingsintoawiki.Ineverassociatenames
andonlyusethepicturesformyreference.Idon’tsharethembecausethenotesmightbe
misunderstoodoutofcontextorsomeonemightrecognizeaparticularindividual’s
handwriting.Partofthefacilitator’sjobistoensurethatwhathappensintheretrospective
staysintheretrospective.
KeystoSuccess
DerbyandLarsensummarizeitbestinAgileRetrospectives:“Productiveteamsjudge
retrospectivesbytheirresults”[DERBY,p.145].Icouldnotagreemore.The
retrospectiveisacrucialpartoftheinspect-and-adaptcycleinScrum.Itenablesteamsto
identifywheretheyarefallingshortandwheretheyaredoingwell.Thoughtheyare
neithereasynortrivial,retrospectivesshouldfeelgood.
Evenwithallthegoodreasonstodoretrospectives,theyarestillthefirstthingmost
teamscutwhenthey’refeelingpressuredtoperformfaster,delivermore,orincrease
quality.What’sironicisthattheyarecuttingtheonethingthatcouldhelpthemincrease
boththeirvelocityandtheirquality.Ifyourteamismakingnoiseaboutcutting
retrospectives,thefollowingactionswillhelpthemreconsider:
Showthemthewhy.
Buildagoodenvironment.
Holdthemwhenyouneedthem.
Treatretrospectiveslikethefirst-classcitizenstheyare.
ShowThemtheWhy
NewScrumteamssometimesdropretrospectivesbecausetheyfailtounderstandthe
purposeofthemeetings—thewhy.Droppingretrospectivescancausecatastrophic
problems—glitchesthatarerelativelyeasytosolvewhentheyarefreshcanstagnate,
fester,andgrowintolargerissues.Commonissuesthatmanifestthemselveswhenthe
retrospectiveiscutincludedropsinquality,avarietyoffailingtests,loweringofteam
morale,andmore.Droppingtheretrospectiveremovestheteam’sopportunitytoreflect.
Withoutreflection,theteamgetslostandwakesuponedaytolookinthemirroratavery
differentpicture.It’salmostlikegroomingyourself—mostpeopleshowerandbrushtheir
teethdailybecause,iftheydon’t,problemsstarttomanifest.Withoutlearningand
continuousimprovement,teamswillfallbacktotheiroldhabitsandbegintowonderwhat
happened—andtheywillusuallyblameScrumfortheirfailure.
BuildaGoodEnvironment
Retrospectivesdriveteamstochangetheirbehaviorovertime.Toensurethatyour
retrospectivesyieldresults,makesurethatteammembersunderstandsomebasic
principles:
Theretrospectiveisfortheteamandnooneelse.Thismeansdon’tgoblabbing
aboutwhatsomeonesaid.
Buildconfidencethroughsmallchanges.Takingontoomuchatonetimecanbuild
frustration.Takingonsmall,calculatedchangesandaccomplishingthembuilds
confidence.
Ownthechanges.Theteam,notjusttheScrumMaster,isresponsibleforits
changes.Getteammembersinvolvedindrivingthechangesidentifiedinthe
retrospective.
HoldThemWhenYouNeedThem
Retrospectivesarenotlimitedjusttosprints.Infact,itiscommontoseeretrospectives
afterarelease,alargeunexpectedissue,anything.Rememberwhyteamsdothem:sothat
theycanstop,lookatthesituation,andthenmakeadjustments.Inapostmortem,
everythinghasalreadyhappened,andtheteam’sabilitytoeffectchangeisgone.Ina
retrospective,teamsusuallyarestillactivelyengagedintheproject.Consideradding
higher-levelretrospectivesforreleases(sayeverythreemonths)and,incaseofsudden
environmentalchangessuchasunexpectedoutages,scalingupforlargerteamswithmore
people.Youcanalsoholdoccasionalretrospectivesbetweentheteamandthecustomers.
Thesecustomerretrospectivescanhelpyouidentifyareasforimprovementbetweenthe
teamandthecustomerandtheproductowner.
TreatRetrospectivesLiketheFirst-ClassCitizensTheyAre
RetrospectivesdeservemoreattentionthanIcangivetheminonechapter.Theyare
similartothesprintreviews,whichdefinethewhat,whiletheretrospectiveismoremeta
andisaboutthehow.Retrospectivesandreviewsareforumsforidentifyingandsolving
problemsandissues,formakingpeopleawareofbothwhatisworkingandwhatisnot
working,andalsoforbuildingtheteam.
Formaximumeffectiveness,youshouldreadProjectRetrospectives:AHandbookfor
TeamReviewsbyNormanKerth[KERTH]andAgileRetrospectives:MakingGoodTeams
GreatbyEstherDerbyandDianaLarsen[DERBY].WhenIfirstreadKerth’sbook,it
influencedhowIranprojectpostmortems.ItwasnotuntilImovedtoScrumandXP,
however,thatIfullyrealizedthepotentialofawell-executedretrospective.
Don’twaittomakeretrospectivesavitalpartofyoursprints.Dothework.Createthe
plans.Helptheteamgettheresultstheywant.Anddon’tsettleforonce-a-sprintvent
sessions.Whentheteammembersbegintounderstandhoweffectiveretrospectivescan
be,theywon’teverconsidercuttingthem.
References
[DERBY]Derby,Esther,andDianaLarsen.2006.AgileRetrospectives:MakingGood
TeamsGreat.Raleigh,NC:PragmaticBookshelf.
[HOHMANN]Hohmann,Luke.2007.InnovationGames:CreatingBreakthrough
ProductsThroughCollaborativePlay.UpperSaddleRiver,NJ:Addison-Wesley.
[KERTH]Kerth,Norman.2001.ProjectRetrospectives:AHandbookforTeamReviews.
NewYork:DorsetHouse.
PartIII:FirstAid
Chapter17.FacilitatingaProductiveDailyScrum
Thedailyscrum,ordailystandup,seemslikeasimpleconcept.Gathertogetherfor15
minuteseachdaytoanswerthreequestionsdesignedtohelpteamssyncupandensure
theyareontrackforcompletingthesprintgoal.Yettoooften,dailyscrumsdegenerate
intodailygrinds,eitherbecausetheydragoninterminably,becausetheylacksubstance,or
worstofall,becausetheyhavebecomefinger-pointingblamesessions.Common
symptomsofdysfunctionaldailyscrumsincludethefollowing:
Deepdives
Latecomers
Inattentiveness
Updatesthataretoolong/toovague
Exceedingthe15-minutetimebox
Recognizingandaddressingtheseall-too-commonproblemsisessentialtofacilitatinga
productivedailyscrum.Thefollowingstoryillustratesanineffectivedailyscrum.Therest
ofthechapterdiscusseswhatwentwrong,howtofixit,andwaystoensurethatyour
team’sdailyscrumbecomesandstaysproductive.
TheStory
TiagowastheScrumMasterforateam.Thoughhehadpreviouslyworkedasasenior
developerforthecompany,hewasrelativelynewtotheScrumMasterrole.Theother
teammemberswereCesar,Pedro,Nuño,Marco,Ricardo,andMariana.LikeTiago,the
teamwasnewtoScrum—bythethirdtwo-weeksprint,theteamwasstillfiguringout
“thisScrumstuff.”
Attheappointedhour,Tiagostartedthedailyscrum.“Goodmorning.It’sgoodtosee
everyone.”Tiagolookedaroundatthefacesinthecircle.“Wait,noteveryone.Hasanyone
seenCesarandMarco?”
“Ithinktheyaregettingcoffee,”saidPedro.
“Yes,Ijustgotbackfromthecoffeeshop,andtheywereinline.Whilewearewaiting,
I’dliketodiscussanarchitecturalissuewe’rehaving,”Nuñoadded.
NuñostartedtodiscusshisissuewithPedro,Ricardo,andMariana.Theyallmovedto
thewhiteboardtostarthashingouttheproblem.
“Holdon,team,weneedtowaittostartthemeeting.CesarandMarcoarenothereyet,”
saidTiago.
Nuñostoppedwriting,glancedoveratTiago,andreplied,“Whentheygethere,I’ll
stop.”Thefouratthewhiteboardthencontinuedtheirarchitecturaldiscussion.
Fiveminuteslater,CesarandMarcoshowedup.
“Hey,guys!Sorrywe’relate.Wehadacrazynightlastnight.ThisbandPivitplayed
neartheuniversity,andman,whatagreatshow!”saidCesar.
Tiagochimedin,“Okay,guys,weareallhere,solet’sstartthemeeting.Who’sfirst?”
Cesarsaidhewouldgo.“YesterdayIworkedonthexmlwebservice.Igotstuckonone
ofthedataelements,butIthinkI’vefigureditout.I’mgoingtoaskPedrotohelpmewith
ittoday,andI’llbedone,”hesaid.
“Thankyou,Cesar.Howmuchtimedoyouhaveleftonyourtask?”askedTiago.
“Ummm,aboutsixhours,”saidCesar.
“Yesterdayyousaidthesamething:sixhours.So,nowyou’resayingyouhavesix
morehours?”askedTiago.
“Yup,it’sjustnotworkingoutthewayIexpected,”saidCesar.
Pedrojumpedin.“Cesar,IthinkIknowwhyit’snotworkingthewayyouexpect.Iwas
lookingatthewebinterfaceyesterdayandsawsomeunexpectedfeaturesintherethatI
thinkweshouldaddress.”
“Doyoumeanbugs?”askedMariana.
“Yes,bugs.Butlook,”Pedrowenttothewhiteboard.“Ifweimplementtheclassinthis
fashion….”Hetrailedoffashestartedtosketchasolution.
Tiagoquietlywalkedtotheotherendofthewhiteboardandwroteanoteonit:Pedroto
workwithCesaronwebserviceissue.
“Okay,guys.I’vemarkedanotetodiscussthisafterthemeeting.Let’smoveon,”said
Tiago.
“I’llgonext!”saidanenthusiasticPedro.“YesterdayIwasworkingontheperformance
testswithMariana,andIcameacrossthisissuethatseemstobeimpactingthewaythe
dataispassedthroughthesequencingsubclass.Westartedlookingintoit,andIthinkwe
havealargerarchitecturalissuethatneedsourattention.Also,Ithinkitmightbe
impactingCesarbecausethesymptomshedescribedshowthatthewaywearehandling
ourclassvariableswillresultinbadormalformeddatabeingpassed.IthinkMarianaand
Ishouldkeepworkingonthesetests,butIalsothinkweallneedtogettogethertoupdate
thesystemarchitectureintheseareas.Idon’twanttojustrefactorit;Ithinkitneedstobe
redesigned.Ifweapproachthislikewedoourotherclassesandmodules,Ithinkwe’llbe
okay—here,letmeshowyoumyidea.”
Withmarkerstillinhandfromhislastsketch,Pedrobegantoillustratethisissueonthe
whiteboardaswell.
Becausetheywereallimpactedbythisillustration;Cesar,Mariana,andNuñoall
jumpedintotheimpromptusession.Ricardo,ontheotherhand,wasbecomingvisibly
bored.HeappearedtoTiagotobetextingwithhismobilephone.Afterafewminutes,
Ricardolefttheroom.Tenminuteslater,Ricardoreturned.Thewhiteboardsessionwas
stillinfullswing.RicardoglancedoveratTiagowitha“pleaseendthis”look.
“Okay,team.Itsoundslikethereisablockingissuehere.I’llmarkitdown,andwecan
moveon,”saidTiago.
Marianaturnedfromthewhiteboardandbeganspeaking,“YesterdayIworkedonthe
performancetestswithPedro.I’dliketokeepworkingonthat.Ialsowanttoworkonthis
architecturalissuethatsurfaced,soI’llbespendingtimeonthataswell.Oh,andIhavea
dentistappointmenttodayafterlunch;Iforgotaboutthat,soI’llbeouttherestoftheday.
Mysonhasonetomorrowmorning,soI’llbeinlate,probablyaroundnoon.”
“Whatwillyouworkontoday,Mariana?”askedTiago.
“I’mnotsure.Imean,I’vegotthisarchitecturalissue,soI’llprobablyjustworkonthat.
Oh,andtheperformancetests.AfterI’mdonewiththose,though,Idon’thaveanywork
lefttodothissprint,”saidMariana.
Ricardojumpedin,“Whatdoyoumean?Wehaveabacklogfullofwork.”
“Yes,butnoneofthetasksaretestingtasks,soI’vegotnothingtodo,”saidMariana.
Ricardonoddedhisunderstanding.
“I’llgonext,”saidMarco.“YesterdayIworkedonthisissuefrommyoldproject.It
tooklongerthanIthoughttofix,butit’saproductionissue,andI’mtheonlyonewhocan
fixit.Ithinkitwilltakemetherestoftheweek.Idon’thaveanyblockingissues.”
“Fine.Who’snext?”saidTiago.
Pedrochimedinagain.“Hey,Marco,Iknowsomethingaboutthatissue.Iwasoneof
thepeoplewhofoundit.Icanworkwithyouonittodayifyouwant.”Onceagain,he
movedtothewhiteboardtoillustratewherehehadfoundtheissue.
Mariana’seyeslitup.“I’dliketohelponthatwithyouaswell,Marco,”sheadded.
ThenextthingTiagoknew,halfoftheteamhadvolunteeredtoworkonaproduction
issueforacompletelydifferentproject.Hisdailyscrumhadturnedintoaproblem-solving
session.
“Now,holdon,”saidTiago.“Wehavecommittedtodoworkonourcurrentproject.We
needtostayfocused.Wecan’tallgooffandworkonthisissue!”
“Well,I’vegottoworkonit,”saidMarco.“Thereisnooneelsetodothework;it’sgot
tobeme.”
“We’llseewhatwecando,Marco,whatI’dexpectis—”saidTiago,buthewas
interruptedbyMarco.
“Tiago,oneotherthing,Ithinkit’stoomuchoverheadforustomeeteveryday—these
dailyscrumsseemtobeawaste.Weshouldmakethismeetingthirtyminutesandonlydo
ittwiceaweek;thatwouldbeamuchbetteruseofourtime.”
“Iagree,”saidCesar,muchtotheshockofTiago.“AndIthinkweshouldbesitting;all
thisstandingisabitofajoke!”
“Holdonaminute.Wecandiscussthisoffline;weneedtofinishthismeetingandstay
focused!”saidTiago.
Withasignofresignation,TiagoturnedtoRicardo.
“Ricardo,you’renext.”
“YesterdayIaccomplishedtheETLfromtheCRMwarehouse.TodayIplanonrunning
thedatavalidationteststoensureeverythingistransferringasexpected.I’llhavethat
finishedtodayandwillupdateallmytasksbeforeIgohome.Ihavenoblockingissues.”
“Thankyou,Ricardo.Thatwasgood,”saidTiago.“Iamabitconcernedontherestof
ourtasks,though.Haveanyofyoulookedattheburndownchartlately?Itlookslikewe
haveflatlined.”
“Sorry,that’smyfault.Iidentifiedsomenewtasksoverthelastfewdays,andIforgot
toupdatethem,sotheburndownsareabitoff,”saidCesar.“I’llgetthemupdatedright
afterthemeeting.”
Theteamstartedtodisperse,someofthemtowardthewhiteboard,otherstowardthe
coffeestation.Ricardomadeabeelineforhisdesk.Tiagoglanceddownathiswatch.“A
45-minutedailyscrum,”hethought.“Unacceptable.”
TheModel
Thedailyscrum/dailystandupisashortmeetingthattheteamholdseachdaytoshare
whateachpersonhasaccomplished,isworkingonnoworwillaccomplishthatday,and
anyimpedimentsorissuesthathavecroppedup.Forteamsnewandold,thisdaily
meetingisoftenastruggle.JustlookattheissuesTiagoencounteredwiththeteaminthis
chapter’sstory:
Peopleshowinguplate
Requeststocondensethemeetingtotwiceaweek
Wantingtositinsteadofstand
Tasksizesincreasingwithoutunderstandingtheirimpactontheoverallsprint
Deep-diveproblemsolving
Ramblingupdatesandlackofclarity
Lackofteamwork
Legacyprojectsconsumingtime
Lackoftransparency
Clearingtheseimpedimentstoaproductivedailyscrumcanseemdauntingbutis
imperative.Thetruthis,allteamscanhaveeffectivemeetings.Itjusttakessomework.
TimeofDay
Iamaskedeveryweek,“Whenshouldweholdthedailymeeting?”Somepeoplelikethe
endoftheday,somepeoplelikethebeginningoftheday,and,yes,somepeoplelikethe
middleoftheday.Irecommendholdingthedailyscrumabout15to30minutesafterthe
lastteammemberarrivesinthemorning.Thismeansthatifyourteamisanearlybunch,
arrivingtoworkby8:00,theteamshouldholdthemeetingaround8:15or8:30.Ifthelast
personcomesinat9:30,havethemeetingat9:45or10:00.Ifindthathavingthedaily
scrumatthebeginningoftheworkdayismuchmoreeffectiveforcoordinatingateam’s
actionsthanhavingitatothertimes.Ithelpstheteamstartfresh,laydowntheplanforthe
day,andgo.
StartandEndonTime
Aneffectivemeetingbeginsbystartingandendingontime.Ifsixpeoplespendanextra
fiveminutesatameeting(whetheritwasbecauseitstartedlateorendedlate),you’ve
wasted30minutes(sixpeopletimesfiveminutes).Ifyoudothateveryworkingday,
you’vewastedtwoandahalfhoursoftheweektalkingwhenyoucouldhavebeen
working.Inthischapter’sstory,Tiago’steamdidnotstartontime,andthereseemedtobe
noconsequenceforbeinglate.Asaresult,hismeetingstartedoutonachaoticnoteand
onlygotworseasitwenton.Luckily,beginningandendingasscheduledissomething
youandyourteamcanmakehappen.
LateStarts
Peoplemustarriveontime.Don’tacceptexcuses.TeammembersortheScrumMaster
shouldpolitelyremindanylatecomersthattherestoftheteamspentxminuteswaitingon
them.Ifsheerpeerpressurewon’tdoit,trypunishment.Seriously,alittlepaingoesalong
waytowardcorrectingbadhabits.Onpastprojects,myteamshavehadtheFridayFund,
wherelatecomersmakeadonationtothefund.Attheendofthesprint,theteamusesthe
fundtogooutforabeerordonatestheproceedstoacharity.Usedcorrectly,thiscanbea
greatmotivationaltool.Still,youmighthavethatoneteammemberwhosays,“I’lljust
paytwentydollarsrightnowandbelateallthetime.”Ifthathappens,you’llhaveto
consideralternativeformsofpunishment.
AnotheroptionthatIlikeisthe“pickyourownmedicine”approach;peoplewillbe
morecommittedtoarrivingontimeiftheychooseapunishmentthatmotivatesthem.
Once,whenourteamdidthis,Ichosetodo29pushupsifIwaslate,anotherteammember
chosetopayadollartotheFridayFund,andanotherchosetogowithoutcoffeeforthe
day.Thedaythatteammemberwaslatewasnotagoodday.Twohoursafterthemeeting,
theentireteambroughthimafullpotofcoffeeandbeggedhimtodrinkittomakehima
tolerableperson.Ifpeerpressureandlightheartedpunishmentsdonotwork,the
ScrumMasterneedstohaveatalkwiththeregularoffender.Bottomline:Everyoneonthe
teammustarriveontime.
Agenda,Rhythm,andLayout
Themeetingmustendontime,everytime.Somepeoplewilltrytodominatethe
conversation,otherswilldodeepdives.Stillotherswillstarttoramble.Topreventdaily
scrumsfromderailing,rememberthatuntildailyscrumsareahabit,rhythmiseverything.
Tostarteveryoneoutonthesamebeat,assoonaseveryonearrives(ontime,seeabove),
theScrumMasterorsomeoneontheteamshouldrestatethepurposeofthemeetingby
saying,“It’sgoodtoseeeveryone.Weareheretoanswerourquestions.Thesequestions
are,Whatdidyoudosincethelastmeeting?,Whatwillyoudotoday?,andWhat
impedimentsdoyouhave?(someteamsaddafourthquestion;seeChapter18,“The
FourthQuestioninScrum”).Whowouldliketostart?”Repeatthismantraeverysingle
day.Repetitionleadstomusclememory.Andgoodmeetingmusclememoryisessential
forhavinganeffectivemeeting.Ifyounoticethateveryoneisstandinginalineorinsmall
groups,suggestthattheteammembersformacircleandlookateachother—theorderof
whoisnextismoreobvious,andthetemptationtotextorhavesideconversationsis
reduced,sothemeetingtendstomovealongfaster.Italsoencouragesteammembersto
reporttoeachotherratherthantotheScrumMaster.
Interruptions
Thenextchallengeistoreduceinterruptions.Everyoneshouldbefocusedontheperson
whoisspeaking.Phonesandcomputersshouldbeputawayandsilent(orplacedintoa
basketbeforethemeetingbeginsiftheycontinuetobeadistraction).Ialsorecommend
usingatalkingobjecttokeeppeoplefromspeakingoutofturn.Thefirstpersontospeak
holdsanobject,whichhewillpassonwhenheisdone.Therulesaresimple:Whenyou
areholdingthetalkingobject,yougettotalk;whenyouarenotholdingtheobject,you
don’t.IliketouseNerffootballsastalkingobjectsbecausetheyarealwaysfuntothrow
andusuallydon’tbreakanything.Ifateammember(ormembers)interruptsanother’s
turn,thepersonspeaking(ortheScrumMaster)cangentlycuttheinterrupteroffby
wavingtoorpointingtotheobjectasavisualindicatorthat“It’smyturntotalk.”Youcan
alsoreinforcetherulebyaskingteammemberstopicktheirownmedicinewheneverthey
interrupt.
RamblesandDeepDives
Manyteamsstruggletocurtailthosewhoramblewhenansweringthethreequestionsor
begintodivedeepintotheproblemstheyarefacing.Ramblinganswersusuallymeanthat
teammembersarecomingtothemeetingunprepared.TheanswerthatPedrogaveinthis
chapter’sstoryisagoodexampleofaramble:
YesterdayIwasworkingontheperformancetestswithMariana,andIcameacross
thisissuethatseemstobeimpactingthewaythedataispassedthroughthe
sequencingsubclass.Westartedlookingintoit,andIthinkwehavealarger
architecturalissuethatneedsourattention.Also,Ithink….
Pedrogoesonandon.Thepurposeoflimitingthenumberofquestionsistokeepthings
movingandtokeeppeoplefocused.Everyteammembershouldcometoeverydaily
scrumhavingalreadyconsideredtheanswerstothethreequestions.Lackofpreparedness
isalmostasbadasbeinglatebecausethoughtheteammembersshowupontime,theyare
notreallythere.
Deepdivestypicallystartaslong-winded,ramblinganswerson“whatIdidyesterday”
andsteamrolloutofcontrolwhenotherpeoplejoinintheconversation.Thedailyscrumis
anopportunitytoexposeimpediments,notatimetocomeupwithsolutionstothe
impedimentsthathavesurfaced.Doyourememberthefirsttimethishappenedinthis
chapter’sstory?Pedrowasattheboard,describingthebugshehadfound.Pedrowasnot
tryingtoderailthemeeting;hewastryingtohelpsolvetheproblemhisteammatewas
having.Withoutinterrupting,Tiagowroteanoteonthewhiteboardtoremindeveryone
thatmorediscussionwasneeded.Hethensignaledtheteamthatitwastimetomoveon.
Byusingthisvirtual“parkinglot”forideas,Tiagodidtherightthing.Hisquietactions
keptthemeetingflowingbutcapturedthefactthataconversationwouldneedtotake
placeoutsidethedailyscrum.ThesecondtimePedroapproachedtheboardandbegan
sketching,though,Tiagoandtheotherteammembersdidnothingtointerruptthedeep
dive—notagoodthingforthemeeting,especiallysinceRicardoactuallyleftwhile
everyoneelsewasproblemsolving.
Assoonasthedeepdivebegins,someoneontheteamcanmakearolling,up-and-down
motionwithhishand(orotherwisesignaltheteam),asshowninFigure17-1;thenthe
ScrumMasterorspeakercanwriteanoteabouttheconversationthreadonthewhiteboard
inanareacalled“discussafterthemeeting”or“parkinglot.”Thisimmediatelyputsthe
teambackonthepathtoresurface.Ifthespeakerstilldoesnotreturntohisthree
questions,theScrumMastershouldremindthewaywardspeakerthattheteamiscurrently
focusingonupdatesandwillexplorehisconcernsindepthafterthemeeting.
FIGURE17-1Tosignaladeepdive,moveyourhandfollowingthispattern.
ExposeHiddenImpediments
Thedailyscrumisaboutmorethanhavingpeoplearriveontime,payattention,andgive
succinctupdates.Ultimately,thedailyscrumisintendedtoencourageteamcollaboration
andexposeanyproblemsstandinginthewayoftheteamreachingthesprintgoal.Daily
scrums,however,caneasilydegenerateintodailyprojectmanagementstatusupdates.
Lookforsignsthatyourdailyscrumisnotaccomplishingitsgoals.Warningsigns
includeteammemberswhoglossoverproblemsandwhoareoverlyvague.Eachofthese
behaviorscanstandinthewayofatrulyproductivedailyscrum.
GlossingoverProblems
Teammemberssometimesgivesuccinctupdatesbutfailtonoticeorhighlightproblems
thattheiranswersreveal.Forexample,inthischapter’sstory,Cesarstucktothescriptbut
glossedoverthefactthathistaskhadgrowninsizesincethedaybefore.Tiagodidthe
rightthingbyinterruptingtopointoutthefactthatthetaskthatwassixhoursyesterday
stillhadsixhourslefttoday.Sometimestasksgrowinsize,sothefactthatitoccurredin
thestoryisnottheproblem.ThedysfunctionisthatCesardidnotknowwhythesizewas
increasingorhowhewasgoingtocompletethetask.
Tiagopointedouttheproblemand,whenPedroofferedapotentialsolution,alsotook
thenextstepofnotingtheimpedimentontheteamwhiteboardforfollow-upafterthe
meeting.Toreachtheirgoals,teamsshouldlookforandhighlightanyissues,thenmake
planstoaddressthemafterthemeeting.
BeingVague
Often,teammembersgiveanswersthatareoverlyvague.Agoodexamplefromthis
chapter’sstorywasMariana’sanswer,whichcouldbeparaphrasedas“YesterdayIdid
sometestingandtodayIwilldosomemoretestingandIhavenoimpediments.”While
thistypeofresponsemighttechnicallyanswerthequestions,itdoesnothelptheteamor
providevaluableinsightintowhatworkisinprogress.Thisresultsinalackof
transparencyandincreasestheriskofhavingtwotasksworkedonsimultaneously.
Tohelpreducethistendency,tryrewordingthequestion“Whatdidyouworkon
yesterday?”to“Whatdidyouaccomplishyesterday?”Thatonewordservestoincrease
specificityandhelphighlighthiddenimpedimentsthatmightwarrantinvestigation.Italso
helpsteammembersfocusmorefullyonwhatthey’veactuallyaccomplishedsincethelast
dailyscrum.
EndwiththeBeginninginMind
Onceeveryonehashadanopportunitytospeak,theteamshouldquicklyreviewanyitems
thatwerecapturedonthewhiteboard.Then,themeetingshouldbeofficiallyclosedsothat
thedailycyclecanbegin.Someteammemberswillreturntotheirdesks;othersmightstay
todiscusstheitemsthatimpacttheirwork.Thereisoftennobetterwaytostarttheday
thantoaddresstheissuesthatsurfacedduringthedailyscrum.
KeystoSuccess
Whenitcomestothedailyscrum,teamswillbattlewithalltheseissuesandmore.The
secrettomakingdailyscrumseffectiveistobuildmusclememorysothatScrumandits
patternsbecomepartofthedailyroutine,likebreathing.Thekeystodoingthis
successfullyaretoputandkeepthedailyinyourdailyscrums,standup,workasateam,
andremainpatientinthefaceofresistanceanddifficulty.
KeeptheMeetingCadence
Dailyscrumsaremosteffectivewhentheyhappendaily.Inthischapter’sstory,Marco
thoughtthatmeetingeverydaywastoomuchoverhead.Heproposedthatthemeetingbe
condensedtotwodaysperweekat30minutespermeeting.Marco’sproposalhasseveral
flaws.
First,themeetingmusthappeninthesameplaceandatthesametimeeveryday.Every
day:noteveryotherday,oreverythirdday,buteveryday.It’snonnegotiable.Teamsmeet
dailytomaintainapattern,amusclememory.Theteamneedstogetusedtosynchronizing
workandmovingtogetherasateam.Itsmembersneedtomakepublic,daily
commitmentstoeachotherinsteadofworkingtowardsomedistantgoal.Theyneedthat
15minuteseveryday—andthey’relesslikelytocomplainifthemeetingactuallylasts15
minutesorless.
Besidesthat,though,valuableinformationcomesoutofdailyscrumsthatcansavetime
andmoney.ImagineifTiago’steamdidmeetonlytwiceaweek.SupposeCesarstarts
workingonhiswebservicetaskonThursday.Herunsintotrouble.Hespendsallday
FridayandalldayMondayworkingonitagain.AtTuesday’smeeting,heshareshis
problems.Pedrosays,“Oh.Ihaveafixforthat.Talktomeafterthemeeting.”Inessence,
Cesarwastedtwofulldays(FridayandMonday)workingonaproblemthatPedrocould
havehelpedhimwithatFriday’sdailyscrum—ifonlytheteamhadbeenmeetingevery
day.Spending15minuteseverydaytopreventwastingtwodaysofwork?Worthitevery
time.
Extendingthemeetingtimeisabadidea.Longermeetingsgivepeopletimetomeander
andramblethatshortmeetingsdonot.Extendingthetimespentinthemeetingonly
increasestheteam’sdissatisfactionandfeelingoftoomuchoverhead.Thedailyscrum
shouldbequick.Theteamwillalsobebuildingmusclememoryforquicklyandprecisely
coveringissues,justastheyarebuildingthemusclememoryofattendingadailymeeting.
Stand—Don’tSit
There’sareasonadailyscrumisalsoreferredtoasadailystandup—it’satimetostand.
Whydoteamsstand?First,itkeepsthemengaged.Whenseatedatatableorhiding
behindadesk,theycanfiddlewiththeirphones,flipapencil,ordoodle.Thesebehaviors
goawaywheneveryonestandsandfaceseachother.Second,standingkeepsthemeeting
short.Whenpeoplearestanding,theyareacutelyawarewhenthemeetingistakingtoo
long.Ifyou’refeelingtheneedtosit,themeetinghassomehowderailed.Third,standing
increasesenergylevelsandawareness,makingteamsmuchmorefocused.Someteamsgo
sofarastonotallowpeopletoleanonwallsordesks.Whetheryouchoosetoleanornot,
theimportantthingistostand,deliver,understandtheday,andthengettowork.
WorkasaTeam
WhenaScrumteamisreallyclicking,itisanintegratedunitworkingtowardacommon
goal,notacollectionofindividualsworkingonspecializedtasks.Onatrueteam,you
shouldneverhear,“MytasksaredoneandIhavenothingtodo”or“I’mtheonlyonewho
canfixthingsonalegacyproject.”Inthischapter’sstory,Marianamentionsthat,despite
theteam’slargebacklogofunfinishedwork,shewouldhavenothingtoworkonaftershe
completedhertworemainingtasks.ThenMarcoannounceshisexpertiseonalegacy
systemwasconsuminghistime.Bothoftheseexamplesarecluesthattheteamisstilljust
acollectionofindividuals.
Youcangaugehowwellpeoplearegellingasateambythinkingintermsoftheteam’s
busfactor.Thebusfactoristhenumberofpeopleonyourteamwhohavetobehitbya
busbeforetheprojectisinserioustrouble.Consideranyoneonyourcurrentproject.If
thatpersonleaves,istheteamabletostepupandcompletetheworkinthebacklog,orare
youleftunabletofunction,therebymissingyoursprintcommitment?Iftheansweris,
“Yikes!We’dbedeadinthewater,”yourbusfactorisone.Prepareforafailedsprintin
yournearfuture.Tiago’steam’sbusfactorisdefinitelyone.
Ifyourteamhasalowbusfactor,thefirststepistohelptheteamunderstandwhy
specializationisaproblemandhowitwillcontinuetohampereffortsunlessitis
addressed.Thesecondstepistorealizethatthereisnoquickfixorsimplesolutiontoa
lowbusfactor.Implementingcertainpractices,though,willhelp:
Workinasharedspace.
Practicepairprogrammingandtest-drivendevelopment.
Haveadedicatedteamfocusedononeprojectatatime.(SeeChapter3,“Using
TeamConsultantstoOptimizeTeamPerformance,”forhowtostructuretheteamto
makethishappen.)
Workinginasharedspacereducesmultiplebarrierstocommunication.Forexample,if
theteamfromthestoryisscatteredinofficesthroughoutabuildingandCesarwantsto
talktoRicardo,Cesarmustgetupandfindhim.Ricardomightnotevenbeinhisoffice
whenCesararrives,orhemightbebusy,butatthispointthedistractionhasalready
happened.Further,Cesarmightgetsidetrackedalongthewaybysomeoneelselookingfor
answers.Intheend,whatwouldhavebeenaquickquestioncaneasilyturnintoa45-
minuteanswersafari.Whenyouputtheteammembersinasharedspace,thesafarigoes
away.Theteamlearnshowtomanagethedistractionsonitsown.
Pairprogramming,orareal-timecodereview,isagreatwaytogetpeopleuptospeed
ontheentirecodebase.Youcantakethisastepfurtherandpracticepairprogramming
patternssuchasPing-Pongpairprogrammingorpromiscuouspairing,bothofwhichtake
theextremeinXPtothesuper-extreme.(ReadmoreaboutthesetechniquesinChapter19,
“KeepingPeopleEngagedwithPairProgramming.”)Test-drivendevelopmentbuilds
furtheronpairprogramminginthatitgivestheteamaloosesetofguidelinesand
documentationonhowthecodeworks.Ifateammembergetsstuck,hecanalwaysread
thetesttoseewhattheintentofthecodeis.
Havingadedicatedteamremovesdistractionsacrosstheboard.Insteadoffocusingon
specificspecializationssuchas“beingtheSQLguy,”theteammembersfocusonhow
theycanworktogethertosolveaspecificsetofstorieseachsprint.
BePatient
ForteamsnewtoScrum,dailyscrumsaregoingtobeachallengeatfirst.That’sagiven.
They’llfacelatecomers,ramblers,anddeepdivers.Correctingthesebehaviorstakestime
andpatience.Rememberthatitcantaketwotothreemonthsforanyonetolearnanew
behaviororpattern.Don’tdespairifitdoesn’tgoperfectlyatfirst.Keepremindingeach
otherwhyyouareallthere.Keepasenseofhumor.AndinvestinsomeNerffootballs.
Onedayyou’lllookupandrealizethatdailyscrumsarejustanotherroutinepartofyour
team’sveryproductiveday.
Chapter18.TheFourthQuestioninScrum
Supposeyouareonateamthatisbothnewlyformedandnewtopracticingagile.Your
dailyscrumsappeartobegoingwell,andyouarelookingforwardtoyourteam’sfirst
demo.Thebigdaycomes,yougetinfrontofthecustomer,andthenthefloordropsout.
Yourappstartsactinginawaythatisdefinitelynotasintended—allbecauseofahidden
blockingissue.
Youwondertoyourself,“Didn’twedoeverythingright?Didn’tweaskalltheright
questions?”Theteamdustsitselfoffandgetsreadytostartthesecondsprint.Howdoyou
ensurethatgoingforwardtheteamuncoverssubtleblockingissuesinthedailyscrum?
Addinthefourthquestion.Tohelpexplaintheconcept,I’llstartwithastory.
TheStory
TheHoppingMadDolphins(HMD)wereanewteam,bothtoScrumandtoeachother.
BecausethemembersoftheHMDteamhadallbeenburnedonprojectsinthepast,they
wereallexcitedbytheprospectofworkingonaScrumteam.Theyknewthebasicsbut
notmuchelse.Still,theyapproachedtheframeworkwithenthusiasmanddovein
headfirst,followingScrumbythebookascloselyaspossible.
Beforetheyknewit,itwastimeforthefirstdailyscrum.TheScrumMaster,Marc,was
thefirsttoarrive.HisfellowteammatesJane,Maria,Cory,andPacocameinshortly
thereafter.
“Okay,guys.Dailyscrumtime.We’resupposedtoanswerthreequestions,”saidMarc.
Hewrotethemonthewhiteboard.
“So,whowouldliketostart?”askedMarc.Theteamjumpedrightin.Theblocking
issuesranthegamutfromstoredprocedureshavingtimeoutissuestothingsthatnever
wouldhavebeenbroughtupinthepast,suchas,“It’stoohotinthislittleteamspace.”
After15minutes,theteammemberseachhadaturnandwerefeelingprettypleasedwith
themselves.
“Thisiskindoffun,”saidCoryonhiswayoutoftheroom.
“We’recommunicating!”agreedJane,half-jokingly.“Justlikeinthebooks!”
Thedailyscrumscontinuedtorunsmoothlyasthesprintwenton.Theteamseemedto
beholdingitselfaccountable,andtheblockingissues,eventheheatissue,werebeing
addressed.“ThisScrumthingisprettyeasy,”remarkedJaneneartheendofonedaily
scrumtowardtheendofthesprint.“We’regoingtoknocktheirsocksoffatthesprint
review.”
Everyonelaughed.Everyone,thatis,exceptPaco,whosimplyshruggedhisshoulders.
Coryslappedhimonthebackandasked,“Paco,whythelongface?”
Pacosmiledandseemedabouttoanswer,whenJaneinterrupted.“That’sbetter,Paco.I
knewyouhadasmileinyou.”EveryonelaughedagainandJanecontinued,“Hey,Marc,
we’veallhadaturn,andit’sbeen15minutes.Mybuildshouldbedone,andI’manxious
togetbacktomydesk.Canweadjournthisthingandgetbacktowork?”
Marcnoddedandthegroupquicklybrokeup,leavingPacofrowning,handsinhis
pocketsandeyesonthetaskboard.Unfortunately,noonenoticed.
Fortherestoftheteammembers,thefirstdemomeetingcouldnotcomefastenough.
TheyfeltlikekidsonChristmasEve.Theyhadaccomplishedmoreworkthantheyhad
imaginedpossibleoverthefirst30-daysprintandwantedtoshowitoff.Theteamworked
hardtomakeeverythingperfect,includingaslidedeck(mandatoryforallpresentationsat
thiscompany,noexceptions)thatlistedtheaccomplishedwork,theworkthatwasnot
accomplished,andwhatitmeantfortheteamtobedone.Theonlyteammemberwho
didn’tseemtoshareintheexcitementwasPaco.Marcnoticedhimatthebeginningofthe
sprintreview,mainlybecausehewasstaringoutthewindowinsteadofatthepresentation.
“IneedtoremembertoaskPacowhat’sgoingon,”Marcthought,turninghisattention
backtothescreen.
Thereviewwentwellatfirst,butwhentheteamwasdemonstratingthelastcomponent,
thesystemcrashedwithnowarning.Whiletheteamscrambledtofixitonthespot,the
stakeholdersreassuredthemthatitwasn’tthatbigofadeal.Oneofthestakeholderssaid
jokingly,“Well,sincethisisyourfirstsprint,we’llletyouoffwithawarning.”
Thoughtheteammemberslaughed,theywereembarrassedanddeflated.Afterthe
stakeholdersleft,theteamsattryingtopuzzleouttheproblemandhowitcouldhave
missedsuchahugebug.Pacosatbackinhischairandsaidunderhisbreath,“Iknewthis
wasgoingtohappen.”
Theroomgotquiet.CorylookedatPacoandasked,“Whatdoyoumean?Howdidyou
knowthiswasgoingtohappen?”
“Well,”Pacobegan.“WhenIwasreviewingoneoftheclassesacoupleweeksback,I
noticedsomeirregularities.Thewayitwaswrittenwasn’texactlywrong,butIsuspected
thatitmightcausesomebugstopopupbecauseI’dhadsomethingsimilarhappentome
before.IthoughtIsaidsomethingaboutit,butIguessIdidn’tsayitdirectlyenough.And
itwasn’treallyblockingme;itwasjustmoreofagutfeeling.Ididn’twanttopointout
someone’smistakeorredosomeoneelse’swork;itseemedpresumptuousandrude.So,I
decidedI’dwaitandseeiftherewasaproblem,andIguesstherewas.”
Therestoftheteamsatthere,lettingPaco’swordssinkin.
“Wedon’tknoweachotherverywell,”saidJane.“But,please,ifyoueverseemedoing
somethingthatisgoingtocauseaproblem,tellme.Iwon’tbeoffended.”
Everyonenodded.
“Andpartofbeingateamisworkingasateam,”saidMarc.“Weneedtogetinthe
habitoffixingproblemsthatwesee,eveniftheyaren’tin‘ourcode’becausewehave
collectivecodeownership.We’reallresponsibleforeverylastline.”
“Iknow,andprobablynexttimeIwilltrytospeakuporask.Thethingis,Iwasn’t
reallysure.LikeIsaid,itwasmoreofagutthing,”explainedPaco.“Butitateatmeand
worriedmeforacoupleofweeks.”
“Iwishyouwouldhavesaidsomething,”saidMarc.“Inoticedsomethingwaswrong
withyou,butnotuntiltoday.”
“Yeah,”saidCory.“Imean,we’veallbeenbusy,andwehaven’tbeenworkingtogether
thatlong.Ifiguredyouwerejustaquietguy.”
“Well,Iamalittlequiet,”saidPaco.“AndI’mnotverycomfortablejustblurtingthings
out.Ireallydon’tlikeconflict,andIdidn’twanttoinsultanyone.”
“Okay,”saidMarc.“Ithinkwe’vebeendoingagoodjobatourdailyscrumsof
mentioningthemoreobviousblockingissues,butwe’remissingsomeofthesneaky,
subtleonesthatwilljumpoutandbiteus.Anysuggestions?”
“Wecouldtrusteachothermore,”offeredMaria.
“Wecouldaskeachother’sopinionsmore,”saidCory.
“Wecouldmakesurethatwe’reallfeelingequallyconfidentaboutthings,”saidPaco.
“That’sit,”saidJane.“Weneedtoaddaquestion.”Shewalkeduptothewhiteboard
wherethethreequestionswerestillprintedintheupperright-handcornerandaddeda
fourth.
“Areweallowedtodothat?”askedMaria.“Canwejustaddaquestion?”
“Idon’tseewhynot.Whatdoyouthink,team?”askedMarc.“Ifwe’dhadthis
question,wouldwehavegoneintoasprintreviewwithabig,system-crashingbug?”
EveryonelookedatPaco.Heconsideredforaminuteandthensaid,“Rightaroundthe
middleofweektwo,myconfidencelevelwouldhavebeenmuchlowerthananyone
else’s.Anditprobablywouldhavegivenmetheopportunitytoairoutmyconcerns
withoutfearoflookinglikeaknow-it-allorbeingtotallywrong.So,yeah,Ithinkthat
wouldwork.”
Theteamaddedthefourthquestiontothelistandstucktoitthroughouttheproject.
Andjustabouteverysprint,someonewouldusethequestionasawaytoraiseanagging
concern.Someproblemswerereal;otherswerejustmisunderstandings.Butallofthem
wereopportunitiestofollowuponahunchthatmighthaveotherwiseremainedunspoken.
TheModel
ThefourthquestioninScrummightbesimple,butitisavaluabletooltoensurethatthe
teamisontrack—especiallyanewteam.Inthedailyscrum,startwiththethreequestions
asyounormallywould:
1.WhatdidIdoyesterday?
2.WhatwillIdotoday?
3.WhatblockingissuesdoIhave?
Thenaddthefollowing:
4.Whatismyconfidence,onascaleofonetoten,thattheteamwillaccomplishthe
goalofthissprint?
Whydoesthisquestionworksowell?Asshowninthestory,noteveryoneisequally
comfortablewithconflict.Andnoteveryoneisusedtoairinganopinion,especiallywhen
trusthasnotbeenestablished.Assuch,thosewhoaremoreoutspokenandopinionated
candominateaconversationorameetingwithoutmeaningto.Inadailyscrumor
retrospective,themoreintrovertedteammembersmightagreewiththestatedopinion
despitetheirmisgivingsbecausetheyfeeltheyarealoneinthinkingdifferently,don’twant
tobewrong,orfeelspeakingupwouldcauseanunnecessarystir.
Often,togetpastthesurfaceissuestothedeeperissues,youneedtoaskaquestion
intendedtostartconversations.Questionfouristhatkindofquestion.Itforceseveryone
toputahardnumberonasoftfeeling.Italsofocuseseveryoneontheoverallproject
ratherthanindividualstoriesortasks.Andjustasdisagreementsaboutestimatesin
PlanningPokercanunearthhiddenassumptionsandmisunderstandings,differencesin
projectconfidencecanuncovercoreissuesthatmightotherwisegounnoticed.Exposing
inequalitiesdrivesdiscussion—thekindthatcanmakeorbreakaproject.
KeystoSuccess
So,whenshouldyouusethefourthquestionofScrum?Irecommendusingitatthe
beginningofanynewproject.Evenifpeopleknoweachother,theymaynothavebuiltthe
trustrequiredtohavetheopenandhonestconversationsthatensureprojectsuccess.I’ve
madeitarequiredelementforanyteamthatisnewtoeachother,toScrum,ortoboth.
Evenifyou’reanestablishedScrumteam,ifyouaregoingthroughyourdailyscrum
withoutuncoveringmanyimpediments,butarefindingthatduringthesprintorthe
review,issueskeepcroppingup,youneedtoaddthefourthquestionofScrum.
Thefourthquestionispowerfulonitsown,butit’sevenmoreeffectivewhenthe
ScrumMasterisalsoonthelookoutforthefollowing:
Nonverbalcommunication—Lookforsmallsignsthatateammembermightbe
unsureoruncomfortable.Inthestory,Paco’slackofenthusiasmwasnoticeablebut
unexplained.Ifyouseeabehaviorthatseemsoutofplace,askaboutit.Andstick
aroundlongenoughtoheartheanswer.
Peoplewhoarenotcomfortablewithconflict—AScrumMasterdoesn’thavetobe
apsychologist,butthebestScrumMastershaveprobablydonesomeresearchon
howhumansworkandinteract.Differentpersonalitiesandculturesoftenrespondto
conflictindissimilarways.Gettoknowthepeopleyouareworkingwith,askthose
whohaveworkedwiththeminthepastabouttheirnaturaltendencies,andspend
sometimechattingwiththeminformally.Establishingtrustgoesalongwaytoward
increasingcommunication.
Opportunitiestopracticecontinuouslearning—Twoelementsthathavehelped
teamsthatIhaveworkedonandcoachedaretheMyers-Briggsassessmentand
workshopandthebookPeopleStylesatWork[BOLTON].Providingteammembers
withthetoolsneededtostretchthemselvespersonallyandtodeveloptheir
competencieswillhavealastingimpactonthecurrentprojectandallfuture
projects.
Asyourprojectprogressesandtheteambondandtrustbecomesstronger,youmight
findthatthefourthquestionisnolongerneeded.MostteamsIworkwithuseitforthe
firsttwotofivemonths,oruntiltheteamreachesaperformingstage.(Formoreonthe
fourstagesofteamdevelopment,seeChapter20,“AddingNewTeamMembers.”)The
teamwillknowwhenthequestionisnolongerneeded,sotheteamshoulddecideifand
whenitshouldbedropped.
Reference
[BOLTON]Bolton,Robert,andDorothyGroverBolton.2009.PeopleStylesatWork:
MakingBadRelationshipsGoodandGoodRelationshipsBetter.NewYork:AMACOM.
Chapter19.KeepingPeopleEngagedwithPair
Programming
Pairprogrammingisoneofthosethingsthatpeopleeitherloveorhate.Whatnoonecan
dispute,however,isthatpairingproduceshigh-qualitysoftwareinarelativelyshort
amountoftime.Indeed,AlistairCockburnandLaurieWilliams[COCKBURN]have
proventhatwhiledevelopmenttimegoesupwhenpairing,thenumberofdefectsinjected
willbefewer.
Theproblemwithpairprogrammingtendstobekeepingpeopleengagedwhentheyare
nottheoneswiththeirhandsonthekeyboard.Thisisexactlywhatwashappeningtothe
teaminthefollowingstory.Theteamwaspairingfaithfullybutwashavingtrouble
stayingfocusedonthetaskathand.
TheStory
LansinglookedoveratColinandsawhimyawn.
“C’monman,snapoutofit!IknowItypeslowly,butwe’veonlybeenatthisanhour!”
saidLansing.
“Sorry,Lansing.It’snotyou.I’mtryingtostayfocusedonwhatyou’redoing,butit’s
hardformetostayawaketoday.Iwasupmostofthenightwithasickkid;Iknowthat’s
notanexcuse.Let’sgogetanothercupofcoffee,”saidColin.
Theirdiscussioncaughttheattentionofanearbypair.“We’llgowithyou.Sittinghere
islikewatchingpaintdry,”saidSean,whowaspairingwithDevon.Theybothrosetojoin
ColinandLansinginthebreakroom.
Asthetwopairssathavingacoffee,Lansingaddressedtheproblemofpairing.
“Guys,wejustdon’tseemengaged.Everytimeoneofusgetstoodeepintothecode,
theotherpersonfallsasleep!”
“Iknow,Iknow,”saidColin.“I’msorry,guys.”
“It’snotjustabouttoday,”saidLansing.“Imeanallthetime.Idoit,youdoit,
everyonedoesit.Weallgetdrowsyorcheckoutwhenwepairfortoolong,evenjust30
minutes!Whatarewegoingtodoaboutit?”
Devonchimedin.“Iwasbrowsingthewebonhowtoincreasetheperformanceofour
buildserver,andIstumbledacrossthispaperfromtheAgileconferencetitled
‘PromiscuousPairingandBeginner’sMind,’byaguynamedArloBelshee[BELSHEE].
Itisprettyinteresting.Heandhisteamincreasedtheirvelocitysignificantlybyhavingone
personfromeachpairswitchpairsandtaskseveryninetyminutes.Noonestaysonone
tasklongerthanthreehourstotal.”
“So,letmegetthisstraight,”saidColin.“You’resayingthatifwewerepairinglikewe
werethismorning—I’mwithLansingandyou’rewithSean—afterninetyminutes,I
wouldswitchplaceswithSeansothatyouandIarepairingtogetheronyourtask,and
SeanandLansingarepairingtogetheronourtask.Thenafteranotherhourandahalf,you
andLansingwouldswitchplacessothatwe’dbebacktoouroriginalpairsbutnotour
originaltasks?”
“Yep,”saidDevon.“Weird,huh?”
“Soundsconfusing,”saidColin.
“So,whokeepsthetaskgoing?”askedLansing.
“Thepersonwhohasbeensittingatthatcomputerthelongesthastogethisnewmate
uptospeed,andquickly,becauseninetyminuteslater,thatsamepersonwillbeleavingto
joinanewpair,”saidDevon.
“That’salottoabsorbinninetyminutes.Doesanycodinggetdone,oristhepair
alwaysgettinguptospeed?”
“That’stheweirdestpart,”saidDevon.“Theteaminthepaperfoundthatthevelocity
washighestwhenthepairsswitchedeveryninetyminutes,evenwiththelearningcurve—
somethingaboutbeginner’smind.Youshouldreadthepaper.”
“Yep.I’dliketodothat,”Lansinganswered.“I’dbewillingtogiveitagotoday,
though,justtoseeifithelps.Wecan’tgomuchslowerthanwearewithColinasleepin
thepassengerseat!”
Colinlaughed.“That’strue.Itwilldefinitelyhelpkeepmeawake.”
“I’mwillingtogiveitatry,also,”saidSean.“Switchingthingsupmightjump-startmy
creativity.Ihaven’tcontributedmuchtoday.”
Devoncontinued.“Well,aslongaswe’refeelingdaring,Ialsocameacrossthisother
webpagebyaguynamedPeterProvostcalledmicro-pairing[PROVOST01].Hismethod
wasmoreextremebutwouldcertainlykeepusengaged.Itlookslikethis.”Devonpulled
outanapkinanddrewupaquickversionofPeter’smicro-pairingidea.
“Devon,I’mallforshakingthingsup,butthatlooksinsane!”exclaimedLansing.
Devonnodded.“Iagreethatit’salotofwork,andIdon’tthinkwecoulddothisallday
but—”
“I’mwillingtotryit,Lansing,”saidColin.“Aroutinelikethatwouldkeepuslaser-
focused.”
“I’mgame,too,”saidSean.
“Okay,okay.Youguyswin.Let’strythis,”saidLansing.“Whatdowehavetolose?”
Thepairswentbacktotheirworkfeelingsimultaneouslyinvigoratedandwary.They
hadtwonewtechniquestotry,whichwouldbesuretokeepthemawakebutthatmight
provedifficulttosustaininthelongterm.
TheModel
People(myselfincluded)havesuchahardtimekeepingtheirheadsinthegame.Mostare
justsoeasilydistracted.In“TheEbbandFlowofAttentionintheHumanBrain,”Trey
HeddenandJohnGabrieliofMITexplorewhathappenswhenpeoplearedistractedby
internalthoughtswhendoingcomplexwork[HEDDEN].TheytellthestoryofDan
Jansen,anOlympicspeedskaterwho,inthe1988WinterOlympics,slippedandfellin
boththe500-andthe1,000-meterevents[WEINBERG].Jansenhadtrainedforyearsand
wasnormallyafocusedchampion.Justsevenhoursbeforehisrace,however,Jansen’s
sisterpassedawayfromleukemia.HeddenandGabrielihypothesizethatwhileJansen’s
focus,hisexternalfocus,wasonskating,hehadsmallgapswhenthoughtsofhissister
wouldintrude.TheseinternalfocusissuestookJansen’sattentionawayfromtheexternal
tasksathandandresultedinhisuncharacteristicfalls.
Intraditionalpairprogramming,youusuallyhaveonepersonatthekeyboard(a
tactician)andonewhoisobserving(thestrategist).Thetacticianisusuallybetterableto
maintaininternalfocusbecausehehastheactivityoftypingtokeephismindfocusedon
thetaskathand.Thestrategist,ontheotherhand,islessactiveandthereforemore
distractedbyinternalfocusissues—theseissuescausegapsinattentionthatultimately
makethepairlessproductivethanitcouldbe.
Eachofthemodelsdiscussedinthischapterhelpkeepbothmembersofthepair
engagedandfocuseddespitethenumerousdistractionsthatmaycrowdtheirminds.
PromiscuousPairing
AtAgile2005,ArloBelsheepresentedandpublishedanexperiencereporttitled
“PromiscuousPairingandBeginner’sMind:EmbraceInexperience”[BELSHEE].
Inthetalkandpublishedpaper,Belsheedescribedmanypractices,allofwhichhaveto
dowithmaximizingthelearningcapacityofthebeginner’smind,amindthatisopento
experimentationandnotjadedbyexperience.Belsheedefinesthebeginner’smindasthis:
“Ifapersonisotherwisecomfortablewithhisenvironmentbutdoesn’tunderstandone
thing,thenhewillusuallytrystuffuntilhefiguresthatonepartout.Thisstateoftryingto
reconcileone’spastexperienceswithanenvironmentthatdoesn’tquitefitisBeginner’s
Mind”[BELSHEE,p.2].
Belsheegoesontoexplainthattheconceptofbeginner’smindisbehindthenotionof
beginner’sluck:“Apersondoingsomethingforthefirsttimeoftendoesitmuchbetter
thanhedoesafterhe’spracticedforawhile.Becausehetriesmoreapproaches,andtries
themrapidly,apersoninBeginner’sMindismorelikelytosucceedatataskthanonewho
thinksheunderstandshowitworks”[BELSHEE,p.2].
Belsheeandhisteamhypothesizedthatiftheycouldstayinastateofbeginner’smind
foraslongaspossible,theywouldincreasetheirvelocity.Theytriedseveralpractices,all
describedinthepaper,buttheonethatstuckoutforme(andtheoneDevondescribesin
thischapter’sstory)combinesteam-ownedtaskswithashortenedpair-cycletime.His
teamhasonepersonfromeachpairswitchpairsandtasksevery90minutes,thereby
keepingan“expert”anda“beginner”oneachtaskatalltimes.
Team-ownedtasksarebestexplainedincontrasttoindividuallyownedtasks.
Individuallyownedtasksarethosewhereonepersonstayswithataskthroughto
completion.So,ifDevonpullsataskfromthebacklog,hewouldworkonitwithany
numberofdifferentpairpartnersuntilthattaskiscomplete.Team-ownedtaskshaveno
designatedowner—ifanything,theworkstationownsthetask.Anynumberofpairsmight
workonataskatvarioustimesuntilitiscomplete.Thereisnoonepersonaccountablefor
itfrombeginningtoend.
Pair-cycletime,orpairchurnasBelsheecallsit,isapredefinedamountoftimethata
paircanstaytogether.Belshee’steamexperimentedwithmanydifferentlengthsbutfound
thatforoptimalvelocity,pairsshouldstaytogethernolongerthan90minutes.Myteam
experimentedwiththisaswellandfoundouroptimaltimetobetwohours.Whetheryour
optimaltimeis90minutes,120minutes,orsomewhereinbetween,movingpairsand
tasksfrequentlytakessomegettingusedto.Thefollowingexampleshowshow
promiscuouspairingworksforatypicalteam.
PromiscuousPairinginPractice
SwitchTeamisasix-personteam(notincludingtheproductownerortheScrumMaster).
SwitchTeamhasestablisheditscoreteamhours(seeChapter10,“CoreHours,”formore
onthistopic)andhasagreedthateveryonewillstartworkasateamat10:30a.m.andwill
stopworkingasateamat4:00p.m.,withaone-hourbreakforlunchbetweennoonand
1:00p.m.Switch’sdailyscrumisat10:00a.m.(seeFigure19-1).
FIGURE19-1SwitchTeampairblockhoursandswappattern
Noticeinthisexamplethattherearethreepairingblocks.Eachblockhasaprimary
personandasecondaryperson(I’venamedthesepeopleA,B,C,D,E,andF).Eachpair
blocklastsfor90minutes,asrecommendedinBelshee’sexperiment.Thepeoplerotate
onespottotheleftinanupwardlyserpentinefashion,droppingtothebottomrightfrom
thetopleft.Thetasks,ontheotherhand,staywiththeworkstationsuntiltheyare
complete.NoticethatevenwhentheoriginalpairsarebacktogetherduringPairBlock3,
theyareonadifferentworkstation(andtask)thanwheretheybegan.Dependingonyour
tasksize,thetaskmightbecompletebeforethepairwhostartedonitreturntothat
workstation.
TheChallengewithPromiscuousPairing
Whenmyteamfirstimplementedthesetechniques,wewerenotready.Wecapturedour
resultsandIpublishedanexperiencereportatAgile2006,“AdventuresinPromiscuous
Pairing:SeekingBeginner’sMind”[LACEY].Inthispaper,Iconcludedthatpromiscuous
pairingisanadvancedskill,bestsuitedforseasoned,well-practicedteams.Newteams
workinginanunfamiliarframeworkwilllikelyfindthishigh-levelpracticetoomuchto
handleatfirst.
Whenyouarereadytotrypromiscuouspairing,considerusingdesignconceptcards.At
thebeginningofeachtask,haveyourpairwritethedesignona3″×5″card,withthe
largerassumptionsandconstraints.Thepurposeofthecardistoallownewpairsto
referencethebriefnotesratherthanaskquestionsoftheoriginalpairwhostartedthetask.
Otherwise,whatI’veexperiencedisthattoomuchtimeisspentrehashingdesign
decisionsateachpairswap.Byhavingdesignconceptcards,thenewpaircaneasilysee
whydecisionsweremadeandeitheracceptorchangethedesign,withoutwastingtime
tryingtounderstandorjustifyparticulardecisions.
Eventhoughbestsuitedforadvancedteams,Ibelieveintheideascommunicatedby
Belshee.Ihavehadsuccesswiththeminpastprojectsbothasawaytokeeppairsengaged
andasatoolforbringingnewpeopleuptospeedonthesystem.Tryit—andmakeagood
faitheffortwhenyoudo.
Micro-Pairing
Micro-pairing,coinedbyPeterProvostin2006,hasitsrootsinagameinventedbyPeter
andBradWilson,ThePairProgrammingTDDGame[PROVOST02].Itissimilartothe
pairprogrammingPing-PongpatternonWardCunningham’sC2Wikisite
[CUNNINGHAM].ThePing-Pongpattern,aswrittenbyCunningham,isasfollows:
Awritesanewtestandseesthatitfails.
Bimplementsthecodeneededtopassthetest.
Bwritesthenexttest.
Aimplementsit.
Andsoon.Refactoringisdonewhenevertheneedarisesbythepersondriving.
Micro-pairingwasbornoutoffrustrationwithtraditionalpairing,specificallywiththe
strategist’stendencytodriftinandoutoffocuswhilethetacticiancompletestasks.It
differsfromthePing-Pongpatterninthateverytransitionofstateinthediagramcausesa
keyboardshift,wherekeyboardshiftsinCunningham’spatternhappenonlyafteranew
failingtesthasbeenintroduced.Micro-pairingforcespairstomakefrequenttransitions
bothinwhohasthekeyboardandalsoinwhatthatpersonisdoingwiththekeyboard.The
rulesoftheoriginalgamearesimple(adaptedfromthedescriptionbyPeterProvost
[PROVOST02]):
Onepersonwritesatestandthenpassesthekeyboardtohispartner.
–Ifthetestfails(thelightturnsred),thepartnerimplementstherequiredcodeto
makeitpass.
–Ifthetestpasses(thelightturnsgreen),thepartnerwritesanadditionaltestor
refactorsthepassingimplementation.
Thekeyboardthenpassesbacktotheoriginalpersontowriteanothertestorto
refactorthecodethatwaswritten.
Thepatterncontinuesuntilthetaskiscomplete.
So,whywouldsomeonewanttowriteapassingtestratherthanafailingtest?One
reasonistocodifyordocumentbehaviorthatalreadyexiststhroughanothertestbutisnot
adequatelydocumented.Writingapassingtestlocksinsomefunctionalityorconfirms
somesuspicionaboutthebehaviorofthecode.Another,perhapsmoreentertaining,reason
isthatwhenyousetouttowriteafailingtestandyouendupwritingapassingtest,
whoops!Thisshouldraiseabunchofquestions:Isthetestthatwasjustwrittenbad?Was
theprevioustestbad?Doesthecodedomorethanyouexpectedittodo?
Moreoftenthannot,youwillwritefailingtestsinthispattern,butdon’tforgetabout
writingpassingonestoconfirmanyexistingsuspicions.Rememberthatyoustillpassthe
keyboardevenafterwritingapassingtest,whichmeansyourpartnerwillprobablybethe
onewritingfailingtests,theonedrivingthedesign,atleasttemporarily.
Inplayingthegame,thepatternlookssomethinglikethis:
1.Lansingstartsinagreenstateandwritesafailingtest,whichtakesabouttwoto
threeminutes.
2.LansingpassesthekeyboardtoColin,whoisinaredstate.Colinimplementsthe
codetomakethetestpass,whichtakesaboutaminute.
3.ColinpassesthekeyboardbacktoLansing,whoisnowinagreenstate.
4.LansingwritesanotherfailingtestandpassesthekeyboardbacktoColin,again
takingaboutthreetofourminutes.
5.Colinimplementsthecodetomakethetestpass,whichtakesanotherminuteor
two,thenpassesthekeyboardback.
6.Thiscyclegoesonforfourorfivemorecycles,anditendsupwithLansing,who
decidestocleanupthecodethatColinhaswritten.
7.Aftercleaningupthecode,LansinghandsthekeyboardtoColinwhonowtakesa
turnatwritingtests.
Intraditionalpairprogramming,thetacticianwoulddoallthesestepsandthestrategist
wouldwatchthedesignandthecodeevolve.Thiscouldgoonforanhourortwowithout
interruption.Byusingmicro-pairingtoswitchcontrolofthekeyboardateverystate
transition,thepairsaremorelikelytohaveaconversationaboutthebitofcodethatwas
justwritten,regardlessofwhetheritisproductioncodeortestcode.Withouttheseforced
transitions,peoplearelesslikelytointerruptsomeonewhoisintheflow,whichcanresult
indegradedcodequality.
Asyoucanseefromthisexample,however,thepairingaspectofthesessioncan
becomefairlyintenseduetothetransitionofthekeyboardateverystatechange.The
personinthegreenstateisultimatelytheonedrivingthedesignbecausethatpersoncan
choosetowriteapassingtest,refactor,orwriteafailingtest.Thepersonintheredstate
hasonlyonechoice:tomakethetestpass(unlesshechoosestoargueatest).Trytovaryit
sothatbothmembersofthepairhaveanopportunitytowritetestsandnooneisstuck
onlyimplementingtests.Andalwayspassthekeyboardaftermakingyour“move.”Ifyou
findthetaskyouareworkingoniscomplete,youcandoavoluntaryswitchofthe
keyboardsinceyouaremovingontoanewtask.Figure19-2showsanexample.
FIGURE19-2Test-drivendevelopmentwithmicro-pairing
EveryarrowinFigure19-2representsamove.Youcancheckinonlywhenyouare
green,andyoucancheckinasfrequentlyasyoulike.
Whenyouimplementmicro-pairing,understandthatthecontroltokenofpassinga
keyboardbackandforthmightnotbenecessary.Icommonlyuseworkstationswithtwo
mice,twomonitors,andtwokeyboards.Inthesesetups,goodteamworkisnecessary,as
onepersoncanjumpinatanytime.
Whetheryouchoosepromiscuouspairingormicro-pairing,frequenttransitionsarea
surefirewaytokeepateamengaged.Bothtechniquesarealsoexcellentmentoring
exercisestojump-startjuniorornewteammembers,sincethedesignisdrivenbythe
personwritingthefailingtests.Whenateamchoosestomixmicro-pairingwith
promiscuouspairing,it’slikeridingonarollercoaster—intensebutalotoffun.
KeystoSuccess
Pairprogrammingshouldbeatiringactivity.Ifyouarrivehomespentbutupbeat,you’ve
probablyhadagooddayofpairing.Expecttodofourtosixhoursofpairingperday,not
eightormore.Ifyougohomefrustrated,yourpairingsessionwaslikelylessproductive
thanitcouldhavebeen.
Understandthatnoteverythingneedstobepaired.Pairingisgreatforcommunicating
designandformentoring,butifyouhaveabunchoffairlystraightforwardbugfixesor
aresimplyimplementingaminorfeatureinanalreadyfleshed-outdesign,pairingmight
notbeworthwhile.
Traditionalpairingisnotabadpractice.Havingastrategistandtacticianisagoodidea
andprovidessatisfactoryresults.Formostpeople,however,stayingengagedforlong
periodsofpairingisdifficult.Explanationsforthisdifficultycouldincludesocialloafing
(thetendencytoslackoffwhenworkinginagroup)orthefactthathavingtointeractwith
anotherpersononwhatisordinarilyasolitaryexperiencecouldbedistractingortiring.
Whateverthecause,therealityisthatpeopletendtocheckout.Whenyoufindthat
happeningonyourteam,youshouldtakeaction.
Thefirstthingyoushoulddoisremoveanyotherdistractions.Distractionscausehuge
issues,andinapair,theseissuesaredoubled.Considerturningoffemail,instant
messaging,andmobilephonesforasetamountoftime.Ifyoumusthaveaccesstoa
phonewhilepairing,putacommonphoneintheteamroomsowhenthephonerings,
everyoneisdistracted.Thatway,thedistractionwillendsooner.Dedicatetenminutes
everysooftenforemailandphonecalls,andscheduledowntime/breaksbetweensessions
sothatpeoplearebetterabletoremainfocusedduringapairingblock.
Next,acknowledgethatswitchingpairsfrequentlyandhavingtheworkstationownthe
taskmaycausearchitecturaldirectionproblems.Usenotecardstokeeptrackofdecisions
andreferenceimportantinformation.
Lookforotherwaystotapintobeginner’smind.Ifsomeonewantstolearnanewskill,
puthimonataskwherehecantrysomethinghehasn’tdonebefore.Belsheedidthiswith
successanddocumenteditinhispaper.Especiallywhenpairedwithsomeonewhois
highlyskilledinaparticulararea,theinexperiencedcanlearnquicklywhilebringinga
freshperspectivetothetaskathand.
Ifyougodownthepathofmicro-pairing,understandthatitisessentialtoswitch
drivers.Nocheating.Controlmustbeshared.Switchingforcesconversationanddrives
thedesign.
Successfulpairprogrammingwithengagedteammembersandcleandesignispossible
withouttheuseofthesetechniques.Ifteammemberscanfocususingtraditionalpairing,
great.Ifnot,trysomeofthepracticeslistedinthischapter.Youmightbesurprisedathow
muchmoreeffectiveandefficienttheteamcanbe.
References
[BELSHEE]Belshee,Arlo.2005.“PromiscuousPairingandBeginner’sMind:Embrace
Inexperience.”ProceedingsoftheAgileDevelopmentConference.Washington,DC:
IEEEComputerSociety,pp.125–131.
[COCKBURN]Cockburn,Alistair,andLaurieWilliams.2000.“TheCostsandBenefits
ofPairProgramming”(PDF).ProceedingsoftheFirstInternationalConferenceon
ExtremeProgrammingandFlexibleProcessesinSoftwareEngineering(XP2000).
http://collaboration.csc.ncsu.edu/laurie/Papers/XPSardinia.PDF(accessed20March
2015).
[CUNNINGHAM]Cunningham,Ward.“PairProgrammingPingPongPattern.”C2
website.http://c2.com/cgi/wiki?PairProgrammingPingPongPattern(accessed20March
2015).
[HEDDEN]Hedden,Trey,andJohnGabrieli.2006.“TheEbbandFlowofAttentionin
theHumanBrain.”NatureNeuroscience9(6):863–865.
[LACEY]Lacey,Mitch.2006.“AdventuresinPromiscuousPairing:SeekingBeginner’s
Mind.”ProceedingsoftheAgileDevelopmentConference.Washington,DC:IEEE
ComputerSociety,pp.263–269.
[PROVOST01]Provost,Peter.“Micro-PairingDefined.”Personalwebsite.
http://www.peterprovost.org/blog/2006/08/07/Micro-Pairing-Defined(accessed20March
2015).
[PROVOST02]Provost,Peter.Personalwebsite.
http://www.peterprovost.org/blog/2005/08/29/The-Pair-Programming-TDD-Game
(accessed20March2015).
[WEINBERG]Weinberg,Rick.“JansenFallsafterLearningofSister’sDeath.”
ESPN.com.http://sports.espn.go.com/espn/espn25/story?page=moments/87(accessed20
March2015).
Chapter20.AddingNewTeamMembers
Establishedteamssometimesneedtoaddnewteammembersinthemiddleofaproject.
Whetherthecauseisattrition,companyreorganization,accelerateddeadlines,increased
workload,orsomethingelse,it’scrucialfornewteammemberstobeaddedthoughtfully
andintegratedquickly.Eveninthebest-casescenario,teamswilltemporarilyslowdown
astheyadjusttohavingnewpeopleintheirmidstandasnewteammemberslearnthe
ropes.Whenthewrongteammemberisaddedortherightmemberisaddedwithouta
plan,thisslowdowncanbelonganddrawnout,provingBrooks’law,whichstates,
“Addingmanpowertoalatesoftwareprojectmakesitlater”[BROOKS].
Thegoodnewsisthatwithsomeforethoughtandplanning,teamscanquicklyreturnto
ahigh-functioningstatebyhelpingnewteammembersassimilatequicklytoanew
environmentandanewsystem.Whatfollowsisthestoryofhowoneteamdealtwiththe
needtoaddmanpowertocomplywithnewdatesinacontract.
TheStory
“Ican’tbelievethey’reaskingustomoveupthereleaseandaddnewfeatures.Areyou
surethereisnothingwecando?”askedCharles.
“Nothing,”saidStephanie,theteam’sproductowner.“Thenewdeadlineandfeatures
areacontractualrequirement,andwehavenochoicebuttocomply.Thebestthingforus
todoatthispointistoreshuffletheproductbacklog,writeanynewstories,anddropthose
thataren’trequired.Weknowourvelocity,andthisnewworkisatleastprettyfamiliarto
us.It’sjustapain.”
“Yeah,you’reright,Stephanie,”saidMary-Kate.“Wejustneedtokeepourheadsup
andgetthroughit.”
ForthenextthreedaystheteamworkedwithStephanietogettheproductbacklogin
order.Severalcallstothecustomerandsomeintensestoryworkshopsresultedinan
updatedproductbacklogandareleaseplantheywereconfidentin.Theyknew,though,
thatmanagementprobablywouldstillnotliketheresult.Stephaniepulledtogethera
meetingwithmanagementandtheteamtogooverthefindings.
Stephaniekickedoffthemeeting.“Thankyouallforcoming.Weallknowthereason
wearehere,soletmecuttothechaseandanswerthequestiononthetable,‘Canwehit
thedateandintegrateallthenewthird-partyfunctionalityyouareaskingfor?’Theanswer
ishardtoexplain.It’saqualifiedyes,withstringsattached.”
Chris,thegeneralmanager,wentfirst,“Greatnews,Stephanie.Team,thanksforpulling
thistogether.I’llletourVPknowthatyoucancommittothis.”
“Holdonaminute,Chris,”saidStephanie.“Youhaven’theardthestrings.Wecan’t
committothenewworkandtheacceleratedtimetablewithoutaddinganewpersontothe
team.Andnotjustanyperson.WeneedDennis.”
“Dennisisnotavailable,Stephanie.Youknowthat,”saidChris.
“Idoknowthat.Butyouaskedmehowwecouldmakethishappen.I’mtellingyouthat
todoit,weneedDennis.Dennishasthecompetenciesweneed.Heunderstandshowwe
work,andhispersonalvaluessystemistherightfitfortheteam.Ifwearegoingtomake
thishappen,weneedDennis.Thisisablockingissueforyoutosolve,”saidStephanie.
After30minutesofnegotiation,Chriscaved.
“Fine,fine.YoucanhaveDennis.We’llfindawaytogettheworkdoneonhisother
projectwithsomeoneelse,”saidChris.
Stephaniesmiled.“Thankyou.He’salreadypreppedandreadytogo.We’vestarted
transitioninghimalready.”
“Yousuredon’twasteanytime,Stephanie.Iknowaddinganewpersonisn’tacure-all,
though.HowwillyougetDennisuptospeedandstillhitthecompresseddate?”asked
Chris.
“Oh,we’vegotourways,”Stephaniesaid,andshelefttheofficewiththeteam.
“So,howarewegoingtogetDennisuptospeed?”askedCharles.“Eachofushas
joinedanotherprojectmidstreamatonetimeoranother.It’sneveraseasyasitseemsand
alwaysslowstheprojectdown.Howarewegoingtoaddressthis?”
Stephaniestoppedtoanswer.“WehadthissameproblemonanotherprojectIwason
recently.Oneoftheteammembersdevisedacreativewaytogetthepersonintegrated,
andit’swhatIproposewedohere.
“First,thenewmemberwasrequiredtopairprogramwithalltheexistingteam
membersinturntolearnthesystemandaskquestions.Wearealreadypairingand
workinginasharedspace,sothiswillbeeasytoimplement.Understand,though,thatit
alsomeanswecan’texpectanythingproductivefromDennisrightoutofthegate.After
all,inpairingsessions,he’llbewatchingmorethanhe’swritingcode.
“Second,tohelpspeedhisdevelopment,wealsoneedtowriteawrittentestand
administeritweekly.I’mtalkingtwentytothirtyquestionsthatillustratethesystemand
ourteammakeup.We’llfocusonthearchitecture,ourprocesses,andourculture.We’ll
expecthimtofailthetestmiserablyatfirst,butbyshowinghimwhathedoesn’tknow
(andwhatweexpectandwanthimtoknow),wecanfocushislearningontheelements
thataremostimportanttous.How’sthatsound?”askedStephanie.
Mary-Katewentfirst.“Soletmegetthisstraight.Weareunderacompressedschedule,
andwecan’texpectanythingoutofDennisforasprintortwo?Andyouwantustowrite
atestforhimandthen,I’mguessing,gradeitorreviewit?”
Stephanienoddedherhead.
“Stephanie,whatbenefitsdoyouthinkthiswillgiveus?”askedCharles.
Stephanieponderedthequestionforaminuteandcameupwithastory.“WhenIwasin
highschool,Ididastudy-abroadprograminSpain.IhadstudiedSpanishforthreeyears
andhadreceivedgoodgrades,soIthoughtthatthistriptoSpainwouldbenoproblem.
Boy,wasIwrong!Istruggledwiththelanguageandbutchereditprettysoundlyforthe
firstmonth.Afteramonth,however,Inoticedthatmylanguageskillshadincreased
exponentially.WhenIcamehomeafterfivemonths,IwasfluentinSpanish.Ilearnedthat
ifyouwanttolearnsomething,youhavetwooptions:Studyitfromthesidelinesor
immerseyourselfinit—allofit—sothatyoupickitupmuchmorequickly.”
Stephaniecontinued,“WewanttogiveDennisthesamekindofimmersiveexperience.
Heneedstogetrightinthethickofthingsandwritecodeinpairswithallofyou.He
needstoknowwhathe’stheretolearnandwhatquestionstoask(andweneedtoknow
thatheunderstandswhatwe’retryingtoconvey):That’swherethetestcomesin.Atfirst,
hewillaskquestionsthatyoumightfindannoying,butafteramonthorso,weshouldsee
himreallyimprovingand,hopefully,contributingtothedesignandwritingmainline
code.”
Theteamnoddedinagreement.
“ThenIguesswehaveatesttowriteandanimmersionscheduletodevelop,”said
Mary-Kate.“Let’sgetstarted.”
TheModel
Scrumteamspridethemselvesonbeingamenabletochange.Theteaminthischapter’s
storywasnodifferent.Theyshowedextremeflexibilityintakingonanewdeadlineanda
differentscope.Theyalsohadthesupportofmanagement,whounderstoodtheimportance
ofteamcultureandhavingtherightfit,andwhowaswillingtomakesacrificestogivethe
teamtherightnewmember.
Evenwithallthatinplace,though,youmustbepreparedforthefactthatwhenany
establishedteamassimilatesnewmembers,thatteamnaturallyfallsbackintotheforming
andstormingstages,asproposedbyBruceTuckmaninhis1965article,“Developmental
SequenceinSmallGroups”[TUCKMAN].
Tuckmanlaidoutfourstages:forming,storming,norming,andperforming.Consider,
forexample,afour-personrowingteam.
Inthefirststage,forming,theteamislearninghowtoworktogether.Everyoneis
“figuringeachotherout.”Peopleareassessingwhotheleaderontheteamwillbe.Politics
andindividualismplayalargeroleaspeopletrytohighlighttheirownperformanceand
establishthemselves.Onarowingteam,formingcouldbeassimpleaslearninghowto
havetheteammembersputtheiroarsinthewateratthesametime.
Inthesecondstage,storming,theteamisnowbeginningtoshowsignsofcohesion—of
gelling—astheinterpersonalissuesstarttoresolve.Thoughconflictstillexists,theteamis
startingtofocusitsenergyonsolvingitscollectiveproblemsandworkingthingsout
together.Onarowingteam,stormingcouldbetryingtogeteveryonetopulltheoarback
atthesametime,regardlessoftheamountofforceeachisusing.
Inthethirdstage,norming,theteamisstartingtoreallyworkwelltogether.Team
processesarebecominginstinctive,andtheteammemberstrustthewaytheyareworking.
Peoplearegrowingfriendliertowardeachother;thecriticismtheyexhibitismore
productivethandestructiveorpolitical.Thereisclarityinwhateachpersondoes,andthe
teamiscomfortablewithit.Theteammembersarefollowingtherulestheyhave
establishedandareontheirwaytobeingaperformingteam.Onarowingteam,norming
couldbewhentheteambeginstoenter,andwin,afewcompetitions.Theteamisnotat
thetopofitsclassjustyetbutisdefinitelyacontender.
Inthefourthstage,performing,theteamhasreachedahighleveloftrust.Thewaythe
teamworksfeelsnaturalandfluid,andtheteamisinastateofflow,in“thezone”ifyou
will.Relationshipsaretightlybonded,andeachpersoniscomfortablewithcriticism.The
teamaddressesandsolvesproblemsasneeded,andindividualsarenotafraidtoaskfor
helporshowweakness.Theteamisimprovingdaily.Intermsofarowingteam,
performingwouldbecompetingforandevencontendingforthegoldmedalinthe
Olympics.Everyone’soarentersthewateratthesametime,eachrowerpullswiththe
sameforce,theoarsexitsimultaneouslyandatthesamepitch,andtherowersshifttheir
weightasonefluidmotioninsteadofmultipledistinctbodies.Tuckmanmodeledhis
theoryinthefollowingway,asillustratedinFigure20-1.
FIGURE20-1Tuckman’sstagesofgroupdevelopment
Althoughreformingandrestormingareinevitablefollowingachangetotheteam,
teamsshouldspendaslittletimeaspossiblegoingthrougheachstage(seeFigure20-2).
Inthischapter’sstory,Stephanieunderstoodthattakingonanadditionalpersonwould
temporarilyslowdowntheteam,asitreformedandstormed,butshealsoknewshecould
shortenthetimetheteamspentinchaosbychoosingsomeonewhosevaluesnaturally
alignedwiththoseoftheteam,immersingthisnewpersonintheteam’sculture,and
acceleratinghislearningthroughfocusedstudy.
FIGURE20-2Ideally,groupsareabletominimizethetimeittakestogothroughthe
phasesagain.
Reformingateamischallengingandhardwork.However,withtherightattitudeand
willingnessoftheteam—again,alignmentofculturalvalues—itcanbedone.Take,for
example,oneofthehigh-performingteamsI’veworkedoninthepast.Afterasuccessful
project,wesplituptoworkonseparateprojects.Sincethen,we’vegottenbacktogether
againforafewnewprojects.Wheneverwe’vereformed,we’vebeenabletoreacha
performingstatewithinaweek.Andwhenwe’vehadtoaddanewpersontothemix
(whomwe’vechosen),it’stakenusalittlelonger,butwe’vestillreturnedtoaperforming
statefairlyquickly.
TheExercise
Thefirststepinfullyintegratingsomeonenewintoanexistingteamistochoosetheright
person.Whenyouaddsomeonetotheteam,lookathispersonalvaluessystemfirst.Isthe
personopentochange?Flexible?Humble?Ateamplayer?Opentolearning?
Ifyouansweryestothesequestions,thenmoveontotheskills.Doesthepersonhave
theskillsneededtodothejob?Valuestrumpskillsnearlyeverytimebecauseskillsareso
mucheasiertoteach.Inthischapter’sstory,theteammemberschoseDennisbecauseof
hispersonalvaluessystemandhiscompetencies.Theyknewhewouldintegrateintothe
teamculturenicely.
Thenextstepistowriteatesttohelpfocusthelearning.Theteamisresponsiblefor
creatingthequestions.Thenewteammemberisresponsiblefortakingthetestattheend
ofeachweek.Someofmyfavoritequestionsareasfollows:
Describethearchitectureofthesystem.
Telluswhypairprogrammingandtest-drivendevelopmentareimportant.
Whoownsthecode?
Mixtechnicalquestionswiththosethatfocusoncultureandvalues.
Thefirsttimetheteammembertakesthetest,manyofthetechnicalanswersshouldbe
blank,andsomeoftheculturequestionsmightbeoffbase.Bytheendofthesecond
week,however,theculturequestionsshouldbeeasyandthetechnicaldetailsshouldbegin
toappear.Youshouldbeworriedifthenewbiecontinuestoanswerquestionssuchaswho
ownsthecodewith,“Iownsuch-and-suchsectionoftheapplicationandBobbyownsthis
othersection”insteadofthecorrectanswerforahigh-performingteam:Theteamowns
thecode.
KeystoSuccess
Thoughkeepingteamstogetherforthelongtermisideal,therealityisthatcompanies
oftenshiftpeoplefromprojecttoproject.ThebestthingaScrumteamcandotoprevent
thisfrombeingarecurringthemeistomakevisiblethepainthatdisruptingteamsbrings.
Thekeytosuccesswhenfacedwithintegratingnewteammembersistoacceptthatsome
backslidingwilloccurandtochoosenewteammemberswisely.
AccepttheDropinVelocity
Achangetoteamchemistrywillresultinsomebacktracking.Ifyouhaveeverplayedor
watchedsports,youknowthatwhenanewplayerjoinstheteam,theteamdynamicsare
thrownoff.Theteam,onceahigh-performingcohesiveunit,revertstoitsbasicformasit
integratesitsnewteammember.Considerwhatwouldhappenifasportsteamchangedits
startinglineupforeverygame.Theresultswouldbedisastrous.Yet,businessesbelieve
thatsomehowtheycanaddorsubtractmemberswithoutaffectingtheteam’soutput.Each
timeateammemberentersorleaves,theteamreturnstoTuckman’sforming,storming,
norming,andperformingstages.Whilethisisinevitable,thegoalistogetthroughthese
stagesasquicklyaspossibleandgetbacktobeingthehigh-performingteamyouwere
beforethechange.Theotherthingtoconsiderwhenaddingnewteammembersisthatit
sometimestakeslongertogetthenewteammemberuptospeedthanthereistime
remainingontheproject.
ChooseWisely
Eachteam’scultureisunique.AsdiscussedinChapter3,“UsingTeamConsultantsto
OptimizeTeamPerformance,”youcanchooseateamconsultanttohelpyouforexpert
advice–typeworkorlimitedengagements.Ifthepersonisgoingtobeacoreteam
member,however,thechosenmembershouldbelike-mindedandabletoadaptquickly.
Forthisreason,teamsshouldtrytoavoidhavingteammembersforcedonthembased
solelyonavailability.Therightpersonisnotnecessarilytheavailableperson.Afterall,
availabilityisnotamindsetoraskillset.Whileitmaytakealittlelongertofindtheright
person,doingsoiscrucialtotheteam’sabilitytoreturntoaperformingstatequickly.
Givepeopletheauthoritytochoosewhowillbeaddedto(orsubtractedfrom)theirteam.
Forcingteamstoacceptateammemberwhodoesn’tmeshwellwiththeexistingteamcan
havecatastrophiceffects,asoutlinedinChapter21,“WhenCulturesCollide.”
RiskyBusiness
Addingnewteammembersisespeciallyriskywhenonlyoneortwoteamsinyour
companyaredoingScrum.Findingnewteammembersfromwithinnon-agilegroupswho
sharethecultureoftheScrumteamisoftendifficultorimpossible.Theteammembersin
thischapter’sstorywereabletofindadesirablenewadditionandhadthesupportof
managementinfreeingthatpersonfrompreviouscommitments.Teamsfacedwith
uncooperativemanagementorforcedtoacceptanon-compatiblenewmemberaren’t
alwayssolucky.Inthesecases,teamsoftencannotfindthewaybacktobeingahigh-
performingteam.Thenextchapterdiscusseswhattodowhencultureclashesforceteams
intoanonproductivestate.
References
[BROOKS]Brooks,FrederickP.,Jr.1996.TheMythicalManMonth,Anniversary
Edition.Reading,MA:Addison-Wesley.
[TUCKMAN]Tuckman,Bruce.1965.“DevelopmentalSequenceinSmallGroups.”
PsychologicalBulletin63(6):384–399.
Chapter21.WhenCulturesCollide
Scrumteamsstrivetocreateanatmospherewherepeoplecanenjoytheprocessof
producingexcellentworkyetstillmeetbusinessgoalsandremainprofitable.Gettinga
teamtoalevelwherethisispossibleisoftenastruggle,butwithmanagementsupport,
Scrumteamscanreachahigh-performing,richmixoffunandaccountabilityanddeliver
atunprecedentedrates.
Thechemistrytheteamhascreated,however,isfragile.Itscultureandabilitytodeliver
effectivelycanbethrownoffbalancewhensomeoneisaddedtotheteaminthemiddleof
aproject.Chapter20,“AddingNewTeamMembers,”showswhathappenswhentheright
teammemberisaddedintherightway:Scrumteamscanquicklyrecoverfromsucha
change.Whathappens,though,whenteamsareforcedtoacceptanewteammemberwho
isnotagoodculturalfit,someonewhoisnotparticularlyflexible,capable,oragilesavvy?
ConsiderthecaseofChemTeam,ahigh-performingteamwho,liketheteaminChapter
20,neededtoaddateammember,butunlikethatteam,wasunabletohandpickthenew
teammember.
TheStory
ChemTeamhadbeenworkingtogetherformonths,andthebondsthathadformedamong
theteammemberswerestrong.Theteamwasdisciplinedandfocused,executingonits
deliverablesanddelightingcustomers.Thepeopleontheteamtrustedeachotherandfelt
comfortablewiththeteam’sbalanceofconflictandcamaraderie.Theywereaperforming
team.
Onefatefulday,inthemiddleofaproject,adirectorinthecompanycametotheteam
withanewbusinessidea.Thisbusinessidea,inhismind,wouldradicallychangethe
team’sworld.Itnearlydid.Butnotinwaysheintended.
“Guys,greatnews!”heshouted.“WehaveanopportunitytoworkwithWirelacorp!”
MostofChemTeamstaredblanklybackatthemanager.“Wirelacorp?I’veheardof
them.Don’ttheyprovideInternetaccessinpubliclocations?”askedKamiar.
“Indeedtheydo,”hesaid.“Andwearethisclosetosigningacontractwiththemthat
willallowoursmallbusinessanduserbasetobeabletoaccesstheWirelacorpnetwork,at
noadditionalcosttoourusers!”
Atthispoint,theteammembers’blanklooksturnedtoonesofslightconcern.The
projectthattheyhadbeenworkingonwasdesignedtofocusonlyontheusersthe
businesshadonitsownnetwork—itdidn’tincludethird-partyinterfacesorintegrations
forthingsliketheWirelacorpnetwork.Notonestobedauntedbysuddenchange,though,
theteamimmediatelybegantothinkofwaysitcouldaltertheserviceofferingtotake
advantageofthisopportunity.
Seeingherteambeginningtogetexcitedaboutthechallenge,Shauna,the
ScrumMaster,spokeup,“Wecandothis,”shesaidconfidently.“Justgiveusacoupleof
weekstocleanupwhatweareworkingon.Thenwecanreprioritizeourproductbacklog
withyouandtheotherstakeholderstogiveyouanewreleaseplan.”
“Notime!”saidthedirector.“WehavetogetthisworkingASAP;thisdealiscriticalto
ourbusiness!”
“Wewouldprefertofinishtheworkonthissprintbeforewechangegears,”explained
Shauna.“Butifthisistrulyurgent,wecanabortthissprint,whichwillessentiallyresultin
usdroppingwhatwe’redoing,andbeginreprioritizingrightaway.Eventhen,ifyouwant
ustoprovideyouaccurate-enoughestimatesforyoutobuildago-to-marketstrategy,we
needatleastthreedays,andyouwillhavetoacceptthepriorityofthebacklog.”
“Icanonlygiveyouuntiltheendoftheday,”saidthedirector.“Dothebestyoucan.”
Fortheremainderoftheday,ChemTeamassesseditsbacklog,thecurrentstateofthe
system,andtherequirementstointegrateandinterfacewithWirelacorp’snetwork.The
teamproposedatimelinethatrepresentedthebareminimumtomeettheneedsofthe
businessandthatitfeltcomfortableindelivering.Theythenpresentedittothedirector
thenextmorning.
“Team,greatwork,butweneedtoshortenthetimelinebythreemonthstomeetthe
contractdatesthatwehavebeendiscussingwithWirelacorp,”explainedthedirector.“I
needyoutomakethishappen—tellmewhatyouneed.”
Theteammemberswerevisiblyfrustrated.Theysaidtheywouldcomebackinacouple
hourswithsomeplansonhowtheycouldpullintheschedulewithouttakingonthe
technicaldebtthatwouldcomefromhackingthesystemtogether.
Itwasnoweightatnight.Thedirectorwasstillinhisoffice,waitingfortheteamto
finishitsdeliberation.Theteamwenttothedirectorwiththeonlysolutionitcouldcome
upwithinthefaceofsomanyimmovableobstacles.Shaunasaid,“Wehaveasolutionfor
howwethinkwecanbringthisin,butitisnotidealandmaynegativelyimpactusoverall
fromateamandtechnicalperspective.Youneedtounderstandthisbeforewecan
continue.”
Asthedirectornoddedeagerly,Shaunacontinued,“Theonlywaytogetthisdonemore
quicklythanweestimatedistobringinmorepeople,butnotjustanywarmbodieswill
do.Wehavelookedattheskillsandcompetenciesweneedonthisprojectandhave
identifiedtwopeoplewho,ifadded,wouldenableustoreachourgoal:JulioandNancy.
Julioisavailable;Nancyisnot.Forthistohavethebestchanceofsuccess,weneedboth
ofthem.”
“YoucanhaveJulio,butyouabsolutelycannothaveNancy.Sheisonaprojectcritical
tothebusinessandcannotbetakenawayfromit.”
“Isn’tthisaprojectcriticaltothebusinessaswell?Wehavebeenworkingonthisfor
eightmonths.Andweareclosetothefinalrelease,”counteredShawna.
“Allprojectsarecritical,”saidthedirector.“AndyoucannothaveNancy.Youcanhave
Katherine;sheisavailable.”
Katherinewasaseasoneddeveloper.Shehadbeenconsideredbytheteamalreadybut
hadbeenreadilydismissed.Whileshehadalltheskillstheteamwaslookingfor,she
lackedthecorevaluestoworkinacollaborativeenvironment.Someoftheteammembers
hadworkedwithheronpastprojectsandhadexpressedconcerns:Shehadfoughtherlast
teamontheideaofdailyscrums,arguingthattheywereawasteoftime,eventhoughthe
restoftheteamfoundthedailycommunicationhighlyvaluable.Shaunahadalsoworked
withKatherineonapastproject.ShehadnoticedthatKatherineoftenrushedandhad
buggycodeasaresult.WhentheQAteammembersonthepastprojectspokewithher
aboutwaystocreatecleanercode,sheattackedthem,sayingtheydidnotknowhowtodo
theirjobs.Itwascleartotheteamthat,whileKatherinehadthetechnicalskills,her
competenciesandattitudeaboutworkingonateamwerevastlydifferentfromthecultural
valuesestablishedbyChemTeam.
Shaunasharedtheseconcernswiththedirector.Hewasnotpersuaded.“Youguyssaid
youneedtwopeopleontheteamtomakethiswork.ThesearethetwoIcangiveyou.We
allagreethatKatherinecandothework.It’suptoyoutofindawaytoworkwithher.”
Theteammemberswerenolessskepticalbutfelttheyhadnochoicebuttotake
Katherine.Theteamreluctantlyagreedtothenewscheduledateandthenewteam
members.BothKatherineandJuliostartedworkingontheprojectthefollowingday.
ChemTeamtriedtodothingstherightway.Itsmembersknewtheyneededtobaseline
newteammembersandusedtheaddingnewpeopletotheteamapproach(seeChapter
20).TheycreatedasurveyandhadbothKatherineandJuliotakeitasabaselinemeasure
oftheirknowledgeoftheteam’sprocess,tools,andproject.Thesurveywastoberepeated
onceaweeksotheteamcouldtrackhowwellthenewmembershadinternalizedboththe
processandtheproduct.BothKatherineandJuliowereenthusiasticatfirst.Astimewent
on,itwasclearfromJulio’sanswersandhisworkthathenotonlyunderstoodthesystem
butalsohadembracedthepracticesoftheteam.
Thoughsheshowedimprovementinsystemknowledge,Katherineshowedlittle
improvementintheareasofprocessandculture.Katherinequestionedthebenefitsoftest-
drivendevelopment(TDD),pairing,andworkinginasharedspaceateveryopportunity.
Moreoftenthannot,workwasinterruptedtodiscussthemeritsofacertainagilepractice
orprinciple.Hervocalskepticismwasbeginningtoerodetheteam.Velocitydecreased,
qualitysuffered,andteammoralefelltoanall-timelow.Shaunadidherbesttohandle
Katherine’scriticismsotheteamcouldstayfocused,butatthebeginningofweekfour,
Katherinesnapped.
“I’veansweredthisstupidsurveyweekafterweek.I’vealreadyproventoyouthatIcan
buildthissystem.IdonotcarethatwearedoingTDD.Ihateworkinginthisspace,andI
wouldbebetteroffjustwritingcodebymyselfinmyoffice.Thishasgottobethemost
inefficientwayofworkingever!Icouldhavewrittenhalfofthesystembynow!”
ShaunatriedonelasttimetobringKatherinearound.“Katherine,aswehavediscussed
manytimes,thereasontheteamisbuildingtheapplicationusingagileprinciplesand
practicesistoavoidthepitfallseachofushasrunintoonpastprojectsthatweremore
traditionalinnature.Wearenotperfect.Thisishardwork,bothfromadevelopment
standpointandfromapersonalgrowthstandpoint.Wearealllearningthisaswego,and
wemakemistakes.”
Shecontinued,“Regardlessofhowwemightfeelindividually,weareateam,andwe
needtocontinuetoworkasateam.ThemerefactthatwearedoingTDDhasloweredour
bugcountbyafactorofninecomparedtotheotherteams.Also,youandJuliohavebeen
abletogetuptospeedonthissysteminlessthanfourweeks,andyounowknowitlike
thebackofyourhand.Withoutworkinginasharedspaceandpairing,thiswouldhave
neverhappened.”
“Icannotworkthisway.Iwillnotworkthisway!”saidKatherineasshestormedoff.
Bythispoint,Katherine’sdestructiveinfluencehadspreadthroughouttheteam.Many
werebeginningtoarguetherelativebenefitsoflong-establishedpractices.Oneday,a
teammemberbecamesofrustratedwiththesuddeninfightingandbickeringthathe
squeezedacanoffreshlyopenedsodawater,sprayingitonthewalls,thecomputers,the
desks,and,ofcourse,theotherteammembers.Hestormedoff,frustratedbecauseofthe
team’sdeterioratingcohesionandtrust.Somethinghadtobedone.
ShaunaandtheteammemberswentbacktothedirectorwithoutKatherine.They
showedhimthesprintandreleaseburndownsthatwereclearlytrendingtowardalower
velocity.Theysharedwithhimnotonlytheirfrustrationsbutalsothefactthatbugcounts
wereupandtheunittestcodecoveragenumbersweredown.Theyrecommendedthatthey
dropKatherinetosalvagetheproject.Theywereturneddownflat.
“Look,Iknowit’sdifficult,butitcan’tbehelped,”saidthedirector.“Myjobistoget
thiscontractsigned.Yourjobistomakethedeadline.You’vetoldmeyouneedtwo
peopletogetthisdone.Sheneedssomethingtoworkon.Workitout.”
Afterthemeeting,Shaunasawthepainonthefacesofeachteammember.EvenJulio
lookedbeaten.Sheaskedherselfwhatwasmoreimportant—theteamanddeliveringthe
projectorthefactthatKatherinedidnothaveanythingelsetoworkon?
Shaunadecideditwastimeforaction.Sheandtheteamoutlinedaplanthatwould
addressboththerequirementofthedirectortogiveKatherinesomethingtoworkonand
theneedfortheteamtoreturntoitsdesiredhigh-performingstate.Allthatwasleftwasto
tellKatherine.
Thenextmorning,ShaunapulledKatherineaside.
“Katherine,look,Iseethatyou’renothappy,andyouseetheteamisnothappy,right?”
saidShauna.Katherinerespondedwithacautiousyes,wonderingwherethiswasgoing.
“Well,Italkedtothedirectoraboutit.Hesaidyoucan’tcomeofftheprojectbecause
youdonothaveanotherprojecttomoveonto.”
Katherinenoddedinagreement.
“Atthesametime,ifthingscontinueastheyare,thisprojectisgoingtoimplode,and
everyone,includingme,willhaveabigblackeyefromit.I’vegotaplanonhowwecan
avoidthisandkeeppeoplehappyintheshortterm,”saidShauna.
TheplanthatShaunalaidoutwastohaveKatherineremainontheteambutonlyon
paper.Katherinewouldworkinherofficeonnon-system-criticaltasks,awayfromthe
teamspaceuntiltheprojectanditsassociatedtimecrunchwasover.Atthatpoint,they
couldalldiscusstheissuesanddeterminealong-termplan.Katherineagreed.
Katherinekeptherdistancefromtheteam,onlycomingtothedailyscrums—justdown
thehallfromthedirector’soffice—toshowthatshewasstillontheproject.Theteamwas
abletostayfocusedanditsvelocityreturned.Thecodecoveragenumbersincreasedand
overallprojectqualitywentup.EvenJulio,whohadabackgroundintesting,wasnow
writingcode.
FourweekswentbybeforethedirectoraskedShaunaintohisoffice.
“Theteamseemstohaveturnedaround,butIdon’tseeKatherine.Whatpartofthe
systemissheworkingon?”
Shaunathoughtlongandhardonhowtoanswerthis.
“Sheisworkingonnon-system-criticaltasks,”shesaid.“Ididthisbecausetheteam
wasnotgoingtomeetthecontractdatesontimewiththequalityourcustomersdemand.
Also,severaloftheteammembershadthreatenedtoquitthecompanyratherthanwork
withher.Itwastherightthingtodo.”
Visiblydisturbedatbeingcountermanded,thedirectorexploded,“YoutookKatherine
offtheprojectevenafterItoldyoutokeepheron?”
“Yes,Idid,”saidShaunawithsteadfastclarity.
Shaunasatasthedirectorexpressedhisdissatisfactionwithher.Herantedandraved
abouthowwhatshedidwasinsubordinationandthatsheshouldbefiredforit.Aftera
fewminutes,Shaunainterruptedhim.
“Letmeaskyouaquestion,”saidShauna.“Whatismoreimportantatourcompany:
keepingpeoplebusyordeliveringforourcustomers?”
“That’snotthepoint,”saidthedirector,poundinghisfistonthedesk.
“Itisthepoint.Pleaseanswerthequestion,”pressedShauna,nowgettingvisibly
frustrated.“Keepingpeoplebusyordelivering—whichisit?”
Reluctantly,thedirectoranswered,“Well,delivery,ofcourse,butthatisnoexcusefor
what—”Shaunainterruptedyetagain.
“Iagree.Ishouldn’thavedonewhatIdid.Notbecauseitwasn’ttherightthingtodo
butbecauseyoushouldhavebeentheonetodoit.Manytimes,Iaskedyoutointervene.
EachtimeIasked,yousaidno.Youwerenotwillingtotakeaction.Itismyjobas
ScrumMastertokeeptheteamhealthyandtheprojectmoving.Youwerelimitingmy
abilitytodomyjob,andyouwerenotinterestedinhearingsolutionsorhelpingmefixthe
issuewithKatherine,soIfixeditasbestIcould,”Shaunafinished.
“It’sabadprecedenttoset,”saidthedirector.
“Aworseprecedenttosetisinactionbymanagement,”respondedShauna.
Shaunawatchedforapainfullylongandsilentmomentasthedirector’sfacewentfrom
adeepshadeofredtolightred,andthentoanalmostnormalskintone.Hetookadeep
breathandsaid,“Look,Iunderstandwhatyoudidandwhyyoudidit.Ihavebeenundera
lotofpressurewiththisdealandamjustthankfulthatit’sover.”
“Whatdoyoumean,it’sover?”askedShauna.
“Didn’tItellyou?Earlierthisweek,thecontractnegotiationscollapsed.Wearenot
goingtodothedealwithWirelacorp.Itmeanswecangobacktowherewewere,”said
thedirector.
“Uhhh,”saidShauna.
“So,howlonguntilwecandeliverontheoriginalfunctionality?Nextmonth?”
Andthatisanotherstoryinandofitself.
TheModel
Inthisstory,theoriginalteamwasawell-oiledmachine.Ithadbecomeaccustomedtoits
ownwayofworkingandhaddevelopeditsownculture.Yet,theintroductionofone
uncooperativeteammemberalmostcausedtheteamtoimplode.Whathappened?
Inanutshell,Katherinefitthecultureofthelargergroupandcompany,butshedidnot
fitthecultureoftheteam—aculturewasformedastheteamhadgrownandworked
togetheronitsown.Juliowasopentotryingnewthingsandhadtherightmindset,but
Katherinedidnot.Shaunaandherteamidentifiedthisdisparitybutwereignoredby
management.IcallthetypeofbehaviorKatherineexhibitedaspartoftheteamsocial
deviance.
Devianceinsocietycanbedefinedasactingordoingthingsinawaythatviolatesthe
establishedculturalnorms.Culturalnormsapplyonmanylevels,frompersonaltofamily
toworktosociety.Forexample,stealingviolateslargersocietalculturalnormsinthe
samewaythat,say,notperformingcertaindevelopmentactivities(suchasTDD)violates
establishedteamculturalnorms.
In1938,RobertK.Mertonpublished“SocialStructureandAnomie”[MERTON].In
thispaper,heproposedstraintheory.Straintheorywasanattempttoidentifyreasonsfor
varyingcrimeratesbetweendifferentsocialclasses.Mertonidentifiedtwosocial
structures:culturalgoals,whichwereculturallyassignedgoalsandaspirations,and
institutionalizedmeans,whichoutlinedtheacceptablemeansforachievingthegoals.
Mertontheorizedthatculturalgoalswerewhatallpeoplewouldwantandexpect
throughouttheirlives,andinstitutionalizedmeanswerethewaystoachievethecultural
goals,suchasobeyingestablishedsocietystandards.Mertonstatedthatabalancewould
occuraslongasindividualsfelttheculturalgoalswerebeingachievedbyadheringtothe
“institutionallyacceptedmodeofdoingso.”Merton’stypologyofdevianceisillustrated
inFigure21-1.
FIGURE21-1RobertK.Merton’stypologyofdeviantbehavior
Mertonidentifiesfiveattributesofstrain:conformity,ritualism,retreatism,rebellion,
andinnovation.Inconformity,peopleacceptboththegoalsandmeansofsocietyandchug
thoughlife.Peoplewhofallintheritualismbuckethavegivenuponthegoalsofsociety
butcontinueconformingtothemeanstoachievethem.Retreatismiswherepeoplereject
boththegoalsandthemeans.Whenpeoplerejectboththegoalsandthemeansand
replacethemwithnewgoalsandnewmeanstoreachthosegoals,theyareinrebellion.In
innovation,peopleacceptthegoalsofsocietybutusedifferentmeanstoachievethem,
traditionallythoseofcrime.
Take,forexample,auniversitystudent.Thestudent’sculturalgoalisgraduation;the
institutionalizedmeanstoachievethisgoalaretoattendclass,study,andperformwellon
examsandpapers.
Ifastudentdoesallthisandgraduates,thestudenthasconformed.
Ifastudentdoesnotstudyanddoesnotpassexams,butkeepsgoingtoclass
anyway,thestudentisbeingritualistic—heisgoingthroughthemotionswithno
hopeofachievingthegoal.
Ifastudentknowsthathewillnotgraduatebecausehisgradesaresubparandasa
resultstopsattendingclassesorstudying,andinsteadplaysvideogamesinhisdorm
room,hehasretreated.
Astudentinretreatmaytireofhisdrifterstatus(orhavehisfundingcutoff)and
foregohiseducationalltogether.Hemaybegintoprotestthateducationisoverrated
andthatthegoalofgraduationisnotonethatpeopleshouldstrivetoachieve.He
willidentifynewgoals(startinghisownbusiness,forexample)anddevisenew
meanstoachievethem.Astudenttakingthisactionisinrebellion.
Lastly,ifthestudentstillhasthedesiretograduateuniversitybutisnotabletodoso
usingtheacceptedmeans(e.g.,studying),hemayresorttoinnovative,even
criminal,methods,suchascheatingorplagiarism.
Businesses,too,haveestablishedsocietiesandcultures.Thesocietalrulesandcultural
normsaretheinstitutionalmeans.Theypropagatethroughoutthecompany,impactingand
directingeachemployee.Inbusiness,culturalgoalsareoftencommunicatedascompany
goals,whichemployeesofthecompanycanachieveorfailtoaccomplish.Therefore,
whenthinkingofbusinesses,IreplaceMerton’sculturalgoalswithcompanygoalsand
hisrejectandacceptwithfailandsucceed,asillustratedinFigure21-2.
FIGURE21-2Shauna’steamexperimentswithinnovativemeans
Withthissmallmodification,considerhowMerton’stypologyappliestoasoftware
organizationwherethecompanygoalistoproducehigh-qualitysoftwarebyacertain
releasedate,andtheinstitutionalizedmeanstodosoaretoworkinatraditional,plan-
drivenway,hittingmilestonesandrelyingonup-frontplanstosteertheproject:
Conformity—Teammembersfollowtheinstitutionalmeansandachievethe
companygoals.
–Leadstohighquality,highmorale,andsoon.
–Employeeshaveboughtintotheprogram.
–Peopleareengaged.
Ritualism—Teammembersaccepttheinstitutionalizedmeansbutfailtomeet
companygoals.
–Lowquality,lowmorale,andsoon.
–Showupanddothejob.
–Peoplearedisengaged.
Retreatism—Teammembersdonotacceptinstitutionalmeansandfailtomeet
companygoals.
–Lowquality,lowmorale,andsoon.
–Showupbutdon’twork.
–Activelydisengagedandknowit.
Rebellion—Teammembersdevisenewmeansanddevelopnewgoalsthatbuildon
thecompanygoals.
–Themeansmightbequitedifferent.
–Companygoalsarestillmet;teamsetsgoalsinadditiontocompanygoals.
–Teamculturecanbequitedifferent.
–Willcausefrictionintheorganization.
Innovation—Teammembersdevisenewmeanstobesuccessfulinachievingthe
companygoals.
–Canbenegative(workingthesystem)orpositive(findingnewwaysto
accomplishthesamegoals).
–Ifitispositive,innovationleadstohighquality,highmorale,andsoon.
–Unlikerebellion,onlythemeansaredifferent,notthegoals.
–Willcausefrictionintheorganization.
Leadersandexecutivesestablishinstitutionalizedmeanstoachievecompanygoals.
Sometimes,however,newmeanscanbecreatedandcancoexistwithinthe
institutionalizedmeansestablishedbybusiness.Inthischapter’sstory,Shauna’steamhad
establisheditsownmeans(TDD,pairprogramming,groupspace)toachievethecompany
goals.Themindsetchangethataccompaniesthesemeans,however,hadalsocausedthem
tothinkandworkquitedifferentlythantherestoftheorganization.
Shauna’steamwasuniqueinsidethecompany.Mostoftheotheremployeeswere
ritualistic:Theyacceptedthemeanstodotheirworkbuttypicallyfailedatachievingthe
goals(e.g.,delivery).Thiswasevidentbylookingatfactorsinsidethecompany.For
example,bugcountswerehigh,qualitywaslow,moralewaslow,therewasamoderate
amountofturnover,peoplehadlittleornopassionfortheirwork,theywerenotproudof
theirwork,andmostteamswerenothighperforming.
Shauna’steamhadgrowntiredoffailureandhaddecidedtotrysomethingnew,sothey
hadadoptedScrumandExtremeProgramming(XP).Byusingnewmeanstoachievethe
samecompanygoals,Shauna’steammovedoutofritualismandintoinnovation,asshown
inFigure21-2.
Astheteamlearnedtoworkandturnedintoahigh-performingteam,itnotonlyused
newmeansbutalsobegantoestablishnewgoalsforitselfbasedonitsheightened
understandingofScrumandXP.Thesegoalsincludedthecompanygoalsofdelivering
softwarebyacertainreleasedate,butalsoincludedotherscenteredaroundpersonal
growthandbecomingbetteratthecraftofsoftwaredevelopment.Theteam’snewgoals
includedthefollowing:
Shiphigh-qualitysoftwarethatmeetscustomerneeds.
Allowtheteamtoberesponsivetochangingbusinessneeds.
Workatasustainablepacewhilestilldeliveringonschedule.
Theseculturalgoalswereuniquetotheteamanddidnotspanthecompany.Theteam’s
meansweresimilarlyuniqueandincludedthefollowing:
Workinacollaborativeworkspace.
Pairprogram.
DesignusingTDD.
Collectivelyestimateandcommittothework.
Asthegapwidenedbetweentheteamandthecompanyintermsofmeansandgoals,
theteammovedoutofinnovationandintowhattherestoftheorganizationmightclassify
asrebellion(seeFigure21-3).
FIGURE21-3Teammovingtorebellion
Theteamisinarebelliousstateatthebeginningofthestory.Theyarehigh-performing
andachievingnotonlythecompanygoalsbuttheteam’sgoalsaswell.Theirdirector,
however,isnotinsyncwiththeteam.Hethrowsanewsetofobjectivesattheteaminthe
middleoftheproject,onesthatthreatentheteam’sowngoals.Thoughtheteamoffersa
planthatwouldallowittocontinuetomeetitsgoalswhileincorporatingthenew
companygoals,thedirectorislessconcernedwiththeteam’sgoalsthanwiththe
company’sgoals.
EnterKatherine.Whenshewasplacedonateamwitharadicallydifferentagenda,she
rejectedtheteam’sgoalsanditsmeanstoachievethem.Heropposingviewpointmadeher
adevianttotheacceptedsocialnormsoftheScrumteam.Inresponse,sheatfirstrebelled,
notbyintroducingnewmeansornewgoalsbutbyattemptingtobringtherestoftheteam
backtothecompanygoalsandinstitutionalmeansthatshewasusedtoworkingwithin.
Herrebellioncreatedsomuchconflictwithintheteamthatitcouldnotfunction.Too
muchoftheteam’sfocusandenergywerespentontryingtoworkwithKatherine.Asa
result,thework,themorale,andtheteam’sperformancesuffered.
OncetheteamrealizedthatKatherinecouldnotfitintoitsculture,theteamtriedto
removeherandherdestructiveinfluence.Whenthatideawasrejectedbymanagement,
theteamaskedKatherinetoretreat,leavingtheothermemberstoworkinpeaceandreturn
to“normal.”
Whilethisworkedfortheteam,itwasariskymoveforShauna.ThestepsShaunatook
toremoveKatherinefromtheday-to-dayteamactivitieswereatruelastresort,onethat
mightnotbeworththerisktomanyofus.Shecould(andshould)havetakendifferent
stepstomoveKatherinetowardconformity.
KeystoSuccess
Thebestwaytodealwithcultureclashistoavoidaddingmemberstoyourteamwho
won’tfitin.Ideally,teamsshouldbeabletocontroltheirowndestinies,sothattheteam
membersthemselvesdeterminethecompositionofeachteam.Thereality,however,isthat
sometimesteamsareforcedtotakeinpeoplewhomtheywouldnothavechosenontheir
own.Whenthathappens,teamsneedtoworkwithwhattheyaregivenandstaythecourse
inthefaceofobstacles.
ControlYourOwnDestiny
Thebestwaytokeepfromhavingadeviantinyourmidstistoensurethattheexisting
teamhasasayinchoosingnewteammembersforyourteam.Itrytofollowthesesimple
rulesofthumbonmyprojects:
Lettheteaminterviewandpickitsownmembers.
Lettheteamdecidethefateofteammembers.
WhenIwasinuniversity,Itookseveralclasseswheretheprofessorrequiredusto
delivertermpapersandpresentationsasateam.Theteamswereself-formingandself-
managing.Lookingback,wewereveryagile.Wepickedourteammembers,we
collaborated,weworkedtogether,andwedeliveredtogether.Wehadteammemberswho
sometimesdidnotpulltheirweight—Ican’tblamethem.Iwenttoaschoolthatwason
thecoastandhadgreatsurfing.Thesepeoplespenttimeatthebeachandcametous,
usuallyattheendofthepaperandpresentationdevelopmentcycle,andsaid,“Hey,canI
getmynameonthis?”
Asyoucanimagine,wesaidno,andtheyoftendidn’tpass.Wewereempoweredto
decidethefateofourteammembers.Ifyoudidthework,youwereallowedtosharethe
credit.Ifyoudidnotdothework,youfailed.Itwasverysimpleatthatpoint.
Whileitcan’talwaysworkthatwayinbusiness,ScrumMasterscaneducate
managementaboutthebenefitsofallowingteamstochoosetheirownteammembersand
askforthepowertomovemembersofftheteamwhoaredoingmoreharmthangood.See
Chapter3,“UsingTeamConsultantstoOptimizeTeamPerformance,”formore
information.
WorkwithWhatYouHave
Let’sfaceit.Teamsarenotalwaysempoweredtodowhattheythinkisrightwhenit
comestotheirownhealth.Sohowcanteamskeepthedamagetoaminimumwhenthey
haveanewteammatethrustuponthem?First,whenanewmemberisadded,takea
momenttoassessthatperson’sconformitylevelandbackground.Ifthenewteammember
iscomingfromaculturethatclasheswithyourteamandhasahistoryofrebellingagainst,
retreatingfrom,orinnovatingtogetaroundthosenorms,youshouldbeencouraged.After
all,rebellionitselfisn’tbad.Onateamwhoisnotyetperformingwell,arebellious
membercanactuallypushtheteamtowarddesirablegoals.
Ifthenewteammemberisinconformancewithorritualistictowardaculturethat
clasheswithyourown,thereisstillnoneedtopanic.Youshouldexpect,however,tohave
todomoretolaythegroundworkonhowyourowncultureworksandwhatwillbe
expectedofanewteammember.ThatmightmeaneducatingheraboutScrumandagile
principlesandhelpinghertoseethereasonbehindthechoicesyourteammakes.Doall
youcantohelpnewmembersconformto,ratherthanrebelagainst,thegoalsandmeans
yourhigh-performingteamhasestablishedorisstrivingfor.
Second,whiledoingyourbesttobringthepersonintoyourculture,beopentonew
ideasandmethodsthepersonmightbringtotheteam.IgobacktomanyofDale
Carnegie’sprinciples,assummarizedinTheGoldenBook:
Becomeafriendlierperson.
Winpeopletoyourwayofthinking.
Bealeader.
BreaktheWorryHabit.
Cultivateamentalattitudethatwillbringyoupeaceandhappiness.
Don’tworryaboutcriticism.
Preventfatigueandworryandkeepyourenergyandspiritshigh[CARNEGIE].
WhenIfirstreadtheCarnegiebooks,Iwasskeptical.ItwasnotuntilIreallyanalyzed
howIapproachandworkwithpeoplethatIunderstoodhowthesebasicprinciplesreally
makeyouabetterpersontobearound—andifyouareabetterpersontobearound,
peoplewillgravitatetoyou.
ThoughIunderstandwhytheScrumMasterinthischapter’sstory,Shauna,arrivedat
herradicalsolutiontotheproblemKatherinepresented,itwasnonethelessrisky:itmight
havecostShaunaherjob,anddefinitelydidleavetheteamshorthanded.IfShaunahad
imploredteammemberstoapplytheCarnegieprinciples,orjustfollowedthemmore
steadfastlyherself,theteammighthavebeenabletoreturntoitsperformingstateeven
withKatherineonboard.HerdecisiontosidelineKatherine,whichwasnecessitatedin
partbythetimecrunchshewasunder,wasatruelastresort,notapracticeIwould
recommendinmostcases.
Instead,ImighthavecounseledtheteamtoallowKatherinemoretimetobecome
accustomedtoTDDasawayofprogrammingbeforeforcinghertopair.Ifallcodewas
writtenbypairs,perhapstheteamcouldhavecompensatedbyaskingKatherinetodo
moretestingworkandlesscodedevelopment.MaybeifChemTeamhadstoppedworrying
somuchaboutwhyKatherinewasn’tadjustingandhadtriedhardertonottakeher
criticismpersonally,itsmemberscouldhavestayedfocusedonthesprintratherthanone
teammember’sdifficulties.Iftheteamhadtriedtowinhertoitswayofthinking
systematicallyandslowly,everyonemighthavebeenlessfrustrated.
Shaunaalsomighthaveaskedotherleadersontheteamtohelpkeeptheteamfocused
onthesprintitselfwhileshepersonallydealtwiththeimpedimentthatwasKatherine.
LettingKatherineassimilatemoregraduallymightnothavehelpedtheteam’svelocity,but
itprobablywouldn’thaveslowedtheteamdowneither.Focusingonthepositiveand
minimizingthenegativemighthavegonealongwaytopreventblow-upsandsodawater
explosions.
StaytheCourse
Whenfacedwithanewmemberwhoisapotentialdeviant,thefollowingguidelinescan
helpkeepyourhigh-performingteamontrack:
Trytokeepthedeviantoffyourteam.Helpmanagementunderstandthatpeopleare
notinterchangeablecomponents.Teammembersshouldbechosenwithcare,not
pluggedinbasedsolelyonavailabilityorevenonaparticularskill.
Educatenewteammembersabouttheteam’scultureandgivethemtimeandspace
toadapt,sothattheyarenotforcedtoinnovate,rebel,orretreat.
Ifanewmemberdoesbegintodeviatefromtheteam’snorms,gobacktothebasics
andgetpeoplecomfortablewithScrum(seeChapter1,“Scrum:Simple,NotEasy”).
Ifthiseffortfails,identifywhatiscausingthepersontoshiftfromtheestablished
cultureoftheteam.
Sometimesanewteammemberbringsadeviantpointofviewthatcanultimatelybe
helpful(ratherthanhurtful)totheteam.Helptheteamtounderstandthatitisnot
onlythenewteammemberwhohastoadjust.Whenteammake-upchanges,existing
teammembersmighthavetochangesomeoftheirestablishednormstoassimilate
newmembers.
Ifdeviantbehaviorpersists(orhasbeenfesteringonyourteamforsometime),it’s
timetoinvolveHR,management,ortheteamitselfinmovingthenonconforming
teammembertoamoresuitableteam.
Whetheryourteamisabletochooseitsownteammembersorhastotakewhatitgets,
returningtonormalafteranewteammemberisadded,isoneofthetoughestchallenges
anestablishedteamhastoface.Sometimesjustknowingthattemporarypainistobe
expectedandthatthereisaroutebacktoaperformingstatecanmakethetransition
quickerandeasierforallinvolved.
References
[CARNEGIE]Carnegie,Dale.TheGoldenBook.
http://www.motivationalmagic.com/library/ebooks/inspirational/GoldenBook.pdf
(accessed27March2015).
[MERTON]Merton,Robert.1938.“SocialStructureandAnomie.”American
SociologicalReview3(5):672–682.
WorksConsulted
Baumer,EricP.2007.“UntanglingResearchPuzzlesinMerton’sMultilevelAnomie
Theory.”TheoreticalCriminology11(1):63–93.
Featherstone,Richard,andMathieuDeflem.2003.“AnomieandStrain:Contextand
ConsequencesofMerton’sTwoTheories.”SociologicalInquiry73(4):471–489.
Monahan,SusanneC.,andBethA.Quinn.2006.“Beyond‘BadApples’and‘Weak
Leaders.’”TheoreticalCriminology10(3):361–385.
Murphy,DanielS.,andMathewB.Robinson.2008.“TheMaximizer:Clarifying
Merton’sTheoriesofAnomieandStrain.”TheoreticalCriminology.November2008,
Vol.12no.4,p.501–521.
Chapter22.SprintEmergencyProcedures
Oneofmyall-timefavoriteTVshowsasachildwasTheA-Team.Myfavoritecharacter
wasHannibalSmith.Hehadasaying,“Iloveitwhenaplancomestogether.”Ido,too,
whichisprobablywhyIloveScrumsomuch.
Sometimes,though,despiteallyourbestefforts,plansfallapart.AndI“pitythefool”
(asMr.Twouldsay)whodoesnothaveacontingencyplanwhenthingsunravel.Too
manytimesontoomanyprojects,I’veseenteamsgetintotrouble.Inthischapter,Ishare
asetofemergencyproceduresforyoutofollowwhenasprintgoeswrong,includingthe
last-ditchoptionofsprinttermination.
TheStory
“Guys,I’vegotbadnews:Therumorsweretrue,”saidMike.“They’vedecidedtosplit
ourdivisionintotwogroupseffectiveimmediately.AndI’mpartofthatshuffle.I’mgoing
tobemovingtothenewgroup;therestofyou,though,arestayinginthisone.Sothe
goodnewsisthatatleastmostoftheteamcanstaytogether.”
Mike’steammatesvisiblydeflated.Mikealwayshadgoodideasonhowtosolve
problemsandissues.Hewasakeypartoftheteam’ssuccessandwouldnotbeeasyto
replace.
“That’sreallybadnews,Mike,”saidBilly.“It’sjustnotgoingtobethesamewithout
you.Idon’tknowhowwe’retofillyourshoes.”
“Well,”saidMike.“Startfiguringitoutquickly.Whentheysaid‘immediately,’they
meantit.Iwastoldtocomebackandpackmystuff.Thereisamovingcrewcoming
tomorrowtorelocatemetoadifferentbuilding,”saidMikewithgreatdisappointment.
“What?”exclaimedBilly.“You’relastdayistoday?”
Mikenodded.
“Whatarewesupposedtodo?”askedMark.“We’rerightinthemiddleofasprint.
Whatdoesthismeanforourproject?”
“Ireallydon’tknowwhattosay,guys,”Mikeansweredresignedly.“Imadethecasefor
amoregradualtransitionbutgotshotdown.Ididn’twanttopressthepointwiththe
generalmanagerinfrontofaroomfullofpeople,”saidMike.
“Well,maybewecanconvincehimifweapproachhimthroughtheappropriate
channels,”saidBilly.“C’mon,Mark,let’sgofixthiswhileMikepacks.”
Thepairapproachedthegroupmanager,Pam.Sheexplainedthatwhilesheunderstood
theirdilemma,therecouldbenodelayinMike’smove.
“Thenweneedtofigureoutwhattodoaboutthissprint,”saidBilly.“We’reabout
halfwaythroughoursprint.There’snowaywecanfinisheverythingwecommittedto
withMikegone.Can’ttheywaitacouplemoreweeksandgiveustimetowrapsome
thingsup?”
“Iknowthisisbadtiming,”saidPam.“Butinthebigpicture,itmakesthemostsense
fortheorganizationtomakethishappennowsoeveryonecanadjustandbeginbeing
productiveagainassoonaspossible.Iknowyou’refrustrated.Ihavebeeninthissituation
before.Whatwedidwaslookathowourworkwasstructuredandhowitalignedtothe
goal.Thenwefoundawaytoapproachthesolutioninadifferentwaysothatwecould
stillachievethesprintgoal.It’saformofimpedimentremoval.Isthereanywayyouand
theteamcandosomethingdifferenttofinishoutthesprintandstillcallitasuccess?”
“Different?Nothingobviousspringstomind,”saidBilly.“Imean,wecanlookatthe
sprintgoalandseeiftherearesomestorieswecanadjustandstillachievethesameresult.
We’llneedtogobackandrefactormorethanexpected,butitmightwork.”
Markthoughtforamoment,thensaid,“I’mnotsosure,Billy.Wecaninvestigateit,but
wewererunningprettyleanthissprintevenbeforewelostMike,whichbytheway,Pam,
isanawfullybigimpedimenttoremedy.Plus,theresearchisgoingtotaketimeandmight
notresultinasolution.Meanwhile,we’llbefallingfurtherbehind.Ourbestoptionwould
betoaskforhelp.BythatImean,maybewecanhavesomeonecomeinandhelpusdo
theworksowekeepoursprintgoal.Pam,doyouhaveanyideasonwhomightbe
available?”
“IwishIdid,guys.Thiswholethingcaughtalotofusbysurprise.WhenItalkedtothe
GMlastweek,theplanwastotransitionslowly,butthiscamedownfromthetop,justlast
night,”Pamexplained.
“HowaboutJavier?Canwegethim?OrAmanda?Theyhavetheexperienceweneed
andwouldfitwellwithourteam,”askedBilly.
“Letmethinkaboutit.Rightnow,Ican’tthinkofasingleteamthatisn’tintheexact
samesituationasyou.So,let’sassumefornowthatIcan’tgetyouanyhelpandyoucan’t
findadifferentwaytoaccomplishyoursprintgoal.What’soptionC?”askedPam.
“Well,”saidMark,“wecanreducethescope,meaningwecutoutafewstoriesfromthe
sprintbacklogandputthembackontheproductbacklog.Thiswillreallyirritateour
stakeholders,though—onegroupofthemespecially.Wetoldthemwe’dhavealotof
functionalitybuiltoutfortheminthissprint,andwetoldthemthatweeksandmonths
ago.Wecan’tgobackonthatnow;they’llkillus!”
“Iagree.Lookingatoursprintbacklog,findingwaystocutscopeisn’tthateasy,”said
Billy.“Thecutswe’dneedtomakearesodeepthatitmightevenimpactsomeofthe
storieswehaveworkedonalready.Iwouldhatetohavetoremovenearlycompleted
storiesfromthesprintdemojustbecausetheyaredependentonotherstoriesthatwehad
tocut.”
“Then,itlookslikeweneedtocancelthesprint,”declaredPam.
“Cancel?”askedBillyandMarktogether.
“Yes,cancel—orabort.Cancellingasprintisrare,butitappliesinthiscase.Youcancel
whenthereisachangetothebusinessthateithernegatesthevalueofthesprintgoalor
leavestheteamincapableofmeetingthesprintgoal.Ithinkwe’reatthatpoint.It’sthe
onlyoptionthatseemstomakesense.”
“Thewholeteamchemistryischanged,”musedBilly.“It’snotjustaboutthetasksMike
wasgoingtodo.It’sthatweneedtorethinkwhowearewithouthim.Iwanttomakeit
veryvisibletothehigher-upshowamovelikethisaffectstheteamandtheproject.”
“Yep,”saidMark.“LikeBillysays,it’snotasmuchabouttheworkMikewasdoingor
Mikeperse.It’sthatwhenyoudisruptateammid-project,there’sacost.We’vegotto
learnhowtoworktogethereffectivelywithouthim.Plus,we’vegotsomelogisticsto
workout.We’llhavetodoacompletereset:Rollbackourcodebasetothelastsprintand
startplanningagain.Andwe’vegottorescheduleallthosemeetings,too.”
“You’reright,”saidBilly,“I’dforgottenaboutallthat.Ihopeit’sworthittothe
company,thelossthateveryoneisgoingtotakeonthis.It’sgoingtotakeusacoupleof
monthstogetbacktowherewewereintermsofvelocity.”
“Agreed.Let’shopethisshort-termpainhasalong-termpayoff,”saidPam.“We’vegot
alotofworktodointhemeantime.Let’sgotalktotheproductownerandmakeour
case.”
TheModel
Cancellingasprintisabigdeal,bydesign.Itisdoneonlyinextraordinarycircumstances
andalwaysasalastresort.Inthischapter’sstory,Billy,Mark,andPamconsideredfour
optionsthatJeffSutherlandaffectionatelycallstheScrumEmergencyProcedures.
Sutherlandhastoldmemanytimesthatwhenhewasafighterpilotflyingsortiesover
Vietnam,hehadstrappedtohislegasetofemergencyproceduresonwhattodoincaseof
anemergencywhileonamission.Scrumteamshaveasimilarchecklistavailabletothem
[LEFFINGWELL].
Whensomethingunexpectedoccursthataffectstheteam’sabilitytorealizethesprint
goal,theteamhasfouroptions:
1.Removeimpediments.
2.Gethelp.
3.Reducescope.
4.Cancelthesprint/abort.
Teamsshouldfirsttryeitherofthefirsttwooptions.Reducingscopeshouldbealast-
ditcheffortthatistrumpedonlybyoutrightsprintcancellation.Theonethingyouneedto
rememberinallthisis,don’tmakeadecisionontheseoptionsblindly.Thisisa
negotiation,soeachstepisworsethanthepreviousonebydesign.Talkthisthrough,work
diligentlytofindoptions,andrememberthatwhentheteamisindirestraits,itsmembers
mightnotbethinkingclearly.Astheteammemberstrytoworkthroughtheoptions,
encouragethemtocommunicate—witheachother,withtheirproductowner,with
management.Thesearenotdecisionstheteamorproductowner(oranyoneelseforthat
matter)shouldmakeinavacuum.Discusstherisksandimpactsanddoeverythinginyour
powertofindasolution.Abortingasprintisthelastresort,usedonlyforaship-is-
sinking-typeevent.Cancellingshouldalwaysbeyourlastchoice.
RemoveImpediments
Scrumhassomethingcalledthesprintgoal.Thesprintgoalcanbesomethingas
overarchingasimproveperformanceoraddadditionalpaymentfunctionalityoras
focusedasallowcustomerstouseanAmericanExpresscard.Whenconsideringwhat
impedimentscouldberemovedtoallowyoutoreachthesprintgoal,focusonanalyzing
theissuesandgatheringideasonhowtocorrecttheproblems.Theblockingissuecouldbe
somethingsimple,likeabrokenbuildmachine,orsomethingmorecomplex,likea
distributedteamimbalance.Regardless,workasateamtoidentifyandremoveor
compensatefortheissueswheneverpossible.
GetHelp
Thisoptionisprettystraightforward.Ifyourteamislookingatitssprintbacklogandfinds
thatitcannotmeetthegoal,andnoclearworkaroundsexist,askforhelp.Thoughthis
worksinapinch,itdoeshavedrawbacks:
Thepersonwon’tbepartoftheteam.Includinganewbieontheprojectforone
sprintcancomewithcomplications.
Yourvelocityforthatsprintmightbeartificiallyinflatedbecausethestorywas
completedbyanoutsiderorateamconsultantinsteadofacoreteammember.
Thebestwaytoovercometheseobstaclesisthroughcommunication.Letyourproduct
ownerknowupfrontwhatisgoingon.Remindtheproductownerthatthevelocitymaybe
inflatedforthatsprintandnottocountonthatlevelofvelocityinfuturesprints.Helpthe
non-teammembergetuptospeedquickly,andkeepyourteaminformedonwhatisbeing
builtandhowitworksbypairingorfindingotherwaystoshareknowledge.
ReduceScope
Teamsoftenmistakenlyreachforthisoptionfirstbecauseitistheeasiest.Here’sthe
thing.Whilesomewouldsaythatthesprintbacklogismoreofaforecastthana
commitment,Ifirmlybelievethatwhentheteamsaysitcandeliveracertainamountof
functionalityinasprint,thatteamhasmadeacommitmenttotheproductownerand
stakeholdergroup,whomightwellneedtoreleasethatfunctionalityattheendofthe
sprint.
Knowthatwhenyouremovescopethatyouhavepromisedtodeliver,stakeholdersare
goingtobemad!Idon’tmeanjustkindofmad,Imeanupset.Tohelpmitigatethis
situation,theproductownerneedstoworkwiththecustomersduringthesprint,telling
themaboutthereductioninscopeandthereasonsthatitwasabsolutelynecessary.And
thosereasonsneedtobecompelling.Acompanyreorgdisruptingasprintis
understandable;someoneoutsickforseveraldays,though,pointstoabiggerissuearound
teamdynamicsandsharingtheworkload.(SeeChapter17,“FacilitatingaProductive
DailyScrum”formoreonworkingasateam.)
Anytimeyouremovescopefromthesprintbacklogandfailtofollowthroughona
commitment,itunderminestrust.It’sdifficultenoughtobuildtrustamongtheteam,the
productowner,andthestakeholder/customergroup.Anydamage,nomatterhowslight,
candegradetrustsignificantly.Ifithappensrepeatedly,allinvolvedwillloseconfidence
intheteam’sabilitytodeliver.
Burndownsmakescopecutsveryvisible.AsFigure22-1shows,whenyoudropscope,
theburndownchartshowsadropintheworkremaining,intermsoftaskhours.
FIGURE22-1Sprintburndownshowingdropinscopebetweendays6and7
CanceltheSprint
Whenoptionsonethroughthreefailoriftheyarenotfeasible,whenyou’vehitrock
bottomandneedtopresstheresetbutton,youcancancel/abortthesprint.Infighterpilot
terms,thiswouldbethetimeforejectingfromtheaircraft,leavingittocrashandburn
whileyoufloattotheground.Assuch,thisisnotsomethingtotakelightly.TheScrum
Guide[SCHWABER]saysthatonlytheproductownermaycancelasprint;however,I
findthatitisalwaysacollectiveteamdecisioninvolvingtheproductowner,theteam,and
stakeholders.
Whenasprintisterminated,therearecosts.Allworkisstopped;someofitmightbe
lost.Allmeetingsarecancelled.Allofthesprintmeetingsmustberescheduled,including
anewsprintplanningmeeting,whichmustbeheldimmediately.Everyoneaffectedis
notifiedandtheteambeginsanew.
TheruleIlivebyisthattheteamresetsthecodebaseaswell,rollingitscodebackto
thelastsprint(thelastknowngoodsprint),andrestartsfromthere.Why?Becausewhile
theteammightwellusesomecomponentsthatithadwrittenbeforethesprintwas
cancelled,theteamdoesn’tknowthisatthetimeofthecancellation.Thebestoption,
therefore,istoreturntothelastknowngoodstateandbeginagain,evenifthecodeis
solidandcouldtheoreticallyremaininthebuild.Withoutproductowneracceptanceand
beingtiedtoasprintgoal,it’sjustextragarbageinthecodeandshouldbedeleted.Afresh
startacrosstheboardispartofthecostoftermination.
Sprintscanbecancelledbytheteamorbymanagementwhen,becauseofextraordinary
circumstances,theteamcannotorshouldnotachievethesprintgoal.Inhighlydynamic
industries,externalbusinessinfluencesoftenleadtodramaticallychangingscope.(Yes,in
theseinstances,theteamshouldprobablyrunshortsprintsofaweek,butsomedon’t.)
Otherreasonstoterminatethesprintmightincludecontractchangeswithathirdparty,
runningoutofmoney,andclient-requestedscopechanges.
Attheendoftheday,terminatingthesprintreducesteammoraleandvelocity.Itbreaks
thestakeholder/teamcadenceandchangesexpecteddeliverypointsandforecastsaswell.
Extracommunicationisnecessarytoexplainwhythesprintwasterminated.Theteam
restartswithnewplanningmeetingsandbuildsanewsprintbacklogwithanewgoaland
commitment.Theteammightneedtoaddressitsestimatedvelocityaswell.
InmyyearsofworkingwithScrumprojects,Ihaveonlyhadtoterminateonesprintin
oneproject.Why?Ateammemberwassuddenlymovedtoanewgroupafterareorg.It
wasneitherpleasantnorfun,andourstakeholderswereupset,butintheend,itwasthe
bestthingtodofortheproject.
KeystoSuccess
It’shardtotalkaboutsuccessduringadiscussionofemergencymeasures.Anypathyou
takewillbepainful.However,ifyoufollowsomegeneralrules,youcanminimizethe
pain.
Communicate—Icannotstressthisenough.ThisiswherethevaluesofScrum
comeintoplay.Ittakescouragetotakethestepsneededtoworkthroughthisrough
patch,butcommunicationneedstohappen—andhappenoften.
Staycalm—Whenthingsgosouth,it’shardnottopanic.Rememberthatasa
ScrumMaster,productowner,orteammember,youneedtomanagethereactionsof
yourcustomersandstakeholders.Takethehighroadandusethisopportunityto
highlightwhyScrumisgood:Itoffersoptionsandflexibility,anditsshortcycles
allowyoutogetbackontrackfairlyquickly,evenafteracatastrophiceventlikea
termination.
Stayfocused—Lookforward,notbackward.Focusoncleaningupthesprintand
gettingonwithgettingon.Thewayyoureactwilldetermineothers’reactions.Ifyou
refusetolingerinthepastbutinsteadsoldieronwithapositiveoutlook,otherswill
followyourlead.
References
[LEFFINGWELL]Leffingwell,Dean.“JeffSutherland’sSprintEmergencyLanding
Procedure.”ScalingSoftwareAgilityBlogwebsite.
http://scalingsoftwareagility.wordpress.com/2008/10/19/jeff-sutherland%E2%80%99s-
sprint-emergency-landing-procedure/(accessed2January2011).
[SCHWABER]Schwaber,Ken,andJeffSutherland.2013.“TheScrumGuide.”
http://www.scrumguides.org(accessed28October2015).Alsoavailableat
http://www.mitchlacey.com/resources/official-scrum-guide-current-and-past-versions.
PartIV.AdvancedSurvivalTechniques
Chapter23.SustainablePace
Ourteamworksatasustainablepace—Youhearitallthetime,butwhatexactlydoesit
mean?Isthereastandardpacethatallteamsshouldstrivetoachieve?Forexample,my
wifeandIarebothrunners.Hersustainablepaceisaseven-minutemile;mine,though,is
morelikenineandahalfminutes.Wecaneachsustainourownpaceformanymiles.But
ifItrytosustainherpace,ifItrytorunaseven-minutemile,forverylong,Igetsick—
physicallyill.ShouldIbetrainingtoreachherpace?Orisminegoodenough?
Whatdoessustainablepacemeanforyourcompany?Whatifateam’ssustainablepace
istooslowtoreachthecompany’sgoals?Shouldtheteambeaskedtodeliver,evenifit
meanspushingitsmemberstotheirlimits,orbeyond?
Tobetterunderstandsustainablepaceanditsimplications,readthefollowingstory,
whichfeaturestwofunctionalprojectmanagersinsidealargesoftwarecompany.Thefirst,
Hugo,isworkingasaScrumMasteronaScrumproject.Thesecond,Joana,isfunctioning
asaprogrammanageronatraditionallyrunproject.
TheStory
“Idon’tknowhowmuchlongerIcantakethis!”exclaimedJoana.Shestoodinthedoor
toHugo’sofficewithanexpressiononherfacethathewasalltoofamiliarwith.He
gesturedtoanearbychair.
“Youmightaswellgetitoffyourchest,”saidHugo.“Moreproblemswithyour
project?”
“Theyneverend,”Joanaexplained.“Thistimeit’sthestakeholders!Wheneverwetry
toclarifyarequirementorgettheacceptancecriteria,wegettherunaround.Butstill,they
expecteverythingtobecompleteintimefortheirbigreleaseandattheendofevery
milestone.Theyvacillate,changetheirmindstwentytimes,andrefusetobepinneddown.
Whentheyfinallydofigureoutwhattheywant,we’reexpectedtomakeithappenin
recordtime!‘Cramitinattheend,’theysay.”
“Sowhat’snew,Joana?”askedHugo,smiling.“Seriously.Iwanttohelpbutdon’t
knowhow.Everytimeamilestonecomesup,youcomeinheretellingmehowridiculous
itistorunaprojectthiswayandthatyou’renevergoingtoletthishappenagain.Every
time,Iagree.Butassoonasthenextmilestonerollsaround,you’rebackinhere,inthe
samesituation,facingthesameproblems.Itcan’tcontinue,butitwillcontinueuntilyou
dosomethingdifferent.”
“Icanalwayscountonyoutotellitlikeitis,can’tI?”saidJoana,smiling.“Andyou’re
right.Thisdoeskeephappeningoverandover.Wejustgetrestedfromourlastwildpush
whenit’stimetoworklikedogsagain!Andit’sgettingworse.We’realreadyworking
overtimeeverynight,andwe’resixweeksawayfromtherelease.Lasttime,asbadasit
was,weonlyhadtoburnthemidnightoilforaboutthreeweeks.Andit’snotlikewe’re
puttingoutaproductwecanbeproudofeither.Theshortcuts,thebugs,andtheproblems
arejustpilingup,butwhohastimetoaddressthem?”Joanasighedandleanedbackinher
chair.Shelookeddownforamomentthencontinued.
“Theteamislosingfaithinmeandeachother.Halfofthemarereadytoquit,again.I
reallydon’tknowwhattodo.”
Hugogotupandsketchedanunsustainablepaceonthewhiteboard.
“Partoftheproblemisthestakeholders.Butlet’stalkaboutwhat’sreallydraggingyou
down.You’renotworkingatasteadypace,”saidHugo,pausingtopointatthedrawing,
whichshowedalargevarianceintheteam’sactivity.
“See?Yousitidleforawhile,waitingforadecisionorrechargingyourbatteriesfrom
yourlastbigpush.Thenyouworkataprettygoodpaceforsomeamountoftime.Then
suddenly,yourampwayupandtrytoholditatthatlevel.Right?”askedHugo.
“That’sexactlyit.Werampuplikewearenow,workingsixtyhoursaweek.Then,once
wehitthemilestone,wecrash.Sure,wemightbeintheofficefortyhoursaweek,but
we’rereallyputtinginsomethingliketwentytothirtyhoursaweek.Afterthisnext
milestone,we’llprobablyneedfoureasyweekstorecoverfromtheburnout.It’snotideal,
butIdon’tknowwhatelsewecando.Wehavetorampupbut,afterwedo,weneeda
break,”admittedJoana.
“Theunevenpaceiskillingyou,though,Joana.Partofthereasonyouhavetoworkan
increasingamountofovertimeisbecauseyou’rerecoveringforlongerperiodsoftime
betweenthespikes.Thevalleyswillbecomewider.It’saviciouscycle.Youneedtowork
atasteady,sustainablepacethewholetime.”
“Itcan’tbedone.Evenifwewantedtoworkmoreduringthosevalleys,we’reoften
blockedbythestakeholders,whocan’tmakeadecisiontosavetheirlives,ortheir
projectsinthiscase,”quippedJoanna.
Hugosmiled.“Itcanbedone,butit’sgoingtorequiretwothings.Thefirstisa
commitmentfromtheteamtorefusetotakeonrequirementsunlessthey’rereadytobe
workedon.Keepthemoffthescheduleuntilthedecisionsaremade—andthisincludes
acceptancecriteria,architecture,everything.Beforeyousayanything,Iknowhowtough
that’sgoingtobe.We’regoingtohavetohelpthestakeholders,andthebusiness,
understandwhat’sgoingonandhowtheirinefficienciescontributetotheseproblems.
They’vebeenlulledintothisfalsesenseofsecurity—nomatterwhattheydo,the
functionalitytheyrequestcomesoutwhentheywantit.Whatthey’regoingtohaveto
learnisthatiftheyhaven’tmadedecisionsaboutfunctionality,theycan’thavethe
functionalitybytherequesteddate.YouandIaregoingtotakethatontogether,”Hugo
reassuredJoanaasshelookedathim,bewildered.
“I’malmostafraidtoaskwhatthesecondthingis,”saidJoana.
“Thesecondthingwe’regoingtohavetodoisteachyourteamhowtoavoidthewall,”
saidHugo.
“Comeagain?”
“RememberyearsagowhenIranmarathons?”askedHugo.
Joananodded.
“Well,oneconceptIdidn’tunderstanduntilIexperienceditmyselfwasthewall.I
thoughtIwasreadyformyfirstmarathon.Ihaddonethetraining,boughttheshoes,had
thevenueallpickedout.ButIdidn’trealizehowimportantitwastofocusonmypace.”
“Goon,”saidJoana.
“MymistakewasthatIhadatimeinmind,agoal.AndasInearedthehalfwaypointof
themarathon,IrealizedIwasn’tgoingtohitthatgoalunlessIwentfaster.Ifeltpretty
goodatthatpoint(plentyofbreath,muscleswereloose),soIpickedupthepace.Eight
mileslater,IknewIwasintrouble.Ibegantofeelsick.Itriedtoeat,triedtodrink,butit
onlymademefeelworse.IknewatthispointIwouldn’tmeetmytimegoal,butIwas
determinedtofinish.Istoppedacoupletimes,caughtmybreath,andtriedtogetmyheart
ratedown.Therestbreaksworked,butonceIstartedrunningagain,Ifeltawfulwithin
minutes.Still,Ifoughtthroughit.
“Then,aboutfivehundredmetersfromthefinishline,ithappened.Ihitthewall.Iwas
disoriented,hallucinating.Ithrewupandcollapsed.AndIcouldn’tgetbackup.Iwas
seeingstars—Ijustcouldnotfunction.ThereIwasinfrontofallthosepeople,withthe
finishlineinsight.AndtherewasnothingIcoulddobutliethere.Ineverfinished.Istill
havethenewspaperclippingfromthatday,andIstillgetahardtimefromtheguysatthe
gymeveryyearaboutit,”saidHugo.
“So,”saidJoana,“ifthisprojectisamarathon,you’retellingmewe’renotgoingto
finish,andwe’regoingtobeembarrassedtrying?”
“Worse,”Hugoexplained.“Yousaidearlierthatthestakeholdersarewaitinguntillate
intheprojecttomakecrucialdecisions.Right?”
“Yes,”saidJoana.
“Whatstateofmindareyouandtheteaminwhenyou’reaskedtomakeyourown
decisionsbasedontheirlast-minuterequirements,decisionsaboutdesign,architecture,
andsoon?”askedHugo.
“We’reprettymuchtootiredtoprocessit.Atthatpoint,we’rejusttryingtoslog
throughandgetitdone,”saidJoana,frowning.“Ihadn’tthoughtaboutthat.”
“Bytryingtohitsomearbitrarymilestoneandincreasingyourpacetosomethingyou
can’tsustain,you’regoingtohitthewallbeforethefinishline,everytime,”saidHugo.
“Unlikeme,youmightfinish,buttheendresultwillbebuggycode,unhappypeople,
brokenbuilds,allconsequencesofbaddecisions.Soundfamiliar?”
“Alltoofamiliar,andIhavefivemoreweeksofthis,thenfivemoremilestonesafter
that.We’llhaveatleastthirtyorfortymoreweeksofsheertorturebeforewe’redone,”
Joanasaid,puttingherheadinherhands.“Thisisimpossible.”
“You’rerightaboutthat.Rightnowitisimpossible,”admittedHugo.“Butthat’swhere
thesecondthingcomesin.Remember?Thereweretwothingsweneededtodo.Thefirst
wastoputthepainofindecisionbackonthestakeholderswhereitbelongs.Thesecondis
tofindapacethatyoucansustain,livewiththatforawhile,andthentraintogofaster.”
“Notfollowing,”saidJoana,frowning.
“Myfirstmarathonendedbadly.ButIrananotheronesixmonthslaterandfinished.I
didn’tgetthetimeIwanted,butIranasteadyten-minutemilethroughouttheraceand,
becauseIdidn’tgobeyondmylimits,IfeltokaywhenIfinished.Stilltiredbutincontrol,
happywithmyperformance,andreadytoraceagainanotherday,”Hugoexplained.
“Soweneedtofigureoutwhatagoodpaceisfortheteam,andstickwithitthroughout
theproject?”askedJoana.
“Yes,butthere’smore.Knowwhatmytimewasonmylastmarathon?”askedHugo.
Joanashookherhead.
“Threehoursflat,”saidHugo.“That’sapaceofsevenminutespermile,threeminutes
fasterthanmysecondmarathonandtwominutesfasterthanthegoalIoriginallysetfor
myself.Throughtraining,practice,anddiscipline,Iincreasedmysustainablepaceand
wentfasterthanI’deverdreamedpossible.
“Joana,ultimately,it’syourjobtodeliver,whetherthepaceissustainableornot.And,
aswasthecasewithmyfirstandsecondmarathons,thepaceyoucansustainnowis
probablynotgoingtobefastenoughtomeetthecompany’sgoals.WhenIrealizedthisa
fewyearsback,Idecidedthattoincreaseoursustainablepace,IneededtouseScrumand
XPengineeringpracticeswithmyteams.Wehaveshorteriterations,usecontinuous
integration,test-drivendevelopment,pairprogramming,andotheragilepracticesto
improveourperformance.Ithinkthat’stheanswerforyou,also—helpyourteams
understandagilepracticesandhowandwhytheywork.You’llbebetterabletogauge
progress,respondtoproblems,andmaintainenergylevels.Plus,yourstakeholderswillbe
betteralignedwithyou,soyou’llhavefewerfirestodealwithasyounearamilestone.”
Hugopausedwhileheletthatsinkin.“So,areyoureadytotaketheleap?Youin?”
Joanasmiled,“Anythingisbetterthanthewaywe’reworkingnow.Showmetheway,
ohmightymarathoner.”
Togethertheysetaboutthehardworkofchangingmindsandhabits.Thefirstthing
theydidwastoappealtomanagementforpermissiontostoptheprojectandbegananew.
TheystartedusingScrum.Theyeducatedthestakeholdersabouttheirroleandthe
expectationsthatweresetforthemtomaketheprojectgoontime.Theyworkedwiththe
teamtoslowlyaddengineeringpracticesthatwouldmakethemamoreeffectiveunit.In
theend,Joana’steamfoundacomfortablesustainablepacethatmorethansatisfiedthe
company’sneeds.
TheModel
Thisstoryisnotuncommon.Alltoooften,thenormforteams,especiallythoseina
traditionaldevelopmentenvironment,istheburnoutscenariothatHugodescribedand
Joanaconfirmed.Withtheadventofmoremoderndevelopmentmethodologies,I’veseen
thenumbersdecreaseoverthepasttenyears,butevennow,atleast50percentofthe
customersIworkwitharestillworkingbeyondcapacityforextendedperiodsoftime.
Thesecompaniesaren’theartless;likeJoana,theywantemployeestohaveagoodwork
environment,butdrivinghardtomeetmilestonesistheonlywaythesecompaniesknow
howtomeettheirgoals.Intheend,though,theyareleftwithhighemployeeturnover,
buggyproducts,andfewercustomers.
Theproblemdoesn’tmagicallygoawaywhenteamsturntoScrum.Whenteamsare
newtoScrum,theyapplythesamepatternsofburnoutandrecoveryinsideasprint.Ina
typicalone-monthsprint,theteammightworkbelowcapacityforaweek,atcapacityfora
weekortwo,andthenbeyondcapacityduringthefinalweek,oncetheyrealizetheyarein
dangerofmissingthesprintgoal.Tomitigatethis,see“ShortenIterations”laterinthis
chapter.
Inthestory,Hugo’ssketchhighlightedthepeaksandvalleysoccurringonJoana’s
project.Valleysareperiodswhentheteamisinrecoverymode(notproductive),andpeaks
aretimeswhentheteamisincrunchmode(overloaded).Joanawasrunningatraditional
softwaredevelopmentprocess,wheretheteammusthitcertainmilestones(codecomplete,
testing,betarelease,andsoon).That’swhyinHugo’ssketch,theverticallinesindicate
theendofvariousmilestones.Ifhehadbeensketchinganagileproject,theverticallines
wouldinsteadcorrespondwiththeendofsprintsorreleasedates.
NoticethatinHugo’ssketchtheteamstartsoutinanunproductivestate,spendssome
timeoperatingatoptimalefficiency,andthenskyrocketsintooverloadwhenitbecomes
obviousthattheteamcannototherwiseachieveagoal.Oncetheteamhitsthemilestone,it
fallsintoanunproductivestateagain,whereitsmemberscatchtheirbreathbefore
rampingupagain.Thishappensforafewcycleswithoutcausingtheteammuchpain
becausetheteamisspendingabouttwiceasmuchtimeinanoptimalstateasitisinthe
unproductiveandoverloadedstates.YoucanseethisinFigure23-1,whichrepresentsthe
typicaldistributionoftimeduringtheearlycyclesoftraditionaldevelopment.
FIGURE23-1Howtimeisdistributedleadinguptoamilestoneintraditional
development
Asthemilestonesprogress,however,thedistributionbeginstoshift.Bugreportsand
changerequestspileup,causingtheteam’sworkloadtoincreaseasitsmembersaretrying
tokeepthesamerateofdevelopmentwhilealsofixingproblemsandintegratingchanges.
Beforelong,theteamisspendingmoretimeinanoverloadedstate,whichmeanstheteam
needsmoretimetorecoverandspendslesstimeinoptimalmode.
Soontheteamiscaughtinaviciouscycle.Themoretimetheteamspendsworking
beyondcapacity,themoretimeitneedstorecover.Andthelongeritspendsresting,the
moretimetheteamneedstospendinburnout/overloadmodetocatchup.Thisis
especiallytrueoncyclesofamonthorlonger.
Itrackedthisonmyownprojects.Myteamfoundthatasworkprogressed,ourdata
lookedmorelikethepercentagesshowninFigure23-2.
FIGURE23-2Overloadincreases,optimaldecreases
Sincethatinitialtracking,I’veobservedthatthistrendisfairlystandardacrossthe
boardandthatthelongerthecycle,themorepronouncedtheimbalance.Tounderstand
whyasteadilygrowingimbalanceissoharmfultoaprojectandtheendproduct,take
anotherlookatHugoandhismarathons.Duringhisfirstmarathon,Hugohitawall—he
reachedaphysicalstateofexhaustionwherehecouldnotcontinue.Long-distancerunners
knowthattheymustmonitortheirheartratessothattheystayinanaerobicstate,where
thebodyiscreatingenergybybreakingglycogenintoglucoseandcombiningitwith
oxygen.Whenthebodyrunsoutofglycogen,itbeginsburningfatandproducinglactic
acidinsidethemuscles,causingtherunnertohitthewall.Tostayinanaerobicstate,the
runnerkeepshisheartrateinacertainrangeandmustexpendenergyataconstantrate
ratherthaninbursts.
Iftheheartraterisesaboveacertaintarget,85percentofyourtargetheartrate,the
bodybeginstoperformanaerobically,burningglucosewithoutoxygenandoperatingfar
lessefficiently.Thisisfineforexercisethatrequiresonlybriefburstsofintenseexertion,
butitisproblematicforlong-termendeavorswhenasustainedlevelofenergyisrequired.
Asoftwareteamshouldstriveforanaerobicstate,workingatasteadypacesothatits
membershaveanalmostendlesssupplyofenergy.Whenforcedtoworkinburstsof
hyperactivity,thereisamuchmorefinitesupplyofenergyavailable.Asaresult,theteam
memberswillquicklybecomefatigued,beginmakingbaddecisions,andastheir
metaphoricalmusclesbegintoburnfromlacticacidbuildup,theywilleventuallystopout
ofsheerexhaustion.Astheseburstsgrowlonger,sodothecorrespondingperiodsofrest,
astheteammembersfeelthemetaphoricalburnoflacticacidintheirownmindsand
bodies.
Maintainingasteady,aerobicstaterequiresthatteamsworkinmultiplesetsofmicro
peaksandvalleystoreachthesprintendormilestonemarker.Figure23-3illustratesthe
peaksandvalleysindicativeofasustainablepace.Whenteamsareinthisstate,theyare
operatingatoptimalefficiency,evenatthepeaksandvalleys.Becausetheteammembers
aren’ttiredorstressedorburnedout,they’reabletobeattheirmostproductive.
FIGURE23-3Shorteriterationsallowforamoresustainablepace
So,howdoyoumovefromahecticpacethatleavesemployeesburnedoutand
overworkedtoasteady,sustainedpacethatisoptimalandproductive?Considerthestory.
WhenHugoranhisfirstmarathon,hetriedtorunbeyondasustainablepace,andinso
doing,burnedoutandcouldn’tfinish.Bythetimeheranhislastmarathon,however,he
notonlyfinishedbutmanagedaseven-minutemile.Thatkindofimprovementwasn’ta
fluke.Ittookpractice,training,discipline,andtime.Likewise,ifteamswanttocreatea
sustainablepacethatisrobustenoughtomeettheirowngoalsandthoseoftheir
stakeholders,theyhavetoimprovethewaytheyrunprojects.Todothis,teamsneedto
shortentheiriterations,increasetheamountoftimetheyspendworkinginateamspace,
andmonitortheirburndownstoensureasteadycompletionratethroughouteachsprint.
ShortenIterations
Thefirstthingtodoifyouareexperiencingburnoutistoshortenthecycletime,breaking
theprojectintosmallerchunksanddeliveringfunctionalitymoreoften.Ifyouareworking
inone-monthsprints,trytwo-weeksprints.Ifyouareusingtwo-weeksprints,tryone-
weeksprints.Byshorteningthesprints,you’llfindit’smucheasiertoconsumeenergyata
constantrateoverashorterperiodoftime.
Toexaminewhythat’sso,revisitthesustainablepacechartshowninFigure23-3.
Noticethatbeforeeachmilestone(orreleasedateonaScrumproject),therearemultiple
peaksandvalleysinactivity,allfallingprimarilyinsidetheoptimalzone.Theseshort
iterationsallowforsmallburstsofenergy,shortrecoveries,andfrequentprogresschecks
andadjustmentsintheformofsprintreviewsandretrospectives.Psychologically,because
theteamcanseethefinishlinefromthebeginningofeachsprint,it’smucheasiertowork
enthusiasticallytowardtheendgoal.Shortiterationsaretheheartbeatofsustainablepace.
Workinginshortersprintsmightmeandecomposingyourstoriesintosmallerbitsand
changinglong-standinghabits,butithasbenefitsthatreachfarbeyondsustainablepace,
includingmorefrequentcustomerfeedback.SeeChapter12,“DecomposingStoriesand
Tasks,”fortechniquesandbenefitsassociatedwithtaskdecomposition.
MonitorBurndownCharts
WhileTheScrumGuidepublishedbySchwaberandSutherlandin2013statesthat
burndownsarenolongerarequiredartifactinScrum[SCHWABER],Ibelievetheyare
stillessentialtoolsinateam’sarsenal.Yourburndownchartcangiveyoumanycluesasto
howyourprojectisprogressing.Whatyouwantisasteadyrateofburndownacrossthe
board.ItshouldlookconsistentwithFigure23-4.
FIGURE23-4Exampleofanidealburndownchart
However,forteamsthatruninanoverloadedstate,theburndowntypicallylooksmuch
different.Fortheseteams,whatyouseeismanydayswhentheworkremainingishigh
(duringthelightactivityassociatedwithrecoverymode)followedbyalargeflurryof
activityandaccompanyingdropinworkremaining(inoverloadmode),asillustratedin
Figure23-5.
FIGURE23-5Exampleofasuboptimalburndownchart,wheretheteamcompletesmost
ofitsworkallatonceratherthanatasteadyratethroughoutthesprint
Highburndownratesattheendofthesprintandslowburndownratesatthebeginning
areindicatorsthatyourteamisattemptingtoomuchworkattheendofthesprint.
Fluctuationsinburndownratesmightoccurbecauseoneteammemberistakingontoo
muchwork,thetasksaresizedinappropriately,theteamisinthehabitofworking
sporadicallythroughoutthesprint,oritcouldbeacombinationoffactors.Spendsome
timeevaluatingwhatiscausingyourend-of-sprintbursts,andtakethestepsnecessaryto
correctthem.
IncreaseTeamTime
Chapter10,“CoreHours,”explainswhycoreteamhoursareimportantandhowto
establishthem.Havingagoodsetofcoreteamhoursisalsocriticaltomaintaining
sustainablepace.Teammembersneedtospendacertainamountoftimeinthesamespace
together.Withincreaseddistance,evendownahall,comesdistraction,andwith
distractioncomesdecreasedteamcommunication—alackoftimely,sharedinformation
contributestothehugespikestowardtheendofasprint.It’sfartooeasytozoneoutorbe
distractedbyothertaskswhenyouareworkingalone—it’sfarbettertohavefourhoursof
teamtimedailythantwo,andifyoucan,sixisideal.
Chapter9,“WhyEngineeringPracticesAreImportantinScrum,”detailshowimportant
engineeringpractices,suchaspairprogramming,test-drivendevelopment(TDD),and
continuousintegration,aretoteameffectiveness.Theyarealsocornerstonesofsustainable
pace.Habitualuseofthesepracticeshelpskeeptheteamfocusedonthestoriesthatneed
tobedelivered.
Ifyouworkinadistributedteam,supplementphysicalspacewithvirtual.Thiscan
includehavingcontinualwebcamsandanopenconferencephoneline.Ondistributed
teams,it’sessentialtohavegoodengineeringpracticesandpairprogrammingsessionsto
helppreventthelossofinformationamongteammembers.
KeystoSuccess
Inthedriveforefficiency,organizationstendtoforgetwhatitmeanstobeeffective.Tom
DeMarcosummarizesitbest,“You’reefficientwhenyoudosomethingwithminimum
waste.Andyou’reeffectivewhenyou’redoingtherightsomething”[DEMARCO,p.
122].Inthesamebook,DeMarcogoesontosaythatwhileeffectiveorganizationswork
towardtheirgoals,albeitsometimesslowly,efficientorganizationscanbecomesofocused
onefficiencythattheyfailtonoticethattheyaremovinginthewrongdirection.
Inthischapter’sstory,Joana’steamwastryingtobeefficientbutwasdoingsoatthe
expenseofeffectiveness.Recognizingthatoverloadingtheteamonlyleavesyourteam
andyourcustomersunhappyisthefirststeptowardadoptingastrategythatleadstolong-
termsuccess.
Asustainablepaceisoneinwhicheachteammemberiscommittedtoworkingator
near85percent,dayinanddayout.Learningtoworkatsuchacontinuouslyhighlevel
takessomegettingusedto,especiallyifyouareusedtotheextremepeaksandvalleys
associatedwithburnout.Besideschangingthewayyouworkinsidethesprint,you’ll
likelyfacesomeorganizationalissuesthatcanderailateam’space,includingteam
longevity,part-timeteammembers,andunrealisticgoals.
Thefirstthingorganizationsneedtorealizeisthattogetteamsworkingatoptimal
levels,managementmustgiveteamstimetogel.Atypicalteamcomestogethertowork
onasingleprojectanddisbandsattheendoftheproject.Managementexpectshighlevels
ofoutputfromthestartbutbythetimetheteamhasreachedaperformingstate,the
projectendsandtheteamdisbands(seeChapter20,“AddingNewTeamMembers,”for
moreinformationonthefourstagesofteamperformance).Thebestteamsareonesthat
havebeentogetherformultipleprojectsandhavehadtimetoformstrongbondsand
workingrelationships.
Thesecondproblemholdingteamsbackisthesplitfocusoftheirteammembers.Tobe
highlyproductive,teammembersmustbefreetofocusonthetaskathand,andonlythe
taskathand.Runningamarathonrequiressingle-mindedness,wheretherunnergivesfull
attentiontohisbody,hisheartrate,andhisenergylevels.Runnersdon’ttrytoruna
marathonanddoothertaskssimultaneously.Yetcompaniesoftenaskteammemberstobe
part-timeonmorethanoneproject,therebysplittingtheirfocusanddegradingtheir
performance.
Next,sustainablepaceisthebestwaytoachieveoptimalproductivitywithoutrisking
teamburnout.Attheendoftheday,however,thereisnooneformulathatIcangiveyou
thatwillguaranteeasustainablepace.Youneedtotalktoyourteam.Findouthoweach
teammemberisfeelingatdifferentpointsinthesprintandoverallintheproject.Listenin
dailystandupsforcluesthatsomeonemightbeoverloadedorsettingherselfupforanend-
of-sprintrushofactivity.Thebestdefenseagainsthittingthewallistokeepaclosewatch
ontheteam’space,toincreasetheamountoftimetheteamspendsworkingtogether,and
toshortenyouriterations.
Onelastkeytosuccesshastodowithcompanygoals.Teamproductivitycanonlybe
increasedtoacertainlevel—eventhefastestrunnerscan’tcompletea26-milerunin26
minutes.Whileultimatelyyou’llbeonthehookforcompletingtheworkyourcompany
requires,ifyoufindthatdoingsorequiresyourteamtoworkatanimpossiblyhighpace,
you’regoingtoneedtohelpmanagementseethedisconnect.Trackyourvelocityand
showthempastreleaseburndownchartsofwhattheteamisabletodoonceitgetstoa
high-performingstate.Finddataandresearchthatshowsthegainsthatothercompanies
haveseen.Inviteagile-mindedseniorleadersinsideoroutsideyourcompanytospeak
aboutsustainablepace,orbringmembersofthemanagementteamtolocalcommunity
eventsorgivethemsomebookstoreadonthetopic—doanythingyoucantohelpwith
theeducationofmanagement.
Buildingsoftwareisateamsport.Alignmentbetweenleadershipandthepeopledoing
theworkhelpsensurethattherearefewersurprises,moresatisfiedcustomers,andmost
important,noteamburnout.
References
[DEMARCO]DeMarco,Tom.2001.Slack.NewYork:BroadwayBooks,p.122.
[SCHWABER]Schwaber,Ken,andJeffSutherland.2013.“TheScrumGuide.”
http://www.scrumguides.org(accessed28October2015).Alsoavailableat
http://www.mitchlacey.com/resources/official-scrum-guide-current-and-past-versions.
Chapter24.DeliveringWorkingSoftware
“Deliveringworkingsoftwareeverysprintisimpossible!”Ihearthatrefrainonaregular
basis,mostoftenfrompeoplewhohaveneverdoneScrum.Iaskthemhowtheyknow,
andtheanswerisalwaysthesame:“Wecan’tdeliverinayear,sowecertainlycan’t
deliverin30days!”
Atonepoint,Ibelievedthisaswell.Nowadays,Iencourageteamstousethatfeelingof
impossibilityasamotivatorforthepossible,tousethatenergytoprovethatitcanbe
done.Moreoftenthannot,teamsfindthatoncetheyadjusttheirmindset,theycandeliver
workingsoftwareeverysprint.Thefollowingstorylooksatoneteamthatfoundawayto
worktowardagoalofpotentiallyshippablesoftwareonasystemthatdidnotatfirstseem
tolenditselftoaniterativeapproach.
TheStory
Paulandhisteamwerebuildingasalesorderprocessingandmonitoringtoolforavehicle
manufacturer.PaulhadworkedwithScrumandExtremeProgramming(XP)onprevious
projects,butfortherestoftheteam,anagileapproachwasfairlynew.Theteamhadlaid
outthefollowinggoals:
Maintainhighcodingstandards.
Haveaclean,manageablecodebase.
Buildknowledgeontheteam.
Bepotentiallyshippableattheendofeverysprint.
Paulandtheteamhadnotroubledevisingaplanforaddressingthefirstthreegoals.
Whenitcametohowtodividetheworksothattheycouldtrulydeliverworkingsoftware
eachsprint,however,thingsgotmoredifficult.Theyinvitedtheproductowner,Matt,toa
kickoffmeeting,wheretheycouldallstudytheproductbacklogandexistinguserstories
totrytofigureoutwheretobegin.
“Ijustdon’tgetit,”saidMatt.“Basedonthesestoriesandwhatyou’retellingme,we’ll
needseveralsprintsbeforewe’vegotanythingthatIcantellthestakeholdersiseven
remotelyshippable.Areyousurethatpotentiallyshippableispossibleatthebeginningof
aproject?”
Paulnodded.Heknewhowhewantedtoapproachtheproblem,buthealsoknewthatit
mighttakesomeconvincingtobringeveryoneelseonboard.Hethoughtforamoment
abouthowtoexplainandcameupwithasimpleexampletodescribetheprocesswithout
boggingdowninthedetailsofthisparticularproject.
“Letmeansweryourquestionwithanotherquestion,”Paulbegan.“Whatisthepurpose
ofslidepresentationsoftware?”
EveryonelookedateachotherlikePaulwascrazy.
“Comeon,whatisit?”pressedPaul.
“Well,”saidDarrenwithadegreeofhesitation.“Iwouldguessitspurposeistopresent
slidesinameeting.”
“Right!”saidPaul,ashepulledupastandardpresentationprogramandpointedtothe
screen.“Sothepurposeofthisistoallowmetopresentmaterialtoyou,likeI’mdoing
rightnow,eitheronthisscreenoronanotherdisplay.Doyouallagree?”
Thegroupnoddedwithhesitation.
“Whatistheonethingthatthisneedstodoflawlessly,everytime,noquestionsasked?”
saidPaul.
“Displayslides;wejustcoveredthat,”saidSara.
“Notjustdisplayslides.Whatelse?”askedPaul.
“Well,tobeuseful,you’dhavetobeabletoaddcontent,”saidDarrenpointingatthe
blanktemplate.
“Right,”saidPaul.“SoIneedtobeabletocreateslidesanddisplaythem.Wecanall
agreeonthat?”
Theparticipantsnodded.
“DoIneedanimations?”askedPaul
“Sureyoudo,”saidDarren.“Yourslidesaresouglythattheanimationsaretheonly
thingthatkeepsusawake!”
Everyonelaughed.
“Okay,okay.Butseriously.DoIactuallyneedtheabilitytodoanimationsorsounds?”
askedPaul.“Whataboutgraphics?Backgrounds?Multiplefonts?”
“Ifyourprimarygoalistojustdisplaycontent,no,youdon’tneedallthatotherstuff.
Whereareyougoingwiththis?”askedMatt.“Wearenotbuildingpresentationsoftware.”
“No,we’renot.Whatwe’retryingtodoisfigureouthowtoapproachourproject.And
todothat,wefirstneedtothinkabouttheminimumfunctionalityweneedtoreachour
goal.Thesimplestthingthepresentationsoftwareneedstodoisenableuserstocreateand
displayslides.Dotheyneedtemplates,animations,orevenmultiplefonts?No.The
absolutebaserequirementisoneslidethatcanbecreatedanddisplayed.Doyouagree?”
askedPaul.
“Well,yes,butit’snotshippable.Noonewouldbuyit!”saidMatt.
“True,butnooneissuggestingweactuallyshipit.That’swhytheycallitpotentially
shippable,right?”saidSara.“Nooneintheirrightmindwouldeverreleasepresentation
softwarethatdidthebarestofbareminimumsintoamarketcrowdedwithrobustoptions.
Tobemarketable,theproductwouldneedtohavemorefeatures,butitwouldbe
potentiallyshippable.”
“Igetit,”saidDarren.“Wewouldneedtoexpandonit,creatingsomefunctionality
overtimetogettoafullyreleasablepoint.Butitcouldbeshipped,theoretically,because
itdoesmeettheabsolutebaserequirement.”
“Good,”saidPaul.“Soourgoalwouldbetobuildthebasiccoreoffunctionalityfully
andtomakesureitflowsthroughthesystem.We’dshowthefirstpiecetoour
stakeholdersandgettheirfeedback.Thenwe’dmakeadjustmentsandaddincremental
piecesoffunctionalitytothepresentationsoftware.Eventually,itwouldgrowintoafull
packagethatwecansell.Wewouldbeabletoaddadditionalfunctionalitytoitthrough
updates,butwe’dreachapointwhereitwasmarketable,andthat’swhenwe’drelease.”
“I’mstillmissinghowthisrelatestooursystem,”saidMatt.“We’renotbuilding
somethingsimplelikeapresentationsystem,we’rebuildingasalesorderprocessingand
monitoringtool.Ithastointegratewithnearlytwentysystems,andthereare,what,ten
differenttypesofcustomers,fromaccountingtoshippingtomanufacturingtosales?This
iswaymorecomplex!”
“Yes.Oursystemismorecomplex,butthatdoesn’tmeanwecan’tusethesame
approach.I’msuggestingthatwefindawaytoproduce,bytheendofthefirstsprint,a
thinsliceoffunctionalityallthewaythroughoursystem.Thisneedstobethecoreslice,
theonepeoplewilluseoverandoveragain.Whatdoyouthinkthatstoryis?”askedPaul.
“Ithinkit’sthebasesalesorderpath.Peopleneedtobeabletoputinanorderand
knowthattherightvehiclewillshiptotherightplaceasaresult,”saidDarren.
“Whatisthemostcommonvehicle?”askedPaul.
“Probablythebasiccabandchassis,”saidSara.
“Waitaminute,”blurtedMatt.“You’retellingmethatwe’regoingtobuildasystem
thatjustprocessesanorderforabasiccabandchassis?Next,you’regoingtosaythe
salespeoplecanonlypickonecolororoneengine,right?”
“That’sexactlywhatI’msuggesting,”saidPaul.“Wewillbuildthemostcommonflow
throughthesystem,whichisanorderforthebasecabandchassis.Overtime,we’lladd
morestoriesthatallowtheentirecompanytodomore.Storieslikeaddingcolorchoices,
configuringfordifferentengines,andtheabilityforthefloortoupdatethestatusonthe
vehiclesoshippingknowswhenitwillbebuilt.Thesethings,thesevariables,areall
dependentonthiscorestory,though.Doesthatmakesense?”
“Igetit,”saidDarren.
Someotherteammembersnoddedyes.Mattdidnot.
“I’mstillconfused,”saidMatt.
“Me,too,”admittedSara.“Whyonearthwouldwedothis?I’dthinkthatweshould
justdoournormalapproach,whichistodesignthebackendintegrationdatabase,getthat
sortedout,buildtheintegrationcomponentsonebyone,andthenbuildtheinterfaceatthe
end.”
“Whatwouldwedeliverattheendofthefirstsprint?”askedDarren.
“Well,”saidSarawithabitofhesitation,“Iguesswe’dshowthemthedatabase
designs.”
“Remember,Sara,we’retryingforworkingsoftwarehere,andthethresholdis
potentiallyshippable,nomatterhowsmallthestory,”saidPaul.
“Yeah,butwon’tthiscreateatonofworkthatjustgetstossedout?”askedMatt.
“Right.Imean,we’reprobablygoingtohackthatfirststorytogetheranditwillhave
mocks,whichwewon’tusefortheactualendproduct.Whatdoesthisgetus?”asked
Sara.
Darrensteppedupforthisone.
“IfIunderstandthisScrumthing,it’sgoingtogetusanopportunitytogetfeedback
fromourusersinthereviewmeeting,right,Paul?”
“Exactly!”exclaimedPaul.“Thefeedbackisthemostimportantpart.Wewillshowthe
usersthebasestory.ThenMattwillworkwiththemtoordertheremainingstories
accordingtotheirhighestvaluetothecustomer:Isitmoreimportantfortheusersto
configureadifferentengineoradifferentcolor?Thesevariablesarethingswecanaddin,
buttheybuildoffthiscorebasestory.We’reshowingworkingsoftwareandgettingreal-
timecustomerfeedback.Sincewe’relimitingourvariables,thestorieswillremainsmall,
whichmakesthemeasiertotestandvalidate.Fromthere,weexpandandthenreleasewith
aminimalmarketablefeatureset.”
“IthinkIgetitnow,”Sarasaid.“Whatyou’resaying,Paul,isthatwebuildeachlayer
ofthesystem—evenifthatmeanscreatingthingswe’llnevershipwith,likeflat-file
databases,mocks,basicallythethingsthatwewillrefactoroutatalaterdate.We’lldothe
basestorytovalidatethatthesystemworksend-to-end,andthenwecanexpandonitas
wego.”
“You’vegotit,Sara,”saidPaul.“We’llbuildathinsliceandexpanditaswego.Itwill
bepotentiallyshippablebecauseitwillmeetourdefinitionofdone.We’llbuild
acceptance,integration,andperformancetests,documentit,doeverythingweneedtodo
toship.”
“Myoneremainingconcernisrework,”arguedMatt.“Itseemslikewe’recreatingaton
ofstuffwe’lljusthavetothrowawaylater.Isn’tthatwasteful?”
“No.It’sriskmitigation.We’llbeabletoshowthestakeholdersasystemattheendof
thefirstsprintthatworksend-to-endforthatbasestory.We’llgettheirfeedbackandmake
changesasneeded.Fromthis,wewillincreasethesystemcapabilities,expandingthe
functionalityandscalingitformoreusers,”saidPaul.
“Okay.I’mwillingtogiveitatry,butIcan’tcommittoitwholeheartedlyuntilIsee
thisinaction.I’llgiveitashotforthreeorfoursprints,butifIdon’tlikeit—”saidMatt.
“Youwill,”saidPaul.“Trustme,youwill.”
TheteamtriedthemodelPaulproposedandlikedit.Asthesystemexpanded,andthe
numberofstoriesexpandedfromtheoriginalbasestory,sodidtheteam’sconfidence.
Afterthreemonthsofrunningtwo-weeksprints,theteamwasabletopublisharelease
planbasedonobservedvelocity.Inturn,therestoftheorganization’sconfidenceinthe
teamincreased—theteam’spredictionswerebasedonhistoricalvelocityrangesand
knowledge,sotheywereveryaccurate.Theteamdeliveredthesystemwithinthe
predicteddaterangewithoutanyoftheusualintegrationglitches.Intheend,everyone
agreedthattheteam’scommitmenttodeliveringworkingsoftwareeachsprintwasthekey
toitssuccess.
TheModel
It’snotalwayseasytoseethepathtoworkingsoftware.Inthestory,Matt,Paul,andthe
teamstrugglewithhowtobuildthesystemwhileachievingthegoalofcreating
potentiallyshippablesoftwareeverysprint.Theteam’sunderstandingofworkingsoftware
was“completed,releasablecomponents”ratherthanaworkingend-to-endscenario.Paul
helpedtheteamandtheproductownertounderstandthevalueinbuildingthesystemend-
to-endfromthebeginning,startingwiththesmallest,mostcommonlyusedpaththrough
thesystem,withallthetestsandsupportintheteam’sdefinitionofdone(seeChapter7,
“HowDoYouKnowYou’reDone?”).
Togetateaminthemindsetofdeliveringworkingsoftwareeverysprint,havethem
focusononeoftwofactors:corestory,whichwesawinthestory,ornumberofusers.
Thenmoveontotacklingsystemelementsintermsofriskand,astheprojectprogresses,
expandingandvalidatingfurtherfeatures.
CoreStory
Onewaytodeliverworkingsoftwareistoidentifyonecorestory,fixingtheother
variablessothatthecorestoryrunsthroughtheentiresystem.Astheprojectprogresses,
thevariablecomponentsreplacethemocksandflatfiles.Throughoutthesprints,theteam
deliversworking,potentiallyshippablesoftwaretoitscustomers,validatingthatwhatis
beingcreatedisindeedwhatthecustomersneed,andrevisestheplanaccordingly.
Thecorestoryistheabsolutebaserequirement,thesmallestsliceofend-to-end
functionalitythatisthemostcommonlyused,asshowninFigure24-1.Inthestory,the
basewasthesliceoffunctionalitythatallowedsalespeopletoenteranorderforasingle
cabandchassisvehicle.Allotherfactorswerelimited.Thecolorwaswhite,theengine
sizepredetermined,thedrivetrainandtransmissionhadonlyonesetting,theshipping
addresswasconstant,andsoon.Thissliceoffunctionalityenabledtheteamtocreatea
singleflowthroughthesystem,validateitwithcustomersandstakeholders,andthenbuild
offthatbasestory.
FIGURE24-1Startwiththebaseend-to-endfunctionality
Constrainingstoriestotheirsmallestpossiblesizemakestestingandvalidationmuch
moremanageable.Asteamsexpandindividualend-userfunctions(userinterfacedesign
changes,updatingamockobject,orevenbringinganewintegratedsystemonline)and
addthemtothecore,workingsoftware,theyalsoaddthetestsanddocumentationthatgo
alongwiththenewfunctions.Theteaminthischapter’sstoryknowsthatuserseventually
needtobeabletoorderacabandchassisinavarietyofcolors.Whilethisisasimple
changeintheUI,thebackendintegrationtomakethishappenismuchbiggerthanadrop-
downbox.Addingthissmallstoryexpandstheinterfaceandalsoaddstotheacceptance,
unit,andintegrationtestframeworks;documentation;andthelike.
NumberofUsers
Anotherwaytoshrinkthesizeoftheinitialstory,whilestillenablingend-to-end
functionality,istolimitthenumberofuserswhocanuseoraccessthesystem.For
example,Ionceworkedaspartofateamtobuildanauthenticationenginedesignedto
supportfivemillionusers.Itwasathree-tiersystemwithaC++plugin,awebserviceto
managethebusinesslogic,aninterfaceforadministratorstoconfigure,andseveral
databasesthathandledvariouscallsbasedontherequestscomingintothesystem.
Ourinstinctwastobuildeachcomponentserially,integratingandtestingeverythingat
theend.Weknew,though,thatweneededtointegrateandtestaswewentalongtobe
potentiallyshippable.Therefore,liketheexampleinthestory,we,too,startedbyfinding
aninitialcorestory.Becauseofthenatureofthisparticularsystem,wechosetolimitone
bigvariable,thenumberofusers,whichchangedthewayweapproachedtheother
variablesinthesystem.
Webuiltathinsliceoffunctionalitywiththeunderstandingthatonlyoneuserwouldbe
abletousethesystemattheendofthefirstsprint.Thinkingthiswayallowedusto
employdevicesthatwouldneveractuallyshipwiththesystem,suchasflat-filedatabases
andmockintegrationpoints.Thisapproachenabledustogathercustomerfeedback
quickly.Weranstressandperformancetestsateverysprint,aswellasanautomatedsuite
ofintegrationandacceptancetests.Becausewewereworkingwithalimitednumberof
users,wewereabletofilteroffselectusersfromthefront-endproductionloadbalancers
basedontheiruserIDandpushthemtothissystem.Asweexpandedthenumberofusers,
thesystemgrew,andsodidthecomplexityofthesupportingobjects.Flatfileswere
replacedwithdynamicdatabases;mockobjectswithrealintegrationpoints.Intheend,we
wereabletoshipacomplexsystem,withinthereleasewindow,withouttheperformance
andintegrationissuesthathadplaguedusonotherprojects.
StartwiththeHighestRiskElement
Regardlessofwhichapproachyoutake,makesureyoustartwiththehighest-risk
elements.Buildingthehighest-risktechnicalsolutionsfirstmaybesecondnatureforan
experiencedScrumteam;however,liketheteaminthischapter’sstory,manyteamshave
troublefiguringouthowtobothaddresstheirhighestrisksandalsofindwaystobuild
incrementally.Intheauthenticationmodeldescribedinthe“NumberofUsers”sectionof
thischapter,forexample,theteamknewitshouldattacktheC++componentfirstbut
struggledwithhowtodothatwhilestilldeliveringworkingsoftwareeachsprint.Limiting
thenumberofusersenabledtheteamtoworkonthehighest-riskcomponentwhilestill
deliveringworkingsoftware.
WhenIthinkofthehighest-riskitems,Iliketoassociatethemwiththephrase,the
windowofopportunity.Ioftentalkaboutthisasanactualwindow—onesobigthat,atthe
beginningofanyproject,youcoulddriveabusthroughit.Eachhigh-riskitemisa
metaphoricalbusthatneedstodrivethroughtheopenwindow.Thebiggerthebus,the
biggerthewindowneedstobe.
Earlyintheproject,it’srelativelyeasytodrivethebigbusesthroughthewindow.As
eachdaypasses,however,thewindowclosesalittlemore.Ifyoudon’tfindthebigbuses
inyourprojectearly,anddrivethemthroughthewindowwhileit’sopenwide,you’re
goingtohaveahugedisasterattheendasthebuscrashesintothenearlyclosedwindow.
That’swhythehighest-riskelementsneedtobebuiltearly,inlockstepwithlimitingusers
anddoingthesmallestslicethroughthesystem.Attheendofthesprint,youare
validatingthatthebushasbeendriventhroughthewindow,showingpotentiallyshippable
software.Aseachsprintgoesby,thebusesyoupassthroughyoursystem,yourhigh-risk
items,getsmallerjustasthewindowdoes.
ExpandandValidate
Astheprojectprogresses,expandonthesmallslice,sprintbysprint,untilyouhavea
minimalmarketablefeaturesetorscenario,asshowninFigure24-2.Theexpansion
shouldbeorderedbytheproductowner,whoworkswiththestakeholderstoidentifythe
highest-prioritystoriessotheorderingintheproductbacklogisaccurate.
FIGURE24-2Expanduntilthesoftwaremeetscustomerneeds.
Eachelementyoubuildmustalwaysfollowyourdefinitionofdone.Asmoreandmore
functionalityisadded,ensurethatintegrationisworkingasexpected.Investin
automation,asthecostsofmanualtestingincreasewitheachsprint.(SeeChapter9,“Why
EngineeringPracticesAreImportantinScrum,”formoreinformationonautomation
testingcosts.)Alwayskeepallunittestsuptodate.Creatingsoftwarethatispotentially
shippablemightappeartoslowdowntheteaminitially.However,thestabilityand
predictabilityitbringsensurestheteamhasastablevelocity,increasingeveryone’s
confidenceintheteam’sabilitytodeliver.
Onepossibleoutcomeofalwaysbuildingworkingsoftwareisthatyoureachapoint
wherethesoftwaresolvesthecustomerneedsinitscurrentstate,beforealltheonce-
desiredfunctionalityhasbeenimplemented.LookingatFigure24-2,astheteamgoes
througheachsprintandfunctionalityisbuilt,thedemonstrableworkingsoftwaremight
reachapointwhereitisgoodenoughtoshipbeforeitreachestheboundariesofthecircle.
That’snotonlyokay,it’sdesirable.Youarebothsavingmoneyanddeliveringwhatthe
customerneeds.
Thebiggestadvantageofworkingsoftwareisthatteamscangetrealfeedbackfrom
theirstakeholdersorusersveryearly.Earlyfeedbackallowsteamstoincorporatechanges,
shiftdirections,andreworkdesignswhenthesystemissmallandrelativelyinexpensiveto
change.Therefore,keepyourcustomersonthepathofincrementallybuildingthesystem
withyou,whichinvolvesgettingfeedback,makingincrementalchanges,andalways
rememberingthatworking,potentiallyshippablesoftwareisthetruemeasureofproject
progress.
KeystoSuccess
OneofthehardestthingsfornewScrumteamstodoistoproducepotentiallyshippable
softwareattheendofeverysprint.Theyarejustnotsurehowtobuildasystem
incrementally.Theytendtojumptobuildinganentirepackage,adatabaseforexample,
first,beforetheyhaveawaytovalidateit.
Creatingworkingsoftwarerequiresteamstochangetheirthinking,accepttherealityof
rework,andfocusonanend-to-endscenario.
ChangeinThinking
AgilecoachandconsultantJeffPattonhasademonstrationwhereheripsupadollarbill.
Hesaysthatthedollarbillisauserstory,andheisgoingtobreakitintosmallerpieces.
Asheripsthedollarintopieces,thosewatchingunderstandthatalthoughthedollaris
gettingsmaller,theindividualpieceshavenovalue.Theyonlyhavevalueasacollective
whole.
Teamsoftenripstoriesintopiecestomakethemsmaller.However,theseindividual
pieces,liketheshredsofadollarbill,oftenhavenovalue;theyarenotpotentially
shippable.Tocreatesmallpiecesthatstillhavevalue,youneedtochangeyourwayof
thinking.AsPattonexplainsit,inthecaseofadollarbill,thismeanssplittingitintocoins,
sayfourquartersortendimes,insteadofshardsofworthlesspaper.Inthecaseof
software,itmeansmakingeachsmallstoryapotentiallyshippablepiece,onethathas
valueandcanmovethroughtheentiresystem.
Historically,teamshavebuiltentiresoftwarecomponentsallatonce,integratingthem
neartheendoftheproject.Theproblemwithintegratingcomponentslateinaprojectis
thateachcomponentonitsown,likeapieceofadollarbill,hasnovalue.Thecomponents
havevalueonlywhentheyareintegrated.Thisisahugeshiftinmindset,onethatisnot
easilybroken.Achallengeyouwillhaveinthismodelisfightingtheurgetobuilda
completecomponentofasystemwhentherestofthesystemdoesnotwork.
Inthischapter’sstory,theteamandthecustomershadtocompletelychangetheway
theythoughtaboutcommunicatingtheirneedsandwishes(e.g.,userstories),andbuilding
anddeliveringsoftware.Theyrealizedtheydidn’thavetoworryaboutwhetherthe
softwarewouldactuallyship—justthatitwaspotentiallyshippableanddemonstrableat
theendofeachsprintandthatitmetthedefinitionofdonecriteria.Thoughthesystem
wouldnotbefullyreleasedtoproductionuntilitcouldsupportmultiplescenariosanda
varietyofusers,limitingthestoriesinitiallyandbuildingontothecorestorywasthebest
routetoworkingsoftware.Itwaslowriskandprovidedhighdegreesoftransparencyto
thebusiness.
Don’tbeafraidtothinkinwaysthatmaynotbefamiliartoyou.Breakingoutofaserial
mindsetrequiressomebasiclogicalthinkingthatwouldapplyintherealworld—don’t
tearthedollar,changeitintocoins.Thinkitthrough,drawitout,andbewillingto
experimentwithwhatworksforyourteamandyourproject.Onceyoustartdelivering
workingsoftware,you’llfinditeasierandeasiertofinditerativesolutionstoyourmost
complexsystems.
Rework
Whenteamsstartsmall,withflatfilesandmockobjectsinplaceofcomponents,they
knowthateventuallythey’llhavetorefactor.Astheyaddmoreandmorebackend
functionality,they’regoingtothrowawaysomecode.They’regoingtohaverework.
Theyjustdon’tknowhowmuch.
Dealingwiththisrealityisdifficultformanyteams.Somereworkisknown—inthe
story,theteamknewthattheflatfilesandmocksitcreatedwouldlaterbediscarded,but
theteamwaswillingtodothatreworkbecauseoftherewards:riskmanagement,
mitigation,andtransparency.Otherreworkismoreunpredictable.Customersarelearning
moreaboutthesystemduringtheprocess,too.Astheyseehowthingsactuallywork,they
areabletobetterexpresstheirneedsorevenchangetheirmindsaboutwhatfeaturesthey
reallywant.Businessdirectioncanturnonadime.Newinformationcanbediscovered
thatcanalterthecourseofaproject.Allthesechanges,whilecrucialtothesuccessofthe
project,meanreworkfortheteam—reworkthatshouldbemanagedandaddedtothe
productbacklogfororderingandprioritization.
Whatteams(andmanybusinesses)thatmakeacommitmenttopotentiallyshippable
softwareunderstand,however,isthatitismuchbettertomakemanysmallchanges
throughouttheproject,andyes,eventhrowawaysomework,thantodelivera“complete”
systemafteroneyearandthenfindoutthatitdoesn’tworkasintended,isnotatallwhat
thecustomerwanted,ordoesn’tmeettheneedsofthecurrentbusinessenvironment.
FocusonEnd-to-EndScenarios
End-to-endscenariosarewhatworkingsoftwareisallabout.Withoutseeinghowthings
willfunctionattheend,eveninabasicstate,yourcustomersandstakeholderswillremain
bewildered,andbig-busriskswillcontinuetohideinthefogofambiguity.
Thevalueofend-to-endscenarioscanbeseenintheexampleofmoldingapieceofclay
intoafigure.Atthebeginning,youhavetwooptions.Thefirstistolistentoallthe
requirementsfortheclayfigure,andthengoofftomold,paint,fire,andglazeit,
eventuallycomingbackwithahardenedfiguretopresenttothecustomer.Thesecond
optionistolistentotherequirementsandfindawaytopresenttheabsolutebase
requirement,shapedroughlyasprescribed.Ifyoudothat,youmakeafirstpassand
presentthestillmalleableclaytothecustomer,asking,Isthisthebasicform?You
continuedownthatpathmakingslightmodificationsandreturningtothecustomerto
validatethechanges.Eventually,youreachapoint(maybesooner,maybelater)whenthe
pieceisreadytobefired.
Focusingonend-to-endscenariosgivesyousomethingtangibletoshowthecustomer
whilekeepingthecodemalleable.Itallowsyoutoeasilyreformyourunderstandingofthe
endproducttobetterfitcustomerexpectations.Itkeepsthewindowofopportunityopen
aslongaspossible,whilestillgivingyourcustomerssomethingsubstantiveandofvalue
—somethingthatcouldbefiredandglazedasis,ondemand.
Workingsoftwareisnotanimpossibledream.Itistheonegoalthatensuresthatyouare
meetingyourcustomer’sexpectationsanddeliveringvalueeachandeverysprint.Whether
youchoosetofocusonacorestoryorlimityourusers,remembertobeginwiththe
highest-riskitems,thosebusesthatcouldotherwiseremainhiddenuntiltoolateinthe
process.Educateyourcustomeraboutwhatyou’llbedelivering,why,andhowtheycan
helpensurethattheyreceivetheproducttheyneed,whichmayormaynotbetheproduct
theyoriginallyaskedfor.Adjustingthewayyouthinkaboutsoftwaremightbehard,but
onceyoudo,you’llfindthatincrementallydeliveringpotentiallyshippablesoftwareisas
easyasexchangingadollarbillfortendimes:Change,please.
WorkConsulted
Patton,Jeff.Personalwebsite.http://agileproductdesign.com/jeff_patton.html(accessed
31March2015).
Chapter25.OptimizingandMeasuringValue
Domore.Domore.Domore.Teamshearitintheirsleep.Buttherecomesapointwhen,
nomatterhowhardtheypush,teamsjustcan’tincreasevelocityunderthecurrent
conditions.Therealityisthatotherthings—companymeetings,documentation,email,
productbackloggrooming,planning,younameit—arecompetingfortheteam’stime.
Unfortunately,customersandstakeholdersdon’tseeallthebehind-the-scenesworkthe
teamisdoing.Theyseeonlytheirownstoriesandthemesbeingdeveloped.Andthey
understandablywantasmanystoriesfortheirmoneyastheycanget.
Whatismissingistransparency.MostScrumteamsusetheburndowntoprovide
visibilityintotheircurrentsprints,whichisagoodvisualindicatorofprogress—mostly.It
turnsout,though,thatifyoureallywanttounderstandhowtheteamspendsitstime,you
needmorethanthat.Byprovidingtransparencyandvisibilityintotheworktypestheteam
isaddressing,productownersandbusinessescanoptimizeforvalue.
TheStory
SabelawasworkingasaScrumMasterwithaproductownernamedReisa.Reisadida
greatjobworkingwiththestakeholderstoprioritizethebacklog.Shewasreadily
availableandhadtheanswerstheteamneeded.Shewasagoodproductowner.Aboutfour
monthsintotheproject,though,ReisastartedaskingSabelaandtheteamiftheycould
improvetheirvelocityalittlebit.Theteamwasstunned.SabelatoldReisathat,for
months,theteamhadbeentweakingthingsandfeltitwasatitsmaxoutput.Not
convinced,Reisacontinuedtoquestiontheteamonaregularbasis,askingiftherewasany
waytheycouldgetmoreworkdoneinthetwo-weeksprints.
Theteamwantedtopleaseitscustomer,soduringitsninthandtenthsprints,theteam
triedtosqueakoutahighervelocitywithoutsacrificingteamvaluesandengineering
practices—anditworked.Thevelocityoutputforthesprintjumpedtojustabovethe
averageforsprintnine,butduringsprintten,itreturnedtotheaverageithadbeeninthe
firsteightsprints.Unfortunately,afteramonthof“pushingit,”theteam’saveragevelocity
wasnotaffectedinanysignificantway.Attheendofsprintten,theteamfinallyasked
Reisawhatwasbehindherprodding.
Reisatoldtheteamthatshewasgettingpressurefromherstakeholdersand
managementtodrivetheteamharder.“Weunderstandthatpressure,”SabelatoldReisa,
“butgiventheconstraintswehave,ourcurrentvelocityisasgoodasit’sgoingtoget.We
can’tgoanyfaster.”
“Whatconstraintsareyoutalkingabout?”Reisaasked.
“Theconstraintsthatwehavewithinthecompany.Noteverytaskwedoisforastory;
sometimesweneedtofigurethingsout,gototheScrum-relatedmeetings,gettogether
withthedatabaseteamtorehashsomeofourongoingproblems,ordosecurity,
compliance,oraccessibilityreviews.Thelistseemstogoonandon.Someofitis
essentialforproducingqualitysoftware;someofitis,frankly,justnoise,”repliedMichel,
ateammember.“Theproblemis,wecan’tgetridofthestuffthatisnoisebecauseit’s
drivenbythepowersthatbe.”
“Idon’tunderstand,”saidReisa.“Pleaseelaborateforme.”
TheteamwentontoexplaintoReisathattheworkitwasdoingbasicallyfellintofour
simplecategories:
1.Stories(featurework)thatdelivervaluetothestakeholdersofthesystem,whatthe
moneyismeanttobespenton.
2.Taxes,whicharethecostofdoingbusiness.Theteamlumpedthingslikemandatory
companymeetings,theScrum-relatedmeetings,securityreviews,andsimilaritems
inthiscategory.
3.Spikes,whichareitemstoinvestigateforthefuture.Thesewereusedtounderstand
andestimatetheambiguousstories.Teamsoftenencounterstoriesthatarequite
largeandhavealotofuncertainty.Insteadofmakingalistofassumptionsand
estimatingthestory,thisteamsetasideacertainnumberofhourstoinvestigatewhat
neededtobedonetodriveouttheambiguityandreducetheriskprofileforthat
storyorsetofstories.
4.Technicaldebt,whichpaysdownofthesloppinessofthelegacysystem.
“IthinkIunderstand,”saidReisa.“So,whatcanwedotoincreaseourvelocity?”The
teamlookedatherblankly.
“Whatdoyoumean,‘ourvelocity’?”Sabelaasked.“Doyoumeanyouwanttosee
morestoriescompletedineachsprint,despitetheconstraintswedescribed?”
“Yes,Iwanttoseeyoucompletemorestoriesthatdeliveruservalue.Whatdoweneed
todotomakethishappen?”saidReisa.
Sabelapickedupaone-literwaterbottlethatsomeonehadonthedesk.
“Reisa,howmuchwatercanfitinthisbottle?”Sabelaasked.Reisalookedperplexed
buttriedtoanswer.
“Well,oneliter?”saidReisa.
“Youtellme:isit?”Sabelaresponded.
Reisaconfirmed.Theone-literbottlethatSabelawasholdingcouldindeedholdone
literofwater.Thetwocontinuedtheconversationandclarifiedthatthebottlecouldalso
holdonehalfofaliterofwater,butitcouldnotholdmorethanoneliter.Tohavethe
bottleholdmorethanoneliter,thebottlewouldneedtobereengineeredorReisawould
needtogetabiggerbottle.Reisaunderstoodthis.
“Reisa,thebottlerepresentstheteam’svelocity.Wecancarryoneliterofworkper
sprint.Storyworkisjustonetypeofworkthatweneedtodoovereachsprint,sofeature
workcannotfilltheentirebottle,takeuptheentirevelocity,becausethoseotherwork
types,thingsliketaxes,spikes,andtechnicaldebt,needtofitinthebottle,too,”Sabela
said.
Reisanoddedinunderstanding.
“So,whatyouaretellingmeisthatfortheteamtogetmoreworkdone,morefeature
work,weneedtoeithergetabiggerbottleorreducetheotherworktypes?”
“Yes!”shoutedtheentireteam,visiblyexcitedthatitwasstartingtoclick.
“WhywouldInotwanttojustfocusondeliveringstoriesnowandworryaboutthe
otherworktypeslater?”saidReisa.
“Because,justdeliveringstoriesdoesnotensurewehavepotentiallyshippable
software.Callingastorydone1dependsdirectlyonsomeofthenon-storyworkwehave
inourbottle—specifically,thedebtandthespikes,”Sabelaexplained.
1.SeeChapter7,“HowDoYouKnowYou’reDone?,”formoreontheteam’sdefinitionofdone.
“So,isthereanywaytomakeroom?”askedReisathoughtfully.“Whatabouttaxes?”
“Mostofthosearesetinstone.Butthereissomewastethere—especiallysomeofthe
unnecessarymeetingswe’rehavingwiththedatabaseteam,”saidSabela.“Ifwecould
comeupwithamoreefficientwaytosharedata,itmighthelpfreeupsomeroominour
bottle—butwe’llneedyourhelpandmanagementbuy-intomakethathappen.”
“Iagree.Idonotthinkwecangetridofmanyofthetaxes,butitwouldhelpme
tremendouslytobeabletoshoweveryoneexactlyhowtheteam’stimeisbeingspentand
wherewemightbeabletohelpclearsomespaceforyou.CanyoumakeachartsothatI
canbringthisupwithmanagement?”
Reisasharedthechartwiththeprojectcustomersandstakeholders.Theywereableto
removetheunnecessarymeetingsandsawanimprovementintheteam’savailabilityand
velocity.Thoughtheysearchedformorewaystomakeroominthe“teambottle,”they
wereunabletoidentifyotheropportunitiesforoptimization.Onceeveryonesawhowfull
theteam’stimereallywas,however,thebarrageof“getvelocityhigher”statementsand
questionsstopped.
TheteamcontinuedtoworkwithReisaonwhatthesefourbasiccategoriesmeantand
howtheteamcouldintegratethemodelintoitsworkmovingforward.Reisalikedthe
visibilityitgaveher,andsodidthestakeholders.
TheModel
Reisa’sissuewasthatshedidnotunderstandwheretheteamwasspendingitstime.She
knewthattheburndownchartsshowedworkgettingdone,andsheknewthatthework
wascomingfromherproductbacklog.Butsheneverunderstoodwhytheteamcommitted
tosuchalowlevelofoutput—atleastalowlevelasshesawit.Shewasgettingpressure
fromstakeholdersandmanagementtohavetheteamcompletemorefeatures,andtheteam
waspushingback.Whatwasmissingwasthetransparencyandinsightintowhattheteam
wasactuallydoingwhenitwasn’tworkingdirectlyonfeatures.
I,too,havebattledambiguitywithcustomers,soIcreatedthismodel.Itisnot
revolutionary,butitssimplicityenablesteamstocommunicatetotheirstakeholdersand
customershowtheyarespendingtheirtime.
Toprovidethetransparencyandinsightneededtomakeintelligentdecisions,breakthe
workdownintothesecategories:
Stories
Taxes
Spikes
Technicaldebt
Otherpotentialcategories
Thesebucketsandtheirdefinitionshavebeenbuiltandrefinedoveryearsofprojects
anddozensofprojectteams.Idonotexpectthemtofityourbusinessperfectly;however,
theywerewrittenbroadlyenoughthattheyshouldworkforeveryone.Ifyouchooseto
refactorthem,remembertokeeptheintentofthedefinitionthesamewhilechangingthe
words.
Stories
Storiesareanyworkthatdeliversactualvaluetothestakeholders.Idefineastoryas
follows:
Functionalitythatwilldeliverbusinessvaluetoeitherauserorpurchaserofasystem
orsoftwarepackage.
Storyworkcanbeattributeddirectlytoreturnoninvestment.Moneyspentonfeature
workshouldprovideanoverallreturnthatishigherthanthecosttobuildthefeature.
Taxes
Taxes,whetherthey’reincometaxesorcorporatemandates,areoftenloathedbythe
peopleandteamspayingthem.Ihavefoundthemtobebothasavinggraceandagnarly
burrinmyside,almostsimultaneously.Withoutthem,infrastructurewoulddegrade,and,
whenbuildingsystems,teamswouldquicklyfalloutofcompliance.Theyareanecessary
evil.Idefineteamtaxesasthefollowing:
Corporateservicesormandatedrequirementsthatposeaburdensomecharge,
obligation,duty,ordemandonteamsorgroups.
Basedonthisdefinition,whenshippingsoftware,youfindtaxesimposedbyboththe
companyandtheteam.Asampleofthetaxespaidontheprojectdiscussedinthestory
follows:
Securityreviews
Legalcompliancereviews
Scrummeetings,suchasplanning,grooming,andthedailyscrum
Companyordivisionalmeetings
Theseareonlyahandfuloftheitemstypicallyrequiredtoreleasesoftware(suchas
securityreviews)orthatcompetefortheteam’stime(suchastheScrummeetingsand
companymeetings).Businesseswanttoseevaluebuiltasquicklyaspossibleand
sometimesforgetthatthereismoretobuildingsoftwarethanjustwritingcodeandtesting
it.Financialsystems,forexample,requireSarbanes-Oxleycompliance,soteamsshould
expectthetaxesinthatenvironmenttobehigherthanthoseofasmall,eight-personweb
developmentcompanythatfocusesonbuildinginteractivewebsites.Themeetingswith
thedatabaseteamsthatSabelamentionsinthestoryareanexampleofextraneoustaxes
thatinflatethecategoryandreduceoverallfeaturedelivery.
YounoticedthatIlumpedScrumoverheadintotaxes.TheScrummeetings,suchas
planning,retrospectives,andreviewsareallnecessarytaxesthatcomewithScrum.
Spikes
Aspikeisusedwhentheteamcannotaccuratelyestimateastoryorfeaturesetinits
currentform.Aspikeisabriefexperimenttolearnmoreaboutanareaofanapplication.
Therearemanydefinitionsforaspike.Iusethefollowing:
Aspikeisabrief,timeboxedactivityusedtodiscoverworkneededtoaccomplisha
large,ambiguoustaskorstory.
Allspikesmustbetimeboxed—acertainamountoftimeshouldbeallocatedtoexplore
andfurtherdefineworkeffortsthatcan’tbeproperlyestimated.Theendresultisastory
ortaskestimatewithahigherdegreeofaccuracy.
Anexampleofaspikewouldbesomethinglike“Investigatewhatthedeployment
processshouldbe.Documentandestimatethestoryintheproductbacklog.”Itwouldbe
impossibletoestimatehowlongittakestodeploywithoutunderstandingwhatthe
deploymentprocessisfortheprojectandforthecompany.
Onecaveataboutspikingwork:Theoutcomeofthespike,whetheritbeasetoftasksor
morestories,cannotbecomepartofthecurrentsprintbacklog.Thenewworkitems
identifiedinthespikeneedtobeprioritizedwiththerestoftheproductbacklogbythe
productowner.Why?Thereareseveralreasons.
Forone,theteamhasalreadycommittedtothegoalandthestoriesonthesprint
backlog,oneofwhichwasaspike.Theteamshouldnothavecommittedtothetasksthat
wouldcomeoutofthatspike,becausetheteamcouldnothaveknownwhatthetasks
wouldbe.Therefore,thesetasksshouldnotbeaddedtothesprint.Iftheteamdidtryto
factorintheoutcomeofthespike,andsetasidetimetospendonsomeoftheworkfrom
thespike,itwouldnegatethebenefitofaspike.
Anotherreasontonotaddspikeoutcomestothesamesprintisthatthenewlydefined
tasksmaynotbeofhigherpriorityorvaluetotheproductownerthantasksthatalready
existinthebacklog.Itisnotuptotheteamtoarbitrarilytakeontheworkfromthespike
withoutconsultingtheproductowner.
Bottomline:Spikeinonesprint;workontheoutcomeinanother,asprioritizedbythe
productowner.
TechnicalDebt
Technicaldebtseemstobeeverywheretoday,fromlargecompaniestosmallstartups.The
debtthataccumulatescankillaproject,Scrumorotherwise.Whatistechnicaldebt?
Wikipediahasagreatdefinition[WIKIPEDIA]:
Technicaldebt(alsoknownasdesigndebtorcodedebt)isarecentmetaphor
referringtotheeventualconsequencesofpoorsystemdesign,softwarearchitectureor
softwaredevelopmentwithinacodebase.Thedebtcanbethoughtofasworkthat
needstobedonebeforeaparticularjobcanbeconsideredcompleteorproper.Ifthe
debtisnotrepaid,thenitwillkeeponaccumulatinginterest,makingithardto
implementchangeslateron.
Thinkaboutyourcreditcard.Ifyoupaythefullbilleachmonth,you’redebtfree.
However,ifyoudecidetomakepayments,you’regoingtobepayingalotmoreinthe
futurebecauseyourinterestchargeswillaccumulate.Thesameistruewithsoftware,
whichiswhygoodScrumteamshaveacleardefinitionofdone;it’smeanttocombatthe
debtthatcanaccumulateandbuild.
Managingtechnicaldebtrequiresdiscipline.Allteamsknowwhattherightthingstodo
are;however,sometimeslazinessandbusinesspressuresdriveteamstomakebadchoices.
You’veprobablyswornatalegacysystemyou’vehadtoworkon.Why?Becausethe
teamfromyearsagowasundisciplinedandchosetoletdebtaccumulate.Debthastobe
paid,soit’sbesttoshowitinthesprintbacklogandpayasyougo.
Lastly,rememberthatifyouendasprintwithnewlycreatedornewlyfounddebt,
you’renotreallyatthepointwherethestoriesarepotentiallyshippable,ordone.
OtherPotentialCategories
WhenIseeteamsutilizethismodel,theyhaveusuallytweakeditabit,whichisgreat.
Often,teamschoosetotrackbugs,unplannedoutages,andanitemIonceusedquiteoften,
preconditions.
Someteamsfinditimportanttotrackdefectsseparatelyfromthefeaturework.Thisis
okay.Ifyourorganizationlikestoseewheredefectsgo,byallmeanstrackthem.You
woulddosobyaddingacategoryof“bugs”or“defects”andtrackit,asillustratedin
Table25-1.Thisenablesyoutoviewitvisually,asillustratedinFigure25-1.Dothesame
foroutages.
TABLE25-1TypeofWorkIncludedintheSprintBacklog
FIGURE25-1Achartmakesitclearwhatproportionofoverallworkisspentoneach
typeofwork.
Thelastpotentialcategoryispreconditions.
Preconditionsareitemsthatarenotpartofastory’stasksyetstillmusthappenby
sprintend.Theseareitemsidentifiedbytheteamandnegotiatedwiththeproduct
ownerandScrumMaster.Withoutcompletionoftheitems,theteamcannotsay,“The
sprintiscomplete.”
Anexampleofapreconditionis“Setupabuildenvironment.”Whilemanyaretempted
tosaythebuildshouldjusthappen,it’sbettertotracktheworkbecauseitisgoingto
contendfortheteam’stime.WhatIhavefoundovertheyearsisthatitemsthatwould
typicallyfallintothisdefinitionarenowaddressedinsprint0,orprojectsetup/initiation.
Settingupabuildenvironmentisasetuptask,anditneedstohappen,butwellbeforethe
firstsprintstarts.IfindthatIusepreconditionslessandlessasmoreandmoreteams
matureandbecomefamiliarwithagileandScrum.Iwouldnotconsidertracking
preconditionsahardrequirement,butteamstodaystillfindpreconditionshelpful.
StructuringtheData
Tostructurethedata,theteamshouldstartwithareviewofeachtaskinthesprint
backlog.Itemsshouldthenhaveanidentifieraddedtothemthatfallsinoneofthebuckets
justdiscussed.AsnippetfromanexamplesprintbacklogislistedinTable25-1.
UsingtheData
Whetherthedataisstoredinaspreadsheetoradatabase,youarenowabletodoanalysis
onthedataandprovideinsightforthestakeholdersofthesystem.
BydrawingachartliketheoneshowninFigure25-1,youcanseewherethetimeis
spentonaproject.
Inthisexample,theteamaveragescloseto60percentofitstimeonfeaturework,about
25percentofitstimeontaxes,andabout10percentofitstimeonspikes.Technicaldebt
isaddressedmoreintermittently.
Fromafinancialperspective,thecustomersarepaying$1.00butaregettingonly$0.60
ofvalue,basedonfeaturework.Somewillarguethatspikesprovidevaluebecausethey
willinformfuturework,buttheydon’tadvanceanyfeaturesinthesprint,whichmeans
theyhavenovalue.Insprint5,thetaxeswerenearly30percent—duringthatsprint,$0.30
ofeverydollarwasspentonthecostofdoingbusiness.
Now,expandingthischarttoincludeoutagesandbugs,notethatinFigure25-2the
valueofthecustomer’sdollardropssignificantly.Inthisexample,theteamaveragesclose
to45percentofitstimeonfeaturework,about20to25percentofitstimeontaxes,about
10percentofitstimeontechnicaldebt(whenitsurfaces),10to15percentonoutages
whentheysurface,andabout10percentonbugsordefects.
FIGURE25-2Expandingtomorecategoriesprovidesmoretransparency.
Atfirst,thisincreasedtransparencycanreallyhurt.I’veseenchartslikethiswherethe
value,thenumberofstoriesdelivered,isaslowas10percent,withbugsanddebttaking
up70percentoftheteam’stime.Untilthenon-story-relatedworkisvisible,though,you
can’tstartaddressingit.
KeystoSuccess
Withmetricsliketheseathand,youcanreadilyprovideyourstakeholderswitha
tremendousamountofvisibilityintowheretheirmoneygoes.Whenimplementingthis
model,remembertoeducatestakeholdersaboutthemodelandworkwiththemtofind
waystomaximizework-producingtime.Ifyoutrackthisdataformultipleteams,youalso
gainanewabilitytospottrendsandpotentialproblems.
EducateStakeholders
Ioftenfindthatstakeholderscareabouttwothings:moneyandtime.Oneimpactsthe
other.Toooften,stakeholdersdon’tunderstandwhysolutionscannotbeimplemented
instantaneously.Acomplete,high-levelcategorizationoftheactivitiescompetingforthe
team’stimegivesthemabroaderperspective:Timespentinonecategoryofworkdirectly
impactstheothers.
Themodeldividestheteam’stimeintobuckets.Thoughthechapterprovidedsome
samplecategories,youmightneedtoidentifydifferentwaystocharacterizetheteam’s
workinyourorganization.It’simportantthatyouexposehowtheteamspendsitstime;
it’smuchlessimportanttoadheretoarigidcategorylist.
Showyourstakeholdersandcustomerswheretheirmoneygoesandhowmuchtimeand
moneytheactivitiesthatsurroundsoftwarebuildingconsume.Thisleveloftransparency
willgoalongwaytowardultimatelygainingtheiracceptanceofreleasedates,value,and
theoverallcostofphysicallydoingtheproject.
WorkwithStakeholders
Oncethechartiscomplete,remindtheteamandstakeholdersthatthisbehind-the-scenes
lookintotheteam’stimeisnotmeanttobeusedagainsttheteam.Onthecontrary,itis
designedtohelptheteamandthestakeholdersidentifywaystoeliminatetime-wasting
activitiesorinefficienciesintheteam’stimethatarebeyondtheteam’scontrol.For
example,thestakeholdersinthestorywereabletoseehowmuchtimecertaincorporate
taxesconsumedandlookforwaystomovethemoutoftheteam’sway.
Youwillnevereliminateallnon-work-producingactivities.Someoverheadis
inescapable.Andinsomecases,youmightnotbeabletoidentifyanyareastostreamline.
That’sokay.Thepointistoprovideatransparentlookatwhytheteammovesatthespeed
itdoes.Theactionsthatcomeoutofthedatawillvaryfromteamtoteamandprojectto
project.
DetermineTrendsandPatterns
Ifyouusethismodelwithmultipleteams,youcanusethedatayoucollecttoidentify
patternsandtrendsinthecompany.Recognizingpatternsandtrendscanhelpyouto
forecastorspottroubleonnewprojects.Establishbaselines,showingthatbyprojecttype,
acertainpercentageofteamtimespentineachcategoryistypical.Whenateambeginsto
ventureoutsidetheestablishedpatternsofthecompany,furtherdiggingintothatteam’s
deviationmightbeappropriate.
Intheend,thismodelisabouttransparencyandaccountability.Implementingitonyour
projectnotonlyenablesyoutohaveinsightintotheworktheteamisdoing,andwhattype
ofworkitis,itundoubtedlyraisesblockingissuesthatarepreventinghigher-performing
teamsfromthrivinginyourcompanytoday.
Reference
[WIKIPEDIA]Wikipedia.com.http://en.wikipedia.org/wiki/Technical_debt(accessed25
February2015).
WorksConsulted
Jeffries,Ron.“EssentialXP:Card,Conversation,Confirmation.”XProgramming.com.
http://xprogramming.com/articles/expcardconversationconfirmation/(accessed1April
2015).
Wake,Bill.“INVESTinGoodStoriesandSMARTTasks.”
http://xp123.com/xplor/xp0308/index.shtml.
Chapter26.Up-FrontProjectCosting
It’sbudgetplanningtime.ManagementhasaskedyouwhatyournewScrumprojectwill
cost,whenitwillbedone,andhowmanypeopleyouneed.Youknowtheanswersyou
givemightbeforeverforgedinstone—howdoyourespond?
TheStory
Lindellwasaprojectmanagerworkingonnewfeaturesforthewebcomponentofthe
company’sonlineserviceoffering.Shehadbeenwiththecompanyformanyyearsand,
liketherestofherteam,hadheardofScrumbuthadneverusedit.Oneday,shewas
approachedbytheleaddeveloper,Brian,touseScrumonhernextproject.The
conversationwentsomethinglikethis.
“Lindell,IthinkweshoulduseScrumonthisproject,”saidBrian.
“I’dliketouseitaswell,butScrumwouldn’tworkforus.Wehavetolockdownour
people,date,andfeaturesupfront.Speakingofwhich,managementisbreathingdownmy
neck.I’vegottogetthemsomethingsoon,”Lindellglancedatherwatchbefore
continuing.“Managementwantsanup-frontplan—andthatkindofplanningisnot
somethingScrumteamsdo.”
“Iusedtothinkthat,too.I’vebeendoingsomereading,though,andthenlastweek,I
metthisguyMikeatausergroupmeeting.Hegaveatalkabouthowtheydidproject
costinginhiscompany.IfIcangethimtohavelunchwithus,wouldyouhearhimout?”
askedBrian.
“Sure.Ifthere’sawaytouseScrum,I’mallforit,”saidLindell.
“Great,I’llschedulesomethingforearlynextweek,”saidBrian.
ThedaycouldnotcomesoonenoughforLindell,whowasgettingmoreandmore
pressurefrommanagementtoprovidedates,thefeatureset,andthemoneyneededforthe
projectintermsofnewhardwareandpeople.Assoonastheysatdowntolunch,Lindell
turnedtoMikeandpleaded,“Mike,I’mstuck!Ihavenoclueonhowtocostthisproject.
I’dliketouseScrum,butIamnotsurewhattodo.IfeellikeIhavemybackagainsta
wall.”
“IthinkIcanhelp,”saidMike.“Brianfilledmeinonsomeofyourchallenges.The
firstthingIthinkweneedtodoisreviewthefeaturesetthatisbeingaskedfor.”
Lindellpulledouta100-pagespecificationdocument.Mikefrowned.
“Doyouhaveanythingelse,oristhisit?”askedMike.
“Thisisit,forwhatit’sworth.Imean,it’sbeensignedoffandapproved,butI’mnot
evensureweallagreewhatitmeansintheend,”admittedLindell.
“Well,it’sastartingpoint.Whatwereallyneedisaproductbacklog.So,let’sreview
thespecificationandcreateuserstoriesfromit.Ievenbroughtindexcards,”saidMike
withasmile.
LindellandBrianwerebothsomewhatfamiliarwithuserstoriesfromacoursetheyhad
attended,sotheyhadlittletroublegoingthroughthedocumentsectionbysectionand
writingstoriesonindexcards.Afterabout15minutes,Lindellnoticedapotential
problem.
“Mike,I’vebeenwriting‘asauser’likeIwastaught,butoursystemhasmanytypesof
users.ShouldIstartwritingthingslike‘asanaccountinguser’forthestories?”asked
Lindell.
“Yes!Please!I’msorryIdidnotexplainthatearlier,”saidMike.
“Noproblem.I’mjusthappytobedoingsomethingproductive,”repliedLindell.
Afteracoupleofhours,theyhadcreatedcloseto50stories.Mikehadtogetbacktohis
officebutofferedtocomebackthenextdaytohelpthemfinishthecosting.
WhenMikewalkedinthenextday,Brianhadputallthestorycardsonthewallofthe
conferenceroom.HeandLindellhadalreadysortedthembyfunctionalarea,withthe
bigger,higher-levelstoriesonthetopandthesmallerstoriesbeneaththem.Todetermine
thesizeandlevelofeachstory,BrianandLindellhadusedtheirbestjudgmentand
discussedeachstorywithoneanother,creatingmultiplegroupingsoffunctionalitythat
theythoughttheycouldship.
“Thisisgreat,”saidMike.“Ournextstepistofigureoutabouthowmanypointsthese
storiesareandwhattheirpriorityis.”
“Thepriorityissimple,”saidLindell.“Theyareallhighpriority.Pointsseemspretty
abstract.Whycan’twejustskipthatandfigureouthowmanyhourstheyaregoingto
take?”askedLindell.
Brianchimedin.“Usinghoursistoorestrictive.Managementwillseeareallyrough
hourlyestimateandholdustothatnumber.Pointsgiveusawaytocommunicatesize
withouttyingittoaspecificnumberofhours.”
“Itseemstomelikewe’rewastingalotoftimethatcouldbespentdoingreal
planning,”saidLindell.
Mikenodded,“Ihearyou,Lindell.We’reonlypartwaythere,andit’shardtoseefrom
herehowitwillallcometogether.OnethingI’velearnedindoingthisisthatweareall
reallygoodatestimatingrelativesize,likedeterminingwhichbagofgroceriesisheavier.
Ifwewereaskedtodeterminetheabsoluteweightofeachbag,however,wewould
struggle.ThisisthereasonIusestorypoints,arelativemeasure,ratherthanhours,which
isanabsolutemeasure.”
“Allright,”saidLindell.“We’vecomethisfar.Pointsitis.”
“Okay,whereisyourteam?Weneedthemheretoestimatethis,”saidMike.
“Myteam?Ha!Idon’tgetassignedateamuntilwegetthroughprocurementand
resourcing.AndIcan’tgetthroughthosestepsuntilIhaveabudgetandaplan.See,I
knewScrumwouldn’twork.”
“Iknowit’snotideal,Mike,”Brianputin,“butallwe’vegotrightnowisthethreeof
us.Isitpossibleforustoestimateit—justforthecostingprocess?Wecanalwaysput
betterestimatesonthestoriesoncewehavetheteam,right?”
“Wecan,yes,butIwantyouguystounderstandthatitworkssomuchbetterwhenthe
peoplewhodotheworkaretheoneswhoestimatethework.Ifthat’simpossible,we’ll
justmakedowithwhatwe’vegot.AreyoufamiliarwithusingpartoftheFibonacci
sequenceastherangeofpointchoices:one,two,three,five,eight,thirteen,andifneeded,
twentyandforty?”
“I’veheardofit.ButIdon’treallygetit.Arethosenumbershours,weeks?What?”
“Theyaremeasuresofrelativesize,”Mikeexplained.“ThinkofitlikeT-shirtsizes
versusdressshirtsizes.T-shirtsarerelative,XS,S,M,L,XL,XXL.Dressshirtsizesare
moreabsolute,withaneckandsleevemeasurement.”
WhenBrianandLindellnodded,Mikecontinued.“SowetreatourstoriesmorelikeT-
shirts.Youpickarepresentativesmalleststory,anXSorone-pointstory,anda
representativebiggeststory(anXXLorthirteen)…maybeevenastorythat’sinthe
middle(anMorL,threeorfive).Thesebecomeyourreferencestories.Thenalltherestof
thestoriescanbesizedrelativetothose.Atwo-pointstoryisabouttwiceasbigasaone-
pointstory.Aneightisabouthalfasbigasathirteen.Getit?”
“Sortof.”
“Let’sstartwiththisgroupinghere.Let’sfindourextra-smallstory,ouronepoint.”
LindellandBrianagreedonaveryeasy,smallstory.“Now,showmeonethat’sabout
twiceasbig.”Again,LindellandBrianhadlittletroubleidentifyingastoryforMike.
Soon,theywerediscussingthestoriesandassigningthempointswithminimaldifficulty.
Ittooktherestoftheday,butLindellandBrianfinallymadeitthroughtheentirewallof
stories.Throughouttheprocess,Mikewasbusytakingnotesoneachstorydiscussedso
LindellandBriancouldreviewthemlater.
“Let’stakeabreak,”suggestedMike.“So,Lindell,howdoyoufeelthisisworking?”
“Wow!Wesureflushedoutalotofassumptions.Imean,thisstorywhereweneedto
callthedataset.Brianthoughtwewereonlyusinghalfofthedataelementsbecausethe
specwasoutofdate.Throughdiscussingthestory,Iwasabletorememberaside
discussionIhadwithsomeoneatlunchrecentlywhosaidthatAPIhadbeenrewrittenand
thatrewritewouldnowallowustouseallthedatawewerelookingfor.I’msuregladwe
caughtthatnow!”saidLindellenthusiastically.
“That’soneofthehugeadvantagestousingthistechnique;theconversationsreally
flushoutassumptionsandhelpgeteveryoneonthesamepageearly,”saidMike.“What
weneedtodonextisdeterminehowmanystorypointsyourteamcanaccomplishper
sprint.Thisisgoingtobeabitofachallenge,sinceyoudon’thaveateamyet.Haveeither
ofyouworkedonaScrumprojectinthecompany?”askedMike.
“No,sorry,Mike.We’retheguineapigs,”saidBrian.
“Notaproblem.Whenwefirststarted,weestimatedourvelocity.Itwasnotperfect—
farfromit—butitgaveusaroughideaofwhatweneededtodotogetthroughtheproject.
Itgivesyouastartingpoint,”saidMike.
“Canyouwalkusthroughit?”askedLindell.
“IhaveapaperonitthatIcansendyoutonight,Lindell,”saidMike.1
1.Formoreonestimatingteamvelocity,seeChapter4,“PredictingTeamVelocity.”
“Perfect.Sooncewehaveourteam’svelocity,what’snext?”
“Wehavetoprioritizethestoriesyouidentified,”saidMike.
“ButMike,theseareall—”
“Toppriority?”saidMike,“Iknow,buthearmeout.Let’slookatthegroupingsonthe
wall.Whichgroupofstoriesisthemostimportant,theonegroupyoucouldabsolutelynot
livewithout?”askedMike.
“Thissethere,”saidLindell.“Butthisoneisaclosesecond.”
Beforesheknewit,Lindellhadchosenaminimalmarketablefeatureset[DENNE].
Mikemovedthecardsintoalistformat,withthemostimportantstoriesonthetopandthe
loweronesonthebottom.
“Great.Inthebacklog,weseewehaveaboutforty-fivestorieswithatotalpointvalue
atabouttwohundred.Andyousaidearlieryouhadaboutsevenmonthsasyourdeadline
—that’stwenty-eightweeks.So,let’ssayyourunyourestimationexercisesanddetermine
thattheteamcanmanagesomewherebetweentenandfourteenpointspersprint.Ifthey
runfourteentwo-weeksprints(twenty-eightweeks),howmanypointscantheydeliver?
“Somewherebetweenonehundredfortyandonehundredninety-sixpoints;so,not
everything,butagoodchunk.”
“That’swhyprioritizingthislistissoimportant.They’lldothemostimportantstories
inthetimetheyhave,”saidBrian.
“Howdowegetthatnumberclosertothetopoftherange,thevelocitynumber?”asked
Lindell.
“Youcanaddresourcesorimproveefficiency,whichisthejobofyourScrumMaster.I
suggestyouhaveone,and,again,ifyouwanttoknowwhy,Ihaveapaperthatexplainsit
prettyclearlyformanagementintermsthey’llunderstand—money,”saidMike.2
2.Tolearnmore,readChapter8,“TheCaseforaFull-TimeScrumMaster.”
“Thankyou,Mike,”saidLindell.“Icannottellyouhowhelpfulthisis.Sowe’ve
narroweddownthescopeandthevelocitybasedontheteamweget,butwhataboutthe
cost?”
“Howmanypeopledoyouthinkwillbeonthisproject?”askedMike.
“I’mgoingforsixpeople,butrealistically,we’llgetfive,”saidBrian.
“Andwhat’sworseisnoneofthepeoplewegetwillbefull-time.Theywillallbesplit
onotherprojects.Ifwearereallylucky,wewillhaveonefullydedicatedperson,butIam
notholdingmybreath,”saidLindell.
“Howmanyhourswilleachteammemberhaveforyourprojectonagivenday?”asked
Mike.
“Somewherearoundsixhours.Thatmaybehigh,butitmatchestheaverageI’vegotten
onpastprojects,”saidLindell.
“Great!Wewillusesixhoursofavailabilityperpersonontheproject.Now,howdo
youdoyourbilling?Thisisaninternalproject,right?”Mikeasked.
“Yes,internal.Ourloadedemployeecostsvary,butlet’sjustgowithonehundredforty
dollarsanhour,”saidLindell.3
3.LoadedemployeecostsarediscussedinChapter8
“Great,soifweuseonehundredfortydollars,andyouhavefivepeople,yourcost
shouldbearoundsevenhundreddollarsanhour.Ifeachpersongivessixhoursadayto
theprojectoveratwo-weeksprint,theprojectwillcostforty-twohundreddollarsadayor
forty-twothousanddollarspertwo-weeksprint,”saidMike.“What’syourbudget?”
“Ourprojectbudgetiseighthundredthousanddollars,”saidLindell.
“Soatforty-twothousanddollarseach,eighthundredthousanddollarswillbuyyou
eighteentonineteensprints.Theproblemis,youonlyhavefourteensprintsbeforeyour
deadline.Ifyourvelocitystaystenpointspersprint,youcanaccomplishonehundred
fortypointsinfourteensprints,anditwillcostfivehundredeighty-eightthousanddollars
[($42,000/10)×140].That’swellunderyourbudgetofeighthundredthousanddollarsbut
doesn’tdeliverallthestories.You’vegotthemoneytoaddapersontotheteam,which
wouldincreaseyourvelocitybutstillmightnotaccomplishtheentirebacklog.Evenso,
we’veaccomplishedwhatwesetouttodo,”saidMike.
Hewenttothewhiteboardandwrotetheinformationthattheyhadobtained.
ThelightbulbwentoffforLindell.Shenowunderstoodhowitalltiedtogether.And
evenbetter,shethoughtshecouldgivemanagementmoreoptionsandbetterinformation
thanshecouldhaveotherwise.
“Thanks,Mike,”saidLindell.“You’vebeenanincrediblehelp.Youknow,Ithinkthis
Scrumthingmightworkforusafterall.”
TheModel
Inthestory,Lindellwasaskedtodeterminewhataprojectwouldcostintermsoftime,
money,andpeople(akaresources).Thisisatypicalrequestthatstumpsmanywhoare
newtoScrum.Luckily,LindellhadhelpfromBrian,whoenlistedMiketotakethem
throughtheprocessofcreatinguserstoriesoutofafunctionalspecification;sorting,
prioritizing,andestimatingthosestories;andputtingthemtogetherwithanestimated
velocitytodeterminewhatcouldbeaccomplishedwithinthecurrentbudgetandtime
frame.
FunctionalSpecifications
Functionalspecificationsexisttogiveusasenseofcertaintyinthesystemsthatwillbe
built.Theycapturenotonlyfunctionalsystemdetailsbutarchitecturalandcomponent
detailsaswell.Unfortunately,functionalspecificationscapturethesedetailsatthe
beginningoftheproject,whentheleastamountofinformationisavailable.Additionally,
thoughtheyoftendescribedesiredsystemstates,suchasthesystemshallhave99.999
percentuptime,theydonotsayhowthatstatewillbeachieved.What’sworseisthat
functionalspecificationstraditionallyoutlineaplanforthesystemtobebuiltbyfunction
(design,development,testing)ratherthanbyfunctionality(featurex,featurey,featurez).
Scrumprojectsforegofunctionalspecificationsinfavorofuserstoriesandproduct
backlogs.Thesehigh-leveldescriptionsoffunctionalitycapturejustenoughinformationto
letusknowwhatwillbebuiltbutdonotfillinthedetailsuntilclosertowhenthefeature
willbeimplemented.Thisemphasisonminimalup-frontinformationandtheteam’s
commitmenttogroomingtheproductbacklogduringeachsprinthelpsthisliving
documentremainpertinentandavoidswastedtimespentonspecificationsthatareoutof
datebeforetheycanbeimplemented.Productbacklogsaremadeupofuserstories.Your
firststepincostingyourprojectistobuildaproductbacklogbywritinguserstories.
UserStories
Userstoriesareabstractions.Abstractionslivelongerthandetails.Unlikefunctional
specifications,storiesleavethedetailsforafutureconversationbetweentheproduct
ownerandtheteam.Inthischapter’sstory,BrianandLindellusedthefunctional
specificationtobuildaproductbacklogbasedonuserfunctionality,notonjobfunction.
Todothistask,theywroteuserstoriesonindexcardsthattheycouldsort,tearup,oradd
toeasily.Reframingrequirementsandspecificationsinthetermsofauserstoryallowsus
touncoverassumptionsandidentifyhiddenrequirementsthatwouldotherwisecatchusby
surprisefurtherintotheproject.Thecardscontainjustenoughinformationto
communicatetheidea,ortheconcept,ofwhatistobedeveloped,buttheydonotcontain
thedetails.Thedetailsareprovidedthroughconversationandareconfirmedjustbefore
thestartofthesprintastheteamandproductownerdiscussthestoriesthatwillbe
implemented.RonJeffriescallsthesestepsthethreeC’sofuserstories—card,
conversation,confirmation[JEFFRIES].
Creatinguserstoriesfromafunctionalspecificationcanbeachallenge.Ifyoucan,get
ridofthefunctionalspecificationandreplaceitwithuserstoriesandaproductbacklog.If
youmustmaintainthefunctionalspecificationthroughoutthelifeoftheproject,
concentrateyoureffortontheproductbacklog.Partoftheprojectoverheadwillbetomap
theproductbacklogtothespecificationand,asfeaturesaredeveloped,communicate
whichpartofthespecificationiscomplete.
EstimatingStories
Therearemanytechniquesforestimation:historicaldata,individualexpertjudgment,
decomposition,top-down,bottom-up,widebandDelphi,analogous,PlanningPoker
[GRENNING],andtoolssuchasCOCOMO[BOEHM].Regardlessofwhichtechnique
youuse,peoplewantaspreciseanumberaspossible.However,asmentionedinthe
“FunctionalSpecifications”sectionofthischapter,teamsdon’tknowenoughatthe
beginningoftheprojecttobeextremelyprecise,whichiswhytheyinsteadprovide
relativesizingestimates.
IpreferPlanningPokerbecauseitprovidesalittlebitofeverything—analogous
estimating,expertjudgment,decomposition—andinvolvestheentireteam.Planning
Pokerhasseveraladvantages.First,itfocusesonrelativesize,notfixedhoursordays.
Next,itprovidesteamstheopportunitytotalkaboutwhatthestoriesmean.Thisdialog
helpsspreadinformationandknowledgeaboutthesystemandpreventsasingleexpert
fromdominatingplanningdiscussions.
Forexample,saythatfouroutoffivepeopleestimateastorytobea2or3,andone
personestimatesthesamestorytobe8or13.Thatlastestimateislikelyindicativeof
somehiddenassumptionsthatneedtobeshared.Theconversationthatensuesmight
increasetherestoftheteam’sestimates,decreasethehighestimate,orevencausethe
teamtobreaktheexistingstoryintotwodifferentstories.Whole-teamparticipationand
knowledge-enhancingdiscussionsarethekeybenefitstoPlanningPoker.Forreallyhuge
backlogs,whereestimationmightseemoverwhelming,IoftenuseatechniquecalledThe
BigWall,whichisdetailedinChapter29,“PrioritizingandEstimatingLargeBacklogs—
TheBigWall.”
Whenestimating,teamstypicallyestimatetheworkusingstorypoints—abstractunits
ofmeasure.Whydotheyuseabstractmeasures?Becauseit’sbettertoberoughlyright
thanpreciselywrong.Inthestory,Mikeexplainedrelativesizingbypointingoutthatit’s
mucheasiertolookattwoshirtsandsay,“Thisoneisabouttwiceasbigasthatone,”than
totrytodeterminetheexactneckandsleevemeasurements.Sure,youcouldfigureitout
ifyouwantedto,buttheamountofeffortrequiredtogetexactorpreciseestimatesis
economicallyprohibitiveandgenerallynotwarranted.Gettingroughlyrightestimatesis
allthatisneededatthistime.
Keepinmindthatyourrelativesizesneedtobesimilarenoughtoenableyouandthe
teamtomakerealisticcomparisons.ComparingT-shirtsizesissimple.Tryingtochoose
betweenatunaandanorcawhale,oracompactfuel-efficientcarandalong-haulshipping
tractortrailer,isnearlyimpossible.Thedifferencebetweentheirrelativesizesistoogreat.
Instead,choosereferenceitemsthataremorealikeandeasilyvisualizedrelativetoone
another:asmallglasstumblerversusapintglassorasmallpropairplaneversusa747.
Evenrelativesizingmuststaywithinacertainscale.
Finally,rememberthatthoughLindellandBrianestimatedbythemselvesoutof
necessity,wheneverpossible,allthepeopledoingtheworkshouldbetheonesestimating
thework.
PrioritizingStories
Atfirst,Lindellmaintainedthateverythingintheproductbacklogwasequallyimportant
andsohadequalpriority.Mikeknewthiscouldnotbetrue.Ifeverythingisthehighest
priority,thennothingisapriority.Thefirststepinprioritizinguserstorieshastodowith
understandingwhatyouruserswant.(Severalexcellenttechniquescanhelpyou
understandwhatuserswant,includingtheKanomodelofcustomersatisfaction,financial
modelssuchasinternalrateofreturn[IRR],andInnovationGames®suchasBuya
FeatureorPrunetheProductTree.I’veincludedreferencesattheendofthechapterto
helpyoulocatetheseresources.)
Onceyouhavethisinformation,youcanuseoneofmanytechniquestoprioritizeyour
list.Onequickwaytodeterminepriorityistojustorderthelistfromtoptobottomand
labelthem1-n.Nothingcanbeequal.Youcanalsothinkintermsofminimalmarketable
feature(MMF)set,atermintroducedbyDenneandCleland-HuanginthebookSoftware
byNumbers.TheydefineanMMFas“unitsofsoftwarevaluecreation.Theyrepresent
componentsofintrinsicmarketablevalue”[DENNE].Ithinkofitasthesmallestgroupof
featuresthatmustexisttoreleasetheproduct.
Inthischapter’sstory,assoonasMikereframedthequestionintermsofminimal
marketablefeaturesets,“Whichgroupoffeaturesdoyouabsolutelyhavetohaveinan
initialrelease?”Lindellwasabletoquicklyprioritizeherbacklogintermsofwhat
featuresandfunctionalitytheenduserswouldabsolutelyneedtohavetousethesystemin
production.OnceLindellidentifiedthefirstset,shecouldthengrouptheremaining
featuresintoadditionalreleasesthatwouldenhancetheminimalfunctionalityintheinitial
release.
DeterminingVelocity
Withoutknowingthevelocityoftheteam,itisessentiallyimpossibletocalculatewhat
workwillbedoneandbywhen.LindellandBriandidnottakethetimetoestimatetheir
velocitywhiletheywerewithMike.Theyjustpulledarandomrangeoutoftheairtohelp
themfigureouthowtocalculatecostslater.
Eventhoughit’snotideal,itispossibletoestimateavelocityrange,evenforateam
thathasneverdoneScrumbefore.ManyofthosetechniquesaredetailedinChapter4,
“PredictingTeamVelocity.”Thekeyistoupdatetheplanwithrealdataassoonasthe
teambeginsrunningsprints.
DerivingCost
InChapter8,“TheCaseforaFull-TimeScrumMaster,”Idiscussedtheloadedcostofan
employee.Usethatapproachhereifyoudon’tbillforyourtime.Ifyouworkina
consultancyorhaveacapitalexpenditurebudgetthatyouareworkingfrom,youknow
whatthebillablerateisperperson.Eitherway,youneedtodeterminewhattheteamcosts.
Tofigureoutthiscost,youmustdeterminethecombinednumberofhourstheteam
thinksitcandedicatetotheprojectpersprintandmultiplythatnumberbythetotalofeach
person’shourlyrate.It’sreallythatsimple.
Forexample,considertheteamoffivepeopleshowninTable26-1.
TABLE26-1DerivingTeamCost
Theteamcosts$700perhour.Eachteammemberaveragessixhoursperdayonthis
projectandrunstwo-weeksprints(tenworkingdays).Theteamsprintcostis$42,000
($700×6hours=$4200×10days=$42,000).
BuildtheReleasePlan
Thefinalstepincostingistobuildareleaseplan.Inthischapter’sstory,Lindelldidsome
quickcalculationstodeterminethatatacertainvelocityandsprintcost,theteamwould
notbeabletoaccomplishallthestoriesintheproductbacklogbythedeadline.Sheand
Mikewerealsoabletodeterminethattherewasroominthebudgettoaddmore
personnel,whichmighthelptheteamaccomplishmoreworkpriortothefixeddeadline.
Thecostingworkshehaddoneallowedhertopresentamorecompletepictureofthe
initialprojectrealitiesandoptionsforgettingmoredoneinthetimeallotted.Chapter11,
“ReleasePlanning,”providesdetailsonhowtobuildyourownreleaseplan,arequired
elementwhenworkingwithaprojectmanagementoffice(PMO)orexternalcustomer.
Pleaserefertothatchapterfordetailsonhowtobuildanagilereleaseplan.
KeystoSuccess
Costingprojectsupfrontischallenging,ScrumornoScrum.Howmuchwillitcost,when
willthesefeaturesbedone,whatwillweget,andhowlongwillittake?UsingScrum
doesn’tchangethesequestions,butitdoeschangehowyoucommunicateyouranswers.
Discussionsofcostarereframedintermsofcostpersprint.Featurespecificationsare
replacedbyestimated,prioritizedproductbacklogs.Thequestionsofwhenandhowmuch
becomequestionsofbudgetandvelocity.Youhavetoeducatemanagementaboutthese
termstosomedegreeforthemtounderstandandacceptyouranswers.Intheend,though,
you’lllikelyfindthat,likeLindell,youactuallyhavemoreinformationandoptionsto
givethemwithScrumthanyoudidwithfunction-basedprocesses.
Asyoureadintheearlierdiscussions,theactualprocessofdeterminingcostisnotthat
complicated.Whatiscomplicatedisbuildingtrust.Afterall,costquestionsareusually
concretewaysofaskingtheabstractquestion,HowdoIknowyou’lldeliverwhatwe’ve
askedfor?Mostpeoplehavebeenburnedinthepastbyteamsthatmissdeadlines,have
less-than-desirablequality,anddon’tdeliver.Fromthesefailurescomethedetailed
questions,Ganttcharts,andproductspecificationsthatdominatesoftwaredevelopment
conversations.Asyouprovideanswers,don’tforgetthatwhatyouarereallyofferingis
assurance.Takethetimetoexplainhowthesoftwarewillbedevelopedanddelivered.
Frameeveryanswerintermsofarangeofwhatyoubelievecanbeaccomplishedrather
thanasadefinitivenumber.Ifyouranswerswerederivedwithoutthehelpofyourteamor
withvelocityestimatesinsteadofrealdata,communicatetheseassumptionsandprepare
managementforhowactualdatamightalteryouranswers.Atthesametime,reassure
themthatyouwillprovidethatdataassoonasitisavailable.Remindthemthatchanges
willoccurandthatyouranswersareaneducatedguess,notapromisesetinstone.
Up-frontcostingispossiblewithScrum.Trytoinvolveyourteamandusehistorical
datatocomeupwithyourprojections.Ifthatisimpossible,youcansimulateteam
discussionsandestimatestoanswerinitialcostingquestions.Assoonasarealteamis
available,though,throwoutthesimulatednumbersandreplacethemwithrealestimates
andactualvelocities.Changeyourprojectionstoreflecttherealityofyourteamand
communicatethenewnumbersassoonasyouhavethem.Ifyou’vedoneyourjobwell
duringup-frontcostingdiscussions,managementwillnotonlybereceptivetonew
numbers,they’llbeexpectingthem.
Agileplanningisstraightforward.Tellyourcustomers,stakeholder,andmanagement
whatyouknowupfront,informthemofanychangesalongtheway,anddeliverworking
softwareattheendofeachsprintsothatadjustmentscanbemade.That’sScrum.That’s
agile.That’saprocessthatbuildstrust.
References
[BOEHM]Boehm,Barry.1982.SoftwareEngineeringEconomics.EnglewoodCliffs,NJ:
Prentice-Hall.
[DENNE]Denne,Mark,andJaneCleland-Huang.2004.SoftwarebyNumbers.Upper
SaddleRiver,NJ:PrenticeHall,p.5.
[GRENNING]Grenning,James.
http://renaissancesoftware.net/files/articles/PlanningPoker-v1.1.pdf(accessed2April
2015).
[JEFFRIES]Jeffries,Ron.Xprogramming.com.,http://xprogramming.com/-
articles/expcardconversationconfirmation/(accessed2April2015).
Chapter27.DocumentationinScrumProjects
We’veallheardthecommonmyth,Agilemeansnodocumentation.Whileotheragile
fallaciesexist,thisisabigone,anditcouldnotbefartherfromthetruth.Goodagile
teamsaredisciplinedabouttheirdocumentationbutarealsodeliberateabouthowmuch
theydoandwhen.Inthischapter’sstory,wefindaduostrugglingtoexplainthatwhile
theywon’tbefullydocumentingeverythingupfront,theywillactuallybemorefully
documentingtheentireprojectfrombeginningtoend.
TheStory
“Hey,youtwo,”saidAshley,stoppingCarterandNoelinthehallwayastheypassedby
heroffice.“I’vebeensensingsomeresistancefromyouovertheinitialproject
documentation.IneeditbynextFridayforprojectsignoff,okay?”Ashleylookedbackat
hercomputerandbegantypingagain,clearlyexpectingaquickanswer.
CarterandNoellookedateachother,thenbackatAshley,theirmanager,before
replying.Theyhadknownthisconversationwascomingbutdidn’trealizethey’dbe
accostedinthehallwaybyanobviouslyharriedAshleywhenitdid.
“Listen,wecandocumenteverythingupfrontlikeyouask,”Noelbegan,assheand
CartermovedtostandclosetoAshley’sdoorway.“Butwedon’tthinkit’sthebest
approach.Thingschange,andwecannotpromiseyouthatthingswillgoasplanned.
Further—”Ashleystoppedtypingandlookedup,interruptingNoelmid-stream.
“Look,Idon’twanttoargueaboutsomethingasbasicasdocumentation.Ijustneedit
onmydeskbyFriday.”
Carterspokeup.
“Ashley,”hebegan.“CanIhavefiveminutestocommunicateadifferentapproach?I
knowyouhaveafullplate,butIthinkit’simportantforyoutotrytounderstandthispoint
beforewetableourdiscussion.”
Ashleyglancedatherwatchandnodded.“Fiveminutes.Go.”
“WhenIwasincollege,Iworkedforouruniversitynewspaper,”Carterexplained.“I
wasasportsphotographerandalwaysattendedlocalfootballgameswiththesports
writers.Iwasonthefield,andtheywereinthestands.
“Itprobablywon’tsurpriseyoutohearthatnotoneofthosesportswriterscametothe
footballgamewiththestoryalreadywritten.Now,theymighthavedonesomeresearchon
theplayers.Theymighthavetalkedtothecoachesabouttheirgameplans.Theymight
evenhaveaskedmetobesuretogetsomeshotsofparticularplayers.Buttheynever
wrotethearticlebeforethegameevenbegan.
“That’skindofwhatyouareaskingustodowiththesoftware.Youwantthecomplete
storyofhowthisapplicationwillunfold,includingthefinalgamescore,beforewe’ve
evenstartedplaying,”saidCarter.
Ashleyreplied,“Well,that’showwegetthingsdonearoundhere.Withouttheup-front
documentation—designdocs,usecases,masterrequirementsdocs,architecturaldesigns,
andreleaseplans—Iwon’tgetprojectapproval,andIcan’tbesurethatyoutwoandthe
restoftheteamunderstandwhatweneedtobuild.”
“Right.Igetthat,”agreedCarter.“It’snotunreasonableforyoutowantsome
informationbeforewegetstarted.Andyoushouldexpecttoreceivefrequentupdatesfrom
usonwhat’sgoingonwiththeproject.Afterall,thereportersIusedtoworkwithwould
takenotesandwritesnippetsofthearticleaboutthegameasitwasunfolding.
Periodically,wewouldgettogethertodiscusstheangletheywereworkingonandsome
oftheshotsIhadcapturedsofar.
“Buttoaskustotellyouwhatthesoftwarewilllooklike,exactlyhowmuchitwill
cost,andpreciselywhenwe’llbedoneislikeaskingustopredictthefinalscoreofthe
footballgame.Wecantellyouhowwethinkit’sgoingtogo,butwhenthingsare
changingandunfolding,it’sdifficulttopredictallthedetails.”
Ashleynodded.“Butthingsaren’talwaysthatvolatilewithourprojects.Weknow
basicallywhatwewantthistolooklike.It’sonlysomeofthefeatureswearen’tsureof.”
“Right,”chimedinNoel.“Andonprojectswherewecannaildownmostofthe
variablesandhaveaclearpictureofthefinalproduct,wecangiveyoumoreup-front
documentation.”
Carternodded.“Togobacktomysportswriteranalogy,thereweretimeswhenone
teamwasclearlydominating—thegamewasablowout.Inthosecases,thereportershad
theirstoriesmostlywrittenbyhalftime.They’dalreadycomeupwiththeheadline,filled
inalotofthedetails,andwerejustwaitinguntiltheendofthegametoaddthefinalstats
andscore.
“Mosttimes,though,thegameswerecloseandtheoutcomeuncertain.Inthosecases,
thereporterswouldkeepfillingintheskeletonofthestorywiththeeventsasthey
happenedinrealtime.Theywouldcomedowntothefieldathalftime,andwewould
discusstheunfoldingstory.We’dstrategizeandsay,‘Ifthegamegoesthisway,we’lltake
thisapproach.Butifitgoesthatway,we’lltakethisotherapproach.’
“Likewise,thelevelofdetailinourdocumentationatanygivenpointintheproject
shoulddependonhowcertainwearethatthingsaren’tgoingtochange.”
Ashleyleanedbackinherchairwithherhandonherchin,deepinthought.Noel
decidedtogoinforthekill.
“Ashley,rememberthe1989SanFranciscoearthquake,orthequakeandtsunamiin
Japan?OrwheneitherReaganorKennedywereshot?”
Ashleynodded.
“Well,youwouldnoticeatrendinalltheseevents.Intheinitialaccounts,themedia
headlinesconveyedthetragicevent,butwithveryfewdetails.Alltheycouldtellusat
firstwasgenerallywhathadhappened(explosion/quake/tsunami/shotsfired),when,and
where.Why?Becausetheeventswerestillunfolding,andthatwasallanyoneknew.As
thereportersonscenelearnedmore,theyaddedthenewfactsandchangedtheheadlines.
“Allthelittleupdates,facts,anddetailswereimportanttocaptureinrealtime,evenif
theylaterhadtobechangedtoreflectnewinformation.Withoutcontinuousbriefings,
manyofthedetailssurroundingtheeventswouldhavebeenforgotteninthechaosofthe
devastation.Thereportersdidn’ttrytowritemoreupfrontthanthefactstheyknew.
Instead,thereportersrecordedwhattheydidknowastheywentalong.Later,afterthe
detailshadsolidified,theywentbackthroughthevariousarticlesandwrotealarger,
encompassingsynopsisthatoutlinedthespecificeventfromtheinitialreportstothe
currentstate,”Noelexplained.
“That’swhatwe’resuggestingwedo:Makeourdocumentationastoryinprogress.Are
wemakingsense?”askedCarter.
Ashleysatforward.
“IthinkIgetitnow.WhatIoriginallyheardyousaywas,‘Ican’tgiveyou
documentation.’Butwhatyou’reactuallysayingisthatyouwilldocumentcertainthings
upfront,mostthingsinrealtime,updatingthemasnecessarytoreflectreality,andsome
thingsafterthefact.Butwhatdoesthatmeanintermsofsoftwareexactly?Ineedcertain
documentstogettheprojectapproved,suchasthemasterrequirementsdocumentation
outliningalltheusecases.”
Noelanswered,“WewillprovidethatMRD,butitwillbeintheformofaproduct
backlog,withstoriesatdifferentlevelsofdetail.Someofthestorieswillbeheadlines
only.Otherswillbefullyfleshedout,maybetothepointofusecases.Forexample,the
teamhasagoodideaofthearchitecturaldirection,whichgivesusinsightintowhatstories
tobuildforthefirstsixorsosprints,butafterthat,thingsgetfuzzy.Weknowthatyou
wantustocapturethedetails,andwehavethosedetailsforthehigher-prioritystories,but
wedon’thavethemforthestoriesoffinthedistance.Forthose,wemightonlyhave
headline-levelinformation.”
Ashleyasked,“Butifwedon’tcapturedetailsupfront,won’ttheygetlost?Anddon’t
youthenhavetospendasignificantamountoftimegettinganddocumentingthat
informationduringtheproject,iftheycanbefoundatall?Howcanyouensurethatwhat
isbeingbuiltiscorrect?”
“Remember,likeanygoodjournalist,wewilldocumentaswego,assoonaswecan,
withoutdoingtoomuch.ThatwaywhenwegettothepointwheretheUIstabilizes,let’s
say,wecancreatethemoredetaileddocumentationthatthecompanyneedsandthatwe
needtodeliverintheformofusermanualsandsuch.Wewon’tloseourdetailsbecause
we’llbegoingoverthosedetailseverysprintwiththeteam,documentingthemaspartof
ourdefinitionofdone,andcheckingwiththestakeholderstoensurewe’reontrack,to
ensurethatwhatisbeingaskediswhatwearebuildingandmatcheswhatthecustomers
meant.Keepingdocumentationuptodatealittleatatimeisjustpartofthecostof
buildingworkingsoftware.Ifthingschangealongtheway,wewillupdatewhatwehave
writtentoreflectthechange.It’sabalancebetweenstabilityandvolatility.Themore
volatilesomethingis,themorecarefulweneedtobeinwhatlevelwedocument.Ifit’s
stable,wecandosomethingsuchasalargedatabasediagrammodelinatool.Ifit’s
volatile,wemightjustdrawapictureonthewhiteboard—again,botharedocuments,
databasemodelstobeexact,buttheyareverydifferentintermsofformality,”finished
Noel.
“So,areweonthesamepage?”askedCarter.
“Yes,”saidAshley.“Igetitnow.IthinkthisisagoodapproachandsomethingthatI
willadvocate,providedyougivemeregularfeedbacksoIcanupdateexecutive
management.ButIstillneedthebigheadlinesbyFriday.Agreed?”
“Agreed,”saidCarterandNoeltogether.
Andthatwasthat.
TheModel
ManypeoplecanquotethepartoftheAgileManifestothatstates,“Workingsoftwareover
comprehensivedocumentation,”buttheyfailtomentiontheveryimportantexplanatory
componentthatfollows:“Whilethereisvalueintheitemsontheright,wevaluetheitems
ontheleftmore”[BECK].Scrumteamsstillvaluedocumentation;theyjustchangethe
timingofthatdocumentationtobemoreconsistentwiththeirlevelofknowledge.
Forexample,imagineyouareinauniversityworldhistoryclass.Yougettothepointin
theclasswhenit’stimetodiscussWesternEuropeanhistory.Yourprofessorsays,“Iwant
eachofyoutobuymynewbookWesternEuropeanHistory:The30thCentury.Come
preparedforanexamonthefirstfivechaptersintwoweeks.”
Youwouldprobablylookaroundtheroom,wonderingifwhatyoujustheardwas
correctandaskafellowstudent,“Didhejustsaythirtiethcenturyhistory?”
Commonsensetellsyouthatwithouttimemachines,itisimpossibletoreadafactual
accountoffutureevents—theyhaven’thappenedyet!Sure,therearepredictorsand
indicatorsthatsuggestwhatmighthappen,butnothingiscertain.Thisthenbegsthe
question:Ifthisapproachiswrongforauniversityclassthatmayreachafewthousand,
whyistheexactsameapproachacceptedwhendevelopingsoftwarethathasthepotential
toreachmillions?
Beforewe’vebegunanyworkonaproject,weareoftenaskedforexactdetailsasto
whatwillbedelivered,bywhen,andatwhatcost.Todeterminethesethings,teamsoften
writevolumesofdocumentsdetailinghowthesystemwillwork,theinterfaces,the
databasetablestructures—everything.Theyare,inessence,writingahistoryofthingsthat
haveyettooccur.Andit’sjustasludicrousforasoftwareteamtodoitasitwouldbefor
yourhistoryprofessor.
Thatdoesn’tmeanweshouldabandondocuments,anditdoesn’tmeanthatweshould
leaveeverythinguntiltheend,either.Acertainamountofdocumentationisessentialat
eachstageofaproject.Upfront,weusespecificationsoruserstoriestocaptureideasand
conceptsonpapersothatwecancommunicateprojectgoalsandstrategies.Whenwesign
offontheseplans,weagreethatwhatwehavedocumentedistherightthingtodo.
Thequestion,then,isnotwhetherweshoulddocument,butwhatweshoulddocument
andwhen.Theanswerhaseverythingtodowithnecessity,volatility,andcost.
WhyDoWeDocument?
Everyprojectneedsacertainamountofdocumentation.Ina1998articleonSalon.com
titled“TheDumbing-DownofProgramming,”authorEllenUllmannoteshowlarge
computersystems“representedthesummed-upknowledgeofhumanbeings”
[ULLMAN].Whenitcomestosystemdocumentation,weneedtorealizethatwe’renot
buildingorwritingforus;wearewritingforthefuture.IthinkUllmansummarizesitbest
withthissnippetfromthesamearticle:
Overtime,theonlyrepresentationoftheoriginalknowledgebecomesthecodeitself,
whichbynowissomethingwecanrunbutnotexactlyunderstand.Ithasbecomea
process,somethingwecanoperatebutnolongerrethinkdeeply.Evenifyouhavethe
sourcecodeinfrontofyou,therearelimitstowhatahumanreadercanabsorbfrom
thousandsoflinesoftextdesignedprimarilytofunction,nottoconveymeaning.
Whenknowledgepassesintocode,itchangesstate;likewaterturnedtoice,it
becomesanewthing,withnewproperties.Weuseit;butinahumansenseweno
longerknowit.
WhyisUllman’sconceptofcodeasaprocessthatchangesstateimportant?Becausewe
needtorealizethat,inahumansense,weusethesystemandweknowthesystem.Thatis
whywedocument.
So,whatisessentialtodocument,andwhatisneedlesswork?Muchofthatdependson
thetypeofsystemyouarebuildingandthewayinwhichyouwork.Teamsthatare
colocatedneedtodocumentlessthandoteamsdistributedacrosscontinentsandtime
zones.Teamsthatarebuildingbankingsystemsneedtosatisfymoreregulatory
requirementsthandoteamsbuildingmarketingwebsites.Thekeyistodocumentasmuch
asyouneedandnothingmore.
WhatDoWeDocument?
Thelistofessentialdocumentsisdifferentforeveryproject.Goingthroughmylistofpast
projects,somefrequentdocumentationitemsincludethefollowing:
End-usermanual
Operationsuserguide
Troubleshootingguide
Releaseandupdatemanual
Rollback/failovermanual
Userstoriesanddetails
Unittests
Networkarchitecturediagram
Databasearchitecturediagram
Systemarchitecturediagram
Acceptancetestcases
DevelopmentAPImanual
Threatmodels
UML(UnifiedModelingLanguage)diagrams
Sequencediagrams
Wedidn’twriteallthesedocumentsbeforetheprojectbegan.Andwedidn’twaituntil
thefinalsprinttostartthemeither.Wedidthemastheinformationbecameavailable.
Manyoftheuserstories,forinstance,werewrittenupfront.Butsomeofthemwere
changed,andotherswereaddedastheprojectprogressedandrequirementsbecame
clearer.Ourunittestswerewrittenaswecoded.Andattheendofeverysprint,we
updatedtheend-usermanualtoreflectnewfunctionality.Weincludedinourdefinitionof
donewhatwewoulddocumentandwhenwewouldwriteit(seeChapter7,“HowDoYou
KnowYou’reDone?”).
WhenandHowDoWeDocument?
So,ifwedon’tdoitallupfrontandwedon’tsaveitallfortheend,howdoes
documentationhappeninanagileproject?Documentation,anydocumentation,costs
money.Themoretimeittakestowriteandupdate,themoreitcosts.Whatagileprojects
strivetodoisminimizewritetime,maintenancetime,reworkcosts,andcorrections.
Let’slookatafewapproacheswecantakewhendocumentingourprojects:
Documentheavilyinthebeginning.
Documentheavilyintheend.
Documentaswegoalong.
DocumentHeavilyintheBeginning
Traditionalprojectsrelyonearlydocumentation.Asyoucanseefromthediagramin
Figure27-1,atypicalwaterfallteammustcapturerequirements,buildaprojectplan,
documentthesystemarchitecture,writetestplans,anddoothersuchdocumentationatthe
beginningoftheproject.Ifweweretooverlayalinethatrepresentedworkingsoftware,it
wouldnotbegintomoveupuntilthegraylinestartedtoflatten.
FIGURE27-1Traditionalprojectwithup-frontdocumentation
Thebenefitofthisapproachisthatpeoplefeelmoresecureaboutthesystembeing
built.Themajordrawbackisthatthissenseofsecurityismisleading.Inpointoffact,
thoughagreatdealoftime,effort,andmoneyhasgoneintowritingthedocuments,no
workingsoftwarehasyetbeencreated.Thechancesofgettingeverythingrightupfront
aremarginalonstableprojectsandnexttozeroonvolatileprojects.Thatmeansfactoring
incostlyreworkandextratime.Chancesaregoodthatthesehigh-priced,feel-good
documentswillturnintodustyartifactsontheprojectbookcase.
DocumentHeavilyattheEnd
Whenwedocumentheavilyattheend,wedocumentaslittleaspossibleasthesoftwareis
developedandsaveallthematerialneededtorelease,sustain,andmaintainthesystem
overtimeuntiltheendoftheproject.Figure27-2illustratesthisapproach.
FIGURE27-2Documentingheavilyattheendoftheproject
Thebenefitsofthisapproacharethatworkingsoftwareiscreatedquicklyandthatwhat
iseventuallywrittenshouldreflectwhatthesystemdoes.
Therearemanyproblemswiththisapproach.Peopleoftenforgetwhatwasdoneand
when,andwhatdecisionsweremadeandwhy.Teammembersontheprojectattheend
arenotnecessarilythepeopleontheprojectinthebeginning;departingteammembers
takemuchoftheirknowledgewiththemwhentheygo.Afterthecodeforaprojectis
complete,thereisalmostalwaysanotherhigh-priorityprojectthatneedsattention.What
usuallyhappensisthatmostoftheteammembersmoveontothenewproject,leavingthe
remainingteammemberstocreatethedocumentationforthesystembythemselves.
Countlesshoursarespenthuntingfordataandtryingtotrackdownandaccessoldteam
members,whoarebusywithnewworkandnolongerhavetimeforsomething“as
insignificantasdocumentation.”
Thoughsavingdocumentationuntiltheendischeaperinthebeginningbecausemore
timeisspentonactualsoftwaredevelopment,itisusuallyexpensiveintheendbecauseit
canholdupareleaseorcausesupportandmaintenanceissues,asitwilllikelycontain
gapsandfaultyinformation.
DocumentasWeGoAlong
Agileprojectsdothingsdifferently.Weacknowledgethatwhilewecan’tknoweverything
upfront,wedowanttoknowsomethings.Wealsomaintainthatdocumentationshouldbe
partofeachstory’sdefinitionofdone,sothatitiscreated,maintained,andupdatedinreal
time,aspartofthecostofcreatingworkingsoftware.Figure27-3illustratesthe
document-as-we-goapproach.
FIGURE27-3Documentingasyougo
Theproductownerworkswiththestakeholdersandcustomerstobuildtherequirements
whiletheteamworkswiththeproductownertoachieveemergentdesignandarchitecture.
Theteamkeepsthecodeclean,createsautomatedtests,andusescodecommentsand
othertoolstoslowlybuildotherrequireddocumentationforthesystem,suchastheuser
manuals,operationsguide,andmore.
Theonedrawbackisthatitdoestakealittlelongertocodewhenyoudocumentasyou
gothanitwouldtoflythroughthecodewithouthavingtowriteacommentorupdatean
architecturaldiagram.Thisdrawbackismorethanoffset,though,bythebenefits.Thereis
lesswaste,lessriskofeleventh-hourholdups,andmoreemphasisonworkingsoftware.
Muchofthedocumentationisupdatedautomaticallyaschangesaremadetothecode,
reducingmaintenanceandreworkcosts.Justasnewsreportscapturethedetailsofastory
forposterity,real-timedocumentationofdecisionsandbehaviorminimizesgapsin
knowledgeandcreatesalivinghistoryofthesoftwareforfutureteamsandprojects.
DocumentinginanAgileProject
So,weagreethatinmostcases,agileteamswillwanttodocumentastheygo.What
exactlydoesthatlooklikeonatypicalsoftwareproject?Toillustrate,let’suseadocument
thatisfamiliartoalmosteveryone:theusermanual.Awaterfallapproachwouldbeto
writetheentiremanualattheend.We’vediscussedwhythismethodisaworkablebut
riskysolution.Themoreagilewaytoapproachausermanualistoinclude“updatethe
usermanual”asoneoftheacceptancecriteriaforastorythathastodowithuser-facing
functionality.Bydoingthat,themanualisupdatedeachtimeworkingsoftwareis
produced.
Let’ssay,forexample,thatI’mwritingtheusermanualforanupdatetoAdobe
Lightroom(mycurrentfavoritepieceofsoftware).I’minsprintplanning,andtheproduct
ownerexplainsthatthestorywiththehighestpriorityis“AsanAdobeLightroomuser,I
canexportaseriesofphotographstoAdobePhotoshopsoIcanstitchthemtogetherto
makeapanorama.”Aswe’retalkingthroughthatstory,Irecommendthatweadd“update
usermanualtoreflectnewfunctionality”asoneoftheacceptancecriteriaforthatstory.
AsIwritethecodeorasIfinishthefeature,Iwouldalsoeditadocumentthatprovides
theuserinstructionsonhowtousethefeature.Dependingonhowstablethefeatureis,I
mightevenincludescreenshotsthatwalktheuserthroughtheinstructionsforboth
LightroomandPhotoshop.Ifthefeatureislessstable,meaningthecorecomponentsare
builtbuttheuserinterfaceteamisstillhashingouttheuserinterfacethroughfocus
groups,Iwoulddocumentthebehaviorbutprobablyonlyincludeplaceholdersforthe
screenshots.Thekeyhereisthatthestorywouldnotbedoneuntiltheusermanualis
updated.
Updatingtheusermanualwouldbeappropriatetodoatthestorylevel,asIdescribed,
butcouldalsobeaccomplishedatthesprintlevel.Forinstance,ifwehaveseveralstories
thatrevolvearounduser-facingfunctionality,wemightaddastoryduringsprintplanning
thatsays,“Asauser,Iwanttobeabletolearnaboutallthenewfunctionalityadded
duringthissprintinmyusermanual.”
WhatIamdoingisbalancingstabilityversusvolatilityofthefeaturetodeterminehow
deepIgoandwhen.Itwouldnot,forexample,beprudenttomakeupdatingtheuser
manualpartofthedefinitionofdoneforatask.Toomuchmightchangebeforethestoryis
complete.Norwoulditbeacceptabletowaittoupdatetheusermanualuntilrightbeforea
release.That’sfartoolatetostartcapturingthedetailsofthenewbehaviors.
Whendeterminingwhentodocumentyourownsystems,youalsoshouldbalancecost,
volatility,andrisk.(Formoreondeterminingyourdefinitionofdone,refertoChapter7.)
StartingProjectswithoutExtensiveDocumentation
Onechallengeyouwillfaceistohelpstakeholdersandcustomersunderstandwhyyouare
notdocumentingeverythingupfront.TellthemastorysimilartotheoneCartertoldatthe
beginningofthischapter(orsharethatstorywiththem).Remindthemthatwhile
documentingheavilyupfrontdrivesdowntheperceivedrisk,youneverknowwhatyou
don’tknowuntilaworkingsolutionisinplace.
Eschewingextensivedocumentationupfrontdoesnotmeanyouareoffthehookfora
projectsignoffpiece.Butitdoesmeanthatthepiecewilllookdifferenttoyour
stakeholdersthanithasonotherprojects.Ratherthangivethemthespecificartifactsthey
request,answerthequestionstheyareaskinginregardtoschedulesandrequirementsin
themostlightweightwaypossibleforyourprojectandsituation.APMOmight,for
instance,askforaMicrosoftProjectplan,butwhatthePMOreallywantstoknowiswhat
willbedonebyaboutwhen.Bythesametoken,astakeholdermightaskyouforadetailed
specification,whenwhatshereallywantstoknowis,“AreyouandIonthesamepage
withregardtowhatI’maskingyoutodo?”
Signoffandapprovalwilloccurearlyandoften.Theproductownerwillholdmany
storyworkshopstobuildtheproductbacklog,workwiththeteamtobuildtherelease
plan,andthencommunicatethatinformationtoallinterestedparties,solicitingenough
feedbacktoensurethattheteamwilldeliverwhatthestakeholdershadinmind(whichis
rarelyexactlywhattheyaskedfor).Thedocumentstheproductownerusesforthesetasks
areonlyamodeoftransportationforideasandconcepts,andadocumentisnottheonly
waytotransferthoseideas.Up-frontdocumentationcanjustaseasilytaketheformof
picturesofwhiteboarddrawings,sketches,mockups,andthelike—itdoesnotneedtobea
largeformaldocument.
Thebeginningoftheprojectiswhenyouknowtheleastaboutwhatyouarebuilding
andwhenyouhavethemostvolatility.Whatyourstakeholdersneedisthepeaceofmind
thatcomesfromknowingyouunderstandwhattheyneedandcangivethemsomeideaof
howlongitwilltaketodeliver.Expendtheleastamountofeffortpossiblewhilestill
givingthemaccurateinformationandreassurance.Atthispointintheproject,everything
canandwillchange.
KeystoSuccess
Thekeystosuccessaresimple:
Decide—Determinewhatyouneedtodocumentforyourprojectandwhenitmakes
themostsensetoproducethatdocumentation.Somethings,suchascodecomments,
areeasytotime.Otheritems,suchasthreatmodels,aremoredifficult.Workasa
teamwithyourproductownertodeterminethemust-havedocumentsateachstage
ofyourproject.
Commit—Onceyouhaveadocumentationplan,sticktoit.Putitinyourdefinition
ofdone.Holdyourselvesaccountable.Documentationisneverfun,evenwhenit’s
brokenintosmallchunks.Remindyourteamthatalittlebitofpainwilleliminatea
greatdealofriskcomereleasetime.
Communicate—Ifthisisthefirstprojecttomoveforwardwithoutextensiveup-
frontdocumentation,thestakeholderswillbenervous.Helpthemout,especiallyat
thebeginningoftheproject,bysendingfrequentupdates,picturesofwhiteboards,
andanyotherdocumentsthatareproduced.Doasyourmathteacheralwaystold
you:showyourwork.Seeingworkingsoftwareandphysicalartifactsgoesalong
waytowardcalmingthefearsofeventhemostanxiousexecutives.
Investinautomation—Documentationiseasierandultimatelycheaperifyou
investalittletimeinautomatingeitherthesystemorthedocumentationitself.For
example,ifyoucancreateanautomatedscripttocompileallthecodecommentsand
parsethemintodocumentation,you’vesavedamanualstepandinstantlymadeyour
documentationmoreinsyncwiththeactualcode.It’salsomucheasiertodocument
acceptancetestresultsandAPIdocumentsautomaticallythanitistodoitmanually.
Ontheflipside,youmightfindthatautomatingthefeaturesthemselvescansave
youalotofdocumentationwork.Forexample,amanualinstallationprocessmight
requirea40-pageinstallationguide;anautomatedinstallationprocess,ontheother
hand,probablyneedsonlyaone-pageguideandisbetterfortheenduseraswell.
Wheneverpossible,automateeitheryourdocumentationorthefeaturesitsupports.
Theresultsarewellworththeinvestment.
Beingagiledoesnotequatetonodocumentation;itmeansdoingtimely,accurate,
responsibledocumentation.Makesurethatdocumentationisequallyrepresentedinyour
team’sdefinitionofdonealongsidethingslikecodeandautomation.Rememberthatwhen
changehappens,it’snotjustthecodethatchanges—theentiresoftwarepackagethatyou
aredeliveringchanges,documentationincluded.Lastly,rememberthatasmuchasyou
mightwishotherwise,documentationisapartofeverysoftwareproject.Whenyoudoa
littleatatimeandautomateasmuchaspossible,you’llfindthatwhileit’sstillan
obligation,it’snotnearlyasmuchofachore.
References
[BECK]Beck,Kent,etal.“ManifestoforAgileSoftwareDevelopment.”Agile
Manifestowebsite.http://agilemanifesto.org/(accessed16January2011).
[ULLMAN]Ullman,Ellen.Salon.com.http://www.salon.com/technology/-
feature/1998/05/13/feature(accessed18November2010).
Chapter28.OutsourcingandOffshoring
Whentheworkpilesupandthecostsmount,shiftingaportionoftheworktooutside
teams,especiallyonesthatcanbehiredatareducedrate,mayseemlikeagoodidea.
However,whatstartsoutasawell-intentionedshiftingofaprojectcanturnouttobea
nightmare.
Outsourcingand/oroffshoring,thoughsometimesunpopular,areheretostay.And
becausethat’strue,toomanyteamsfindthattheyhavetomakethebestofaless-than-
optimalsituation.ThatwasthecasewithJonathan,anewlyhiredleaddeveloperata
SiliconValleysoftwarecompanyknownforitsagiledevelopmentpractices.Here’sthe
storyofhowhereactswhenhelearnsthatheistomanagetwonewlyoutsourcedteams.
TheStory
Jonathanhadbeenworkingathisnewcompanyforthreemonths.Hewashappytobe
associatedwithanemployerwhovaluedagileprinciplesandwithteammateswhoknew
howtodeliverqualitysoftware.Itwasafarcryfromhislastjob,withanemployerthat
justdidn’tgetitandateamstuckinScrumBut[GUNNERSON].Thatallchangedtheday
hismanagerEmmainvitedhimforcoffeeandatalk.
“Iwon’tbeataroundthebush,Jonathan,”Emmabeganastheywalkedtowardthe
coffeestand.“I’vegotsomegoodnewsandsomepotentiallybadnews.I’llstartwiththe
bad.Ijustgotoutofanexecutivemanagementmeeting.Wearegoingtoshiftyourproject
toouroverseasoffice.Thegoodnewsisthatyou’vebeensuchagreatleaderthatthe
seniorvicepresidentaskedspecificallyforyoutorunit.Areyouupforit?”
Jonathantriedtokeephisfaceneutral,buthisfirstthoughtwasthatmovingtheproject
overseaswouldbetheworstpossiblesolution.HeturnedtoEmma.
“Emma,isthereanywaywecankeepitlocal?”hepleaded.“Theteamherehasmade
greatheadwayoverthelastfewmonths.”
“IwishIcouldsayyes.Theproblemisn’tyourteam;it’sthatourflagshipproductis
slippingbehindschedule.Togetitbackontrack,we’redivertingalllocalresourcesto
help,includingthepeopleonyourteam.Atthesametime,wewanttokeepyourproject
moving,sowe’vedecidedtocontractwithsomeoverseasgroups.Asidebenefitisthatthe
movewillsaveusmoney;we’llbeabletogettheoutsourcedteamforeightypercentless
thanwearepayingthelocalteams,”Emmaexplained.
Jonathanhadexperiencedoutsourcingandoffshoringworkbeforeathislastcompany,
whichwasoneofthemainreasonshehadchangedjobs.Hewasnothappyaboutthe
news,butbeinganewhire,heunderstoodhisplaceandwhatheneededtodo.
“I’vebeenthroughthisbefore,Emma,andIhavetobehonest.I’mnotsureit’sagood
idea,”saidJonathan.
“Whichpart—havingyouleadtheeffortorshiftingtheproject?”sheasked.
“Both,”saidJonathan.“NowI’mnotsayingeverycompanyislikethis,butIdowantto
sharewithyoutheoutsourcingexperienceatmylastcompanysothatyoucanunderstand
myhesitation.Acoupleofyearsago,Iwasleaddeveloperforanin-houseteam.Wewere
crankingalongprettywell.Becausewealwaysmetourtargets,managementthoughtwe
hadextracapacity,sotheyaddedmoreworktothepile.Weeventuallybecamestretched
sothinthatasinglesnapwouldhaverippleeffectsthrougheveryprojectinthecompany,
delayingeveryone.Itwasnotaprettysight.
“Management’ssolutionwastohavelocalteamspartnerwithoffshoreteams.Itseemed
harmlessenough,right?Itworkedlikethis:Thearchitectontheteam,acoupleother
guys,andIwouldbuildaprototype,documentit,andthensendeverythingtotheoffshore
teamtobuildandimplement.Weranitallinsprints,sotheoffshoreteamdeliveredstuff
backtouseverycoupleofweeks.
“Theproblemwasthatthevastmajorityofthecodetheysentbackwasinsuchbad
shapethatitwaseasiertorewriteitthantoexplaintotheoffshoreteamshowtofixit.
Thiswentonformonths.Onebyone,myteammatesquitoutoffrustration—thefunhad
goneoutoftheirjobs,andtheyrightlyorwronglyfeltthattheyweretheclean-upcrewfor
theoffshoreteams.WhenIleftafewmonthsago,Iwasthelastoriginalteammember.
Sincethen,thatcompany’sbeenhavingsupportnightmares.There’snooneleftlocally
whoknowsthecode,andbecausetheturnoverwasjustashighontheoffshoreteams,
thereisnoonelefttheretoaskquestionsofeither.Allthey’vegotisbuggycodeandmad
customers.
“Managementatthatcompanylookedattheinitialfiguresanddecideditwasacheap
andeasysolution.Andsure,itcostlesshourperhour,buttheworktheygotbackwas
inferiortowhatwewouldhavedeliveredhadweallbeenlocal.Anditcostthemawhole
lotofgoodemployees,includingme.Ifyoudecidetodothis,I’llgiveitmybestshot,but
Ifirmlybelievethisisthewrongthingtodo.Outsourcingthisworkwillnotgetusto
whereweneedtobefaster.Infact,itwillslowusdownandcostusmoreintime,money,
employees,andcustomersatisfaction.”
Emmastoodtherestunned.Sheusedthedistractionoforderingcoffeetogivehertime
tothink.Whentheypickeduptheircoffeeandtooktheirseats,shelookedupatJonathan
andcautiouslyreplied.
“Youbringupsomegoodpoints,Jonathan,pointsIhavenotconsideredbecauseIhave
neverhadtooutsourcework.Canyouexplaintomehowyouhadthingssetupandwhat
youwoulddodifferently?IaskbecauseIknowthedecisionhasbeenmadetopushwork
out.But,andthisisabigbut,wecanstillinfluencehowwestructureit,”Emma
explained.
“Yes,therearesomethingswecandotohelpmakeitlesspainful.Wewerebasedhere
inthevalley,andtheoutsourceshopwasinIndia.Theywere13hoursaheadofus,which
madeituncomfortableforbothpartiestocommunicate.Eitherwehadtoworklateorthey
hadtoworklate.Itjustdidn’twork.Itwouldbemuchbettertostayclosertothesame
timezone,whichinourcasewouldmeancontractingwithanoverseasteamtooursouth
—Argentinaisonlyfourhoursaheadofus,socommunicationwouldbemucheasier,”
saidJonathan.
Emmawasfranticallywritingnotes.“Whatelse?”sheasked.
“Well,nextIwouldembedteammemberstogether.WhatImeanbythisisthatifwe
outsourcethework,weneedtoplanahead.Weneedtohavetheoutsourceteamcometo
ourofficestoworkwiththelocalteambeforethoseguysshiftovertotheflagshipproject.
Thisenablesustohaveagoodknowledgetransferandgivestheoverseasteamachance
toaskthelocalteamquestionsarounddirection,architecture,andstrategy,”Jonathan
explained.
Emmawasabitstunned.“Youwantmetoflythewholeteamin?”sheasked.“Isn’tthat
abitexpensive?”
“Ofcourseitis,”saidJonathan.“It’soneofseveralhiddencostsofoutsourcing.Andit
won’tjusthappenonce.Aftertheinitialtransition,we’llneedtospendtimedoingthe
deeperknowledgetransferthatcanonlyoccurwithpairingandtime.Thatwillmean
flyingafewpeopleatatimefromoverseastohereforaone-monthstay,thensending
themhomeandflyingafewmorepeoplein.Eventually,everyonedowntherewillhave
theopportunitytoworkwiththelocalteambeforethelocalteamswitchestotheflagship
project.Itcostsalotmorethanjustwhatyouseeonpaper—andyouwon’thavethelocal
teamswitchedovertotheflagshipproductasquicklyasmanagementthinks.The
alternative,though,istotaketheapproachmyothercompanytookandjustswapthetwo
teamsoutnow.Butbelieveme,we’llpayforitlaterinbuggy,unintelligiblecodeand
unhappycustomers.”
Emmawasvisiblydiscouraged.
“Whatelsedoweneedtoconsider?”sheasked.
“Well,ifI’mgoingtorunthis,weneedtoplanonaheftytravelbudget.Iwillbe
travelingtoArgentinaonceamonthataminimumtospendtimewiththenewteam.
Communicatingonlyviaemailwillonlyslowthetransitiondownandmakethingsworse.
Oncetheygetontheirfeet,wecanlookatlesstravel,butitwillbealongwhilebefore
thathappens,”saidJonathan.
“Whatifwecan’tmakethishappenthewayyouwant?”askedEmma.
“Idon’tknow.Ican’timaginethisendingwellifit’ssetuppoorlyandisunderfunded.”
“Allright,”saidEmma.“WhatIneedyoutodoiswritethisupformeandpreparea
presentationformanagementsowecanpresentourcase.Canyoudothat?”askedEmma.
“I’llhaveitreadybynextweek,”saidJonathan.“Icametothiscompanybasedonits
reputation.Themanagementteamisknownforputtingpeopleandcustomersfirst.I’m
confidentthatoncemanagementseesthepotentialproblems,itwillfundthesolutions.Of
course,I’dstillratherkeepitlocal,butifwedohavetooutsourceit,atleastwecandoit
correctly.”
TheModel
Mostsoftwarecompaniesuseoutsourcingnotbecausetheywanttolaypeopleoffbut
becausetheyhavetoomanyconcurrentprojectsandnotenoughstafftosupportthem.
Managementlooksattheoptionsanddecidesthatoutsourcingistheirbestsolution.What
theymightnotconsiderorunderstandiswhatthatcouldmeanfortheteams,thecode,the
finalproduct,andthecompany’sbottomline.
Outsourcingandoffshoringarechallengingusinganyframework,butwhenyouadd
agilepracticessuchaspairing,coreteamhours,anddailyscrums,offshoringanagile
projectquicklybecomesacomplicatedandsomewhatexpensiveendeavor.Thatbeing
said,I’mnotsuggestingyouuseatraditionalapproachforoffshoring.Onthecontrary,
youneedtobeverydisciplinedandhavefrequentcheckpoints,greatertransparency,and
otheragilepracticestomakesuccessmorelikely.
Beforeyoumakethejumptooutsourcing,considerthetruecosts.Ifyoustilldecideto
doit,orthedecisionhasbeenmadeforyou,worktostructuretheoutsourcingforthebest
chanceforsuccess.Thefollowingsectionslookfirstatthecostsandthenatstrategiesfor
dealingwiththerealitiesofoutsourcing.
ConsidertheTrueCosts
Whenitcomestocost,it’stemptingtolookonlyattheper-hourcostperperson.Ifa
developerinyourcountrycosts$100perhourandanoverseasdevelopercosts$20per
hour,itappearsmuchcheapertouseanoverseasteam.ButasJonathanpointedoutinthe
story,managementmustfactorinthehiddencostsofoffshoring.Theseincludetransition
costs,increasedoverhead,long-termretention,managingthework,andmakingsurethe
overseasteamfollowsthedevelopmentpractices/processesthatarevaluabletothelocal
team.
TransitionCosts
Onebighiddencostofoutsourcingistransition.AsJonathanhighlightedinthestory,it
willcosttimeandmoneytogetthenewteamuptospeed.Peopleneedtogettogetherto
transitioneffectively,whichmeansbringingsomeoralloftheremoteteamintoacentral
officetoworkwiththeoriginalteam.Andit’snotjustaone-timeknowledgedump.
Gettingthenewteamtounderstandasystemindepthinvolveshavingatleastsomeofthe
newteammemberstraveltopairwiththeoldteammembers.Itisbesttohavepeople
cyclebackandforth—meaninghavesomeofthelocalteamtraveltotheremotesiteand
havesomeoftheremoteteamspendtimeatthelocalteam’slocation.Dependingonthe
sizeofyourproject,theknowledgetransfercantakeweeks,months,orevenayear.
Researchshowsthattransitioncostscanrangefrom5to57percentofthetotalproject
cost[OVERBY][ROTTMAN].Studiesalsoshowthatwhiletheremightbeacostsavings
inusingoverseasteams,itisoftenoffsetbytransitioncosts.Infact,theoutsourcing
venturemightultimatelycostmorethanusinglocalteams[YU].
IncreasedOverhead
Chapter8,“TheCaseforaFull-TimeScrumMaster,”discusseshowtodeterminethe
loadedcostofanemployee.Itdescribeshowmostemployeesspendonly40to60percent
oftheirtimeontaskduringanygivenproject,meaningtheyspendthreetofivehoursper
daydoingproductivework.Therestofthetimeispartnecessaryoverheadandpartnoise
andwaste.
Outsourcedprojectshavebuilt-incommunicationchallengesthattakeawayfrom
availableproductivetime.OnteamsI’veworkedon,andmanyoftheteamsI’vecoached,
peopleonoutsourcedprojectsspendatleastanextrahoureverydayonadditional
meetings,email,andphonecalls,asteamsstrugglethroughculturalandlanguagebarriers.
Thisincreasedoverheadmustbefactoredintotheoverallprojectcost.
Long-TermRetention
Overseasteamstendtohavehighturnoverrates.Oncetrained,peopleunderstandtheyare
morevaluableandwilloftenleaveforabetterpayingjob.OnecompanyIworkedwithin
Indiasaiditsturnoverratewasninemonthsforeachemployee.Iftheorganizationyou
workwithhasasimilarturnoverrate,thatmeansthatbythetimeyoufinishtransitioning
theproject,you’regoingtohavetoabsorbthecostsassociatedwithbringinginnewteam
members.Turnoverratesmightincreaseforyourlocalteamsaswell.Thelongerthe
transitiontime,themoredifficultitwillbeforyourlocalteammemberstostayengaged
withtheproject.
CulturalChallengesandManagingtheWork
Day-to-dayworktendstobemoredifficultonanoutsourcedproject.Youhaveto
coordinatemeetings,phonecalls,andvideoconferences.Youneedtobemoreexacting
withyouremailcommunications.Alltheseobstacleshavesolutions,butthesolutions
comewithincreasedcosts.
Culturalchallengesalsocomeintoplayasyouattempttoworktogether.InSucceeding
withAgile,MikeCohnsuggeststhatteamscreatecoherence—or“sticktogether”[COHN]
(calledgroupcohesioninthesocialsciences).Todemonstratethedifficultiesof
understandingothercultures,CohnrelateshowGeertHofstedefromIBMidentified“five
keydimensionsalongwhichculturesvaried”[HofstedeascitedinCohn,p.359].Thedata
isilluminatingandrevealswhatmostpeopleknowalready.Chineseteamsarelesslikely
tochallengeauthorityinapublicforum.Americanteamsaremorelikelytocallitlike
theyseeit.Indianteamsmightsayyeswhentheyreallymean,“Ihearwhatyouaresaying
butdon’tagree.”Overcomingculturalobstaclestakestime,training,andpatience,which
ultimatelymeansmorecosttotheproject.
DevelopmentPractices
Afriendtoldmethatmanagementonceoutsourcedakeycomponentinoneofhisprojects
toacompanyinanothercountry—supposedlyatop-notchconsultingfirm.Hemanaged
theremoteteam(aswellasthelocalteam),butknewtherewastroublewiththeremote
team.Afterweeksoftheremoteteamcodingwithnocodedrops,hefinallyinsistedthat
theysendhimthepartiallycompletedcode.Whatwasreceivedwasasolutionwitha
singleProgram.csfile,whichcontainedasinglemethod,Main().Thatmethodwas
hundredsuponhundredsoflineslong.Theout-of-countryteamhadadozendevelopers
workingonthatsinglemethod,amethodthatdidn’tevenwork.Hesubsequentlyfiredthe
remoteteam.
It’snotasifthesepracticesarelimitedtooneortwocountries;I’vewitnessedthis
behaviorinUS-basedfirmsaswellasfirmsinEurope,China,Japan,SouthAmerica,and
India.Butifyou’vegotahigh-performingScrumteam,you’velongsinceabandoned
thosekindsofpracticesinfavorofagileengineeringandbestpracticesthatmakesoftware
morereliableandeasiertomaintain.Youexpect,therefore,toworkwithteamsthat
developsoftwaretothesamestandardsandusingthesamebeneficialpractices.Those
teamswon’tcomecheap.
Youalsoneedtodecidehowyouwanttoworkwiththeoffshoreteams.Youmight
choosetouseworkpackagedeliverables,whereeachteambuildsasetoffeaturesonits
own.Oryoumightchoosetoworkaroundtheglobe,wherethecodemovesfromsiteto
sitetosite.Theseaddedcomplexitiesmakehavingstandardengineeringpracticessuchas
continuousintegrationandtest-drivendevelopment(TDD)evenmorecriticaltoyour
success.Trainingteamsthatdon’talreadyusethesepracticestakestimeandmoney.
DealingwithReality
Ifyourcompanyisconsideringoutsourcing,showthemthelistofhiddencostsdescribed
inthischapter.Maybeyou’llbeluckyandthepowersthatbewillreconsider.It’smore
likely,however,thatyou’llbestuckwithanoutsourcingsituationatsomepointinyour
career.Itdoesn’thavetobethedeathknellforyourproject,aslongasyoudon’tgoin
blindly.Whenconsideringoffshoringaproject,youneedtofactorinallthecosts,
includingvendorselection,transitioncosts,travel,lossesinproductivity,andincreased
projectmanagement.So,howdoyouaccountforallthesecostsandgivemanagement
someideaofwhattheirrealcostwillbe?
BudgetandCost
Whenestimatingthebudgetforanoutsourcedproject,I’vefoundithelpfultoshowonly
thebest-andworst-casescenarios.Idonotguessatamostlikelycostbecausetoomuch
informationisunknownabouttheentireexperience(suchasteamdynamics,codequality,
andtheproductitself)todeterminewhatismostlikely.
Yourfirstinputisthetotalprojectbudget.Thebudgetmightbeanexternalnumberif
youaresolicitingabid.Oritmightbeaninternalnumberifyouhaveafixedcostyouare
tryingtohiteitherintermsofloadedcostsofateamoractualfundsthatyouhaveto
manage.
Listthebroadcostcategoriesinonecolumn.IntheexampleinTable28-1,thefirstcost
itemIlistisvendorselection.Ifyoualreadyhaveavendor,thatcostwouldn’tapplyto
yourproject.Fromthere,Ilistwhatpercentage(variable)ofthatiteminthebudgetis
usuallyspentonthatcategory.Forexample,thebest-casescenarioforvendorselectionis
0.5percent;intheworst-casescenario,thepercentagegoesupto1.5percent.Idothisfor
eachpotentialhiddencostitem.
TABLE28-1OffshoreCostVariableModel
You’llnoticethatsomeofthecategoriesinTable28-1areidenticaltothecostsI
highlightedinthemodel,suchastransitioncosts,forinstance.Others,suchasincreased
projectmanagementandlossesinproductivity,arelargebucketsforseveralrelatedcosts.
Forexample,lossesinproductivityincludeincreasedoverhead,theneedtogetupto
speedondevelopmentpractices,andananticipationofhighturnoverrates.Youcanuse
thesecategoriesorchooseonesmorespecifictoyourownsituation.
ThepercentagevariablesIlistarebasedonmyexperience.Isuggestyoustartwithyour
owndataand,ifneeded,lookatindustrytrends.Forexample,ifyouhaveabigteam,your
travelvariablecouldbesignificantlyhigherthanthatshowninTable28-1.Somecosts,
suchasbugsandturnover,arehardertoestimate,asyou’dneedtohavesomeideaofthe
qualityyouwillreceivebackfromtheoffshoreteamandwhattheirturnoverislikelyto
be.That’swhyit’simportanttocontinuetotrackthesecostsovertime,sothatyounot
onlycandemonstratetherealcostofoutsourcingbutalsocancreateabetterestimatefor
thenextproject.
Ofcourse,youmightalsofindthatwithtightengineeringpracticesandagoodeffortat
engineering,youactuallydosavemoney.Ifso,great.It’sstillimportanttoknowfor
certainwhattherealcostofoutsourcingisinyoursituation.Gatherdataandgointoeach
newoutsourcedprojectwithasmuchinformationaspossible.Thatway,youwon’tbe
caughtoffguardbythecostofdoingbusinessoffshore.
Time,Distance,andCulture
Thecountrywheretheoutsourcingtakesplaceisnottheissue.Youcanfindhigh-
performingteamsallaroundtheworld.ThereasonIamnotaproponentofoutsourcing,
asidefromthehiddencosts,hastodowiththefundamentalchallengewithdistributed
teams.Successfulagileprojectsdependonhigh-performingteams.It’sdifficultto
functionasaunitwhenteamsarespreadoutacrosstimezones,don’tsharethesame
culture,andfeeldisconnectedfromoneanotherbydistance.
Thatbeingsaid,IhaveseensuccessfulScrumteamsthatoutsource.Justnotmany.The
onesthataresuccessfulhaveinvestedtime,money,andpatienceincreatinglong-term
relationshipsbetweenthelocalandoutsourcedteams,havetakenthetimetounderstand
eachother’scultureandpoint-of-view,andhavebeenwell-versedinagileprinciplesand
practicesbeforetheybegin.
Thenextsectionlooksatwaystoensurethatyouroutsourcedprojectrunsassmoothly
aspossible,bothforthelocalteamsandtheoffshoreteams.
KeystoSuccess
Thekeystosuccesslistedinthissectionareallbasedononeunderlyingideal:Nomatter
howfarapartyouare,doallyoucantomakeitfeellikeyou’reasingleteam,workingin
onecolocatedspace.
ChoosetheRightOffshoreTeam
Therearetwodimensionstochoosingtherightoffshoreteam:hiringagileteamsand
goingnorth/south.
HireAgileTeams
Whenyououtsourceanagileproject,hireateamthatisalreadyagile.Otherwise,you’ll
spendagreatdealoftimeandmoneytryingtoeducatetheoutsourcedteamonagile
practicesandprinciples,whichmightbeveryforeigntotheteam’sowncompanyculture.
Anddon’tdependonalineiteminamarketingpresentationthatsays,“We’reagile.”Go
watchtheteaminaction.Seehowtheteammemberswork,andmakesuretheirideaof
beingagileisconsistentwithyourlocalteam’sphilosophy.Andmostimportant,talkto
formerclients.Thiseffortoftengivesyouanideaofhowagiletheoutsourcedteamreally
is.
GoNorth/South
Ifyou’regoingtogooffshore,trytochooseateamwithinthreetimezonesofyourlocal
teams.Thisusuallymeansgoingnorth/southinsteadofeast/west.IfyouareintheUnited
Kingdom,outsourcetoPortugal,Spain,anAfricannation,ormaybeevenEasternEurope.
IfyouarebasedinNorthAmerica,looktoSouthAmerica.North/southreallysavesonthe
late-nightconferencecallsandthetimelostwaitingforanswers.Whentimezonesarea
factor,quickquestionsthatcouldbeansweredinfiveminutesinasharedteamspace
couldpotentiallyderailtheprojectforaday.
AllocatetheWorkintheLeastPainfulWay
Whenlookingtobreakupyourwork,youhaveseveraloptions.OnecommonapproachI
seeisforteamstouseworkpackagedeliverables.Aworkpackageisanallocationofa
specificfeatureorsetoffeaturesthattheoutsourcedteamworkson.Thoughyouneeda
localteamcoordinatortofacilitateandcoordinatethework,thisapproachgenerallydoes
notrequireentireteamsfrombothlocationstomeetonadailybasis.
Anotheroptionthatmanycompaniesuseistohavemultipleteamsworkonthesame
codebase.Whenconsideringthisoption,understandthatcertainengineeringpracticesand
standardsneedtobeinplace,suchasqualitygatesorcheck-inrequirements,notations,a
commondefinitionofdone,andmore.Byhavingteamsworkingonthesamecodebase
duringtheirnon-overlappingworkdays,youcanostensiblyaccomplishmoreworkinless
time.
AthirdapproachI’veseencompaniestryistooffshorespecificfunctionssuchas
testing,UI,datamodeling,orvaryinglayersofasystem.Thisisprobablytheworstofthe
optionsoutlinedherebecausethefeedbackloopsandintegrationloopsarehorrendously
long.
Thebestapproachwhenallocatingtheworkistogivetheoffshoreteamindependence
byprovidingthemwithacompletefeatureorfeaturesettoworkonandbuildendtoend.
It’salsothemostexpensiveoption,mainlybecausestakeholdersandcustomers(or
productowners)willhavetotravelandspendsignificantamountsoftimeworking
offshorewiththeteam.
StickwiththeScrumFramework
Whenyououtsourcework,don’tabandontheScrumpracticesthathavebeensuccessful
foryourcolocatedteam.Whileitmightbemoredifficulttocoordinatetheseactivities,
theyareessentialforlong-termsuccess.
DailyScrum
Continuetohavedailyscrums,preferablyviateleconference.Ifyou’vefollowedmy
adviceandstayedwithinthreetimezones,thisshouldberelativelypainless.Ifyou’reup
againsta12-to13-hourtimedifference,findingatimeduringworkinghourseachdayfor
bothteamstomeetisgoingtobedifficult.Sharethepainbyalternatingweeks:Oneweek,
onepartoftheteamisondutylateandthenextweek,it’stheotherpartoftheteam.In
otherwords,ifyouareinAsiaandyourteamisinCanada,you’llbeworkinglate,sayat9
p.m.,foroneweek.Thenextweek,you’llgettomeetatthemorenormaltimeof9a.m.,
whileyourteaminCanadagetsthelate-shiftdailyscrum.It’sessentialthatyoukeep
havingthedailyscrums,evenwhenit’sdifficultorpainful.
SprintReviews
Keepsprintreviewsliveviavideoconferencing.Don’tdemoviaemail.Scheduleatime
forademo,andthenalternatewhichteamhastohavetheawkwardtime.Useastaggered
approachaswiththedailyscrums.Foronesprint,thereviewisat9p.m.onaFriday,and
forthenextsprint,it’sat9a.m.onaFriday.Yes,peoplewillcomplain,especiallythe
stakeholders.Kindlyremindthemtheirsupportisneeded,thattheteamdealswiththis
daily,andthatthisisthebestwaytodealwiththerealityofoutsourcedprojects.
Retrospectives
Don’tskimponretrospectives.Infact,theymightbethemostimportanttoolyouhaveto
buildacohesiveteam.Youstillneedtoalternatethetimesandhaveatleast30to60
minutesofoverlappingretrospectivetime.Youmightalsochoosetohaveeachteamdoa
localretrospective,usinggrouptimetoreviewwhatwasdiscussedandtodeterminehow
toremovetheidentifiedimpediments/issues.
BuildaOne-TeamCulture
Inthischapter,I’vetalkedabouttheimportanceofawell-runtransition.Butifyourteam
isgoingtoremaininvolvedintheprojectwiththeoffshoreteam,youhavetofindwaysto
worktogetherlongterm.Starttheprojectwithanin-personkick-offmeetingwiththe
entireworldwideteam(don’tforgettoaddthisexpensetoyourgrowinglistofsoftware
costs).Youneedtoplanteam-buildingactivitiesandregularvisits.Youshouldalsobring
offshorepeopletothecentraloffice,andviceversa,onarotationalbasistobuildteam
cohesion.Atthesejointmeetings,worktowardbuildingacommondevelopment
environmentandcodebase.Thishelpswithpackingandunpackingthecode.
UsePairingtoCreateSharedUnderstanding
Itisessentialthattheteammembersknoweachotherandunderstandthecodebasethey
arecollectivelybuilding.Thebestwaytoaccomplishthisistohavededicatedpairing
sessions.Manytechnologiesexisttoenablelong-distancepairing(VNC/Live-
Meeting/WindowsLiveSharedDesktop,tonameafew).Usethem.Letmanagement
knowupfrontthatyouwillneedtoworkinpairsandthatthesepairswillbebasedon
work,notproximity.
Aswitheverythingelse,pairingismorechallengingwhenteamsdonotsharemany(or
any)commonworkhours.Evenifit’spainful,findtwohoursatleastthreetimesperweek
whenpairscanworktogetheronthecode.Aswithmeetings,sharethepainofstayinglate
orarrivingearlybyalternatingthetimeseachweek.
ContinuousIntegration
Pairingisagoodstart,butyoualsoneedawaytoknowhowthecodebaseisdoinginreal
time,whetheryourcodespanstheglobeorisaworkpackagedoneinasingleoffice.To
helpensurethateveryoneknowsthecurrentstateofthecode,maintainacontinuous
integrationserver(andinstituteagoodbranchingstrategy).Continuousintegration
becomesmoreofachallengewhenyoufactorinbandwidthissuesbetweencountries,
availablehardware,andthenumberofpeopleinvolved,sodon’tgointothislightly—
thinkthestrategythroughandfactorinthethingsthatcolocatedteamsmighttakefor
granted,suchasagigabitInternetconnection,readilyavailabletelecommunication
equipment,andelectricity.
KeepinConstantContactwithTechnology
TwothingsIinsistonwithoffshoreteamsareadedicateduplink,suchasan“alwayson”
phoneline,instantmessaging,orvideoportal.Theuplinkallowsaliveconnection
betweenteamsthroughouttheday.IfsomeoneinSeattlehasaquestionfortheteamin
Brazil,sheneedonlypickupthehandsetandaskthequestion.Itmimicsrollingbackin
yourchairandshoutingoutthequestiontothelocalteam.It’squick,it’seasy,andit’s
alwayson.
Thewebcamisanothermust-have.Eachteamshouldbeabletoseetheotherteam’s
members.Ifyouneedtotalktooneanother,it’seasytoseewhichpeopleareattheirdesks
andtotalktotheminstantly.
Anotheroneofmyfavoritetechnologytoolsisthepush-to-talkmobilephone,which
worksgreatforteamsthatareinthesamecountrybut,forwhateverreason,don’thave
webcamsoradedicatedphoneline.Itenablesthemtohavethatinstantaneous,
spontaneousconnectiontotheirteammates.
Ifyourcompanyfightsyouonimplementinganyofthesetechnologies,pushbackhard.
Remindthemthatitwastheirdecisiontooutsourcetheworkandthatthesetoolsare
essentialtotheproject’ssuccess.
BePreparedtoTravel
Travelisimperativewhenoutsourcingwork.Ifyourcompanyhasnotallocatedatravel
budget,don’tdooutsourcing.Itwillneverwork.Frequentin-personmeetingsaretheonly
waytoestablishatrueteamenvironmentandbuildtrustamongteammembers.
Travelisalsoanessentialtooltonipproblemsinthebud.Everycultureisdifferent,
evenwithinyourowncountry.Whenyouaredealingwithanothercountry,culture
differencesmagnify.Words,mannerisms,andevenbodygesturesinonecountrycanmean
somethingentirelydifferentinanother.Nothingdegradesteamtrustfasterthanhavingtwo
teamsthatarenotversedineachother’scultures.Whentrustissuessurface,takeaction
immediately,andtakeitinperson.Youmightevenhavetomovethewholeteam—pick
one—totheotherlocation.Getthemtoworktogetherforaweekortwotorebuildtrust.
Dothisfrequently.Yes,thiswillbeexpensive,butthetimewastedwhentwoteamsdon’t
trusteachotherisfarmorecostly,especiallywhenyourcustomersaretheonestosuffer
theconsequences.
HaveaProject/TeamCoordinator
Ifyoudodecidetohavemultipleteamsworkingwiththesamecode,helpreduce
problemsbyassigningeachteamaprojectcoordinator.Iworkedwithoneteamwhohad
itscodecircletheglobedaily.ThecodewentfromEasternCanadatoChinatoWestern
Europeovera24-hourperiod.Ineachlocation,adifferentteamworkedonit.Each
location’steamhadacoordinatorwhowasresponsibleforunpackingthecodeinthe
morningsotheteamcouldworkonitandthenrepackingitattheendofthedayto
transitionittothenextteam.Thecoordinatoralsorelayedthebacklogtothenextteam
andtrackedhisorherteam’sprogress.
Whilenotoptimal,havingaprojectcoordinatoroneachteamresponsibleformanaging
theflowofcodeenabledtheseteamstospendacombined14productivehoursonthe
codeperday.
NeverOffshoreWhen…
Therearecertaintimeswhenacompanyshouldnotoffshoreaproject,nomatterhow
carefultheattempt.
Don’toutsourcethefirstreleaseofcomplex,high-technology-riskprojects.
Don’toutsourceifyourlocalteamstruggleswithdevelopmentpractices(TDD,
continuousintegration,refactoring)orisundisciplined.
Don’toutsourcewithoutcorporatesupportforthetravelbudgetandtechnologies
necessaryforsuccess.
Don’toutsourcewithanoffshoreagilecompanyifyouhavenoexperiencewith
agileyourself.Youcan’tsucceedwithanoffshoreagileprojectifyouhaveyettodo
anonshoreagileproject.
AsmuchasIdislikethepracticesofoutsourcingandoffshoring,Iamrealisticenough
tounderstandthatitislikelysomethingeverycompanywillface.Doallthatyoucanto
fightit.Demonstratetherealcosts.Lookforpeoplewithinyourownorganizationwho
cansharetheirstruggleswithoffshoring.Presentawell-reasonedcaseforkeepingthe
projectlocal.
Whenallthatdatafailstoswaymanagement,don’tdespair.Withaheftytravelbudget,
awillingnesstousetechnology,afaithfuladherencetotheScrumframeworkandagile
engineeringpractices,andtherightoffshoreteam,youcansucceed.Itjusttakesalittle
moreeffort,alotofextrapatience,andseveralmillionfrequentflyermiles.
References
[COHN]Cohn,Mike.2010.SucceedingwithAgile.UpperSaddleRiver,NJ:Addison-
Wesley.
[GUNNERSON]Gunnerson,Eric.2006.“ScrumBut.”
http://blogs.msdn.com/b/ericgu/archive/2006/10/13/scrumbut.aspx(accessed3April
2015).
[OVERBY]Overby,Stephanie.2003.“OffshoreOutsourcingtheMoney:MovingJobs
OverseasCanBeaMuchMoreExpensivePropositionthanYouMayThink.”CIO
Magazine16(22):1.
[ROTTMAN]Rottman,Joseph,andMaryLacity.2006.“ProvenPracticesforEffectively
OffshoringITWork.”ITSloanManagementReview47(3):56–63.
[YU]Yu,Larry.2006.“BehindtheCost-SavingsAdvantage.”MITSloanManagement
Review47(2):8.
WorkConsulted
Saran,Cliff.“Badly-ManagedOffshoreSoftwareDevelopmentCostsFirmsMillions.”
ComputerWeekly.http://www.computerweekly.com/Articles/2004/06/15/203138/Badly-
managed-offshore-software-development-costs-firms.htm(accessed3April2015).
Chapter29.PrioritizingandEstimatingLargeBacklogs—
TheBigWall
You’vejustfinishedagreatstory-writingworkshopwithyourstakeholders.You’re
excitedaboutthenewproductandareanxioustogetstarted.Until,thatis,yougetbackto
yourofficeandlookatthemountainofstoriesthatyounowsomehowhavetoestimate
andprioritize.Allyourexcitementdisappearsasyoufacethehardtruth:Youhavenoidea
wheretostart.
Thesheersizeofnewproductbacklogscanbeoverwhelming.That’swhyyouneedto
attackthemwithsomethingevenbigger.I’mtalkingaboutaclubthat’sabout14feetwide
and10feethigh.Youneedtofindyourselfabigwall.That’swhatthefolksinthis
chapter’sstorydid—withexcellentresults.
TheStory
Gordonwastheproductownerforanew,two-yearproject.HehadworkedwithScrum
beforeandhadagreatrelationshipwiththeteamanditsScrumMaster,Lynne.The
problemwastheproductbacklog.Alreadyat300storiesandgrowing,andwithmorethan
20competingstakeholders,thebacklogwasgoingtobeabeartoestimateandprioritize.
AsGordonbegantoworkthroughthestories,hequicklybecameoverwhelmed.Hesought
outLynneforsomeadvice.
“Lynne,you’vegottohelpmeoutonthisone,”hesaid.“Ican’tgetahandleonthis
backlog.Ineedtoinvolvethestakeholders,butI’mafraidaprioritizationworkshopwith
alltwentyofthemwilljustdeteriorateintoabigfightthataccomplishesnothing.I’d
tackleitmyself,butit’ssohuge,Idon’tevenknowwheretostart.Anyideas?”
“Idohaveanidea,actually.WehadasimilarproblemontheJetproject,”saidLynne.
“Wehadhundredsofstories,acoupledozenstakeholders,andalotofquestions.”
“Whatdidyoudo?”askedGordon.
“Well,thefirstthingIdidwasprintoutallthestorycardsontosomethingthatwecould
use.You’vedonethatright?”askedLynne.
“Yup,gotthemallrighthereonindexcards,”saidGordon,pattinghisbag.
“Great.Next,weneedtofindaroomwithagreatbigwall,”saidLynne.Shegestured
towardtheconferenceroomdownthehall.“Thatonewilldo.”
Togethertheywalkeddowntotheemptyconferenceroomandlookedatthespace.
“Well,ithasabigemptywall,”saidGordon.“Andit’sfairlyeasytobook.DareIask
whyIneedabigwall?”
“You’llsee,”saidLynne.“Assoonasyoucangetitcoordinated,booktwomeetingsin
thisroom,preferablythesameday.Inthemorning,it’llbejustyou,me,andtheteam.In
theafternoon,youcanbringinthestakeholderstojoinus.We’llmeetbackherethen,and
I’llwalkusallthroughthesession.Bringyourstorycards.Ipromiseyou’llwalkoutof
herewithallthestoriesprioritizedandestimated.”
Aweeklater,Lynne,Gordon,andtheteammetintheconferenceroom.
LynneaskedGordontoreadthefirstfivestoriesaloudthenpassthemouttotheteam
members.Theteammembershadafewquestionsaboutacoupleofthestories,which
Gordonfielded.OnequestioncameupthatGordondidn’tknowhowtoanswer.
“That’sokay,”saidLynne.“Markitwithareddotandnotethequestionontheback.
We’llcomebacktoitthisafternoon.”
“Okay,”Lynnecontinued,addressingherteam.“Iwantyouguystoputthosefivecards
onthewall,bigstoriesgototheright,smallstoriestotheleft.Don’tworryaboutexactly
howmanypoints.Justthrowthemupthere.Ifithasareddot,makeyourbestguessfor
now.We’llrefinethislateranyway.”
Afteralittlehaggling,thecardswereinplace.Gordonreadthenextfivecards,andthe
teamplacedthemonthewall.Theteamsoongotintoarhythmandwasabletoquickly
placethecardsonthewallinspotsthateveryonefeltcomfortablewith.Afterthreehours,
allthecardshadfoundtheirwayontothewall.
“Okay,”saidLynne.“Wehaveanhourleftbeforelunch.Iwantustocorrelatewhatwe
seeonthewalltopoints.Let’sstartwiththesestoriesontheleft.Ifyouweregoingto
separatetheone-pointstoriesfromthetwo-pointstories,wherewouldyoudrawtheline?”
Theteammovedafewstoriesleftandafewtotherightuntiltheyfeltcomfortablethat
thestoriesonthefarleftwereone-pointstories.Lynnethentapedalinefromtopto
bottomtoseparatetheonesfromthetwos.
“Okay,”shesaid.“Howabouttwo-pointversusthree-point?Where’stheline?”
Again,theteamdebatedfortenminutesorso,andthensettledonadividingline.Lynne
addedanothertapedline.Theydidthisagaintodividethethree-pointstoriesfromthe
five-pointstories,thefive-pointstoriesfromtheeight-pointstories,andtheeight-point
storiesfromthethirteen-pointstories.Lynnethenaskedforonefinaldividingpoint.
“Whichstoriesarebiggerthanthirteen?”sheasked.
Theteamhadlittletroublegroupingallthebigstoriestogetheronthefarright.Lynne
addedonefinaldivider.ShereleasedtheteamforlunchandturnedtoGordon.
“So,we’vegotbasicpointsassignedtoallthestoriesinyourbacklog.That’simportant
becausehowmuchastorycostscouldwellaffectitspriority.Someofthehugestories
mightneedtobebrokendown,butthatwilldependonwhatwefindoutinparttwoofour
meeting,whenthestakeholderscomein.We’llhavethemmovethecardsintopriority
orderandaskthemanyquestionsthatyoucouldn’tanswer.”
“Won’titmessupourestimatesoncewestartmovingthesearound?”askedGordon.
“Nope,”saidLynne.“Wedeterminedsmallesttolargestbymovingcardsfromleftto
right.We’llillustrateprioritybymovingcardsupanddown.High-prioritystorieswillbe
highonthewall;low-prioritystorieswillbelowonthewall.We’lljustaskthe
stakeholderstokeepthecardsinsidethetapedlineswheretheyfoundthem.”
“Whatwilltheteamdoduringallofthis?”askedGordon.
“Theyshouldwatchandseewhatthestakeholdersmoveupanddown,andlistenasto
why.Someofthoseanswerswillinfluencethesequenceofthework.Theyshouldtake
notesaswell,andmightasksomeofthequestionsonthereddotcards.YouandI,though,
willhavetolimitthediscussiontoacoupleofminutespercard,tomakesureweget
everythingprioritizedbeforeweleave,”Lynneexplained.
“Andwhatdowedoifthestakeholdersdisagreeandgethunguponastory?Orkeep
movingthesamestoryupanddown?”
“YouandIwillwatchforthosetypesofdiscussions.Ifwecan’tresolvethedispute
fairlyquickly,we’llmarkanystoriesthathaveseriousconflictswithayellowdot,notethe
twostrongestdissentingstakeholdersontheback,andmoveon.Youcantakethe
conversationaboutthesemorecontroversialstoriesofflinewiththetwostakeholderswho
feltthemoststronglyaboutanyparticularstory.Thereshouldn’tbetoomanyofthose.”
Thestakeholderscameinafterlunchandwerebriefedontheprocess.Aftersomeinitial
hesitancyandafewtensemoments,theysoonweremovingcardsupanddowninsidethe
tapedlines.Whenstoriesweresimilar(butnotidentical),Lynneencouragedthemto
groupthestoriesneareachotherandmovethemtogether.Whenanyduplicateswere
found,Lynnetoldthegrouptofeelfreetodisposeofone,orhideonebehindtheother.
Afterafewhours,theprioritizationwascomplete.LynneandGordonthankedeveryone
fortheirparticipationandpromisedtofollowupontheyellowdotcards.
“Justonemorethingbeforewego,”saidLynne.“Doyouseethesehigh-prioritystories
overhereonthefarrightofthewall?”
Everyonenodded.
“Thesearestoriesthatyouwanttohavedoneveryquickly,yettheyaresobigwecan’t
evenputestimatesonthemyet.Iproposewespendalittletimewiththesestories,
breakingthemintomoremanageablepieces.Wecanthenre-estimateandreprioritizethe
smallerstories.Thereareonlyabouttwentyofthem,soitshouldn’ttaketoolong.
Twohourslater,GordonandLynnesatbackintheirchairsandstaredatthewall.The
lastofthestakeholdershadjustleftandtheywerefinallyalone.
“Well,Ipromisedyouanestimated,prioritizedbacklog,”saidLynne,smiling.“What
doyouthink?”
“Ican’tbelievehowwellitworked!”saidGordon.“Everyonefeelslikethey’vehada
say,Ionlyhaveafewoutstandingissuestodiscuss,andwe’rereadytogetstartedwith
ourfirstsprint.Ifeelliketheweightoftheworldisoffmyshoulders.HowcanIthank
you?”
“Don’tthankme,”saidLynne.“Thankthewall.Thereallybigwall.”
TheModel
Workingonbigbacklogscanbechallenging.There’sjustsomuchtokeeptrackof!And
whenthebacklogisraw(notestimatedandprioritized),itcanbehardtoknowwhereto
begin.WhilePlanningPokerisafantastictoolforestimatinguserstories,it’sdifficultto
imaginetacklingapileofhundredsofstories,oneatatime.That’sexactlywhatGordon
wasfacinginthestory.Hewasoverwhelmedbythesheersizeoftheeffort.
Lynne’ssuggestion,andmyproposal,istouseatechniqueIcallTheBigWall.It’s
similartoLowellLindstrom’sAffinityEstimation[MAR],whichIwasintroducedtoin
2008,withonedistinctdifference:TheBigWalltechniqueallowsyoutoconsiderboth
sizeandpriority.
TheBigWallisaquickandeffectiveestimationandprioritizationtechniquethat
enablesteamsto,atleastinitially,riseabovediscussionsoftwoversusthreeandfive
versuseightandjustgroupitemsinapurelyrelativemanneralongacontinuum.Itenables
stakeholderstoassignageneralprioritizationtoalargegroupofstorieswithoutgetting
hunguponwhetheronestoryisslightlymoreimportantthananother.Theonlytoolsyou
needareastackofuserstoriesfromtheproductowner,sometapeandmarkers,andabig
emptywall,about14feetlongby8to10feethigh.Thenyouneedtogatheryour
stakeholdersandteamandplantospendadayorsogettingphysicallyinvolvedwiththe
stories.
Widthisreservedforsize.Storiesontheleftaresmallerandstoriesontherightare
bigger(orifyou’rein,say,Japan,youcanreversethisandmovefromrighttoleftifit
seemsmorelogical).Teammembersshouldaskthemselveswhere,relativetotheother
stories,doeseachonefit?Onceallofthestoriesareonthewall,they’llworktogetherto
addabitmorespecificity,dividingthestoriesintovariouspointgroupsusingverticallines
oftape.
Heightisreservedforpriority.Storiesatthetoparehigherpriority;storiesatthebottom
arelowerpriority.Afterthestoriesareestimated,thestakeholdersdeterminetherelative
priorityofthestoriesbyplacingthemhigherorloweronthewall.Astory’sprioritycan
bebasedonreturnoninvestment,businessvalue,orsomethingasvagueas“it’sjust
importantandIdon’tknowwhy.”
Ifyoualreadyhaveanestimatedbacklog,justdotheprioritizationportionofthis
exercise.Iftheproductownersandstakeholdershavealreadygivenyouaprioritized
backlog,justdotheestimationportion.Yourproductownerwilllikelywanttorevisitthe
prioritizationaftertheestimatesaredone.Afterall,costhasabigimpactonpriority.If
youareliketheteaminthischapter’sstory,youcandobothpartsoftheexercise.The
nextsectionsexaminehowthisexerciseplaysout,beginningwiththeteam’srole.
Team
Givenarawproductbacklog,youshouldstartwithestimation.Onewaytoperform
estimationistoemulatetheteaminthischapter’sstoryandstartwithablankslate.
Instructtheteamthatthefarleftofthewallshouldholdthesmalleststories,andthefar
rightshouldcontainthebiggeststories,withoutregardtonumbers.Theteamputsstories
somewhereonthewallbasedonthosetwopoles.Theadvantagetodoingitthiswayis
thatthereisnopreconceivednotionofwhatatwo-orthree-pointstoryis;it’struly
relativebasedonhowbigthewallis,whichiswhyit’snicetohaveareallybigwall.
Theteammembersmightfeelmorecomfortableiftheygivemorestructuretothewall.
Theycandosobyidentifyingreferencestories.Asktheteamtoidentifysomereference
stories:arepresentativeone-point,two-point,three-point,five-point,eight-point,and
thirteen-pointstory.Don’tworryaboutcreatingabiggerreferencestory;anythingbigger
willusuallybebrokendownasitrisesinpriority.Findingreferencestoriesshouldtake
lessthananhour.
Oncetheteamhasidentifiedthefivereferencestories,placethosestoriesonthewallin
thelocationsthatreflecttheirsizes—smallertolarger,movingfromlefttoright.Leave
someroomontherightsideofthewallforstoriesbiggerthaneight.Askteammembersto
taketheremainingstoriesandcomparethemtothesizeofthereferencestories,andthen
placethemintheappropriateareaonthewall.Forinstance,ifastoryseemstobeabit
smallerthanafivebutmuchbiggerthanathree,putitjusttotheleftofthefive-point
referencestory.Ifitseemsbiggerthanafivebutmuchsmallerthananeight,putitjustto
therightofthefive-pointreferencestoryandleftoftheeight-pointreferencestory.The
closerinsizeitistothereferencestory,thecloseritshouldbeplacedtothereference
story.Thereisnorealsciencetothis,it’sjusttheteamlookingandsaying,“Thisismore
thanatwo,butit’snotathree,soitbelongsinthemiddle.”Thatisgoodenough.
Whicheverwayyouchoosetogetthere,onceyou’veplacedthestoriesonthewall,
discusswiththeteamwherethelogicalbreaksarebetweenstorysizes.AsLynneand
Gordondointhestory,putalineoftapewheretheteamdeterminesthestoriesseparate
intopointgroups.
Atthispoint,you’llhaveawallthatlookssomethinglikeFigure29-1.Note,however,
thatyoumighthavemoreverticalsections,dependingonthenumberofstory-point
categoriesyou’veidentified.Nowthatyou’veestimatedyourstoriesintermsofsize,the
stakeholders,withyourassistance,willneedtoplacethestoriesinpriorityorder.
FIGURE29-1Storiesarrangedbysize
Stakeholders
Yourcustomersandstakeholderswilllookatthestoriesdifferentlythantheteamdid.
They’renotasinterestedinhowmanypointsastoryhas.Theyaremostinterestedin
findingthestoriesthatrelatetothemandmakingsurethosestoriesgetdone.Thismayput
thestakeholdersinconflictwithoneanother.Thisisokay—infact,it’sidealbecauseit
reflectsrealityandhelpstheproductownermaketoughdecisionsaboutstoryordering.
ThefirstthingIdoisexplaintothestakeholderswhywearehereandwhatwewantto
accomplish.Thislittlespeechgoessomethinglikethis:“Ihavemetwitheachofyouand
havecreatedasetofstoriesthatreflectyourwantsandneeds.Theteamhasspentthe
morninggroupingtheseintorelativesizes,whichgivesyousomeideaoftherelativecost
ofeachstory.Thestoriestotheleftaresmallest(andcheapest).Thestoriestotherightare
thelargest(andmostexpensive).WhatI’dlikeyoutohelpmedoisdeterminetherelative
priorityofallthesestories.I’mgoingtoaskyoutomovethesestoriesupordowninside
thetapedlines.Thehigherupastoryisonthewall,thehigheritsprioritytothebusiness.
Ifyouplaceastoryatthetop,Iwillaskyouforjustificationastowhyit’supthere.You
mayalsoaskeachotherwhyonestoryismoreimportantthantheother.Intheend,we’ll
knowhowallthestoriesrelateandhaveanideaofwhatfunctionalitywe’llhaveearlyin
theprojectandwhatfunctionalitywillhappenlater.Theteamwillobserveandmayask
questionstohelpthembetterunderstandastory.”
Remindthestakeholdersthatiftheymoveastoryloweronthewallthansomeoneelse
did,theyneedtomarkitwithayellowdot.Thisalertseveryonethatthegroupmightneed
tohaveaconversationaboutthestory’struepriority.Ialsoencourage/expectpeopleto
ask,“Whomovedthisonedownorup?”Ortosayaloud,“Ithinkthisoneneedstomove.
Whowantstodisagree?”Thisenablesaconversationtohappenbetweentheinterested
partieswithoutfacilitation.Ifadiscussiongoesontoolongwithoutresolution,the
facilitator(usuallytheproductowner),shouldcollectthecard,notethetwostakeholders
whocannotagree,andmakeanotetomeetwiththemprivatelylater.
Althoughtheteamisintheroomforthisexercise,theyarenotactiveparticipants.They
areobservers,takingnotesonthebehavior,theinteractions,andthereasonsthatcertain
storiesarerisingorfallinginpriority.Theycanalsoansweranystakeholderquestions,if
needed.Iftherearestoriesthattheteamcouldn’tsizewithconfidencebecausethose
storiesrequiredanswersfromaparticularstakeholder,theteamcanalsoaskquestions
aboutthosestories,astimeallows.
Thegoalofprioritizingasagroupistohelpallstakeholdersunderstandtherelative
prioritiesofvariousstoriesandhowtheothersviewthemaswell.Theexerciseshould
takebetweentwoandsixhours,dependingonthenumberofstoriesandthenumberof
stakeholders.TheendresultshouldlooksimilartothepictureshowninFigure29-2.
FIGURE29-2Prioritizedandestimatedbacklogonthewall
Whenallofthesizingandprioritizationisdone,usetapeorstringtodividethewall
visuallyintofourquadrants,asshowninFigure29-3.
FIGURE29-3Thefourquadrants
Thestoriesinthetop-leftquadrantarehighpriorityandsmall.Thesestorieswillendup
inthetopoftheproductbacklog.Thestoriesinthetop-rightarehighpriorityandlarge.
Youshouldbreakmostofthosedownintosmallerstoriesbecausethey’llbecomingupin
thefirstseveralsprints.Asthestoriesarebrokendown,youmayfindthatsomearehigher
prioritythanothersandthatafewmightevenmovetoadifferentquadrant.Thisshiftis
expectedandnormal.
Thebottomquadrantsarelessurgent.Thelower-leftquadrantismadeupofsmall
storiesthatarelowerinpriority.Thelower-rightquadrantisfilledwithlargestoriesthat
arelowerinpriority.Thesestoriesaretypicallyyourepicsorthemes.They’lleventually
needtobebrokendownintosmaller,moremanageablestories,butnotuntiltheyrisein
priority.
Leavesometimetoreexaminethewallasawhole.Ifstoriesareinthewrongquadrant,
movethem.Ifafewstoriesneedtobebrokendown,trytodoitwitheveryoneinthe
room.Rememberthatthisexerciseisjustafirststepindevelopingthepriorityorderingof
thebacklog.Itcanandwillchange.What’simportantistoestablishtheinitialorderand
thenallowtheknowledgeyougainthroughsprintsandfeedbacktoinfluenceandalterthe
orderasneeded.
Bythetimeyou’refinished,thestakeholderswillbeabletoseethestartofarelease
plan.Ifyouknowtheteam’shistoricalvelocity,youcanevengivethemaroughrangeof
whichstoriesintheupper-leftquadrantwillbefinishedandbywhen(seeChapter4,
“PredictingTeamVelocity,”andChapter11,“ReleasePlanning”).
Don’tbeoverwhelmedbyarawproductbacklog.Youcanestimateandprioritizeeven
themostout-of-controlproductbacklogfairlyquicklyifyouhavetherightpeople,alotof
inkandpaper,sometape,alittletime,andareallybigwall.
KeystoSuccess
Usingabigwallseemslikeafairlyeasytechnique,anditis,butitdoeshavesome
drawbacks.Topullthisexerciseoffsuccessfully,youneedtoconsiderandbepreparedfor
preplanning,settingtimelimits,parkingunresolvedissues,creatinganddestroyinguser
storiesonthefly,andremindingeveryonethatnothingissetinstone.
PreplanningIsEssential
Noonelikestobesurprisedbyameeting’sagenda.Ifyourstakeholderswalkintoaroom
unpreparedforawallfullofindexcards,theymightjustturnandwalkrightbackout
again.Whenyou’rewritingthemeetinginvitation,explainthatitisaworkshopofsorts.
Takethetimetocommunicatethepurposeandinteractivenatureofthemeeting,the
desiredoutcome,anynecessarypre-work,thescheduledbreaks,andthegeneralrules.A
littlepreparationgoesalongwaytowardminimizingtheamountofhandholdingneeded
duringthemeetingandensuresthatallparticipantscometotheworkshoppreparedfor
whatthey’llencounterthere.
FocusDiscussionsandSetTimeLimits
Oncetheygetstarted,peoplelovetotalk.Whenestimating,developerscangooffintothe
weeds,discussingthemeritsofthisortheotherapproach.Whenprioritizing,stakeholders
canspend30minutesarguingoverwhyaparticularstoryisofmorevaluethananother,
especiallywhentheyarepoliticallymotivated.Yourjobistonoticethesetangentsand
redirecttheconversationtothetopicathand.
Whenteammembersgetstuckonimplementationorarchitecturalissues,bringthe
focusbacktotherelativesizeofthestory.Remindthemthatyou’relookingforarelative
size.Ifastoryistoobigortooconfusingtoestimate,theyshouldgiveitthelargestvalue
availabletothem,whichisappropriateforstoriesthatneedtobebrokendownor
discussedpriortobeingworkedon.
Whenstakeholdersstartdebatingmarketingstrategiesorcustomersatisfactiongoals,
remindthemtoconsidertheproductasawhole.Itmighthelptostructurethediscussionin
termsoftheminimalmarketablefeaturesets—theminimumamountoffunctionalitythat
anenduserneeds—forspecific(requested)deliverywindows(asintroducedinChapter
26,“Up-FrontProjectCosting”).ConsideraswelltheMMR,orminimalmarketable
release[ANDERSON].Theminimalmarketablerelease,asidentifiedbyAnderson,is“a
setoffeaturesthatarecoherentasasettothecustomerandusefulenoughtojustifythe
costsofdelivery.”
Don’tlethealthydebateturnintoatime-eatingtangent.Keeppeopletalking,butmake
surethey’restillfocusedonthetaskathandandthattheirconversationsaremovingthe
processforward,notstallingitunnecessarily.Ifallelsefails,useaneggtimer(notaphone
becauseitwillcausedistractionsforcheckingemail/texts/orplayingSimCityinsteadof
watchingtheparticipants).
UseaParkingLotforUnresolvableDisagreements
Aparkinglotisaplacetoputstorieswhenthediscussionsseemtobeendlessand
unsolvable.Afterall,it’snotthatthediscussionsorargumentsarebad;mostofthemare
valid.Theproblemisthatpeoplehaveahardtimeleavingtheargumenttofocusonthe
largerbacklogorbiggerpicture.Whenthishappens,Iliketoputcontentious
discussions/storiesintheparkinglottorevisitlater.Aparkinglotistypicallyrepresented
byaflipchartorsmallsectionofawhiteboardthatisdesignated“ParkingLot.”Itisa
temporaryplacetostoreindexcardsorwritetopicsthatneedtobeparkedfornowand
pickeduplater.Useitanytimeaprocessbogsdownoveranargumentoroff-topic
discussion.
Iftheteamissimplyunabletosay“it’salittlebigger”or“it’salittlesmaller,”andits
membersreallyneedtimetogointotheweeds,parkthestorybyphysicallyrelocatingthe
indexcard.Ifthestakeholdersneedtositdownwithyoutoreallydigintotherelative
priorityoftwostories,parkthosestoriesaswell.Ifatopiccomesupthatisinterestingbut
doesn’tmoveprioritizationforward,parkitandmoveon.
Thestoriesandideasintheparkinglotcanbehandledattheendofthemeetingoreven
afterthemeetinghasbeencompleted.Afewunresolvedstoriesshouldnotdistractfrom
thehugepileofunsortedstories.
BringExtraCards/PaperforStoriesCreatedintheRoom
Bringingsufficientsuppliesmightseemlikeano-brainer,butwhenyou’refocusedonthe
bigpicture,youtendtoforgetthelittlethings—extracards,pens,stickynotes,andpaper.
Understandthatyou’regoingtoberippingupstoriesandwritingnewonesduringthis
exercise,somakesureyouhavetheextrasuppliesneededtodojustthat.
Aboveall,makesurepeopleknowthatit’sokaytojustdisposeofastoryifitneedsto
berewrittenorbrokendown,duplicatesanother,orthegroupdecidesitjustdoesn’tneed
tohappen.Storiesaren’tgold.They’reideasonpaper.Thegroupshouldfeelfreeto
destroysomeandaddothers.
TheBigWallwithRemote/DistributedTeams
TheBigWalltechniqueisnotlostondistributedteams.Ithasbeenturnedintoanonline
InnovationGame®andcanbeaccessedathttp://www.innovationgames.com/mitch-lacey-
team-prioritization/.
RemindEveryoneThatThingsWillChange
Remember,astory’ssizeorpriorityisnotalife-or-deathdecisionandnothingiscarvedin
stone.Storiesmightberesizedandprioritiesmightshift.Theproductbacklogisaliving
documentthatmustbegroomedeachsprint.Initialestimatesandprioritiesareimportant
tocapture,buttheyareonlystartingpoints.Iftheteamandstakeholdersgethungup,
remindthemofthiscrucialfact:Thisisapreliminaryestimate(orpriority).Wecanrevisit
itlateraswelearnmoreandastheproductdevelops.
Theworkloominginsidealarge,rawbacklogdoesn’tneedtoweighyoudown.Witha
littlecooperationfromtheteammembersandstakeholders,awholelotofpatience,anda
wallbigenoughtoholdallofthestories,youcanbringthatbacklogdowntosizeinjust
oneday.
References
[ANDERSON]Anderson,DavidJ.2010.Kanban.Sequim,WA:BlueHolePress,p.152.
[MAR]Mar,Kane.2008.“ScrumTrainersGathering(4/4):AffinityEstimating.”
Personalwebsite.http://kanemar.com/2008/04/21/scrum-trainers-gathering-44-affinity-
estimating/(accessed4April2015).
Chapter30.WritingContracts
Everyoneknowsthestandardcontractingmodel—customerswanttoseeinwritingwhat
they’llget,atwhatcost,andbywhatdate.ButhowdocontractsworkwithaScrum
deliverymodel?Afterall,Scrumteamsdeliberatelyavoidfixedscope,cost,anddate
contracts,preferringtofixonlyoneofthoseelementsandleavetheothersvariable.So
howcanScrumteamsworkwithclientswhoexpecttohaveallthreeelementsfixedina
contract?Inthefollowingstory,twocolleaguesfaceacustomerwhohasbeenburnedby
fixed-everythingcontractsinthepast,andtheyproposeapay-as-you-gosolutiontoearn
thatcustomer’sbusiness.
TheStory
JulioworkedforalargeconsultingcompanybasedinLosAngeles.Julio,astheaccount
manager,helpedprocurebusinessandwritecontractsbutwasnotthatexperiencedin
Scrum.Onenightatadinnerparty,Juliometandchattedwithapotentialnewclient
namedRichard.HediscoveredthatRichard,adirectorofalargetalentagency,wasdeeply
dissatisfiedwithhiscurrentbookingsoftwarebutwasleeryofpayingforyetanother
systemthatdidn’tmeethisneeds.
“Julio,”Richardsaid,“I’msureyou’vegotagreatshop.Butyoucan’tguaranteeme
thatIwon’tbewastingmymoney.EveryonesaystheyknowwhatIneed.Butintheend,I
spendasmallfortuneforsoftwarethatisn’tright.We’rejustgoingtolivewithwhat
we’vegot.”
“Iknowwhatyou’retalkingabout,andIknowhowfrustratingitistopayfor
somethingthatdoesn’tworkthewayyouwantitto.Wedothingsdifferently.Whatwe
candoisdeliverasmallportionofyoursystematatime,andofferyoutheopportunityto
payonlyforthesoftwareyouaccept.Becauseyou’repayingasyougo,yourriskis
minimal.CanwemeetatyourofficenextweeksomycolleagueandIcanexplainwhat
weofferthatothersdon’t?”askedJulio.
Richardagreedtothemeeting.Thenextday,Juliospoketohiscolleague,Rob.Robhad
manyyearsofexperiencerunningScrumprojectsandhelpeddelivertheworkJulio
broughtin.
“IthinkIknowhowwecanapproachthis,”saidRob.“I’vejustfinishedworkingwitha
cruisecompanyinLongBeachonabookingwebsite.They,likeRichard,hadbeenburned
beforeandwereprettyskepticalthatwecoulddeliverwhatwepromised.Wewereableto
delivernewfunctionalityeverytwoweeks.Aftereverydelivery,theclientsreviewedthe
workinprogressandhadtheopportunitytoadjustwhatwedevelopednextbasedonthe
newestfeaturesontheircompetitor’swebsite.Theywereextremelyhappy.”
“That’swhatI’vetoldRichard,”saidJulio.“Wecangivehimimmediateresults.I’d
likeyoutogowithmetomeethimandhelpexplainhowScrumworks.We’lltellhim
abouthowwithScrumwecandelivertwiceasfastforhalfthecostasanyoneelse.This
willreallyhelphimsolvehisproblems,”explainedJulio.
“Whoa!Holdup!”exclaimedRob.“Wecan’ttellRichardthatScrumisgoingtosolve
allhisproblemsbecauseitisn’ttrue.Scrumisaframeworkthathelpsusdeliverworking
softwaretoourclients,sure,butitisn’tsomemagicpixiedustthatcanbesprinkledon
yourproblems.Ifanything,ithighlightsproblemsfasterandforcespeopletofixthem—
that’sthebeauty.”
“Idon’tthinkIcansellRichardonhowwe’regoingtocomeinandexposehis
problems!He’sstressedoutenoughasitis,”repliedJulio.
“That’snotwhatImeanbyhighlightsproblems.Forexample,Iwasabletoshowour
managementhowhavingpeoplesplitacrossmultipleprojectswaswastingresourcesand
loweringquality.Asaresult,Igotadedicatedteam—andourqualityandproductivityare
improving.WeneedtobringthissamesortofawarenesstoRichard.However,tosay
Scrummakesthingslikethishappenisavastoversimplification.Atthesalescall,letme
clarifyhowScrumworksandyoujustclosethedeal,”saidRob.
“Okay.Iseeyourpoint.I’llletyouexplainhowwewillrunthisprojectusingScrum.If
Icangethimtocommit,Imightneedyoutohelpmeputtogetheracontractthatwill
assureRichardhe’llactuallybegettingsomethingforhismoneywhilegivingussome
reassurances,too.”
Thedayofthemeetingarrived.Aftertheinitialpleasantries,Richardgotrighttothe
point.“Let’scuttothechase.Whatcanyoudothatthislistofothervendorshasfailedto
do?”askedRichard,pointingtoawhiteboardwherehehadwrittenthenamesofthepast
consultingcompanieshehadworkedwith.
Robjumpedrightin.“Wecangiveyouthefreedomtochangeyourmindwithout
chargingyouafortune.”
“How’sthat?”askedRichard.
“Iworkedforoneoftheconsultingcompaniesonyourwallforacoupleofyears.Ileft
becauseInoticedthatwealwaysfailedtomentiononekeycomponentintheproject:It’s
software—thingschange!Andcontractsshouldallowforthatchange.Instead,most
contractsdotheopposite.Mostcompaniesbidthelowestpossiblepriceandthenchargea
fortuneforeverychange.Thesearesoftwarecompanies.Theyknowthatchangeis
inevitable.Infact,theycountontheircustomerschangingtheirminds(andtheresulting
scopechangesandchangeorders)tomakeadditionalmoneybeyondtheoriginalcontract
price.Butyouarenotasoftwarecompany;you’reatalentagency.Youexpecttopay
somethingclosetothecontractpriceforsoftwarethatworksthewayyouwantitto.
Instead,youenduppayingalotmorethanyouthoughtyouweregoingto,andyou’retold
it’syourfaultbecauseyou’retheonewhoaskedforallthechanges.”
“You’vehitthenailonthehead.So,howareyoudifferent?”askedRichard.
“You’vegotabunchofsoftwarethatdoesn’tworkthewayyouenvisioned,right?”
askedRob.
“Right.”
“Inthepast,atthebeginningofeachproject,youdescribedhowyourcurrentsystem
wasnotworking,talkedaboutafewthingsyouwantedinstead,andansweredawhole
bunchofquestions.Thenthecompaniesonthiswallwentaway,didsomesoftwaremagic,
anddeliveredyouabrandnewsystemseveralmonthslater.Isthatprettyaccurate?”
“Yep,”Richardanswered.
“Andonceyoustartedusingthesystem,itdidn’tdowhatyouneededinthissection,
didn’tworklikeyouhadpicturedinthisplace,andjustdidn’tmeasureupoverall.You
paidforsomecostlychangestotrytofixit,butintheendwhatyougotwasaclunky,
buggysystemthatresembledwhatyouaskedforbutisnothinglikewhatyouwanted,or
meant.”
“Exactly.Everytime,oncewesawtheendresult,weknewitwasn’tquiterightand
couldthendescribewhatweneededinstead.Butwewereoutofbudgetandfranklyoutof
patience,”Richardexplained.
“You’renotalone.Peopleknowwhattheywantwhentheyseeit—butit’sdifficultto
describeandevenmoredifficulttothinkofalltheangles.”
Richardnoddedinagreement.
Robcontinued,“OneofourcustomersisabigcruisecompanyinLongBeach.
Everyonethereknewthecurrentbookingsystemwasnotworking,butnoonecouldagree
onwhatexactlyneededtobebuiltoreventheproblemsthesolutionwassupposedto
solve.Itwasabigmess.Wecouldhavetakenamoretraditionalapproachtobuildingthe
system,whichIthinkyou’refamiliarwith,Richard.”
“Oh,youmeanmepayingandthevendornotdelivering?I’mveryfamiliarwiththat
model,”saidRichard.
Juliolaughedandinterjected,“Whatwedidinsteadwaswriteatwo-stagecontract.The
firststageissimilartoatraditionaldiscoveryphaseinaproject.Thesecondisapay-as-
you-godevelopmentplan,withdeliverieseverytwoweeks.”
“Whatdoyoudointhediscoveryphase,andhowlongdoesitlast?”askedRichard.
“It’safixed-costphasethatlastsabouttwoweeks,”answeredJulio.
Robexplained,“Duringdiscovery,Iwillcometoyoursitefortwoweeks.I’llbring
withmetwopeoplefromtheteamwhowilldotheworkforthesecondphase—ifweget
thework,thatis.Duringthattime,thethreeofuswillintervieweveryonewithanysayin
howthesoftwareshouldwork:users,stakeholders,accounting,younameit.”
“Accounting?Whyintheworldwouldyouneedtointerviewthemforanengagement
bookingtool?”askedRichard.
“Theyholdthemoneyandwanttoforecastrevenueonfutureengagements,”explained
Julio.“Thatmakesthemastakeholder.Wealsoneedtounderstandtheirexpectationswith
regardtocapitalexpenditureandtheirimpressionofITshops.”
“Interesting,”ponderedRichard,scratchinghischin.
“Oncewe’vemetwithalltheinvestedparties,we’lldevelopasetofuserpersonasand
usethattobuildaproductbacklogwithuserstories.Thelastthingwedobeforetheendof
thefirstphaseisestimatetheproductbacklog,”saidRob.
“Backup,”saidRichard.“Userpersonas?Userstories?Productbacklogs?Idon’tknow
whatthesethingsare.”
RobandJuliospentsometimeexplainingtheseconcepts,thebasicScrumartifacts,and
theconceptofagileestimation.Afterawhile,Richardcouldseetheconnectionbetween
endcustomervalueandstories,andhowtheychangebasedontheemergingfunctionality
ofthesystem.
“So,”concludedRob,“attheendofthediscoveryphase,we’llgiveyouanestimated
productbacklogandapotentialreleaseplan.Theplanwillbebasedonhowmuchwork
theteamwilllikelydelivereachsprint,atacertaindegreeofconfidence,and,ofcourse,
theinformationyouhavegivenustodate.We’llofferyoutheabilitytochangeyourmind
withoutpenalty,aslongasyouswapitemsofequalsize.You’llhavetheestimatedcostof
thefunctionalityandadetailedplanforwhatwebelievewecandeliver.”
“Whathappensattheendofthatdiscoveryphase?IstherearequirementthatIuseyour
companyforphasetwoifIdon’tlikewhatIsee?”askedRichard.
“Absolutelynot,”answeredJulio.“Ourjobinthefirstphaseistobuildabacklogwith
userpersonasandstories.Thatisthedeliverable,alongwithourreleaseplan.Youcan
takethatbacklogtoanumberofothervendorstogettheirbidsifyoulike.Atthatpoint,
wearesimplyanothercompanycompetingforyourbusiness.Youmayfindourcoststo
bealittlehigherupfront,butwedon’tgougeyouforchangesintheend.”
“Whathappensifyou’regreatatmakingaplanbutnotsogoodatfollowingthrough?
WhatifIchooseyoutobuildtheproductandyoudon’tdeliver?”askedRichard.
“Ithinkthisisthebestpart,”explainedJulio.“Thecontractwillsaythatifyouarenot
happywiththequalityofwhatwebuild,yougiveusonesprint’snoticeandwearedone.
Itwillalsosaythatifatanytimeyoudecideyouhaveenoughfunctionality,yougiveus
onesprint’snoticeforustoclosetheprojectdown.It’strulyapay-as-you-goplan.Oh!
Onemorething.Ifyoudonotacceptthework,meaningyoudon’tagreethatwhatwe
builtinthesprintwaswhatyouaskedfor,youdon’thavetopayforthatsprint.”
“So,whatstopsmefromsayingIdon’taccepttheworkandgettingitalldoneforfree?”
askedRichard.
“Trust,”saidJulio.“It’sassimpleandascomplicatedasthat.You’llgointophasetwo
knowingthatweareworkinginyourbestinterests—we’veworkedtounderstandyour
businessandaretheretohelpyousucceed.Wewantlong-termpartners,notone-off
customers.Wemaynothavethelargestcustomerbase,butwehavethemostloyal.Weare
strategicaboutwhowehireandthecustomerswetakeon.Ifpotentialcustomersdonot
aligntoourcompanyvalues,wereferthemtoalargerconsultingshop.It’sthatimportant
tous.”
Intheend,RobandJuliowereabletoworkwithRichardtostructureacontractthathe
wasconfidentinandtheycouldalltrust.
TheModel
Customerswantmaximumfunctionalityandqualityatthelowestpossiblecostandinthe
shortestpossibletime.What’sworseisthattheyexpectITprofessionalstoguaranteeall
thesefactors,with100percentconfidenceatthebeginningoftheproject.Whocanblame
them?LikeRichardinthischapter’sstory,customershavebeenburnedtoomanytimesby
failedprojects,costoverruns,missingfunctionality,andbug-riddencode.
Sowhycan’tteamsdeliverwhattheypromise?Tworeasons:changeandtiming.
Beforegettingintowritingcontractstosupportagileprojects,itwillhelptounderstand,
andcomecleanabout,howalotoftraditionalcontractsarewritten.I’vebeenguiltyinthe
pastofdoingwhatI’vewritteninthefollowingsection,“TraditionalContractsand
ChangeOrders,”andI’veseenthisrepeatedhundredsoftimes.
TraditionalContractsandChangeOrders
CustomersarenodifferentthanyouandI.Peoplewanttoknowwhattheyaregettingfor
theirmoneyandwhenitwillbedelivered.Becauseofthis,Iseemosttraditionalcontracts
startinglikethis:
1.Companies/customerspublishorsendarequestforaproposal.ThisRFPlistsallthe
features/workthatthecustomerdesires.Inmostcases,theworkwillnotbein
priorityorder—itallneedstobedone.
2.Thereisashortquestion-and-answerwindowwhereyou,thecontractor,canask
detailedquestions,fleshoutassumptions,andattempttodiscoverthebudget.Asyou
workonyourproposal,youwillnaturallytrytofityouroverallprojectcostintothe
customer’sbudgetrange—ifyoudon’tknowtherange,you’llmakeyourbestguess.
3.Yourfinalproposalwillcomeinascloseaspossibletothelowerendofthebudget
(or,ifyoudon’tknowthebudget,lowerthanwhatyouthinkyourcompetitors
wouldsubmit)andwillincludeeverythingthecustomersthinktheywant.Youdo
thissothatyoucanwinthebusiness.Figure30-1depictsanactualcustomer
contract—imaginetryingtoreadit,letaloneunderstandit.
4.Yourproposalmightincludeyourchange-orderprocess.Itshould,asit’sakey
componentofhowyou’llhandlechangeswhentheysurface.Youknowgoinginthat
changeordersmightbetheonlywaytobreakevenormakemoneyoffthelowbid
youofferedtosecurethebusiness.Why?Becauseeveryoneknowsthatthere’sno
waytouncoveralltheassumptionsandfigureoutwhatthecustomerreallymeantto
askforinthesmall,pre-executionRFPwindow.Thingsaregoingtochange.
Now,changeorderprocessesareinterestingbeasts.Theyareabsolutelynecessaryona
traditionalcontract,yetthemereactofhavingtoaskforchangesandgoingthrougha
complicatedprocesswillfrustratecustomersand,overtime,degradetrust.
FIGURE30-1Thistraditionalcontractweighedmorethan100pounds/50kilos.
Whyusechangeordersiftheydegradethetrustofyourcustomers?Becausewithout
them,mostsoftwarecompanieswouldgooutofbusiness.OnoneprojectIran,for
example,thestatementofworkwas“loose”becauseofthesheersizeofthesystem.We
hadspentsixmonthstryingtotightenupthestatementofwork,butthesystemwasso
immense,itwasimpossibletocaptureeverythingitwoulddo.Eventually,theaccount
managergaveuptryingtonaildownallthedetailsandtoldustojustcompletethe
statementofworkasis.Bythetimetheprojectranoutofbudget,thecustomersaidthey
neededsignificantchangesforittobeaccepted,eventhoughitmetthelooserequirements
laidoutinthestatementofwork.Asaresult,ourcompanydidacouplehundredthousand
dollars’worthofworkforfree.BelievemewhenIsaythatournextcontractincludeda
detailedchangeorderprocess.
Ifvendorsagreetomakeallchangesforfree,thesheervolumeofchangesislikedeath
byathousandcuts.Bytheendoftheproject,allthosesmallchangeswilladduptoa
staggeringcostthatthevendorhastoeat.Toprotectagainstthat,vendorswhowantto
stayinbusinessdevelopagoodchangeorderprocessandsticktoit.
Agoodchangeorderprocessisgoingtobefairlylengthyandverydetailed.For
example,theWashingtonStateDepartmentofTransportationConstructionChangeOrder
ProcessGuideis80pages[WSDOT].Itisaverywell-thought-outdocumentthatclearly
illustrateshowtheprocessworksandhowtodoit.Whyisthestatesometiculousabout
itschangeorderprocess?Toprotectitsinterests.AuthorDaveNielsenwroteanarticle,
“HowtoControlChangeRequests”[NIELSEN],inwhichhedescribesseveraldetailed
waystohandlechangerequests.ThepointI’mtryingtomakeisthatagoodchangeorder
processwilllikelybelargeanddetailed.Tryingtohandlechangesmoreinformallyona
traditionalprojectisarecipefordisaster.
Considerthissimpleexample:makingcookies.Mywife,Bernice,asksthekidsandme
whatwewantfordessert.We’reonourwayoutthedoor,soweyellback,“Cookies!”
Bernicecallsout,“Whatkind?”
Oneofthekidsrespondswith,“Thekindwithsugar.”(Sheisatotalsugarfreakand
doesn’twantanyhealthy-typefood.)
Nowmywifeisconfused.We’vejustleftthehouseandmymobileisoff.Sheknows
ourfavoritecookieischocolatechip,butweseemedtohavejustaskedherforsugar
cookies.So,aimingtoplease,andassumingwemustbeboredwiththeoldstandby,my
wifebakesthesugarcookies,completewithsprinkles.Wereturntofindthemcoolingon
thecounter.
“Yuck,”complainthekids.“Whatarethese?”
“Sugarcookies,likeyouaskedfor,”answersBernice,perplexedandfeelingabit
unappreciated.
“Wedidn’taskforsugarcookies,”whinethekids.“Wejustmeantwewantedthe
chocolatechipthatyoumakewithsugar—notthehealthykindyoumakewithapplesauce.
Youshouldhaveknownthat’swhatwewanted.”
“Well,sugarcookiesiswhatIheardyousay,sothat’swhatImade,”statesmy
frustratedwife.“IfImakemore,it’llcostyou.”Andoffwego.
Thefailureherewaspoorcommunicationbetweenthecustomers(thekidsandme)and
theteam(mywife).Thekidsthoughtthatwhentheysaid“thekindwithsugar”for
cookies,Bernicewouldunderstandthatwemeantchocolatechipmadewithsugar,as
opposedtothekindshemakesusingapplesauceasasweetener.Butwhatsheheardusask
forwassugarcookies.Withnofurtherwaytoclarifyourneeds,Bernicedidthebestshe
couldandmadesugarcookies.Thekidsweredisappointed.Shewasfrustrated.Noone
gottheresulttheywanted.
Intheworldofsoftware,peopledon’thaverelationshipslikefamiliesdo—trustis
fragileandeasilybroken.Buildingthewrongthingforcustomerserodestheirtrustin
softwarecompanies.WashingtonStateDOTandNielsendon’tlovehugeprocesses,but
theyimplementthesecomplicatedchangeordersinanattempttoimprovecommunication
andtogetitright.Plus,iftheydogetitwrong,theycanprovethattheyfollowedthe
agreed-uponprocessandthatwhattheybuiltaccuratelyreflectstheapprovedchange
order.
IneveryinstancethatIcanthinkof,changeordersarenecessaryontraditionalprojects.
Andlarge,detailedchangeordersarethebestwaytoensuresomesortofaccuracyand
accountability.Yet,nomatterhowgoodyourchangeordersystemis,youwillstilllose
trustwithyourcustomersassoonastheygothroughtheprocess.Intheirminds,you
shouldhaveunderstoodtheirrequesttobeginwith.(Likemykidsexpectedtheirmotherto
understandtheymeantchocolatechipcookies,yourcustomersfeelthatyoushould
understandtheirrequestsbecause,tothem,theyareclearandeasytofollow.)Instead,
yourcustomershavetospendtimere-explainingwhattheywantthroughalengthychange
order,it’scostingthemmoremoney,andthereleasedateislikelytoslip.Worseyet,the
finalproductstillmaynotbeexactlywhattheyenvisioned,evenaftergoingthroughthe
process,buttheyhavetopayforitnonetheless.
Anotherreasonchangeorderprocessesdegradetrustisthatmostcompanieshave
figuredoutthattheycanbidonthelowerendofthecostscale,bettingonthefactthatthe
customerswillchangetheirminds.Achangeorderprocessisnecessaryontraditional
contractsbecausevendorsaretakingalltherisk.Thisbusinessstrategy,onethattakes
advantageofthefactthatallcustomerswillchangetheirrequirementsastheydiscover
whattheyreallywant,canleaveareallysourtasteinthecustomer’smouth,though.Take
thepictureshowninFigure30-2asanexample.
FIGURE30-2Twoboats:OriginalContractandChangeOrder
Noticethenamesoftheboatsinthepicture.ThebigoneiscalledChangeOrder;the
smalleronebehinditiscalledOriginalContract[BOAT].Fromthecustomer’s
perspective,thevendorpromisedacertainproductatacertaincost.Whatthecustomergot
wasaproductthat’skindofclosetowhattheywantedbutcostthemmuchmorethanthe
originalbidandprobablycameinbehindschedule,too.Meanwhile,thevendorissailing
offintothesunsetwithabigwadofcash.Andpeoplewonderwhytrustisaproblem!
Change,though,isn’ttheonlyreasonteamsstruggletomeetcustomerexpectationson
traditionalcontracts.Theyarealsohandicappedbytiming.
Timing
Customersaskfordates,cost,andcommitmentrightfromthestart.Thetimingofthese
requests,atthebeginning,putsboththevendorandacustomerinabadpositionfromthe
onsetofthecontract,becausethereisnowaytobeprecisethatearlyintheprocess.
Vendorsandteamsaskcustomerstospelloutpreciselywhattheywantbeforethey’re
exactlysure,whilecustomersrequirevendorstoguaranteeasetoffunctionality,ata
certaincost,andbyacertaintimebasedonasetofrequirementsbuiltonasoft
foundation.
Thefactis,vendorsjustdon’tknowenoughattheproject’sstarttomakethosekindsof
promises.Tounderstandwhythetimingissuchaproblem,considertheConeof
Uncertainty,whichBarryBoehmintroducedin1981[BOEHM]andSteveMcConnell
reintroducedin1997inhisbookSoftwareProjectSurvivalGuide[MCCONNELL].Itis
picturedinFigure30-3.
FIGURE30-3TheConeofUncertainty
Theconedemonstratesthatatthebeginningofanyproject,teamshavethemost
uncertainty,withavarianceof0.25xto4xinrange.Thismeansthatwhatateamestimates
tobeasix-monthprojectcouldactuallyenduptakinganywherefrom1.5to24months!
Yet,whenteamsareinthisstateofheighteneduncertainty,theyareforcedtowritea
contractsayingexactlywhatwillbedelivered,bywhen,andatwhatcost,withno
variance.It’snowonderpeopleendupwithhugeboats!However,veryfewprojectsstart
atthefarleftofthecone.Inacontractingsituation,thecustomerwillhavesomenotionof
theirrequirements,whichputsthemsomewherebetweentheApprovedProductDefinition
andRequirementsCompletemilestones.
Scrumprojectstrytomovefromuncertaintytocertaintyinasshortacycleaspossible.
AsexplainedinChapter1,“Scrum:Simple,NotEasy,”teamslearnasmuchaboutthe
systemandhowitshouldbedesignedasearlyaspossiblebyprovidingworkingsoftware,
sprintaftersprint,basedonemergingrequirements.Designandrequirementassumptions
areflushedoutearly,allowingteamstomovetocertaintyquickerandwithmore
confidence.
EvenScrumprojects,however,wouldbeunabletopredictondayoneexactlywhatthe
finalcost,functionality,anddeliverydatewillbe.That’swhyadifferentcontracting
modelisneeded,onebuiltforchangeanddesignedtolimitpromisesmadeinthetimeof
greatestprojectuncertaintywhilemaximizingearlylearning.
Ihavenamedthatcontractingmodelrangesandchangesanddescribeitindetailinthe
nextsection.Whilethemodelmightnotbeappropriateforallclientsandinallsituations,
itdoesworkextremelywellwithcustomerswhoareengaged,understandtheproblems
withtraditionalprojects,andareinterestedinrefiningtheproductasitisbeingdeveloped.
Ifyouhavecustomerswhoarenotengagedorarelooking“forjustthedate”or“justthe
cost,”thismodelwillnotworkforyou.
RangesandChanges
TherangesandchangesmodelwasbornfromaprojectmyteamandIbidin2007.The
resultingcontractwasjustliketheoneJulioandRobdescribedtoRichardinthestory.
TheprojectwasanoverwhelmingsuccessandprovedtomethatthecontractIhad
developedcouldworkforotherclients.Andithas.Althoughwe’vetweakeditabitover
timeandtosuitcertainsituations,mycolleaguesandIhaveusedtherangesandchanges
modelatvariouscompaniesacrossNorthAmericaandEurope.Themodelsharescertain
featureswith“MoneyforNothingandChangeforFree,”presentedbyJeffSutherlandat
theAgile2008conferenceandupdatedattheAgile2013conference[SUTHERLAND].In
fact,afterhearinghispresentation,Iadjustedmyownmodelbasedonsomeofhisideas.
Inthissection,Idescribetherangesandchangesmodelanddiscusssomekeytake-
awaysfromSutherland’spresentationaswell.Beforestructuringyourowncontract,I
recommendyoureadtheSutherlandpapertounderstandthedifferencesbetweenwhathe
proposesandthemodelIpresenthere.Afterall,themodelyouultimatelyusewilllikely
beahybrid,asyouincludeelementsfromyourcompany’sownpolicies.
Traditionalsoftwareprojectcontractscontainthingslikemilestones,thedescriptionof
work,projectgoalsandobjectives,deploymentschedules,andwarranties.Theseitems
mayliveinarangesandchangescontract,buttheygetthereinaverydifferentway.The
projectiscoveredbytwodifferentcontracts:adiscoverycontract,whichgivestheteam
theinformationitneedstowriteameaningfulcontract,andtheprojectcontract,which
includesclausestoprotectboththecustomerandalsothecontractororoutside
developmentshop.
Discoverycontract
–Ascertaintheranges
–Determinecostandtimeline
Projectcontract
Thefirstcontractcoversonlythediscoveryphase,whichhelpstheteamdeterminethe
rangesnecessarytowriteameaningfulcontract.Thesecondcontractincludesthe
parametersforexecutionandtheall-importantchangesclause.
DiscoveryContract:AscertaintheRanges
Thediscoveryphaseiscoveredbyafixed-fee,fixed-lengthcontractwithonegoal:Allow
theteamtogatherenoughinformationtodiscoverandcommunicatewhatitcandeliver(a
productbacklog),bywhen(apreliminaryreleaseplan),andforhowmuch(costper
sprint).Thediscoverycontracttypicallycoverstwoorthreeteammembersforabouttwo
weeks.Ifthecustomerchoosestogowithanothervendorattheendofthefixed-length
discoverycontract,thecustomerisallowedtoreusethesedeliverables.
Thefirstdeliverableistheproductbacklog.Buildingtheproductbacklogfora
customerisathree-stepprocess:identifyusertypesorpersonas,writeuserstories,and
estimatethestories.
IdentifyUserTypesorPersonas
Thediscoveryteamshoulddeterminewhotheusersofthesystemare,whattheirinterests
are,howtheywillusethesystem,andtheirexpectations.Thishelpsthepeoplewritingthe
storiesbetterunderstandtheendusersandcreatesamoreuser-centricviewofstories.
WhenIidentifyuserpersonas,Iinterviewasmanyusersofthesystemaspossibletolearn
abouttheircurrentinteractions,whattheylikeanddislikeaboutthecurrentsystem,and
whattheirnumberonethingwouldbetofix.
WriteUserStories
Thediscoveryteamisresponsibleforwritinguserstoriesfollowingthe“Asa<user>,I
can<action>sothat<result>”template.Iinvitethecustomertotheuserstoryworkshop
alongwiththeteam.Theteammembersandcustomerrepresentative(s)canbrainstorm
ideasandwritestories.Evenifthecustomerchoosesnottowriteanyactualstories,
havingsomeonefromthecustomersideisbeneficialbecauseitgivestheteaman
opportunitytoaskcustomer-specificquestionsregardinguserstoriesthathelpclarifyand
makethestory-pointestimatesmoreaccurate.Theuserstoriesmakeuptheproduct
backlog.
EstimatetheStories
Thediscoveryteamestimatestheresultinguserstoriesinstorypoints.Storypointshelp
thestakeholdersandcustomersunderstandthestorycost,aconsiderationwhen
prioritizingthestoriesinthebacklog.Makesurethecustomerisavailableforquestions
duringestimation.(Italkmoreaboutensuringcustomeravailabilityinthe“Keysto
Success”sectionofthischapter.)
DiscoveryContract:DetermineCostandTimeline
Oncetheteamhasworkedwiththecustomertogatherinformation,theteamworksonits
owntodeterminethecostandtimeline.Thisisafour-stepprocess:determineteam
velocity,calculatecostpersprint,buildareleaseplan,andestablishpaymentoptions.
DetermineTeamVelocity
Ifyouhaveanestablishedteamandarebiddingonaprojectthatissimilartoworkthe
teamhasdonebefore,youshouldhaveagoodideaoftheteam’sexpectedvelocity.Ifyou
don’t,youneedtoforecastorestimatetheteam’svelocity.Eitherway,youcommunicate
velocitytoyourcustomerasapotentialrange.Understandthatyourcustomerwillexpect
youtoachievesomewherenearthetopofthatrangeonmostsprints.RefertoChapter4,
“PredictingTeamVelocity,”formoreinformation.
CalculateCostPerSprint
InChapter8,“TheCaseforaFull-TimeScrumMaster,”Idiscussedtheloadedcostofan
employee.Usethatapproachhereifyoudon’tnormallybillforteammembers’time.If
youworkinaconsultancyorhaveacapitalexpenditurebudgetthatyouareworkingfrom,
youknowwhatthebillablerateisperperson.Eitherway,usetheloadedemployeecosts
tocalculatethecostpersprint.Sumuptheloadedcostperpersonperhour,multiplyitby
theestimatedhourspersprinttheteamwillburn,andyouhaveyourcostpersprint.
BuildaReleasePlan
Thereleaseplanisyourbreadandbutter—itlaysoutalowrangeandahighrangeofwhat
youandyourteamcandeliver.Inthereleaseplan,youshouldhaveaguaranteednumber
ofpointstheteamwilldelivereachsprintandbytheendoftheproject,sothatanything
youachievebeyondthosenumbersisperceivedasadditionalvaluetothecustomer.Fix
thevariable(budget,time,scope)thatmattersmosttoyourclients.Iftheyhaveaset
budget,youcanshowthemhowmanysprintsthatmoneywillbuythemandtherangeof
functionalitythatwillbedeliveredbythetimethosesprintsarecomplete.Iftheyhavea
setdate,youcanshowthemhowmanysprintsyoucanrunbythatdate,howmuchitwill
cost,andtherangeoffunctionalitythatcanbedeliveredinthesetnumberofsprints.If
theyaremostconcernedwithfunctionality,youcanshowthemtherangeofhowmany
sprintsitwilltaketocompletethefunctionalityandtherangeofcosttheywillincur
dependingonwhetheryoursprintcountisintheloworhighendoftherange.Read
Chapter11,“ReleasePlanning,”formoreinformationonbuildingareleaseplan.
EstablishPaymentOptions
Thefinalstepistodeterminethepaymenttermsthatworkbestforyouandyourcustomer.
Youcanoffertheoptiontopaybypoint(potentiallymoreexpensive)ortopaybysprint,
withaguaranteedminimumnumberofpointsdeliveredeachsprint.
ProjectContract:DeliveringtheGoods
Attheendofthediscoveryphase,youhaveseveraldeliverablesforyourpotentialclient:
asetofuserpersonas,anestimatedproductbacklog,apreliminaryreleaseplan,anda
projectcontract.Theresultingcontractreferencestheestimatedproductbacklogand
preliminaryreleaseplan.Itguaranteesthatyouwilldeliverarangeofpointseachsprint,
basedonyourpredictedvelocity,foracertaincost,oratacertaincostperpoint.It
includestheteam’sdefinitionofdoneandanexplanationofhowtheworkwillbe
deliveredtothecustomer.Italsoincludeslanguagethatindicatesunderwhatconditions
thecustomercanrejectasprint’sworthofwork.Youalsoneedaclauseinthecontractto
discusschange,thechangessectionofthecontractmodel.
Changes
EvenaScrumcontracthastoincludeaclauseaboutchanges.Mycontractsallowthe
customertoreshuffletheworkontheproductbacklog,oraddnewstories,aslongasthe
totalnumberofpointsontheproductbacklogremainsthesameandthetotalforeach
sprintstillfitsinsideapredeterminedrangeofwhattheteamcandeliver.Attwodifferent
Agileconferences,JeffSutherlandpresented,“MoneyforNothing,ChangeforFree.”In
thosepresentations,hedescribedhisapproachtochangewiththecleverlynamed“change
forfree”clause.(Ilookmorecloselyatthe“moneyfornothing”clauseinthe“Keysto
Success”sectionlaterinthischapter.)
Sutherland’schangeforfreeclausestipulatesthatthecustomermustbeworkingwith
theteamandthattheproductownerisresponsibleforprioritization.Underthose
conditions,thecustomercanaddnewstoriestotheproductbacklogatnochargeaslong
asastoryofequalorsimilarvalueisremoved.AnexampleIusetohelppeople
understandthisconceptisinrelationtoaone-literbottle.Thisbottle,whichwould
representtheproductbackloginScrum,isdesignedtoholdoneliterofanyliquid:water,
oil,vinegar,softdrinks,cream,younameit.Atthebeginningoftheproject,thecustomer
decidestoaddoliveoilandwhitevinegar.Changeforfreemeansthecustomercanchoose
toaddsomeamountofwatertothebottleaslongasheremovesanequalamountofolive
oilorwhitevinegar.Otherwise,thebottlewilloverflow.
Atthispoint,yourcontractisnearlycomplete.Youjustneedtoaddfourcrucial
elements:customeravailability,acceptancewindow,prioritization,andtermination
clauses.Idiscussthesefourelementsinmoredetailinthenextsection.
KeystoSuccess
Acontractbasedongoodinformationandcustomerinteractiongoesalongwaytoward
buildingtrustwithyourcustomer.Butbesidescostrangesandallowancesforchanges,a
goodScrumcontractincludesstipulationsfortheelementsthatmakeorbreakaScrum
project.Theseincludecustomeravailabilityforfeedback,acceptance,andproductbacklog
maintenance,aswellasaclearunderstandingofhowtomaximizetheScrumadvantageof
keepingtheprojectinapotentiallyshippablestate.Thefollowingsectionsdescribehowto
includethesefactorsinaScrumcontract.
CustomerAvailability
Includeasetnumberofhoursperweektowhichthecustomercancommittobeing
available.Youdon’twanttofallintothetrapofhavinganactivecustomerupfrontonlyto
havethemdisappearastheprojectprogresses.Shortsprintsrequirealmostdailycustomer
availability.Longersprintshavetheadvantageofofferingmoresporadiccustomer
availabilitybutalsoruntheriskofhavingallthecustomertimefocusedtowardthe
beginningandendofthesprint.Eventhoughthedemandsonthecustomerarehigherwith
shortersprints,Iprefershortersprintdurationsbecauseitkeepsthecustomermore
engagedthroughouttheproject.
AcceptanceWindow
Theacceptancewindowistheamountoftimethecustomerhastoacceptfunctionality
deliveredinagivensprint.Youcanchoosetohavethewindowbeasshortas“duringthe
reviewmeeting”or“xhoursordaysafterthereviewmeeting.”Theclauseshouldalso
illustratewhatthedefaultbehaviorisifthewindowpassesandthecustomerhasneither
acceptednorrejectedthework.Inmostcases,thedefaultshouldbeautomaticacceptance.
Ipreferashortacceptancewindow,ideallyadayorless.Longeracceptancewindows
starttointrudeonthenextsprint,assomeworkcannotmoveforwardwithoutacceptance
ormodificationoftheworkthathasoccurredtodate.Ifafeatureisreviewedoutsidethe
acceptancewindowandfoundtobelackingsomething,thisnewworkgoesontothe
productbacklogasanewstorytobeprioritizedandwillcostthecustomermoney,which
iswhycustomeravailabilityisessential.Featuresreviewedinsidetheacceptancewindow
thatarelackingaregenerallyrejectedandcostthevendormoneytofix.
Ialsoincludethefollowingincontracts:Ifthecustomerrejectsanyworkdeliveredina
sprint,thecustomerdoesn’thavetopayforthatwork.Willpeopletrytotakeadvantageof
this?Yes.Ifyouchoosetoincludethisprovision,chooseyourcustomerswisely.Learnto
saynotocustomersyoucan’ttrust.
Prioritization
UnlikeatraditionalRFP,aproductbacklogmustbegroomedregularly—keptinpriority
orderandestimatedinpointsonaregularbasis.Makesurethatthecontractincludestime
spentwiththecustomerestimatingnewstories,breakingdownlargestories,andhelping
determineanyshiftsinpriority.Helpthecustomerunderstandthatadjustingthepriorityof
thestoriesintheproductbacklogisallowedbutthatdoingsomayhaveanimpactonhow
muchtheteamcanaccomplishbyacertaindate.
Theteammaywanttosequencetheworkacertainway,toachieveoptimalflowfor
development.Thismayormaynotalignwiththeorderinwhichthecustomerwouldlike
toseefeaturesdelivered.MikeCohnwritesinUserStoriesApplied[COHN]thatthe
customerwinseverytime.Iagree.Itistheteam’sjob,though,tohelpcustomers
understandthattherequestedsequenceoftheworkmayincreaseestimates,thereby
causingcertainfeaturestonotgetdone(e.g.,lessfunctionalityforthesameamountof
money).Atthesametime,though,thereisnocostforaddingitemstotheproduct
backlog,aslongassomethingcomesoutinitsplace.
TerminationClauses
Mycontractsstatethatthecustomercanterminatethecontractatanytimeandforany
reason.Theteamfinishesoutthecurrentsprintandhasoneadditionalsprinttoclean
thingsupasneeded,suchastransfercodebases,paperwork,andthelike.Thisbuildstrust
andforcestheteamtodrivetowardpotentiallyshippableeachsprint.
Similarly,Sutherland’smoneyforfreeclauseacknowledgesthatthesoftwaremayreach
ashippablestatebeforeallfeatureshavebeencompleted[SUTHERLAND].Assuch,the
customercandeterminewhethertostopdevelopmentafteranysprintandshipthe
softwareasis.Monetaryprotectionisbuiltinforboththecustomerandthevendor.So,
howdoesthiswork?Thecustomerhasaprioritizedlistofwork.Atsomepointduringthe
executionofthecontract,thecostofdoingtheremainingworkwillbeprohibitiveand
returnlittleornovalue.SutherlandcallsthistheROIcutoff.Oncethecustomer
determinesthattheprojecthasarrivedattheROIcutoff,theremainingworkisscrapped.
Becausetheprojectis,inessence,shippingearly,moneywillbeleftoverinthecontract.
Withthemoneyforfreeclause,thecustomerkeepsthemajorityoftheremainingfunds,
butthevendoralsoreceivesapercentage(inSutherland’scase,20percent)tocoverthe
costsofendingthecontractearly.
Trust
Onceyougetusedtoworkingwiththem,contractsthattakeintoaccounttherealitiesofa
Scrumdeliverymodelmightjustbethebestcontractsforvendorsandcustomers.
Structuringyourcontractstoallowforchangeisthebestwaytodobusiness.Giving
customersthefreedomtopayforeachsetoffunctionalityasit’sdelivered,toaddor
changestoriesaslongasastoryofequalvalueisremoved,andtocutthecontractshortor
terminateitaltogetheratwillpaysoffintermsofrepeatcustomersandreferrals.Afterall,
intheend,trustisthemostimportantthingasoftwaredevelopmentcompanywillever
build.
References
[BOAT]ConstructionMarketingIdeaswebsite.
http://constructionmarketingideas.blogspot.ca/2009/03/change-order-boat-mystery-
partially.html(accessed24July2015).
[BOEHM]Boehm,Barry.1982.SoftwareEngineeringEconomics.EnglewoodCliffs,NJ:
PrenticeHall.
[COHN]Cohn,Mike.2004.UserStoriesApplied.UpperSaddleRiver,NJ:Addison-
Wesley.
[MCCONNELL]McConnell,Steve.1997.SoftwareProjectSurvivalGuide.Redmond,
WA:MicrosoftPress.
[NIELSEN]Nielsen,Dave.“HowtoControlChangeRequests.”PMHutwebsite.
http://www.pmhut.com/how-to-control-change-requests(accessed24July2015).
[SUTHERLAND]Sutherland,Jeff.“Update2013:AgileContracts:MoneyforNothing
andChangeforFree.”JeffSutherlandwebsite.
http://jeffsutherland.com/Agile2008MoneyforNothing.pdf(accessed24July2015).
[WSDOT]WashingtonStateDepartmentofTransportationwebsite.
http://www.wsdot.wa.gov/NR/rdonlyres/7C35B04E-F736-47BA-891E-
0B9F7021FDF3/0/ChangeOrderProcessGuide.pdf(accessed8April2015).
PartV.WildernessEssentials
Chapter31.DrivingtoDonethroughCollaboration
ForScrumtowork,teamsneedtobeconfidentintheirpredictionsandabletodeliveron
theirpromises.Toooften,however,teamsthatarenewtoScrumstruggletofinishthe
storiesinthesprint,leadingtostakeholdermistrustandadeclineinteammorale.A
contributingfactorthattypicallyhidesbehindthepilesofhalf-finishedworkisalackof
collaboration,fromestimationthroughexecution.
ThatwasthecasewiththePNIteam.Despitetheteam’sbestefforts,itfailedtodeliver
onpromisedfunctionalitysprintaftersprint,leavingeveryonefrustratedandsearchingfor
answers.
TheStory
Overthepastsevenmonths,thePNIteamhadbeenslowlyintroducingScrum.Asaresult,
theteamandthecompanyhadnoticedbettercustomerengagementandmorevisible
projecttracking.Yet,thebiggestproblem—andoneofthemainreasonstheteamhad
adoptedScruminthefirstplace—hadyettobesolved:theteam’sinabilitytodeliver
completedfunctionality.Aftersevenfrustratingsprints,theteamstillwasnotableto
deliver.
Thesituationcametoaheadduringonesprintreviewmeeting,whentheteam’sproduct
owner,Rafael,confrontedtheScrumMaster,Sam,aboutthelackofcompletedwork.
“Whatdoyoumeanthisfunctionalityisstillincomplete?Youtoldusweeksago,or
sprintsago,orwhateveryoucallit,thatit’dbedone!”exclaimedRafael.
“Iknow,Iknow,we’resosorry,”answeredSam.“TheteaminParisreleasedsomething
anditbrokealotofourtests.Wetriedtogetitfixed—”
Rafaelinterrupted.“Enoughwiththeseexcuses.Weekafterweek,sprintaftersprint,
there’ssomereasonwhyyoucan’tgiveuswhatyoupromised.Ican’ttellyouhowmuch
thishurtsallofus!Wehavetogobacktothevariouscountrymanagersandtellthemthat
thefunctionalityyoustatedwouldbedeliveredtwomonthsago,thenonemonthago,then
twoweeksagoisstillnotdone!Wecan’trunabusinesslikethis.Ithoughtyousaidyou
werepracticingScrum?Shouldn’tthatmeanthattheseissuesarebeingresolved?What’s
therealproblemhere?”
Thatwastherightquestiontoask:Whatwastherealproblem?Samhadhersuspicions,
butwithonlysevensprintsundertheteam’sbelt,andwithonlyafewofthoseeven
remotelysuccessful,theteamwasstillworkingthroughsomanyissuesthatitwas
difficulttopinpointthecauseofthisparticularproblem.“We’reunhappyaboutittoo,and
we’recontinuingtomakechangestoimprove.We’regoingtofigureitoutasquicklyas
possible,”Samreplied.
Samknewthatinherglobalcompany,manyteamsweredoingScrumanddoingitvery
well.Shealsoknewthatshehadaccesstoagileexpertsinsidehercompanywhowere
onlyaplanerideaway.SoSamdecidedtoreachouttomanagementtoaskforhelpinthe
formofatravelbudgetandsomeconsultingtime.Twoweekslater,oneofthecompany’s
internalagilecoaches,Olivia,wassittingintheofficewithSam,discussingthesituation.
“So,Sam,youtoldmeonthephoneyourstakeholdersarereallyangrybecausethey
feelthattheteamisjustnotdelivering.Canyoutellmemore?”askedOlivia.
Samnodded.“Yes,we’veadoptedScrumandhavejustfinishedourseventhsprint.
Overall,theteammembersfeelbetteraboutwherewearenowthanwherewewere
before.Atonofissueshavesurfacedthatwehavebeenabletoaddress.However,atthe
endofeverysprint,wejustcan’tseemtofinishandIhavenoideawhy.Itisreally
embarrassinggoingtothereviewandsayingweonlycompletedaboutfortypercentofthe
storieswesaidwewould.It’snowonderourstakeholdersdon’ttrustus.We’reamess!”
“What’sthenumber-onecomplaintthestakeholdershave?”askedOlivia.“Isitquality,
isitthatthefinishedworkdoesn’tfunctionastheyimagined,orwhat?”
“Simple,”statedSam.“Theysayweareconsistentlyfailingtodeliverthefunctionality
wepromisedtogivethematthebeginningofthesprint.Mostofthestoriesjustaren’t
donebythesprintreview.”
“Okay.Let’svalidateorinvalidatethatclaimbylookingatthedatawehave,”said
Olivia.
Shewentthrougheachsprint,puttingthetasksintothreecategories:completed,stillin
process,andnotstarted.
Theresultswereshocking.Oliviaputthechartonthedisplayintheconferenceroom.
“Well,Sam,thestakeholdershaveavalidpoint.Inyourfirstandsixthsprints,your
teamcompletedonlytwenty-fivepercentofwhatitcommittedtofinish.Eveninthebest
sprint,whichwassprinttwo,youcompletedonlyseventy-fivepercent.Youraverage
percentagecompleteisaroundforty-five,lessthanhalfoftheworkyousaidwouldbe
done,whichisprettybad.AnotherthingI’mnoticinghereisthatyouhaveatonofwork
inprocess.That’ssomethingIwanttolookatcloser,too”saidOlivia.
“Yep.Wow.Thatisoneuglychart.It’snowonderthestakeholdersaresoangry!”
exclaimedSam.“Ihavenoideawhattodoaboutit,though.Iknowtheteamisworking
andtrying—it’snotaquestionofeffort.”
Oliviathoughtforamoment,thensaid,“Sam,tellmehowtheteamderivesitstask
estimates.Whentheteammembersestimatetheirwork,dotheydoitcollaboratively?As
anexample,alotofourteamsareusingTaskPoker—somethingverysimilartoPlanning
Pokerbutfocusedonestimatingtasksinsteadofstories.”
“TaskPoker?No.Whenwewanttoestimateaparticulartask,wejusthavetheperson
whoisgoingtodotheworkfigureouthowlongthattaskwilltake.Thatworks,butifthe
personwhoestimatedthetaskcan’tdoit,thingsgetmessedup.Usually,nooneelsecan
dothework—oratleastnotasquicklyastheoriginalestimate,”saidSam.
“Okay.Ithinkwe’vefoundafewoftherootcausesofyourproblem.Letmeclarifya
fewthingstobesure.First,itsoundslikeyouarerelyingsolelyonexpertanalysisfor
yourestimates.Anditalsosoundslikeyouaren’tpracticingpairprogramming.Areboth
ofthosethingstrue?”
“Yes,that’sright.Themanagersherethinkpairprogrammingisawasteoftime.They
believethatsinceeachofusisanexpertinoneortwothings,eachofusshoulddowhat
heorsheisgoodat.And,cometothinkofit,thatattitudebleedsoverintohowwe
estimatetoo.Thepersonwhoisanexpertatthattypeofworkestimatesitandthendoes
it,”saidSam.
“So,youhaveabsolutelyzerocollaborationwhenitcomestoestimatingtasks?NoTask
PokerorWidebandDelphioranything?It’sjustexpertanalysis?”askedOlivia.
“Exactly.Thepersonwhoisgoingtodotheworkestimatesit—noteamdiscussionor
collaboration,”saidSam.
“Whathappensifsomeonegetsblockedonatask?Orifsomeoneispulledoffataskto
workonsomethingelse?”Oliviainquired.
“Well,ifsomeoneisblocked,thatpersonmovesontothenexttaskinthelist,”answered
Sam.“See,weallsignupforthetaskswe’regoingtoworkonduringsprintplanningso
wecanmakesurethatwehaveenoughcapacitytodothework.So,wehaveplentyof
taskstochoosefromifwegetblocked.Andifsomeoneispulledoffaparticulartask—
theygetsickorthereisanemergencyonanotherproject—well,thattaskjustsitsthere
untilthatpersoncanreturntoit,Iguess.Noonehastheavailablecapacityorexpertiseto
pickuptheslack.”
Oliviashookherhead.“So,you’reallmaxedoutonassignedtasksfromthebeginning
ofthesprint.And,ifsomeonegetsblockedonmorethanonetask,thatpersonjustkeeps
startingnewonesuntilhefindsonehecancomplete?Thatexplainsalloftheworkin
processattheendofthesprint.
“Whataboutyoursprinttaskboard?Whenyoulookattheboardduringthedaily
scrums,areyoueveralarmedbythenumberoftaskslistedasinprocess?”
“Umm,confessiontime.Wedon’treallyusethetaskboard.Itseemedlikewaytoo
muchworktokeepituptodate.That’sprobablywhyit’sbeensohardtofigureoutwhois
workingonwhat,”admittedSam.“Youmusthaveagreatimpressionofourteam.”
OliviaassuredSamthatherteamwasnodifferentfrommostnewteamsshehadrun
across.Theteamwasnotfullyformedyet.Instead,itwasagroupofexpertsworkingon
subjectmattercomponents,lackingtrueteamcohesion.
“Sam,I’vegotanideathatIthinkwillhelpfixyourteam’sissues.Wearegoingtorun
atestoverseveralsprints.We’regoingtodoTaskPoker,limittheworkinprocessand
activetaskstonomorethantwoperteammember,useavisualtaskboard,andpair
programwheneverpossible,”saidOlivia.“Oh,andIsuggestwemovetotwo-weeksprints
fornow.Theteamneedstoworkwithsmallertasksandneedstogetmorefrequent
stakeholderfeedback,potentiallyallowingthestakeholderstomakechangesbasedonthe
progressoftheproject.Theteamalsoneedstopracticeplanningandestimationmore
oftenthanonceamonth.”
Samagreed.Andsodidherteamonceshepresentedthemwiththechartdetailingthe
problemsalongwithOlivia’ssuggestionsforimprovement.Afterfoursprints,thingshad
beguntoimprovesubstantially.SamandOliviasatdowntoreviewthedata.
“Olivia,you’renotgoingtobelieveit.Thingsaregoingsowell!”exclaimedSam.“Just
lookatthisdata.We’redelivering!”
“Sam,thislooksfantastic!Whatwasthedrivingforce?WasittheshifttoTaskPoker
thathelpedthemost?Pairing?Thetaskboard?Shortersprints?”inquiredOlivia.
“Allofit!ButwethinkthatTaskPokerhelpedthemostbecauseithashadthebiggest
impactonteamcollaboration.Thesmallertasksizeshelp,whilepairingmeanswehave
fewerblockedtasks.Plus,now,ifsomeonehasanemergency,theteamismorelikelyto
beabletopickuptheslack.However,noneofthiswouldhaveworkedwithoutTask
Poker.Thevaluethattheteamgetsfromthesprintplanningmeeting—frombeingonthe
samepageandknowing,collectively,whathastobedone—hasbeeninvaluable,”stated
Samexcitedly.“Wecan’tthankyouenough.”
TheModel
Inthestory,Oliviaidentifiedseveralissuesthatplaguedtheteam.Fortunately,though,she
foundsomesolutions,asillustratedinTable31-1.Thetabledetailstheproblemsandthe
fiveproposedsolutions:TaskPoker,pairprogramming,limitedworkinprocess,shorter
sprints,andavisualtaskboard.Thesectionsthatfollowdiscusseachsolutionindetail.
TABLE31-1CommonScrumTeamProblemsandSolutions
TaskPoker
Scrumteamslivebythemantrathatthepeoplewhoestimatetheworkshoulddothework.
Butthatdoesn’tmeanthateachteammemberestimatesonlytheworkheorsheplansto
do.Instead,themantraisintendedtoconveythattheteamthatdoestheworkestimates
thework.Theestimatesshouldnotbecalculatedbysomegrouporsubgroupactingonthe
team’sbehalf.Inthischapter’sstory,theteamreliedonindividualexpertstodetermine
howlongeachtaskmighttake—leavingmostteammembersinthedarkaboutthework
involvedinthesprintoutsideoftheirindividualcontributions.Tohelptheteamlearnhow
toestimateitsworkcollaboratively,OliviarecommendedTaskPoker.
TaskPokerisavariationofPlanningPoker,atechniquethatwascreatedbyJames
Grenning[GRENNING]in2002andpopularizedbyMikeCohninAgileEstimationand
Planning[COHN].AsdetailedinTable31-2,TaskPokerandPlanningPokerbothusea
limitednumbersettorepresentestimates,andbothemploywholeteamdiscussionanda
collaborativewayofarrivingataconsensusestimate,butthat’swherethesimilaritiesend.
InTaskPoker,theteamisestimatingsprintbacklogtasksintermsofhours,asopposedto
PlanningPoker,whereteamsareestimatingproductbacklogstoriesintermsofstory
points.DistinguishingTaskPokerasadifferentestimationtoolhelpstopreservethemore
abstractnatureofPlanningPokerasameasureofeffort,wherethestorypointsdonot
correlatetohours(seeChapter32,“HowStoryPointsRelatetoHours”).
TABLE31-2KeyDifferencesbetweenTaskPokerandPlanningPoker
WhileteamscanusePlanningPokercardsandtheirmodifiedFibonacciseriesforTask
Poker,theymightprefertouseadifferentsetofnumberstodistinguishbetweenrelative
sizingandtasktimeestimation.Oneeasysettorememberconsistsofbinarynumbers:1,
2,4,8,and16.Ifyouusebinarynumbers,theestimationcanbedonewithoutcards.Just
voteusingonehand,rock–paper–scissorsstyle.Thenumberoffingersapersonholdsup
istheestimate,whereonefingeris1hour,twofingersis2hours,threefingersis4hours,
fourfingersis8hours,andfivefingersis16hours[SLIGER].Itmayfeelawkwardatfirst
toholdupthreefingersandhaveitmean4hours,buttheteamwillsoonlearntomentally
convertittobinaryandadjustquickly.Witheithersetofnumbers,thebenefitofusinga
limitednumberofpossibleestimatesisthatitencouragesteamstobeaccuratewithout
beingoverlyprecise,tobeabletoreachconsensusandmoveontothenexttaskwithlittle
ornoargumentorwastedeffort.
LikePlanningPoker,TaskPokerdrivescollectiveteamownershipoftheestimation
processsothattheteamhasahigherdegreeofconfidenceinitsestimates.Italsofosters
discussion,whichhelpstomitigaterisksandremoveassumptions.Whole-team
discussionsencouragememberstoaskforhelpwithataskortovolunteertolearnanew
typeoftask,bothofwhichcanbeaccomplishedthroughpairprogramming.
PairProgramming
Pairprogrammingiswhentwoteammembersworktogethertoaccomplishasingletask.
Amongpairprogramming’smanybenefitsarecollectivecodeownershipandanincrease
inknowledgesharingamongteammembers.Duringapairprogrammingsession,each
paircontinuallydiscussesanyproblemanditssolution,essentiallycreatingareal-time
codereview.Aspeoplemovefromtasktotask,ideallyswitchingpairsandtypesofwork,
alloftheteammembersbecomefamiliarwiththevastmajorityofthecodebasewhilealso
broadeningtheirownskillsets.
That’swhyoneofOlivia’sproposedsolutionsforthePNIteamwastostartusingpair
programming.Theteammembersweredisconnectedfromoneanothernotonlyduring
estimationbutalsoduringtheactualsprints.Allofthetaskswerepre-selectedby
individualswhoseskillsbestmatchedaparticulartask.Theproblemis,whenteamswork
thisway,individualsmovethroughthesprintinavacuum,workingandthinkingabout
onlytheirindividualtasksandnotmakingconnectionstothetotalityofworkforthesprint
—andworseyet,theprojectasawhole.Acommonsideeffectofthiswayofworkingis
thattheteamlosestrackofitsprogress—tasksareblocked,peoplegoonvacationorget
sick,orqualityissuesgounnoticeduntiltheendofthesprint.Soon,theteamisleftwith
manyopentasksandalackofvisibilityintowhatiskeepingthosetasksfrombeing
complete.
Pairprogramminghelpssolvetheseissuesbycreatingcollectiveownershipforthe
team.Theresultisthateveryteammemberhasvisibilityinto,andanactiverolein
completing,alloftheworkofthesprint.(Chapter9,“WhyEngineeringPracticesAre
ImportantinScrum,”offersanin-depthlookatpairprograminganditsbenefits.)
LimitOpenWorkItems
AnotherchallengethePNIteamfacedwasthatteammembershadtoomanyopenwork
items.Notonlydidtheyself-selectalloftheavailabletasksduringsprintplanning,they
alsotendedtofireandforgetwhentheygotblocked.Inotherwords,alloftheteam
membershadafulldocketofplannedtasks.Whentheygotblockedonatask,theymoved
ontothenexttaskintheirlists,rarelyreturningtothetaskstheywereoriginallyworking
on.Thislefteachteammemberwithmultipletasksopenbutunfinished,resultingina
sprintfilledwithhalf-donework.Tohelpaddresstheseissues,Oliviarecommendedthat
theteamsetanactivetasklimit(ATL)byrestrictingthenumberofself-selectedtasksto
twoperperson.Shealsosuggestedthattheteamcreateaworkinprocess(WIP)limitof
twotasksperperson.WithaWIPlimitoftwo,ifsomeoneisblockedontwotasksand
cannotmoveforward,thatpersonisnotallowedtoopenanewtask.Instead,theblocked
personisexpectedtoeitheraskforhelportohelpotherteammemberswiththeirtasks.
Thereasontheselimitsworktodrivetaskstowardcompletionisthattheyreduce
multitasking,ahabitthatoftenleadstoreducedfocus,lowerproductivity,andlessoverall
output.Thenegativeimpactsofmultitaskinghavebeenproven.In1992,KimClarkand
StevenWheelwrightfoundthatonceapersonhasthreeormoreactivetasks,timespenton
eachtaskdramaticallydecreases.Twotasksyieldedthebestresults[CLARK].
Similarly,in2009whileatStanfordUniversity,EyalOphira,CliffordNass,and
AnthonyD.Wagnerstudiedtheimpactofmultitasking.Theresearcherscreatedtwo
groups.Thefirstgroupwasmadeupofself-describedheavymediamultitaskers(HMM),
whiletheotherwascomposedofself-describedlightmediamultitaskers(LMM).Those
whoclassifiedthemselvesasHMMwerepeoplewhoconstantlyjuggledtasks.They
preferredtodomultipletaskssimultaneously,suchaswritingemailwhilealsohaving
instantmessengerconversationsorreviewingwebsiteswhilewatchingTV.TheLMM
group,however,preferredtoremainfocusedonthetaskathandandwouldpurposefully
reducethenoisesurroundingthem.Amongotherthings,theresearcherswantedto
discoverwhethertheHMMgroupweremoreadeptatswitchingtasksthantheLMM
group
Toevaluatetheparticipants’skillsatswitchingtasks,theresearcherscreatedaseriesof
images.Eachimagecontainedbothanumberandaletter,suchasA4orV17.Priorto
showingtheparticipantstheimages,theresearcherswouldinstructtheparticipantsto
focusonlyontheletteroronlyonthenumber.Iftheparticipantswereaskedtofocuson
theletter,they’dhavetodeterminewhethertheletterwasaconsonantoravowel.Ifthey
wereaskedtofocusonthenumber,they’dhavetodeterminewhetherthenumberwasodd
oreven.Surprisingly,theHMMgroupdidnobetterthantheLMMgroupatswitching
tasks.Infact,theydidworse.Whendiscussingtheresults,theresearchershypothesized
thathabitualmultitaskers(likethoseintheHMMgroup)haveatendencytofocus
simultaneouslyonalloftheinformationinfrontofthem.Consequently,theystruggle
whenaskedtoseparatethenoisefromthetaskathand.Thebottomlineisthateven—and
maybeespecially—thosewhoclaimtothriveinamultitaskingenvironmentactuallydo
betterwhenfocusedononetaskatatime[OPHIRA].
Whetherit’sapersonworkingonadedicatedteamhavingmultiplestreamsofworkor
tasksinprocess,orsomeoneworkinginaspecialtyareaacrossmultipleteams,it’s
impossibleforanyonetokeeptoomanyballsintheair.Focusingononestreamofwork
anddrivingittocompletionisfarbetterthantryingtodoalittlebitofworkonthree
differenttasks,evenifthosetasksareassimpleaswritinganemailoransweringaphone
callwhiletryingtofinishalineofcode.Whateverisleftincompletewillloomoverthe
personlikeadarkcloudscreaming,“Finishme!”It’smuchmoreefficientinthelongrun
tofinishonethingbeforestartinganother.
AfunexperimentIliketoruntodrivehometheimportanceofasingle-mindedfocusis
tohavepeopleworkonaseriesoftaskstoaccomplishthesamegoal,butindifferent
ways.Thegoalisacompletelistofthenumbers1through26,alphabetAthroughZ,and
RomannumeralsIthroughXXVI.Isplitthegroupintwoandcreateimbalance,typically
75percentforGroupAand25percentforGroupB.
GroupAisrequiredtocompletetheworkbyrows.Itsmemberswriteonpaperfromleft
toright,startingwithnumber1,andcontinueacrosstherow,writingtheletterA,andthen
RomannumeralI.
Theycontinueworkinginrows,goingrowbyrow,untilrow26,Z,XXVIisreached.
Thepatternshouldlooklikethis:
GroupB,however,isrequiredtocompletetheworkbycolumns.Itsmemberswriteon
paper,startingwithnumber1andcontinuedownthecolumnuntilnumber26isreached.
Oncethefirstcolumniscomplete,theybeginanewcolumn,whichisthealphabet,
fromAtoZ.
Oncethealphabetiscomplete,theymovetothelastcolumn,Romannumerals,Ito
XXVI.
BothfinaloutputsfromGroupAandGroupBwilllooklikethisintheend,butthetwo
groupscreatethesetablesinvastlydifferentways.
Irecordhowlongittakesthefirstpersontofinish.Over90percentofthetime,thefirst
finishercomesfromGroupB—thegroupthatfocusedononecategory(column)atatime.
Imakeanotationofthetimeittookthefirstpersontofinish,aswellasthetimesofall
completingparticipantsuntileveryoneinonegroupisfinishedwiththetask.What
typicallyhappensisthatbythetime100percentofGroupBisfinished,onlyabout33
percentofGroupAhascompletedthetask.Whenaskedhoweachparticipantfeelsabout
theexercise,GroupAcomplainstheyhaveaheadachefromthetaskswitching.GroupB
complainstheyhaveasorehandfromtherepetition.
Thisexerciseprovesapoint:Peopletendtofinishfaster,andwithlessofaheadache,
whenfocusingononlyoneortwoitemsatatime.That’swhy,inthischapter’sstory,
OliviarecommendedtheteamsetanATLandaWIPlimit.RecallthatATListheactive
tasklimit,whereactivemeansself-selectedbyateammember.WIPistheworkin
process,workthathasbeenstartedbuthasnotyetbeencompleted.
WithoutanATL,peopletendtosignupforallofthetaskstheyknowhowtodountil
theyhavefilledalloftheirplannedtimeforthatsprint.Theunintendedconsequenceis
thatwhensomeonehasanissue,nooneisavailabletohelp.Andifsomeonefinishes
early,therearenoavailabletaskstosignupforbecausetheywerealldistributedinthe
beginningofthesprint,duringsprintplanning.Similarly,whenateamdoesnothavea
WIPlimit,itsmembersarefreetohaveasmanyin-processtasksastheywant.Whattends
tohappen,then,isthatassoonassomeoneisblocked,thatpersonstartsworkingon
somethingnew.Thisresultsinanexcessofin-processwork,allindifferentstates,andall
assignedtotheparticularpersonwhobeganworkonthetask.
EstablishingbothanATLandaWIPlimiteliminatestheseissues.HavinganATL
meansthatthemajorityofthetaskswillremaininactiveatthebeginningofthesprint.
Thiscreateslearningopportunitiesforteammembersbyallowingthemtoselectany
inactivetask,whethertheyhaveexpertiseinthattaskornot.Meanwhile,ifsomeonegets
stuckonanytask,teammembersarelikelyavailabletopairontheworkanddriveitto
done.Thisbehaviorcreatesalearningculturethathelpsreinforcetotalteamaccountability
andcollectiveownership.Additionally,theselimitshelpteammembersbreakthehabitof
workingonlyonthetaskstheypreferbasedontheirskillsetorhoweasythetasksare.
Lastbutbynomeansleast,strictWIPlimitsforcetheScrumMaster,orevenstakeholders,
togettasksunblockedasquicklyaspossible.
Twohelpfulbooksforthosewhowanttoknowmoreaboutthemanyadvantagesof
WIPlimitsareKennyRubin’sEssentialScrumandDavidAnderson’sKanban.
Two-WeekSprints
AnotherchallengethatSam’steamfacedwasthattheteamwasworkinginafour-week
sprint,thelongestrecommendedsprintlength,yetstillcouldnotcompletethework.
Sometimeswhenthishappens,teamsactuallytrytoextendtheirsprints,saying,“Justgive
usonemoreweek!”Instead,ascoveredinChapter6,“DeterminingSprintLength,”what
mostteamsneedtodowhentheystruggletofinishasprintistoshortentheirsprintsrather
thanlengthenthem.
Thoughcuttingthesprintlengthmightseemcounterintuitive,shortersprintsprovide
moreopportunitiesforlearninghowtoworkatasustainablepaceandforimproving
decomposition(seeChapter23,“SustainablePace,”andChapter12,“Decomposing
StoriesandTasks”).ThereasonOliviasuggestedtwo-weeksprintswastoencouragethe
teamtoworkwithsmallertasksoverall,tobeginwiththeendinsight,toincreaseits
opportunitiestointeractwithstakeholders,andtopracticeestimation.
Shiftingtotwo-weeksprintsforcesteamstoreducethesizeofeachtask.Teamsmust
decouplesmallerserialtasksthattheymightnormallylumpintoasinglelargertask,
whichprovidesmoregranularity.Itallowsteamstoseetheworkremainingwithgreater
clarity,likegoingfromastandarddefinitionscreentoa1080-pixelhigh-definitionscreen.
Implementingtwo-weeksprintsalsoallowsteamstobeabletoseethefinishlinefrom
thestartinggate.IfateamstartsonaMonday,thefollowingFridayfeelslikeitisright
aroundthecorner.Theendisinsight—andtheentiresprintcaneasilybevisualizedviaa
taskboard,awhiteboard,orsoftware.Teammembersfeelaconsistentlevelofpressureto
finisheachtask,ratherthanhavingthefaultyimpressionthattheyhavemanyweeks
beforetheyneedtocompletetheirwork.Shortersprintsincreasefocusanddecreasestress
becauseeachtaskis,bydesign,smaller.
Lastly,atwo-weeksprintgivesteamsmoreopportunities—bothtointeractwith
stakeholdersandtopracticeestimation.Inaone-yearprojectwithone-monthsprints,a
teamwouldhaveonly11opportunitiestointeractwiththestakeholders,andmore
important,thestakeholderswouldhaveonly11chancestoinfluencechange:11!Ateam
wouldalsohaveonly11chancestoimproveitsestimationskills.Ontheotherhand,using
two-weeksprintsonayear-longprojectgivesateamatleast25chancestogetfeedback
andmakeadjustments,givingtheteamanditsstakeholdersmoreconfidenceinthesprint-
over-sprintplan.Fornewteamsespecially,theseopportunitiesaretoovaluabletopassup.
CreatingVisibilitywithaTaskBoard
AnotherchallengethatSam’steamhadwasvisibilityintothework.Althoughtheteam
membershadfaithfullyattendeddailyscrums,theyhadfailedtocreate,utilize,and
maintainataskboardintheteamspace.Thislackofvisibilityintotheworkofthesprint
contributedtotheteam’slackofurgencyandfailuretounderstandtheworkremainingin
thesprint.
Agoodtaskboardhelpsfocusinteractionduringthedailyscrumandcreate
opportunitiesfordiscussionthroughouttheday.Additionally,itcreatesvisibilityintothe
workofthesprintanddrivesteamcommitmentandaccountability.Whenateamisusing
taskboardseffectively,youwillseeteammates,stakeholders,andevenmembersofother
teamsstopbytheboardfordiscussionthroughouttheday.Thisbehaviorincreasesgreatly
iftheboardisprominentlydisplayedneartheteam.
AgoodboardtypicallylooksliketheoneinFigure31-1.Noticethatallofthestories
theteamhaschosenforthesprint,andtheircorrespondingpointestimates,areshownin
theleftmostcolumn.Then,tothesideofeachstory,theteamhaspostedthatstory’s
plannedtasksandtheestimatesforthosetasks,inhours.Thetasksareplacedinthe
correctcolumnaccordingtotheirstatus.ThoseintheToDocolumnhaven’tbeenstarted.
ThoseintheInProcesscolumnarecurrentlybeingworkedonbyoneormoreteam
members.ThosetasksthatarereadyfortestingorfeedbackareintheToVerifycolumn.
AndthosetasksthatarecompleteareintheDonecolumn.Whenallofastory’stasksare
Done,thestoryisconsideredcomplete.
FIGURE31-1Atypicaltaskboard
Healthytaskboardshaverelativelyfewtasksinprocessorwaitingforverificationat
onetime.WhenOliviaarrived,however,thePNIteam’staskboardwouldhavelooked
morelikeFigure31-2,withapileoftasksshowinginprocess.ThoughhavingaWIPlimit
helpspreventtasksstackingupinonecolumn,havingatasksitinprocessformorethana
dayortwoshouldstillbeawarningsigntotheteam.Theteamcanhelpkeeptasks
movingfromlefttorightbyaskingquestionsofoneanotherduringthedailyscrum—
questionsintendedtohelpuncoverhiddenissuesandtofindareaswhereteammembers
mightneedhelp.Tasksthataretoolargewillalsositfortoolonginonecolumn.When
thishappens,trybreakingthetaskintosmallertasks.
FIGURE31-2Toomanytasksinprocess
EspeciallyforteamsnewtoScrum,Irecommendusingataskboard—evenifyouuse
softwaretotracktheworkofthesprint.Teamsneedavisiblereminderofhowthesprintis
progressingandatactiledisplayaroundwhichtohavefocused,collaborativediscussions.
KeystoSuccess
Ofthefiveitemsexploredinthemodel(TaskPoker,pairprogramming,ATLandWIP
limit,two-weeksprints,andtaskboard),theonewiththemostimmediateimpactona
team’sabilitytodeliveronitscommitmentistheswitchfromexpertanalysisestimationto
collaborativeestimationusingTaskPoker.TaskPokerisacollaborativemethodthatI
havebeenusingfortaskestimationwithteamsforyears.
Let’slookmorecloselyatthisvitalpieceofthepuzzleandhowithelpsteamsdrive
storiestocompletion.
Theteam-levelbenefitscommonlyseeninTaskPokerareasfollows:
Eachvoiceontheteamisheard.
Theteamhasasharedunderstandingoftheworkofthesprint.
Everyteammemberisinvestedintheoutcomeofthesprint.
CommonmistakeswhendoingTaskPokerincludethefollowing:
Theteamchoosestoaveragetheestimatesratherthantakingthetimetoreach
consensus.
Theteambecomesoverlygranularwithitsestimates.
EachVoiceIsHeard
Someteamsestimatetasksbyhavingeachpersonestimatetheworkthatheorsheismost
familiarwith.I’vealsoseencaseswherepeopleestimateworkintendedforothers,which
isworse.ShiftingteamstoTaskPokerfortaskestimationencouragesteammembersto
discussthetasksassociatedwithastoryandtheplanforcompletingthem.It’sokayif
somedon’tunderstandtheworkorhowtocompleteit.Adiversesetofinitialestimates
makesforgreatconversations—andsomeoneunfamiliarwithataskmighthaveavaluable
pointtomakeormightaskaquestionthatpromptsotherstorememberacomponentofthe
tasktheyhadforgottentoincludeintheirestimates.Atesterandauserexperience(UX)
person,forinstance,mightnotbefamiliarwithhowtowritethecodethatmakesadialog
box,buttheyaregoingtobeabletoprovidesomepertinentinsightintothetasksinvolved
incompletingitnonetheless.Everyvoiceshouldbeheardbecausethosevoicescan
exposehiddenassumptions,promptexplanationsthatremindpeopleofworktheymight
otherwiseoverlook,andhelpfurtherclarifytheoverallstory,allofwhichminimize
surprisesandplan-sizegrowth.
SharedUnderstandingoftheWork
Forateamtoreachconsensusonataskestimate,itsmembersmustfirstdiscusshowthey
intendtoapproachtheworkandwhatexactlyneedstohappenbeforethestorycanbe
considereddone.Thesediscussionsalsohelpidentifyneworpreviouslyhiddenissuesor
problems.Whentheteamcollectivelyunderstandsandacceptstheworkitneedsto
accomplish,itsmembersaremorelikelytoseeopportunitiesforchoosingtasksinwhich
theycanexpandtheirskillsbypairingwithsomeonewithmoreexpertise.Theoutputof
theplanningmeetingshouldbeastrategyforwhatwillbepairedversusnotpairedanda
listofconcernsandideastomitigatethem.
EveryTeamMemberIsInvestedintheOutcome
TaskPokerallowseveryvoicetobeheardandeverypersontounderstandtheworkthat
needstobedone.Oncethishasbeenaccomplished,theteamreachesconsensusonan
estimateforeachtask.Theseelementscombinetogiveeachteammembertheconfidence
tocommittotheworkofthesprintaswellastohaveastakeintheoutcomeofthesprint.
Don’tAverageTaskEstimates
Sometimesateamwillcometoanimpasseinestimation.Halftheteamestimatesfive
hoursforanitemandtheotherhalfsayseighthours.Often,Iseeteamsdecidetoendthe
stalematebyaveragingthecompetingestimatesandassigningthataveragetothetask(in
thiscase,a6.5).Don’tdoit.Theproblemwithusinganaverageestimateisthatit
circumventstheconsensus-reachingdiscussionsthatleadtomostofthebenefitsdetailed
above.Teamsthatfailtohavetheseconversationsalsofailtosurfacetherisksand
concernsorfailtotalkabouthowtheymightapproachthework.Inessence,theyare
missingthepoint.Thepointofcollaborativeestimationistheconversation,notthe
number.
AvoidExcessivelyGranularTaskEstimates
Anothercommonproblemtoavoidisbeingover-granularwithestimates.Someteams
haveproudlyboastedtomethattheyare95percentaccuratewhencomparingtheirtask
estimatestotheiractuals.They,too,aremissingthepoint.Don’tmistaketheneedfor
accuracyinestimationtobeaquestforprecision.Thegoalofestimationisnotto
preciselypredicthowlongtheteamwilltakeonanyonetaskbuttobeabletocommit
withconfidencetodeliverasetofstoriesandtheircorrespondingtasksduringasprint.
Theteamisn’tbeingheldtoitsestimates,it’sbeingheldtoitssprintforecastto
completingacertainincrementoffunctionality.Thisdistinctionissubtlebutimportant.
Onceagain,thepointofcollaborativeestimationisthattheteammembersactuallytalk
abouttheworkthatneedstobedoneinthesprint,whichwilluncoverassumptions,hidden
tasks,andareasofambiguity.
ScrumIsBuiltonTeamwork
AllScrumteamsfunctionattheirbestwhentheyareworkingtogether—fromestimation
toexecutiontodelivery.IfyourScrumteamisfailingtodeliver,considertryinga
collaborativetaskestimationtechnique,suchasTaskPoker.Encouragetheteamtopair—I
can’temphasizeenoughhowvitalitisforanyagileeffort.Considershorteningyoursprint
lengthandcreatingtimeboxestodevelopasteady,sustainablepacewithmultiple
opportunitiesbothforteampracticeandforcustomerandstakeholdercoursecorrection.
Limitworkinprocess.Itnotonlyeliminatesmanyoftheproblemsthatcomewith
multitasking,italsogivespeoplethemotivationandthecapacitytobandtogetherto
completetheworkofthesprint.Trackthatworkusingavisibletaskboard,prominently
displayedintheteamspace.
Inallthings,though,makeeveryefforttomovepeoplefromworkinginavacuumto
participatingasamemberofacooperativeteam.Whenyoudo,thedrivetodonewon’t
seemnearlyasdaunting.
References
[CLARK]Clark,KimB.,andStevenC.Wheelwright.1993.ManagingNewProductand
ProcessDevelopment:TextandCases.NewYork:FreePress.
[COHN]Cohn,Mike.2005.AgileEstimationandPlanning.UpperSaddleRiver,NJ:
PearsonEducation.
[GRENNING]Grenning,James.2002.“PlanningPoker,orHowtoAvoidAnalysis
ParalysiswhileReleasePlanning.”
https://renaissancesoftware.net/files/articles/PlanningPoker-v1.1.pdf(accessed25August
2015).
[OPHIRA]Ophir,Eyal,CliffordNass,andAnthonyD.Wagner.“CognitiveControlin
MediaMultitaskers.”ProceedingsoftheNationalAcademyofSciencesoftheUnited
StatesofAmerica.106(37):15583–15587.
http://www.pnas.org/content/106/37/15583.full?sid=da6eb41a-a4dc-4f40-ace7-
425c75c22789(accessed16August2015).
[SLIGER]Sliger,Michele.Personalcommunication,3September2015.
http://www.sligerconsulting.com
Chapter32.HowStoryPointsRelatetoHours
“Howmanyhoursdoesastorypointequal?”Igetsomeversionofthatquestionona
regularbasis.Nomatterthecompany,industry,orgeographiclocation,thedeeperconcern
isalwaysthesame:Whatexactlyistherelationshipbetweenstorypointsandhours?
Tohelpanswerthatquestion,IintroduceyoutoRachel,theproductowner,and
Jackson,theteam’sScrumMaster.WefindRachelandJacksoninameetingpriortoa
productbackloggroomingsession.Rachelissharingherconcernsaboutstorypointsin
generaland,morespecifically,abouthowtheteam’sstory-pointestimatesseemwildly
inconsistent.
TheStory
“I’mjustnotentirelysoldonstorypoints,”saidRachel.“Theyseemtoosoft.Ilikedthe
certaintythatestimatinginhoursgaveus.”
Jacksonwasabitperplexed.Overthepastfiveyears,hehadbeenonseveralteamsat
thiscompanythathadestimatedworkinhours.Everyoneofthoseprojectshadended
late,goneoverbudget,andultimatelyfailedtofullysatisfythecustomers.
“Rachel,”Jacksonsaidpolitely.“Itmightseemlikehoursgaveuscertainty,butIcan
assureyouthattheydidn’t.Imean,canyoutellmeoneprojectthatfinishedanywhere
nearitsestimate?”
“Well,no,”answeredRachel.“Butatleastwhenyougavemeanestimateinhours,I
knewwhatitmeant.Ithoughtstorypointsweresupposedtobemoreaccurate.Instead,it
feelslikefuzzymath.”
“Igetit.Iusedtofinditeasiertothinkinhours,too.Thewholeconceptofstorypoints
wasastruggleformeatfirst,butthenoneday,whenIwasmountainbikingwithafriend,
itallclicked.Yousee,mybuddyisareallygoodcyclist—I’monlyokay.Onthis
particularday,hepickedmeupandtoldmehewastakingmetohisfavoritetrailandthat
itwouldbeachallengingbutquickride,about30minutes.Whenwearrived,Irealized
thatIhadriddenthissametrailacoupleoftimespreviously.Iturnedtohimandsaid,‘30
minutes?Thisisatleastanhour-longride.’”
“Right.Youweretalkingintermsoftime,likenormalpeopledo,”saidRachel.“How
didthishelpyoufigureoutstorypoints?”
“Bearwithme—I’mabouttomaketheconnection,”explainedJackson.“Sowestood
thereatthetrailheadarguingabouthowlongtheridewasgoingtotake.Heexplainedthat
itwasahard,six-mileride,butthatitalwaystookhimabout30minutes.Iexplainedto
himthatthisparticularsix-mileridewasalmoststraightuphill,withatonofswitchbacks,
whichwasgoingtorequireeverybitofphysicalfitnessandskillthatIpossess.Iadded
thatthelasttimeIhadriddenthetrail,ithadtakenmeeverybitofanhour,ifnotmore.
Somewhereinthemiddleofourargument,itdawnedonmethatthiswaswhyteamsneed
storypoints—pointscapturetheamountofeffortastoryrequires,asopposedtothespeed
atwhichateamcanaccomplishthestory.Wecouldbothagreethatitwasarelatively
toughsix-mileride—whatwecouldn’tagreeonwassomethingmoreabsolute—exactly
howlongthesixmileswouldtake.”
“Sowhynotjustaverageitout?Whynotcompromiseandcallita45-minutetrail?”
askedRachel.
“Becausetimeisvariable.Howlongittakesanyriderwillvaryfromdaytoday,
dependingontheweather,therider’sfitnesslevel,howcrowdedthetrailis,andmore.
However,theeffortforthemostpart,willremainthesame.Andonceweagreeonan
effortestimateforthetrail,let’ssaywecallitatwo,wecanquicklydeterminehowthe
othertrailsintheareacompare.Asimilarsix-mileloopwouldbeatwo,butsomighttrails
thatarefifteenmileslongbutflatter,right?Oncewehaveacommonmeasuringstick,it’s
mucheasiertostartcomparingthevarioustrailsinthearea.”
“Thatmakessense—Icankindofseehowmeasuringeffortislesscontentious,and
ultimatelymoreconsistent,thandetermininghours.Plus,Irealizethatstory-point
estimatesarelessdependentonwhichteammemberisdoingtheactualworkandother
variables.Butifallthatistrue,whydoestheteamfinishforty-fivepointsonesprintand
thenstruggletocompletetwentypointsthenext?”Rachelasked.
“Ithinkthatisourbiggestissuerightnow—andissomethingweallagreeneedsto
improve.Becausetheteamisnewtostorypoints,we’restillprettyinconsistent.Someof
ourone-pointstoriesaretakingaslongasourthree-pointstories.Andatthesametime,
we’recompletingsomeofoureight-pointstoriesinthesametimeittakesustodoathree-
pointstory.
“Ithoughtweweren’tsupposedtoworryabouthowlongastorytakes,”Rachel
interjected.
“We’reonlysupposedtoworryaboutitinsofarasithelpsusmakemoreaccurate
estimates.LetmeshowyousomethingI’vebeenworkingon,”Jacksonadded,pullingup
adiagramonthescreen.
“Youaskedmeseveralsprintsagowhattherelationshipbetweenstorypointsandhours
was,”Jacksoncontinued.“Ithoughtitwasagoodquestion,sosincethenI’vebeen
chartingtherelationshipbetweenthenumberofhoursitactuallytakestheteamto
completeastoryandthatstory’sestimate.Ithoughtbydoingthis,I’dbeabletogiveyou
adefiniteanswer.Whatdoyousee,though?”
“Well,”saidRachel.“Itlookslikeaone-pointstoryhastakentheteamanywherefrom
onehourtofiftyhours.Butwhat’sweirdisthattwo-pointstoriesseemtotakebetween
twentyandfortyhours.Sothereisalotofoverlap.Atthesametime,though,thebigger
stories,asawhole,areconsistentlytakingmoretimetocomplete.”
“Exactly,”saidJackson.“Iseetwothings:first,astorypointimpliesarange—inthis
casearangeofhours—whichiscompletelyconsistentwitheverythingI’veread.Also,I
seethatrightnow,thoserangesoverlapalot,whichgivesussomeinsightintowhysome
sprintsweseemtogetatondoneandsomesprintswestruggletocompleteanything.”
“So,whatI’mhearingisthattheteamisbadatestimatingandthatstorypointsaren’t
helping.Sothatbringsusbacktomyoriginalquestion:Whycan’twejustusehours?”
“Ihearyou,”saidJackson.“ButIwouldsaythatweareallnoworseatestimatingstory
pointsthanwewereatestimatinghours.Infact,Isuspectwemightbealittlebetteratit.
Andkeepingthingsabstracteliminatesalotofthefearthatbogsotherteamsdown—fear
ofbeingheldtoacertaintimeestimate,oroftakingonastorythatmighttakeusalittle
longerthanitwouldanexpert.Usingstorypointsalsomakesestimationsessions,likethe
onewe’reabouttohave,gomuchmorequicklybecausewearen’tgettingcaughtupin
pointlessargumentsabouthowlongbutinmorerelativediscussionsabouthowmuch
effort.Thoseconversationsalmostalwaysleadtoouraskingquestionsthatuncover
assumptionsanddetails—stuffthatusedtostayhiddenwhenwejustaskedhowlong.”
“You’vegotapointthere,”agreedRachel.
Jacksonnoddedandcontinued.“Basically,storypointskeepusmovingandgetustoan
estimatethatisaccuratewithoutbeingoverlyprecise.However,trackinghowmanyhours
eachstorytakes,asI’vebeendoing,alsogivesustheopportunitytobemoreaccuratein
thefuture—whichwillbebetterforallofus.Withthisdata,wecanaskthreecritical
questions:Whichofthestoriesthatwehavecompletedsofarappeartobeoutsidethe
range—tohavehadaninaccuratestory-pointestimate?Areanyofthestoriesremainingin
theproductbacklogsimilartothemisestimatedstories?Ifso,shouldwechangethe
estimatesonthosestories?”
“IthinkIgetit,”saidRachelexcitedly.“Toreturntoyourbikeanalogy,let’ssayyou
gaveatrailafivebecauseitwasreallylongandyoufeltlikeithadalotofuphillportions.
Thenyouactuallyrideitanddiscoverthattheuphillportionsweremoregradualthenyou
thoughtandthatmostofthetrailispaved.Aftertheride,youdecidethatyoushouldhave
estimatedthattrailasathree.Youthentakealookattheothertrailestimatesyouhad
madeandrevisesomeofthem,keepinginmindthattheinclineandthesurfacearegoing
toimpacttheamountofeffortthetrailrequires.Youalsokeepthosefactorsinmindwhen
evaluatinganynewtrails.”
“Right.Becausetheyimplyarange,storypointsgiveusaccuracy.Wethentakewhat
welearnastheprojectprogressesandcombinethatknowledgewithourhistoricaldatato
improveonthataccuracyovertime,andtherangesgetsmaller.Aftereverysprint,we
needtoanswerthosethreequestionsImentionedearlier,andthenadjusttheplan
accordingly.Oncewegetinarhythm,Ithinkyou’llseethatwe’llbefarmorepredictable
thanteamsthatestimateeverythingupfrontusingdaysorhours.”
Andtheywere.Sprintaftersprint,theteamwasabletotightenitsaccuracyonstory-
pointestimates,enablingtheteammemberstobefarmoreconfidentintheirsprint
commitmentsandtoworkatamoresustainablepaceduringeachsprint.
TheModel
LikeRachel,manyproductowners,ScrumMasters,andteamsstrugglewiththe
correlationbetweenstorypointsandhours.Theyhavetroubleunderstandinghowtasks
relatetostorypoints,orwhyit’sadvisabletoevenusepointsinthefirstplace.Theywant
toknowwhatthepointsstandfor:time,complexity,orsomethingelse.
Icompletelyunderstand.WhenIfirstbeganusingScrum,ittookmeaboutsixmonths
towrapmyheadaroundusinganabstractmeasureasopposedtosomethingmorefamiliar,
suchashoursordays.WhatIlearned,though,wasthatstorypointshavedistinct
advantagesthatmakelearninghowtoworkwiththemcompletelyworthwhile.Hereare
thetopthree:
1.Thebuilt-inrangeshelpalleviatethefearofanestimateturningintoacommitment.
2.Therelativelywiderangethatstorypointsofferismoreaccuratethananarrowor
preciseestimate.
3.Talkingaboutstorypointsfocusestheconversationontheworkitselfinsteadofthe
timeinvolved.
FearFactor
Toomanytimes,asoftwareteam’sestimatesareturnedintocommitments—some
managerupthechainaddsupallofthehourestimatesandsays,“Okay,then,you’llbe
doneinxhours.”Then,whentheteamtakeslongerthanthatprecisenumberofhours,
managementdemandsanswers.Imagine,forexample,thatonMonday,youaskmehow
longit’lltakemetowritethischapter.Ithinkforamomentandthensay,aboutfivedays.
YoushowuponFridayanddemandafinishedproduct,whichIdon’thave.That’sbecause
whenIsaidaboutfivedays,Imeantsomewherebetweenfourandsevendays,depending
ontheotherprojectscompetingformytimeandwhatobstaclesIrunintowhen
researching,writing,andediting.Youtranslatedmy“about”into“exactly”andmadeita
commitment.IfI’dknownyouweregoingtoshowupinfivedaysandexpectafinished
product,Iwouldhavepaddedmyestimate—probablydoublingittosomethingliketen
days.Why?BecauseIknowI’mgoingtobeheldaccountabletoapreciseestimate.
Teamsdothesamething.Whentheirestimatesareroutinelyturnedintocommitments,
whichiseasytodowhentheestimatesareconcretemeasuressuchastime,teammembers
tendtohaggleovertheexactnumberofhours,outofafearofbeingwrong.They,likeme,
alsotendtoprotectthemselvesbypaddingtheirestimates.
Storypointshelpdiscouragethisbehavior.Whenteamsadoptstorypoints,their
estimatesaremoreabstract,partlybecausetheyarebasedonoverallteameffort.Inother
words,astory-pointestimateisnotaguessathowlongit’sgoingtotakeMaryorBobto
completeatask,it’sanestimateofhowmucheffortisinvolvedincompletingthetask,
regardlessofwhodoesthework.Determiningastory-pointestimaterequirestheteamto
talkaboutwhattasksareinvolvedindeliveringthecompletedstory,andwhennecessary,
togetclarificationfromtheproductowneraboutthestoryitself.Then,wheneveryone
understandsthestory,theteamcomestoconsensusonthestory’sestimate.Everyone—the
team,theScrumMaster,andtheproductowner—understandsthatthenumberofpoints
representsanaccuraterangeofeffortratherthanaprecisecommitmentastohowlongthe
storywilltaketofinish,resultinginteamsoftenarrivingatanestimatewithmuchless
argumentoraugmentation.
WideRanges
AsJackson,fromthischapter’sstory,showedRachelwithhischart,storypointshave
built-inranges.Rangeshelpteamsbemoreaccuratewithoutbeingoverlyprecise.After
all,wearen’tnearlyasgoodatguessingaswethinkweare,especiallywhenwehave
limitedknowledgeorunderstandingaboutwhatwe’reestimating.InthebookSoftware
Estimation:DemystifyingtheBlackArt,authorSteveMcConnelldiscussesashort10-
questionquizthatheusestotestaperson’sestimationaccuracy.Thequestionsrequire
peopletoprovideahighandlowestimaterangeforquestions—suchasthenumberof
milestothesunorthetotallengthofthecoastlineofthePacificOcean—withoutlooking
upthecorrectanswers.Heaskspeopletoprovidearangewideenoughthattheyare90
percentconfidenttherightanswerwillfallsomewhereintheestimatedrange.
I,liketheotherrespondents,didn’tdowellonthequiz.Myanswerforthetotallength
ofcoastlineforthePacificOceanwas5,000miles/8046kilometersto50,000miles/80,467
kilometers—arangeIthoughtwasalmostcomicallywide.Itturnsoutitwasn’twide
enough.Thecorrectansweris84,300miles/135,663kilometers.Ifailedtofactorinallthe
islands,theAntarcticcontinent,andmore.That’sjustoneofmanyquestionswhereI
guessedwrong.McConnellreportsthatmostpeoplewhentakingthisquizcorrectly
answeranaverageof2.8questionsoutof10.Fromtheseresults,McConnellconcludes
thatwhenapersonbelievestheyare90percentconfidentinanestimate,weshouldexpect
thattheestimatewillonlyberightabout30percentofthetime.Otherstudiesconfirmhis
findings[JØRGENSEN][ZULTNER].
RememberhowIthoughtmyrangeofanswerswasridiculouslywide—yetitturnedout
tobenotwideenough?McConnellassertsthatnoone’srangesarewideenough,because
(asofthiswriting)noonehasscoreda10/10.Fromthis,McConnellconcludesthatpeople
areconditionedtobelievethatnarrow-rangeestimatesaremoreprecisethanwide-range
estimates.Widerangesseeminaccuratebecausetheyaren’tprecise,butifyourknowledge
andunderstandingonlysupportawide-rangeestimate,thenawiderangeisactuallymore
accuratethananarrowrange.Narrowrangesoftenrestonanunsupportedprecision,
whichmiscommunicatesthedegreeofaccuracyintheestimate[MCCONNELL01,02].
Storypointsgiveteamsthefreedomtocommunicateawiderrangethantheymightbe
comfortablewithwhenusingameasureaspreciseastime.Inthischapter’sstory,Jackson
knewthisinstinctivelybutwashavingtroublecommunicatingittoRachel.Tobetter
understandtherelationshipbetweenstorypointsandhours,Jacksonbegancollecting
sprintdatathatcomparedhoursperstorytothestory-pointestimates.Tounderstandhow
thisdatahelpsteamsimprovetheirestimationaccuracy,andhowstorypointshavebuilt-in
ranges,lookfirstatasinglesprint’sworthofdatafromJackson’steam,showninFigure
32-1.
FIGURE32-1Asinglesprintviewintostory-pointandhourdata
Figure32-1showsthetotalactualnumberofhoursthatallofthetasksassociatedwitha
givenstorytooktocomplete.Forexample,aone-pointstorytooknearly50hourswhile
anotherone-pointstorytookabout10hours.Thismeansthat,forthissprint,theteam
completedtheirone-pointstoriesinarangeof10to50hours.Theteam’scompletiontime
fortwo-pointstoriesrangedfrom35to70hours.Thethree-pointstories,however,both
werecompletedatrightaroundthe75-hourmark,at73and75hours,whichisavery
narrowrange.Theoneandonlyfive-pointstorywascompletedin123hours,andthelast
story,aneight-pointstory,tooktheteamaround110hourstocomplete.
Thisinitialviewprovidesvaluableinformation.It’smuchmorepowerful,though,when
combinedwithdatafromseveralsprints.Atthatpoint,biggerpatternsstarttoemerge,
creatingamuchclearerpictureoftheteam’sestimationtrends.Figure32-2shows
multiplesprintsworthofdata.
FIGURE32-2Patternsemergewithdatafrommultiplesprints
InFigure32-2,youcanseethattheteam’shistoricalone-pointstoryrangeis~5to~50
hours.Similarly,thethree-pointrangeis~40to~85hours.Theadvantagetocompiling
thedatafrommultiplesprintsisthatitallowsateamtoseehowitisdevelopingand
improvingintermsofitsestimates.Awiderdistributionindicatesmorevolatilityand
inconsistencywithestimationaccuracy.
Inthekeystosuccess,Iexplainhowtocollectandusethisdata,butinessence,thegoal
isthatovertime,asateambetterunderstandsitsownestimationtendenciesandthe
projectitself,therangeswillnarrowandtheteam’sestimateswillbecomemore
predictable,likethepatternsinFigure32-3.
FIGURE32-3Astheteamlearnsmoreabouttheproject,therangesnarrow.
It’sHow,NotHowLong
Besideshelpingtoeliminatefearandhavingbuilt-inranges,story-pointestimatesalso
improveaccuracybecausetheconversationstheyfosterhelpteamsbetterunderstandthe
workofthesprintandthestoriesthemselves.Thinkingintermsofstorypointsallows
teamstostopfixatingonatimeestimate—decidinghowlongitmighttakesomeoneto
completethework.Instead,theteamsareabletofocusondeterminingwhattasksare
involvedincompletingastory.Todeterminethis,theteammemberstalkwitheachother
andtheproductownerabouthowtheymighttackleaparticularstory,whattechnologyis
involved,whathiddenissuesmightpopup,andwhatassumptionsmightneedtobe
fleshedout.Theseconversationsincreasetheteam’sknowledgeabouteachstory,which
willhelptomaketheestimatesmoreaccurate.
Forexample,ifI’dbeenworkingwithafewfriendstoestimatethelengthofthePacific
coastwhenItookMcConnell’squiz,someonemighthaveaskedaquestionIhadn’t
thoughtofonmyown,suchas“WhataboutAntarctica?”or“Whatabouttheislands?”
Theinformationwewouldhavegainedfromcollaborativelythinkingthroughtheproblem
wouldhaveimprovedourestimates—wemighthavesaid,“Fivethousandmilesis
ridiculouslylow.It’sgottobesomewherebetween25,000and100,000miles.”Andyou
knowwhat?Ifthathadbeenouranswer,togetherwewouldhavebeenright!
Thesamethinghappenswithteamswhentheycollaborativelyarriveatastory-point
estimate.Theconversationstheyhavemaketheirestimatesbetter.And,asIdiscussinthe
nextsection,knowingmoreabouttheirestimationtendenciesandhavingconversations
aroundwhycertainstoriesweremisestimatedcanhelpmakefutureestimatesmore
accurateaswell.
KeystoSuccess
Tosucceedwithstory-pointestimation,youneedtobeabletodotwothings:collectthe
rightdataandusethatdataintherightway.
CollectingtheRightData
Collectingthisinformationiseasy—itrequiresonlytwopiecesofdata:thestory-point
estimateatthetimethestorymovedfromtheproductbacklogtothesprintbacklogfor
completionandthesumofthehoursforallofthetasksperformedtoaccomplishthat
story.
First,createaspreadsheetsimilartotheoneinTable32-1.(Youcandownloadmineat
www.mitchlacey.com.)Then,pickastory-pointgroup.Foreachofthestoriesinthat
group,reviewandsumupthetotalhoursforeachtaskthatwasneededtoaccomplishthe
story.Recordthedatainthecorrespondingrowand,optionally,capturethesprintduring
whichthestorywasaccomplished.Thenmovetothenextstoryinthestory-pointgroup.
TheresultwilllooksimilartoTable32-1.Note,youdonotneedtousetheFibonacci
sequencetoestimatestoriesforthistowork;anystory-pointvaluesworkforthisexercise,
aslongastheyareconsistentsprintaftersprint.
TABLE32-1SampleDataCollectionTable
Iwanttoaddacautionarynoteaboutthisdata.Youarecollectingthisdataforone
reason:sothatyoucanshowtheteampatterns—timesthatastoryfelloutsidetherange
andopportunitiestoincreaseaccuracybycomparingfuturestoriestoonesthathavebeen
completedinthepast.Youshouldneverusethedatatoreprimandtheteamabouthow
longaparticularstorytookortochastisetheteamabouthowwrongtheestimateshave
been.
UsingDatatoMakeImprovements
Onceyouhaverecordedtheteam’shistoricaldata,youcancreateachartsimilartothe
oneshowninFigure32-4.Thechartisessentialinhelpingteammembersseetheoverall
patterns.Theycanthenidentifystoriesthattheymighthavemisestimatedandhavevital
conversationsaboutwhytheestimateswereoffandhowthatmightchangethewaythey
estimatefuturestories.
FIGURE32-4Overlayofcurrentsprintstoriestothehistoricaldatatodate
Thetimetohavetheseconversationsisintheproductbackloggroomingmeeting,with
theproductownerpresent.Startbylookingatdatafromyourlastsprint.Howdothis
sprint’sstoriesrelatetothehistoricalrangesforthesamesizestories?WhatIprefertodo
iscreateachartthatoverlaysthelastsprint’sworthofdataontoallofthehistoricaldata,
asillustratedinFigure32-4.
InFigure32-4,youcanseethatoneofthissprint’stwo-pointstoriesisoutsidethe
team’shistoricaltwo-pointrange;itseemsinsteadtobelonginthethree-pointstoryrange.
Similaroutliersexistinthefive-pointandeight-pointgroupingsaswell.Whenyousee
storiesthatfalloutsidethetypicalrangelikethese,asktheteammemberstotalkabout
thosestories.Askthemtodiscusswhythesestoriesrequiredmore(orless)effortthanthe
teamhadanticipated.Askthemwhethertheywouldhaveestimatedthemdifferentlybased
onwhattheyknownow.Thenask,Dowehaveanystoriesthatarestillintheproduct
backlogthataresimilartotheseandthatweshouldrevisitandre-estimate?
Theteammaydecidethat,whiletherearesimilarstories,thestory-pointestimates
shouldremainthesame;ortheymightdecidetoadjustcertainstories’estimates.You
shouldexpectre-estimationtobecommonatthebeginningoftheproject,becausethe
teamhasnotyetacquiredmuchknowledgeabouttheprojectitself.Asprojectsmature,
however,thepatternwillbegintostabilizeandlookmorelikethenarrow-rangeidealseen
inFigure32-3.
Remember,thevalueinestimatingorre-estimatingstoriesisnotfoundinthenumber
theteammembersassigntoanygivenstory,it’sintheconversationstheyhave,amongst
themselvesandwiththeproductowner.Theseconversationsfocusonuncoveringrisks
andremovingassumptions,notonfindingtheperfectestimate.
Onelastwordofadviceaboutusinghistoricaldata:Neverchangetheestimatesofwork
thathasalreadybegun.Onceastoryispartofasprint,regardlessofwhetherithasbeen
completedornot,itsestimateislocked.Onlyadjusttheestimatesoffuturestories,not
onescompletedorcurrentlyinproduction.
REFLECTonStoryPoints
Ihopethischapterhashelpedyourealizethatthereisnotadirectone-to-onerelationship
betweenstorypointsandhoursandthattherenevershouldbe.Whenworkingwithstory
points,IadvisepeopletoalwaysREFLECT:range,estimate,focus,learn,effort,
communicate,team.
Range:Storypointsimplyarange.Rangesprovideaccuracyandhelpkeepfearat
bay.
Estimate:Estimateandre-estimate.Reviewingtheproductbacklogandestimating
anynewstoriesis,atleast,aonce-per-sprintactivity.Remember,too,thatstory-point
estimatesarenotcommitments.Ifnecessary,teamscanchangetheestimateofany
storythatisnotcompleteorinproduction.
Focus:Onimprovement.Trackingtherelationshipbetweenstorypointsandhours
highlightsareaswhereteamscanfocustheirenergytohavemoreconsistent,
predictableranges.
Learn:Teamswillmakemistakes,andlearningfromthosemistakesiscritical;
applyingthelearningtofocusonimprovementisevenbetter.
Effort:Storypointsarebasedoneffort—notcomplexity,nothours,notdays.They
areanestimateoftheeffortneededfortheteamtodeliverthestory.
Communication:Storypointsfosterconversations,whichhelptoidentifyrisksand
removeassumptions.Communicateoften.Talkingthroughthestoriestogetherkeeps
everyoneonthesamepage,allowingtheretobeaone-teammindset.
Team:Continuouslyworktobuildaone-teammindset.Storypointshelpdrivecross-
functionalityandcreatecollectiveownershipacrosstheteam—elementsthatare
essentialforsuccessfulScrumteams.
Storypointsprovideenormousvalueand,whenusedcorrectly,offerarhythmical
cadencethatallowsteamstodeliversmoothlyandconfidently,timeaftertime.
References
[COHN]Cohn,Mike.2005.AgileEstimationandPlanning.UpperSaddleRiver,NJ:
PearsonEducation.
[JØRGENSEN]Jørgensen,M.,andD.I.K.Sjøberg,2004.“TheImpactofCustomer
ExpectationonSoftwareDevelopmentEffortEstimations.”InternationalJournalof
ProjectManagement22(4):317–325.
[MCCONNELL01]McConnell,Steve.2006.SoftwareEstimation:Demystifyingthe
BlackArt.Redmond,WA:MicrosoftPress.
[MCCONNELL02]McConnell,Steve.2015.Personalcommunication,8September
2015.
[ZULTNER]Zultner,RichardE.,1999.“ProjectEstimationwithCriticalChain:Third-
GenerationRiskManagement,”AmericanProgrammer,12(7):4–12.
Chapter33.ImmersiveInterviewingandHiring
Yournewhire,theoneyouspentweeksrecruitingandscreening,hasturnedouttobea
badfit.You’renowinatightspotandhavetodecidewhethertocutyourlossesorinvest
infixingtheproblem.Meanwhile,theworkyouhiredthepersontodoispilingupand
yourprojectissuffering.
Thegoodnewsisthatyoucanavoidthissituationifyouscreenforthecorrectthings
whenhiring.Thestorythatfollowsillustratesanimmersiveinterviewingstylethatcan
helpensureyouhiretherightpeopleforthejob.
TheStory
ScottandJaneen’sbusinesswasgrowing.Their100-personcompanyhadasolid
reputation.Thetroublewasthattheywerehavingahardtimekeepingupwithallthenew
businesstheyweregenerating.Itwastimetohiresomenewpeople.
PriortopartneringwithJaneen,Scotthadworkedatalargesoftwarecompany,where
hefeltthehiringpoliciesgotinthewayoffindingthebestpeople.Hewantedtoestablish
betterproceduresattheircompanywhileitwasstillsmallenoughtodoso.Hebroughtup
hisideasduringameetingwithJaneen.Shewasnotimmediatelyconvinced.
“Scott,I’mnotfeelingthesameurgencyyouareaboutrevampingourhiringprocess.
We’vebuiltagreatcompanybyhiringpeoplethisway.Whyfixwhatisn’tbroken?”
Janeenresponded.
“Iagreethatwe’vehadsuccesssofar,”repliedScott.“Aswegrow,though,youandI
won’tbeabletopersonallyscreeneverynewhire.Weneedtohaveascalablemodelthat
buildsonwhatwevalueasacompany.Thinkaboutit.Evenwithyouandmepersonally
screeningpeople,we’veendedupwithafewbadhires.RememberVictor?Thatguywas
awfulwhenitcametoworkingwithpeople.Sure,heknewhisstuff,buthemade
everyonearoundhimmiserable.”
“Ugh.Iremember.Hejustnevermadeaconnectionwiththeteam.Butwhenwegave
himsomethingtodothatdidn’trequirecollaboration,hewasfocusedandwegotalotout
ofhim.Wehiredhimforhisskills,nothispersonality,”JaneenremindedScott.
“Iagree,althoughevenwhenhewasfocused,hisattitudetowardpeoplewasawful.I’d
ratherhaveateamplayerwhocanworkwithothersandhavehalftheskillofVictorthan
gothroughthatagain,”explainedScott.
Janeennodded.“True.Andwhat’sworseisthatbadhirestendtosnowballintoother
losses.Whichremindsme,didIevertellyouthatVictorwastherealreasonwelost
Mark?”
“Nooooo!”exclaimedScottinsuddenrealization.
“Yep,heleftbecauseofVictor.Hedidn’ttellmostpeopleVictorwasthereason,buthe
didtellme.Hesaidlifewastooshorttodealwithpeoplelikethat,”explainedJaneen.
“Thatwasahardloss.We’restillfeelingtheimpactfromthat,”Scottsaidwithasigh.
“Iknow.Itwas.Butthethingis,Scott,wegetitrightnineoutoftentimes.I’mnotsure
wecandobetterthanthat.We’regoingtomakemistakes,”Janeenreplied.
“True,butremember,evenifweacceptthatwe’regoingtomisssometimes,we’restill
goingtoneedtoscalewhatwearecurrentlydoing.Right?Imean,we’regoingtohaveto
makechangeseventuallyanyway.I’dliketotrysomethingnewwhileyouandIcanstill
beinvolvedintheprocess,sowecanpersonallyseewhetheritworksornot.Now,I’ve
beengivingthisalotofthought,andIthinkIhaveawaytohirethatwillscreenforthe
pertinentitemsandisscalable,”Scottsaidenthusiastically.“Wanttohearaboutit?
“I’mintrigued,”saidJaneen.“Goon.”
“Weneedtostartwithwhatwevalueasacompany,”Scottexplainedashewrotethe
companyvaluesoutonthewhiteboard.
Scottcontinued,“Rightnow,asyousaid,we’refocusedoneverycandidate’stechnical
capabilities.Andweassumethatbecausewe’rehiringmostlysenior-levelpeople,they
haveasimilarsetofvalues.Butwedon’thaveanywaytoscreenforthem—nothingwe
dohelpsusdetermineifacandidatesharesourcompanyvalues.”
“True,true,”agreedJaneen.
Scottcontinued,“I’mconvincedweneedtolookatcompetenciesandvaluesasmuch
aswelookattechnicalskills.Technicalskillscanbelearnedinmonths,buttheabilityto
workwellaspartofateamissomethingthattakesaverylongtime.”
Janeennoddedinagreement.“Okay,sowhatdoyoupropose?”
“Well,Ithinkweshouldtakesomelessonsfromhowtheteamshavebeenpairingand
buildonthat.”
“Teaminterviews,then,”Janeensaidthoughtfully.“So,you’resuggestingthatthe
candidatespairwithexistingteammembers—doimmersiveinterviewing.Iliketheideain
concept,buthowwouldwestandardizethat—keepitobjectiveandmakesurewe’rebeing
consistentfrompersontoperson?”
“First,Ithinkwewouldneedtospendsometimetrainingtheteammembersaboutthe
processitselfandhowtheinterviewisakindofforecast.Weareforecastingwhetherthe
personwillbeagoodfitandhowlongit’lltakethatpersontobecomeacontributingteam
member,”saidScott.
“Yes,andwe’dneedtoshowthemthethingsweneedtoconsiderwhenmakingthat
forecast:thecompetencies,skills,andtheleveloftherolewe’rehiringfor—”added
Janeen.
Scottinterrupted,“Thelevel?”
“Yes.Doweneedsomeonejunior-level,senior-level—oriseitheroneacceptable
dependingonthecandidate?Thelevelofmaturityandexpertisewe’llexpectwillbe
differentdependingonthepositionwe’rehiringfor.”
Scottnoddedandadded,“Goodpoint.Andwe’dneedtounderstandandcommunicate
whywe’rehiring.Arethereorganizationalissuesthatareblockingprogress?Istheskillset
thecompanyislackingonethatcouldbefilledbyateamconsultant,ordowetrulyneeda
newcoreteammember?Doestheteamneedsomeoneinalearningroleoraleadership
role?Theanswerstothesequestionsdirectlyaffectwhowe’llhire.”
“Lastbutbynomeansleast,weneedawaytomakethesedeterminationsina
repeatable,consistentwayacrossthecompany,”addedJaneen.
“Yes.Absolutely.Weareinconsistentwithourapproach,andwetreatitmorelikean
individualburdenforanhourortwoonanygivendaythantheteameventitshouldbe,”
agreedScott.
“You’vesoldme,althoughIhaveanideaforhowtomoveforward.We’vegotthatone
teamwho’sbeenclamoringforanewmember.Maybewesitdownwiththeexistingteam
members,explainourgoals,andworkwiththemtocreateapreliminarymodel?Thenwe
runatrialwiththatteam—theycanbethefirsttohiresomeoneusingthenewprocess,”
suggestedJaneen.
“Ilikeit,”saidScottenthusiastically.“Let’sgetstarted!”
TheModel
Hiringpracticescomeinmanydifferentshapesandforms.Somecompaniesconduct
whiteboardteaminterviews,othershaveformalone-on-oneinterviews,whileothersstart
withcoffeeandasocialchat.
Howevertheinterviewisstructured,mostorganizationsfailtoaskthemostimportant
questions,notofthecandidatesbutofthemselves:
1.Whyarewehiring?Whatproblemsarewereallytryingtosolve?
2.Shouldweevenbehiring?Istherepotentiallyanimpedimentinsidethecompany
thatisgoingunresolved,andanewhireisonlythemask?
3.Whatarethemostimportantskillsandcompetenciestoconsider?
4.Howdowedecidewhotohire?
However,beforedelvingintothesequestions,Iwanttobeginwiththedefinitionofan
interviewthatScottputsforthinthischapter’sstory:aforecast.
Forecast
Toforecastistomakeapredictioninadvance.Whenhiringpeople,companiesmake
predictionsbasedonwrittenandverbalcommunicationastowhetheragivenpersonwill
beagoodfit,bothculturallyandskillfully,inthecompanyoveraperiodoftime.
Althoughsomeorganizationsdothisbyhavingatrialperiod,mostcompaniesinthetech
industryhiresolelyonthebasisoftheinterview.Thatmeanstheforecastmadeinthe
interviewhastobefairlyaccurate.Yetthisforecastisonlyasgoodasthecommon
understandingofwhyandhowtheorganizationishiring.
HirefortheRightReasons
Sowhyareyouhiring?Tomeetcurrentneeds?Becauseyouanticipateanewcontract?Or
areyoupartofahierarchicalorganizationandfeeltheneedtobuilduponeparticular
departmentorhavemoredirectreports?Somecompaniessimplygoonhiringbinges,
wheretheystartouthiringfortherightreasonsbutthencan’tseemtostop.Ithinkofthis
likeanicecreamorchocolatebinge.Youoverdoitbecauseittastesandfeelssogoodat
thetimethatyoucompletelyforgethowmuchpainyou’regoingtobeinlater.
Brooks’lawstatesthat“addingmanpowertoalatesoftwareprojectmakesitlater”
[BROOKS].Oneofthereasonsthishappensisthatsomecompanieshirepeopletomask
orcoveruplargerissues,whenwhattheyreallyneedtodoissolvetheunderlying
problem.Forexample,complaintsofnotenoughtestersisoftenasignofpoortest
automation.Reportsofnotenoughdeveloperscansometimespointtoalackof
collaborativeteams.I’veevenseenpeoplehireadditionalprojectmanagerstobringa
projectinontimewhentherealproblemislackofteamcommitment,goals,andbusiness
alignment.
Beforecreatingajobdescriptionormakinganexistingonepublic,makesurethata
newhireaddressestherealproblemyouaretryingtofix.Itdoesn’tmatterwhoyouhire,
goodorbad,ifyoudon’tunderstandthereasonyou’rehiringinthefirstplace.
CostofaBadHire
Everyonehasbeenpartofabadhire—eitherasthepersondoingthehiring,theperson
workingwithoronboardingthenewhire,orthepersonwho’sjustbeenhired.Everyone
knowsitfeelsbad,butwhatdoesitcost?Careerbuilder.com,alargeonlinecareersite,
runsfrequentsurveystofindoutjustwhatabadhirewouldcost.Theresultsare
staggering,asshowninTable33-1[CAREERBUILDER01,02].
TABLE33-1Percentofrespondentsthatidentifiedfactorsofabadhire
RememberVictorfromthischapter’sstory?Victorwasclassifiedasabadhirebecause
hecouldnotworkwellwithotheremployees.Tomakemattersworse,hisconductcaused
oneofthegoodhirestoleavethecompany.It’snowondercompanyculturesufferswhen
thewrongpeoplearehired!
Let’slookatthefactorstoconsiderwhenhiringthatmighthelpyouavoidthesecostly
disasters.
Skills,Competencies,orBoth
Let’ssayyourreasontohireisthatyouneedmoreC#programmers.Mostcompanies
wouldscreenprospectivecandidatesbasedonthatparticularskill:DotheyknowC#or
not?Whilethatmakessenseonthesurface,itmightnotbethebestwaytomakeahiring
decision.Skills,afterall,areeasytolearnandcanbepickedupinamatterofmonths.A
developerwhoknowsJavabuthasnevertriedC#willbeabletopickupC#afterafew
months,especiallywhenpairedwithanexperiencedC#programmer.Thisassumes,of
coursethatthedeveloperhasanabilityandwillingnesstolearn,toaskquestions,toshare
strengthsandweaknesses,toputegoaside,andtohavethecouragetolearnnewskillsin
ordertobesthelptheteam.
Let’sreversethat,then,andimaginewearelookingforaveryexperiencedC#
programmerbutallofthecandidatesareuncomfortableworkinginpairs.Thismeansthey
maynotwanttoimproveorbranchoutfromC#,resultinginlostopportunitiesfor
learningnewlanguages.Further,theymaynotenjoymentoringothersorpairing,meaning
nooneelsebenefitsfromtheirextensiveskillset.Nomatterhowseniororexperiencedthe
candidatesare,theywouldnotbeagoodfitonanyagileteam,becauseofthe
misalignmentofvalues.
Basedonthosetwoscenarios,screeningcandidatesshouldalwaysinclude
competencies—howtheythink,howtheyinteractwithothers,andwhattheyvalue.These
willimpacthowquicklytheycanassimilatewithateamandhowrapidlytheycanlearn
newskills.Afterall,youmightneedthesesamepeopletolearnyetanothernewlanguage
afewprojectsdowntheroad.Ingeneral,youshouldhireforlong-termfit,notforshort-
termexpertise.
HowtoHire
Inmytimeworkingforandconsultingatcompaniesbothlargeandsmall,I’veseenjust
abouteveryhiringstyleandjustabouteveryinterviewquestion.Somequestionswere
quiteobscure:Tellmeyourfavoritesportsteam.Otherquestionsweremorebasic:Doyou
haveafive-yearplan?
ScottandJaneenwantedtofindaninterviewstyleandquestionsetthatwasrepeatable,
consistent,andhelpedscreenforthecompanyvalues.Whattheydecidedtodowasto
applyagiletechniquestohiring—morespecifically,pairing.
Immersiveinterviewingisahiringandforecastingtechniquethatreliesonscreening
ratherthanaseriesofquestions.Itsharessomecharacteristicswithatechniquefrom
MenloInnovationscalledExtremeInterviewing[MENLO]butisnotnearlyasscripted.In
immersiveinterviewing,teamshavealistofitemstovalidatewithcandidates.Arethe
candidatesagoodculturalfit?Aretheyabletogiveandacceptcriticism?Arethey
comfortablewithconflict?Thesequestionsarebestansweredthroughimmersion.
Recallthataninterviewisaforecastandhowimportantitistomakeanaccurateone
whenhiring.Remember,too,thatintheAgileManifesto,thepreferenceisforindividuals
andinteractions,workingsoftware,andrespondingtochange.Why?Theyallleadto
success.Agilehiringshouldbenodifferent.Youneedtoseehowacandidateworks,
thinks,andbehavesintherealworld—themetaphoricalequivalentofworkingsoftware.
Youwanttoknowhowacandidatewillinteractwiththeothermembersoftheteam.And
youwanttounderstandwhetherthecandidateiscomfortableinanenvironmentof
feedback,continuousimprovement,andchange.
Whenitcomesdowntoit,traditionalapproachestohiringonlydemonstratehowwell
thecandidatecananswerthequestionsoftheinterview.Whilethoseanswersmightbean
honestreflectionofacandidate’sabilityandbeliefs,theyareanalogoustoanartificial
shell,showingtheintendedfunctionalityofasoftwarefeature,withoutanythingbacking
it.Idon’tknowaboutyou,butIalwaysfeelmuchbetteraboutmyhiringforecastifI
interactwiththepersoninreal-worldscenarios,withallteammembers.Research
confirmsthatIhaveagoodreasontofeelthisway.Ina2002whitepaper,Menlocitesthat
outof20peoplehiredthroughtheirExtremeInterviewingmethod,theylostonlyone
throughattrition[MENLO].Myexperiencehasbeensimilar.CompaniesthatIam
workingwithtodayreportmuchlowerattritionlevelsforteammembersthatarehired
withanimmersiveinterviewapproachasopposedtoastandardinterviewapproach.
ThepotentialforamoreaccurateforecastiswhyteamsI’veworkedwithpreferto
immersecandidatesinoneormoredaysinthelifeofaprojectteam.Thehiringmanager
oranHRpersonworkswiththeteamtoidentifywhatcompetenciesandskillsthe
candidateshoulddisplay.Theteamaskseachcandidatetopairwithteammembersinthe
teamspace.Theteammembersthenlookforevidenceofthenecessaryattributeswhile
workingwiththecandidate.Atthesametime,thecandidateisencouragedtowritecode,
test,anddocument—toactasateammemberfortheday.Candidatesshouldbeinvitedto
thedailystanduporthedemoorretrospectiveifthetimingworksout.
Thepointoftheimmersiveapproachistoputcandidatesintheenvironmentinwhich
theywillbeexpectedtoperform.Thisenablesteammemberstoevaluatethecandidates
beforereachingconsensusonwhichcandidatewouldbethebestfit,anditenables
candidatestoevaluatewhetherthecompanyisatruefitforthemaswell.
CandidateScreening
Inthestory,JaneenandScottagreedthattheircurrenthiringprocessdidnotscreenfor
competencies.Victorwasanexampleofsomeonewhohadslippedthroughthecracks.He
hadtherightskillsbutlackedcompetenciessuchasbeingopen-minded,beingableto
learnnewthingswhileteachingothersnewskills,andvaluingcollectiveownership.Many
organizationstrytosolvethisproblembydevelopingastandardquestionsettouse
companywide.IprefertoscreenbasedontheactionsandbehaviorsIobserveratherthan
onhowacandidateanswersasetofquestions.
Duringanimmersiveinterview,it’seasytogaugeacandidate’sskillsandexperience.
Codingnotuptosnuff?It’seasytosee.Troublewithtest-drivendevelopment(TDD)?
That’sobvious,too.There’snohidingwhenyourhandsareonthekeyboardandsomeone
iswatchinghowyouwork.Competenciesmightseemalittlehardertojudge,butthey
reallyaren’t.It’sjustamatteroflearningwhattolookforandcommentonthe
competenciesapersonexhibitswhileworkingwithyou.
Forexample,let’ssayScottandJaneen’scompanyishiringanewcandidateforafive-
personteam.Thecandidatewillbescheduledtoworkinapairwitheachmemberofthe
teamthroughouttheday.Duringthattime,theteammembersarenotgoingtobeasking
typicalinterviewquestionssuchas,“Tellmeaboutatimeyoudemonstratedawillingness
tolearn”or“Whenmakingmistakes,whatdoyoudototakeaccountabilityforthem?”
Thatwouldbeawasteoftime.Instead,theteammemberswillbeabletowitnessfirsthand
whetherornotthecandidateseemswillingtolearnandwillobservewhathappenswhen
thecandidatemakesamistake.
PreparationandSetup
Ifthisisthefirsttimeyouhavedonethistypeofimmersiveinterviewing,ensureyou
reviewyourcompanyvalueswiththeteammemberswhowillbepartoftheinterviewas
wellaswithanyfunctionalhiringmanagerorHRmanager.Makesureeveryoneisonthe
samepageastowhyyou’rehiring,whateachofthecompetenciesyouarehiringfor
mean,andwhichspecificbehaviorstheteamshouldlookforduringtheinterview.Having
acompanyleaderworkwiththeteamtoclarifythehiringrequirementswillhelpreduce
riskandallowtheteamtoaskquestionsandgetclarificationonwhattheyarelookingfor,
ensuringthescreenisthesameforallcandidates.Thissetofcompetenciesandbehaviors
mightevenbesomethingyoustandardizeandwritedown,sothattheteammemberscan
refreshbeforeeachcandidate.Then,priortotheinterview,eachteammemberwillpick
oneortwoskillstopayspecialattentionto.OnepersonmightchooseC#skills,while
anothermightchooseTDD.Eachteammember,though,shouldbelookingfor
competenciesandhowthecandidatefitsthecompanyvalues.
Alsokeepinmindthedifferencebetweenhiringsenior-levelpeopleandjunior-level
people.Wheninterviewingseniorpeople,youwouldexpectacertainlevelofmaturity,
expertise,professionalism,andinsight.Theprimaryfocusofthescreen,then,istoexpose
momentsofunprofessionalism,discomfortwhenworkingaspartofateamorwithothers,
andsimilarsignsofimmaturity.You’llalsohaveachancetoseetheirtechnicalexpertise,
includingthewaytheyapproachproblemsandtheiropennesstofeedback,duringthe
pairingsessions.Ifindthatmosthigh-performingcompanieswouldratherteach
knowledgeableandexperiencedpeoplenewtechnicalskillsthantrytochangetheir
mindsets,nomatterhowtalentedtheymightbe.
Ontheotherhand,ifyouareinterviewingjuniorpeople,perhapsfreshoutofschool,
youshouldbemoreforgivingoflackofexpertise.Whatyouwouldwanttoseetechnically
isabasiclevelofcompetencecoupledwithawillingnesstolearn,abilitytocommunicate,
andexcellentproblem-solvingskills.Whileyouwouldgenerallyexpectmoreimmature
individuals,youwillstillneedtomakesuretheyareabletoacceptfeedbackandworkas
partofateam.
Thatbeingsaid,rememberthatthereisactuallysomeadvantagetothelackof
experiencefoundwithjuniorpeople.Experiencecansometimesblindpeopleandforce
peopleintoruts.Ofcourse,senior-levelpeoplecanalsobeopen-mindedandattack
problemswithoutlettingexperiencegettingintheway.That’swhyitissoimportantto
hireindividualswhopossessthewillingnesstolearn,theabilitytoteach,theopennessto
shareknowledge,andthedesiretohelpcreatealearningorganization—nomattertheir
experiencelevel.
ScoringCandidates
Throughouttheimmersiveinterview,thecandidatewillworkwithmultiplepeople,each
ofwhomislookingspecificallyatafewchosenskillsandatalloftherequired
competencies.Attheendoftheday,theteamwillmeettodiscussthecandidate(s),butit’s
criticaltocaptureeachinterviewer’simpressionsinrealtimeaswell.
Tocollectallofthedatafromthevariousinterviewers,useanonlinetable,suchasthe
oneshowninTable33-2,andascaleof1through10,where1istheleastoftenobserved
skillorcompetencyand10isthemostoftenobserved.Thesearesubjectivenumbers.It’s
agoodideatodiscusswhateachvaluemeanssothattheteamhasarelativelyconsistent
understandingofwhatconstitutesa9ascomparedtoa2.
TABLE33-2ExampleofScoringCandidates
Asthecandidateleavesoneteammemberandmovestothenext,theinterviewershould
takeamomenttojotdownscoresthatcorrespondtohowthecandidateperformed.The
chartwillnotbecompletelyfilledinattheendoftheinterview,butyouwillhaveenough
datatohaverelevantdiscussionstodetermineifthecandidateshouldbehiredornot.
ThinkofthislikePlanningPokerbutnotinrealtime.Eachteammembershareshisor
herownestimateofthecandidate’sfitineachofthekeyareas.Then,oncetheinterview
ends,theteamgetstogethertohavetheall-importantconversation.Ifthereareoutliers,
suchasJaneen’sratingof2forC++inTable33-2,theteamshouldtalkaboutthereasons
fortheextremedifferenceofopinionandtrytoreachaconsensus.Ifalltheratingsare
similar,suchastheonesforArchitectureanddesignandDoesn’thaveasenseof
entitlement,thediscussionwillbemoreaboutgeneralobservationsthanreaching
consensus.
Regardlessofwhetherthenumbersmatchorarewildlydifferent,thepointofthetable
istoserveasareal-timedatacollectorforfutureconversations.Donot,underany
circumstances,usethetabletomakeahire-or-no-hiredecisionwithoutalsohavinganin-
depthteamdiscussion.Otherwise,youmightmisssomethingvital.
HiringManagersandNontechnicalPeople
Itmightseemasthoughthisapproachworksonlyfortechnicalhires,butitalsoworks
whenselectingnontechnicalpeople.
Whenusingimmersiveinterviewingtohirenontechnicalpeople,ensureyouhavea
well-definedsetofcompetenciesthatareunderstoodbyall.Animmersiveexperiencefor
anontechnicalhirewouldincludeattendingmeetings,participatinginbrainstormor
creativesessions,ormaybeevendoingasmallportionofapresentation.Yes,thereare
riskshere,butremember,wearelookingforanimmersiveenvironmentthatwillhelp
everyonearriveatanaccurateforecastofeachcandidate’scompetenciesandskills.
KeystoSuccess
WhenValveSoftwarepublisheditsemployeehandbookinMarch2012,itsettheInternet
abuzzwithhowValveoperatedasacompany.Thecompanyisflat,everyoneis
accountableandresponsible,andpeopleowntheirownfates.Infact,Valveputssomuch
emphasisonhiringthatthehandbookevenstates,“Hiringwellisthemostimportantthing
intheuniverse.Nothingelsecomesclose.It’smoreimportantthanbreathing”[VALVE].
That’spowerfulstuff!
Hiringpeopleforanyrole—manager,ScrumMaster,productowner,teammember,or
architect—ischallenging.Hiringmistakesnegativelyimpactthevaluesand/orcultureof
thecompany.Tohelpmitigaterisks,aimtobuildarepeatableprocessthatfocuseson
findingstrongpeoplewithcertainkeycompetencies,developingskillswhennecessary,
andmakingsmartinvestments.
BuildaRepeatableHiringProcess
Havingaconsistenthiringprocessiscrucial.Theprocessshouldalwaysstartby
answeringthequestion,“Whyarewehiring?”Beginningwithwhycanhelpeliminate
costlymistakes,suchasaddingpersonnelinanattempttosolveadeeperproblemthat
mightexistintheorganization.Forinstance,iftheteamisfailingtodeliver,itmightnot
bebecausetheteamneedsmorepeople;itmightbethattheteamneedstoworkon
collaboration.Whenyoubeginwithwhy,youensureyouarehiringfortherightreasons.
Arepeatableprocessisalsorootedinthecompanyvaluesandprinciples.Thelargera
companygets,however,theharderitmaybecometoensurethatnewhiresalignwiththe
companyvaluesandprinciples.Consequently,it’sagoodideatoholdregular“Howto
hireatourcompany”trainingsessionstohelpkeepallteamsfreshanduptodate.This
helpspreventbadhiresandalsorefreshesandrefocusescurrentemployeesonthe
companycultureand,moreimportant,thecompanyvalues.
Finally,arepeatableprocessisoneinwhicheveryoneunderstandswhattherating
systemmeans,andthat,aswithPlanningPoker,theinitialratingsaremeanttobethe
basisfordiscussion,notastandalonemeasureofacandidate’sworth.Thevalueofthis
kindofimmersiveinterviewcomesfromtheconversations—thosethathappenamongthe
teammembersandthosethathappenwiththecandidate.Immersiveinterviewshavethe
addedbenefitofensuringthatpeople,oncehired,arefamiliarwiththecompany’shiring
process,astheyhavegonethroughit,whichaddstoitsrepeatability.
FocusonCompetencies,NotonQuestions
Interviewquestionsaregreat,andhavingagoodsetofthemisdefinitelyhelpful.
However,runninganinterviewbyjusthavingthecandidatesanswerquestionsandsolve
problemswillonlytellyouhowwelltheycaninterview.Whatyoureallywantisan
accurateforecastastoanindividual’sfitasateammember.Screeningcandidatesinan
immersiveenvironmentandfocusingoncompetencieswillhelpincreaseyourconfidence
inthatforecast.Makesureallteammembersknowandunderstandthecompetenciesthey
shouldbelookingfor,bothatateamlevelandatacompanylevel.Youmightevenwant
todomockinterviewworkshopswhereanexistingteaminterviewsanotherteam(and
viceversa)toteachpeoplehowtofocusoncompetencies.
SkillsAreEasytoLearn,CompetenciesAreNot
WhenIwasachild,Iwasdeathlyafraidofpublicspeakingoranyformofpresenting.I
wasgood,however,atresearchandanalysis.ThepapersIwroteforschoolalways
receivedgoodmarks,allthewaythroughmycollegecareer.ItwasnotuntilIwasinmy
late20s,however,thatIbecamecomfortablewithpublicspeakingandpresenting.To
becomeabetterpresenter,Ireadnumerousbooksonhowtoovercomefearsandspent
yearspracticing.Publicspeakingisacompetency—ittakesagreatdealofpracticeanda
certainmindsettobeabletodoitwell.
Skills,ontheotherhand,canbetaughtratherquickly,especiallygiventhatmost,ifnot
all,ofyourhireswillhavesomesortofbackgroundinthehiringarea.Forexample,
perhapsthecandidateknowsOraclesystemsandyouarehiringforaMicrosoftstack—the
candidatewon’thavetoovercomeachildhoodfearoradeeplyseatedhabittolearn
Microsoft.Forsomeonewithatechnicalbackground,thisskillwouldberelativelyeasyto
learn.Ofcourse,thatassumesthatthepersonhasthecompetencyofwillingnesstolearn.
Havingtheabilityandcouragetosay,“Idon’tknowhowtodothat,willyoupleasehelp
melearn?”isacompetencythattakesyearstomasterandcanonlybepracticedsafelyina
culturethatsupportsit.
SeeChapter3,“UsingTeamConsultantstoOptimizeTeamPerformance,”formore
aboutcompetenciesandskills.
FindPeopleStrongerthanYou
Insecurityandpoliticscansometimesmakeitdifficulttohireaperson,evenwhenyou
knowtheyhavegreatskillsandcompetencies.I’veseenpeoplethrowcommonsenseout
thewindowandignoreacandidatewhoseems“toostrong”orcomesfrom“so-and-so’s
department.”Why?Perhapsthepersonposedathreattooneoftheinterviewers,wasn’t
specializedenough,orcamefromadepartmentwherethetwodepartmentheadsdonot
getalong.Don’tfallintothistrap.Hiringstrongpeoplewillallowyou,yourteam,andthe
restoftheorganizationtogrow.
UnderstandtheCostsandInvestHeavily
AsValveSoftwaresaid,“Hiringisindeedmoreimportantthanbreathing”[VALVE].A
badhirecansucktheairrightoutoftheroomandhavedamagingeffectsonthe
environment.Investingoodhiringpracticesandcreateimmersiveexperienceswherethe
teamcandetermineifthecandidateisagoodfitforthecultureoftheteamandcompany.
Pushaccountability.Haveaunitedapproachsoyoucanavoidthemistakesthattraditional
companiesmake.Hiringispainstakinglydifficult,butusingthisapproach,oneofmany
outthere,willdefinitelyhelpincreaseyourchancesofagreathire.
References
[BROOKS]Brooks,Frederick.P.,Jr.1996.TheMythicalMan-Month,Anniversary
Edition.Reading,MA:Addison-Wesley.
[CAREERBUILDER01]CareerBuilder.com.2011.“TrueCostofaBadHire.”
http://thehiringsite.careerbuilder.com/2011/12/16/true-cost-of-a-bad-hire/(accessed1
August2015).
[CAREERBUILDER02]CareerBuilder.com.2012.“WhatBadHiresReallyCost
Companies(Hint:It’sMorethanJustMorale).”
http://thehiringsite.careerbuilder.com/2012/12/13/cost-of-a-bad-hire/(accessed1August
2015).
[MENLO]MenloInnovations.2002.“ExtremeInterviewing.”MenloInstitute,LLC.
http://www.menloinnovations.com/by-reading/PDF/Extreme-Interviewing-Final.pdf
(accessed1September2015).
[VALVE]ValveSoftware.2012.HandbookforNewEmployees.Bellevue,WA:Valve
Press.http://www.valvesoftware.com/company/Valve_Handbook_LowRes.pdf(accessed
1August2015).
WorkConsulted
Rothman,Johanna.2012.HiringGeeksThatFit.Vancouver:Leanpub.
Chapter34.AligningIncentiveswithOutcomes
Weseeitallthetime.Thebusinessgetsagoodleadandthenextthingyouknow,thesales
teamiswavingasignedcontractintriumph.Thedevelopmentteams,ontheotherhand,
areleftwithstatementsofworktheydidn’tagreeto,moreprojectsthantheycanhandle,
anddeadlinesthatareimpossibletomeet.There’sadisconnecthere—onethatputs
differentgroups,companydepartments,andpeopleatodds,andultimatelyresultsin
dissatisfiedclients.
Thesolution?Reconnectthedotsbybringingdevelopmentintheprocessearlier,
keepingsalesinvolveduntiltheworkiscomplete,andfindingnewwaystomanagethe
entireprojectportfolio.
TheStory
BrentandAllywereknownas“TheClosers.”Asaccountmanagers,theybothexcelledat
gettingcustomerstosignonthedottedline.Overthepastfewweeks,BrentandAllyhad
beenworkinghardonaparticularlyfinickynewclient.Thecontractcyclehadbeengoing
onforwhatseemedtobeaneternity,nearlytwomonths.Theyeachhadsalesquotasthey
neededtohitandtheendofthefiscalquarterwasapproachingquickly.
“Whatarewegoingtodo?”askedAlly.
“Don’tworry.We’regoingtomakeit.Icanfeelit—theyaresoclosetosigning,”Brent
statedreassuringly.
“That’snotwhatImean,”explainedAlly.“Togetthisdealclosed,we’vehadtobe
reallyaggressivewithourestimatesandwe’vepromisedacoupleofthingsIdon’tknowif
thedevelopmentteamcanevendo.Thelasttimewedidthat,thedevelopmentteammade
uspromisewe’dneversellthemoutagain.Ifeelasiftohitournumberswehavetothrow
thesoftwareguysunderthebus.It’llbehardtofacethemlater.”
“Ithinkyouworrytoomuch,Ally,”Brentanswered.“Theyjustliketocomplain.They
aregoodatwhattheydoandhavedonethiskindofworkahundredtimesover.They
mightwhineabit,buttheycanhitthesedeadlineswithoutaproblemiftheypushit.”
AtAlly’sskepticallook,Brentcontinued,“Look,weallhavetoworkhardaroundhere.
YouandIhaveputinatonoftimewooingtheseclients.Heck,weevenwenttheextra
mileandtalkedtoourdirectorofsoftwarearchitecture.Heconfirmedthattheestimates
wehaveinthecontractwork,andheshouldknow,right?Pushingtheenvelopeisjust
whatwehavetodotobringinthebusiness.Ninetypercentofoursalaryiscommission—
noneoftheirsis.Wehavetodowhatwehavetodo.”
Allynoddedreluctantly.“You’vegotapoint.Plus,wedon’twanttopromiselessthan
ourcompetitors.Let’sgetthisdoneandaskforforgivenesslater.”
BrentandAllygotthecontractsignedwithinweeksofthisconversationandreceived
theircommissionchecksshortlyafterward.Whenthecontractedstatementofworkwas
senttothedevelopmentteamfordelivery,thereactionwaslessthanfavorable.
“Ican’tbelieveit!”exclaimedPatrice.“Doyouseethedatestheyexpectustohit?”
“Theynevereventalkedtous!”agreedRick.“Nowwehavetoworkontheirdeadlines,
whichwillbeimpossibletohitunlesswework,what,sixtyhoursaweekforthenextsix
months?Plus,thismeansyetanotherprojecttoaddtoourgrowingbacklog—I’m
multitaskingsomuchthathalfthetimeIcan’trememberwhichproductI’meven
building.IthinkI’mgoingtohavetoquit.”
Justthen,Allyhappenedtowalkby.
“Ally!”calledPatrice.“Comehereaminute.”
Allybackedupandpokedherheadintheteamspace.“What’sup,Patrice?”
“Ally,howcouldyoudothistousyetagain?Wejustgotassignedtotheprojectyou
andBrentwereworkingon.We’vealreadygotmoreprojectsthanwecanhandle—and
thiscontracthasestimatesanddatesthatarejustplainwrong!Whodidyoutalkto?”
“Weranitbythedirectorofsoftwarearchitecture,andhethoughtitwasreasonable.I
knowyou’vegotalotonyourplates,butit’sourjobtobringinthework,”explainedAlly.
“Iknowsomeofthedeadlinesareaggressive,butIalsoknowyouguys.Yourteamis
awesome.Iknowyoucanpullthisoff.”
“Ally,”saidRick.“Youcan’treallybelievethat.Youknowthere’snowaywecanwork
onathousandprojectsatonceorhitthesedateswithoutburningourselvesout.Thishasto
stop.Wecan’tworklikethis!”exclaimedRick.
Brentwalkedbyandheardthechatter.HestoppedandcametoAlly’sdefense.“You
knowwhat?Ifwedidn’tgetthiscontract,youguyswouldn’tbeworkingatall.Wedid
whatwehadtodotogetthedealdone,wesecuredthebusinessthatthecompanyaskedus
tosecure,andweevenrantheestimatesallthewayuptothedirectorlevel.Idon’tknow
whatmoreyouwantustodo.”
“Iwantyoutogobacktomanagement,thestakeholders,andthecustomerandtellthem
thiswon’twork,”snappedPatrice.
“Sorry,nocando.Onceit’ssigned,it’soutofourhands,”Brentreplied.Heturnedto
Allyandcontinued.“Ally,thereasonIstoppedbyisthatIneedyoutohelpmeprepfora
callIhaveintwentyminutes.Canyoucometomyoffice?”
“Neitherofyouisgoinganywhere!”shoutedRick,ashepoundedhisfistonthetable,
hischeeksredwithfrustrationandanger.“Youcan’trideoffintothesunsetwithyourbig
fatcommissioncheckswhileweslaveawaymeetingcrazydeadlinesonyetanother
projectyou’vepiledonourbacks!”
“That’scrazytalk,Rick,”retortedAlly.“Weearnedeverydollarofourcommissions
justlikeyouearneverydollarofyoursalary.”
“Rickhasapoint,”Patriceinterjected.“I’mnotsuggestingyoudon’tdeserveyour
commissions—youdo.You’reouttheresellingandmakingthepromisesyouneedto
maketoclosethedeal—that’syourjob.Butthatleavesusonthehookforfulfillingthat
contract—evenwhendoingsoisimpossibletodo.Somethinghastochange.Doyouguys
rememberlastmonthatourbigcompanymeeting?TheVPsaskedallofustocometo
themwithideasforwaystomakeourorganizationmorecohesive—fromdevelopmentto
salestoaccountmanagement.Ourdisagreementheremightbeanopportunityindisguise.
Ally,wouldyoubewillingtomeetwithmelatertodaytoseeifwecanfigureoutsome
sortofmodelthatcouldstarttobridgethedisconnect?Maybetogetherwecancomeup
withaplanthatwecantaketotheCTOandtheVPofsales.”
Allythoughtforamoment,thensaid,“Yes,Patrice,Iwould.I’mnotsurethatthere’s
muchwecando,butI’mwillingtositdownandhashitout.”
“Youtwogoforit,”saidBrent.“Ihaverealworktodo.”
Afewweekslater,AllyandPatricemetwiththeVPofsalesandtheCTOtopresenta
modeltheythoughtmightwork.
Whentheywerefinished,theCTOsaid,“So,letmeseeifIunderstandthis.Rightnow,
wehavetwoproblems.One,thedevelopmentteamisn’tinvolvedatallinthesale.Two,
thesalesteamisn’tinvolvedatallinthedelivery.Youareproposingthatweinvolvethe
developmentteamintheprocessearliertohelpmakeestimatesmorerealistic—as
opposedtorelyingononeortwoexperts.Youalsowanttomodifythesalesteam’s
commissionstructure,whichrightnowissolelyfocusedonmakingasale,ratherthanon
delightingourcustomers,whichweallagreeisourultimategoal.Youwanttotiemoreof
thecommissiontocustomersatisfactionandholdbackalargepartofthepayoutuntilthe
projectisdelivered.Bydoingthis,you’dthinkthesalesteamwouldhavemoreincentive
tosellsolutionsthatthedevelopmentteamcandeliver—notjusttopromisewhateverit
takestoclosethedeal.”
“Right,”saidAlly.“We’reateam—andourprocessandincentivestructureshould
reflectthat.”
“Exactly,”agreedPatrice.“Weknowyouneedtoclosecompetitivedeals.Andyou
knowthatwecan’tmeetthemwithoutworkingcrazyhours,andeventhen,whatweturn
outisusuallynotthequalitythatwe’dlikeittobe.”
“Andthatleavesuswithunhappycustomers,”addedAlly,“whichmakessellingrepeat
businessandgettingreferralsharder.”
“Anotherreasonourcustomersareunhappy,”continuedPatrice,“isthatwehavean
unrealisticprojectload—sowecan’tevertrulyfocus.We’reconstantlybeingpulledoff
oneprojectforafiredrillsomewhereelse.Orwe’remultitaskingsomuchthatwenever
getanythingdone.That’snotthesalesteam’sfault—butit’ssomethingthatneedstobe
addressedattheexecutivelevel.Thismodeladdressesthatproblem,too.”
Ultimately,theVPofsalesandCTOagreedtoinstitutethenewmodelasatrial.After
someinitialresistance,andwithtime,education,andpatience,thenew,alignedmodel
eventuallybecamethewaythecompanydidbusiness—evenBrent.
TheModel
Inthestory,wesawalackofalignmentbetweenseveralgroups,eachwithitsown
incentives,andassuchitsowndefinitionsofsuccess.Thesalesandaccountmanagement
peoplewerefocusingononething—thesale.Managementwantedtobringinmoreand
morebusinesstoseethecompanythroughtheinevitableleantimes.Thedevelopment
teams,ontheotherhand,wererewardedfor,andthereforeworriedabout,successful
deliveries.Thegroupsweremisalignedbecausetheydidnotsharethesamegoalsor
motivations.
PatriceandAllysuggestedanend-to-enddeliverymodelthatwouldaligntheentire
organizationaroundonecommongoal—delightingcustomersandstakeholders.
SettheFocus
Forabusinesstobesuccessful,itmustbefocused,internallyaligned,andbalanced.To
achievethisidealstate,thebusinessneedsanincentivesystemthatrewardsoutcomes(the
waythingsturnout)morethanoutputs(whatwasproduced).Inthestory,thebusiness
desiredmanydelightedcustomersandhighlevelsofcustomersatisfaction,bothofwhich
areoutcomes.Yetthesalespeople,BrentandAlly,wererewardedsolelyforreelinginnew
customersandgettingcontractssigned,bothofwhichareoutputs.Likewise,the
developmentteamwasexpectedtodeliverafixedscopeofworkontime,oftenfor
multiplecustomerssimultaneously,whicharealsooutputs.Moreoftenthannot,theend
resultwasdisjointedprojectoutput,poorquality,andalopsidedburdenonthe
developmentteam.
Inessence,despitethecompany’sdesireformanydelightedcustomersandhigh
customersatisfaction,thebusinessofferednoincentivestoanyoneforachievingthose
goals.Asyoumightexpect,then,thebehaviorofindividualswastiedtotheirown
incentivesratherthantothoseofthebusiness.Whenthissituationoccurs,everyoneloses.
AlignaroundCustomerSatisfaction
Businessesareinthebusinessofsatisfyingcustomers.Greatbusinessesareinthebusiness
ofdelightingcustomers.Thismeansemployeesneedtobealignedtotheoverarchinggoal
ofcustomersatisfaction,andtheirincentivesandoutputsneedtoreflectthatalignment.
Inpracticalterms,salespeopleneedtoberewardednotjustforclosingdealsbutfor
closingdealsthatresultinasatisfiedcustomer.Thedevelopmentteamsneedtobe
rewardednotjustforshippingproductsontime,butforshippingproductsthatdelight
customers.Leadershipteamsneedtoberewardedformanagingaprojectportfolioinsuch
awaythateverycustomerfeelsvaluedandreceivestheproducttheydesire.Inthis
chapter’sstory,PatriceandAllycreatedamodeltoincentivizebehaviorsthatwillleadto
customersatisfaction.Partoftheirsolutionwastotiesalescommissionstocustomer
satisfaction.Theyalsoproposedinvolvingthedevelopmentteamearlyinthesalesprocess
sothattheycouldhelpthesalespeoplemakemoreaccurateforecasts.
Involvingdevelopmentteamsduringthesalescycleallowsthosedoingtheworktoget
abetterunderstandingofcustomerneeds,wants,anddesires.Duringthesalescycle,
developmentteamsworkwiththecustomertobuildaproductbacklog.Theyhave
conversationsdesignedtorefineunderstandinganduncoverhiddenrequirementsand
stories.Forthisarrangementtowork,thedevelopmentteamsshould,ideally,havedirect
accesstothecustomers,andmorespecifically,totheendusersofthesystem.Ifdirect
accesssimplyisn’tpossible,eithergetthemascloseasyoucanormakesureyouhavea
productowner,oraproxy,whohasin-depth,detaileddomainknowledgeofthebusiness.
Andalwaysremember,thefurthertheteamsarefromthecustomersandstakeholders,the
moreambiguitytheremaybeintheproductbackloganditsstories.
Oncetheproductbacklogisbuilt,thedevelopmentteamthenestimatesitandprovides
thesalespeoplewitharangeofcostandtimeestimates(seeChapter11,“Release
Planning,”formoreonthistopic).Thecustomerneedstobeeducatedastowhatrangeof
theproductbacklogwilldefinitelybedelivered,whatmightbedelivered,andwhat
certainlycannotbedelivered,andtobegiventhecorrespondingcostanddeliverydate
ranges.Thegoalofeducatingthecustomeristoreduceriskandambiguityacrossthe
board.
Atcontractsigning,thesalespeoplearepaidsomecommissionforcraftingthedeal,but
themajorityofthepayoutistiedtothecustomer’ssatisfactionwiththeincremental
releasesattheendofeachsprint.Dependingonhowyourcontractsarestructured,the
customersmightpayaftereachsprintonlyfortheworktheteamaccomplishesand
deliversormightpayatlargermilestones.Regardlessofthecontractstructure,the
salespeoplereceivesomepercentageoftheircommissionsateachcustomerpayment
installment,accordingtothelevelofcustomersatisfaction.IpreferaNetPromoterScore
(NPS)formeasuringcustomersatisfaction,andIliketoincludeanNPSmodelmeasure
[NPS]inthecontractsothatthecustomersunderstandhowthebusinesswillbemeasuring
theirsatisfaction.
Oncetheprojectiscompletedandfinaldeliveryismade,thesalespeoplegettheirfinal
commissioncheck.Thiswouldalsobeagoodtimetogivethedevelopmentteamabonus
aswell,shoulditbedeserved.
Thismodeliseffectivebecauseeveryonehasavestedinterestinthecompany’sdesired
outcome—many,highlysatisfiedcustomers.
PrioritizeandShift
Aligningincentivestooutcomesisonewaytoensureyourbusinessisinbalance,but
othertypesofmisalignmentsalsointerferewiththeoverallbusinessgoal—todelight
customers.Aprojectportfoliothatismisalignedwithteamcapacityisoneofthem.You’ll
knowthereisaproblemwhenindividualgroups,especiallydevelopmentgroups,
complainofbeingoverworkedandnothavingenoughtimetodotheworktheyneedto
do.Anunbalancedprojectportfolioleadstomultitasking,aspeoplejumpfromprojectto
projecttotrytomakesomeprogressoneach.Italsogoesagainsttheagileprincipleof
sustainablepaceinagileteams.Unchecked,multitaskingquicklyleadstoburnoutand
fatigue.Soon,teamsbegindoingwhateverisneededtorelievethepressure,whichoften
leadstopoorqualityandlowcustomersatisfaction(seeChapter23,“SustainablePace”).
Itispossibletohavemultipleprojectsorcustomersandstillalignteamstothebusiness
goalofcustomersatisfaction,allwhileremovingtheproblemsbroughtonbymultitasking
acrossmultipleprojects.Thefirststepistoprioritizeyourprojectportfolio.Youcandoso
inoneoftwoways.Thefirstistocreateastack-ranked,prioritizedlistoftheimportance
ofallprojectsinsidethecompany.Startbycreatingalistofallcompanyprojectsandput
themonaboard.Next,determinethekeyperformanceindicatorsorbusinessobjectivesof
eachproject.Isoneofyourbusinessobjectivestoincreasemarketshare?Maybeaproject
thathasbetterpotentialforincreasingmarketshareshouldtakeprecedenceoveranother
withalessimportantbusinessobjective.Isbeatingthecompetitioninaspecificmarket
segmentabusinessobjective?Isthatbusinessobjectivemoreorlessimportantthan
increasingmarketsharegenerally?Usethisinformationtohelpmovetheprojectsaround
inthestack.
Anotherwaytoprioritize,ifyouareavendorformultiplecompanies,istoconsider
whatismostimportanttoeachcustomer.Timetodeliverymightbeofutmostimportance
toonecustomer,whilecostmightbethedrivingfactorforanother.Onewaytoprioritize
thesecompetingproducts,then,couldbetoofferthesecondcustomeralowercostbuta
longerdeliverywindow.Orchargethefirstcustomerapremiumforanexpediteddelivery.
Orevenboth.Inotherwords,findoutwhatmattersmosttoyourcustomersandhave
conversationswiththemupfrontabouthowyourteamscandeliverthatoutcomeforthem.
Buildintheflexibilityyourleadershipteamneedstoprioritizethemosttime-sensitive
projectsfirst.
Onceyouhavetheprioritizedprojectportfolio,thecompanycanbegintomake
decisionsabouthowmanyprojectsareinprogressatonetimeandalsoaboutwhoworks
onwhatbasedonbusinessvalue.Forexample,supposetherearetwo,multiteamprojects
goingon;oneisaveryhighpriority—theotherlessso.Suddenly,anurgentissuecomes
uponahigher-priorityproject.
Usingamoretraditionalapproach,managementmightshiftrandomindividual
resources,otherwiseknownaspeople,toaddressspecificproblemsthatthehigher-
priorityprojectisencountering.However,byhavingaclearpictureofwhatisimportant,
thelower-priorityprojectinsteadcanshiftanentireteamtohelpthehigher-priority
project,leavingtheotherteamstocontinueonthelower-priorityproject.Thissolutionis
similartotheconceptofswarming.Typically,whenateamswarms,alloftheteam
memberssimultaneouslyworkonthesametaskuntilthattaskiscompletethenmoveon
tothenext.Theteamrepeatsthisuntilthestoryiscomplete.Shiftinganentireteam
allowstheteamtofocusondeliveringadditionalstoriesforthehigher-priorityprojectthat
maybestruggling.It’sswarming,scaledtotheteamandprojectlevelversusthetask
level.Justremembertoalwayskeeptheteamtogether.
Becausetheprioritiesareclearlyvisibletoallinvolved,theteamsadjustwithout
incidentorcomplaint.Plus,bymovingentireteams,youavoidtheslowdownscausedby
shiftingindividualsontoandoffofexistingteams.(SeeChapter20,“AddingNewTeam
Members,”formoreontheeffectsofdisruptingteams.)
Whathappens,then,tothelower-priorityproject?Perhapsit’sputonhold,orperhaps
thescheduleshiftstoaccommodatethelossofoneoftheteams.Itcouldbethatthe
scheduleremainsthesamebutlessisdeliveredinthesametime—thescopeisalteredto
fitreality.Theimportantthingtounderstandisthebusiness,fromendtoend,hasaclear
pictureofwhichprojectsaremostimportanttorealizingbusinessobjectives,whichteams
areworkingonwhichprojects,andhowmuchavailableteamcapacitytheoverallbusiness
has.
SideBenefits
Imentionedearlierthatanunbalancedprojectportfolioleadstomultitasking,causedby
peoplejumpingfromprojecttoprojecttoputoutfires,fixemergencies,andusetheir
specialized,subject-matterexpertiseacrossallprojectsinanunscheduled,ad-hocfashion.
Creatingaprioritizedportfolioandfocusingonshiftingteams,ifdoneright,can
eliminatethemultitaskingthatisalltoocommonwhenthenumberofprojectsoutweighs
teamcapacity.Bydoneright,Imeanbyshiftingteams,notpeople,usingteamconsultants
tohelpgrowdedicatedteamskillsanddomainexpertise,andprioritizingonthebasisof
theendgoal,whichisdelightingthecustomers(seeChapter3,“UsingTeamConsultants
toOptimizeTeamPerformance”).
KeystoSuccess
Aligningabusinesscanbeachallenge.Manybusinessesarematrixed,segmented,
disjointedmesses.Peopleintheseenvironmentswilldowhatisnecessarytoachieve
successintheirownjobs,evenifthatmeanssacrificingothersinthecompanyoreventhe
customers.Afocusonalignmenthighlightsthewaysinwhichcompanyincentivesdonot
matchstatedgoalsandvalues.Creatingincentivesthatalignwiththesevaluesandgoalsis
goingtotakeexecutive-levelsupport,organization-widecoordination,andacommitment
todedicatedteams.
IntegrateSalesandDevelopment
Alignmentstartswithintegratingdevelopmentteamsintothesalescycle.Youmaynot
haveafullteamthatisabletoparticipateinthesalescycleinthebeginning,butthat
shouldbethegoal.Startbybuildingaproductbacklog.Havetheproductownerbringin
theteamtoworkhand-in-handwithcustomersandstakeholderstowritethestoriesthat
makeuptheproductbacklog.Thenallowtheteamtoestimatethosestoriesandhelp
createaninitialreleaseplan.Then,andonlythen,havetheproductownerorthe
salespeopleusethisinformationtodeterminearealisticprojectcostandtimeframe
(providedyouknowtheteam’svelocity),whichtheycanthencommunicatetothe
customerasarangeofoptions.
StopSacrificingPeopleandQuality:HaveaPrioritizedProjectPortfolio
Don’tassumethatallprojectsarecreatedequalandhaveequalimportance.Theyarenot.
Havediscussionsaroundwhatisimportanttothecustomersorstakeholders.Isittime?Or
scope?Orcost?Wouldthecustomerbewillingtoletthescheduleslideifitmeansalower
projectcost,astheteamsareabletofitintheworkaroundwhenthehigher-prioritywork
isdoneorisnotunderactivedevelopment?Maybethecustomercanonlyburnasmall
amountofbudgeteachmonth,resultinginalessaggressiveschedule.Whicheveritis,
workwiththecustomertodeterminewhatismostimportanttothem,allowingyouto
prioritizetheworkintheprojectportfolioaccordingly.
AddDedicatedTeams,NotRandomPeople
Whenthingsstarttoheatup,it’stemptingtoaddindividualstoaproject.Don’t.Ifyou’re
goingtoadd,thenaddteams.Ateamhasalreadyestablishedhowitsmemberswork
together.Addingateammeansitsmemberscantakeonproductbacklogitemsendtoend.
Iftheteamisnotfamiliarwiththeproject,addateamconsultantortwotohelpthemget
uptospeed(seeChapter3).
Forthismodeltobetrulyeffective,developmentteamsmustbeallowedtostay
togetherratherthanbeingswitchedoutwitheverynewproject.Theresultingstability
makestheteammorepredictableandbetterabletorespondtoissues,suchasshiftingtoa
newprojectunderduress.Ithastheaddedbenefitofcreatingalong-termrelationship
betweenthesalesanddevelopmentteams,whichhelpstogetworkintothequeuewitha
higherdegreeofconfidenceintheoverallabilitytodeliver.
Organization-wideCoordination
Oncethemodelisupandrunning,you’llneedtoimplementthescrumofscrummeetings
forcross-coordinationbetweenteams,aswellasdailymeetingsamongproductowners.
Theaccountmanagementandsalesteamsactasstakeholders/customerstothe
developmentteamsandshouldbeinvitedtojoinbothmeetings.Ifthereisanissueora
concernaroundtimeordeliverables,theproductownerscanworktoaddressit.
ExecutiveSupport
Becausethismodelrequireschangeinmanygroupsacrosstheorganization,it’s
imperativetogetleadershipandexecutivesupport.Startbyhighlightingtheproblemof
misalignedincentives.Explainhowteamsburnouttryingtohitunrealisticdeadlinesthat
wereoftenmadebywell-intentionedsalespeople.Discusstheimportanceofworking
withinteamcapacity,evenifthatmeanssacrificinglower-priorityproducts.Itmighttake
aslongassixmonthstogetthebuy-inyouneed,butthealignmentthatoccursasaresult
iswellworththeeffort.
Reference
[NPS]TheNetPromoterCommunity.“TheNetPromoterScoreandSystem.”
http://www.netpromoter.com/why-net-promoter/know(accessed20August2015).
Chapter35.RiskManagementinScrum
Ontraditionalprojects,theprojectmanagerisresponsibleforriskmanagement.Scrum
projectsdonothavethattraditionalrole,butthatdoesn’tmeanthatrisks,ortheneedto
managethem,disappear.Instead,asprojectmanagerLeelearnsinthischapter’sstory,
everyoneontheScrumteamhasaparttoplayinmanagingandmitigatingrisks.
TheStory
Leehadhistoricallybeenoneofthemostrespectedprojectmanagersinthecompany.His
workethicandhisabilitytohitthetripleconstraint—getprojectsinontime,onbudget,
andonscope—waslegendary.Hewasalsoagloballyrecognizedriskmanagementexpert
—andhadthecertificationstoproveit.Overtheyears,hehadspokenonthetopicofrisk
insoftwareprojectsatconferencesaroundtheworld.Despitehissuccess,Leewouldbe
thefirsttoadmitthathisprojectstendedtohavealargenumberofchangeordersthat
cameinlateinthegame.Leealsoknewthathisstandingwiththedevelopmentteamsleft
somethingtobedesired.Hewasnotoriousforburningpeopleoutinhisquesttodeliveron
thetripleconstraint.Theseshortcomings,however,wereovershadowedbythenotionthat
hehadsavedthedayforthecustomersandstakeholderstimeandtimeagain.Or,atleast
thathadbeenthecaseinthepast.
Timeschanged,though.Andaspeoplebecamemoreburnedoutandcustomers
demandedmorefrequentreleaseswithaneasierchangeorderprocess,Leeandother
projectmanagerslikehimbecamelessdesirable.Thecompanyhadshiftedalmostallof
thedevelopmentteamstoScrumandexpectedanyonewhowaspartoftheorganizationto
workinamoreagilemanner.Leeremainedoneofthelastholdouts.Hesawquiteabitto
likeaboutScrumbutwasconcernedbywhatheviewedasScrum’slaissezfaireattitude
towardrisk.Forhim,thethoughtofgivingupresponsibilityandcontroljustseemed,well,
toorisky.
HiscolleagueTerryhadbeenoneofthefirstpeopleatthecompanytoadoptScrum.
LeerespectedTerryandwantedtoknowwhyithadbeensoeasyforhimtogetonboard
withagiledevelopment.So,onedayheinvitedTerryoutforacupofcoffeesotheycould
talkitover.
“Ijustdon’tgetit,”saidLee.“Onatraditionalproject,I’mtheoneresponsiblefor
managingrisk,forensuringthatthingsaregettingdonecorrectly,andformanagingthe
overallriskprofilefortheproject.Agileprojectsseemtothrowallformalrisk
managementoutthewindow.”
“Iknowitprobablyseemsthatway,butthatisn’tthecaseatall,”explainedTerry.“For
example,givemethreetypesofriskyoumanageonaregularbasis.”
“Hmmm.Well,thebiggestoneisprobablykeepingthecustomersandstakeholdersin
check.Theyalwayswanttoslipfeaturesintoaprojectwithoutchangecontrol,andI’ve
gottolockthatdown.Otherwise,we’relateandoverbudget.”
“Okay,”saidTerry,“Whatelse?”
“Thenextbiggestoneisensuringtheteamisontrackwithdeliveringtheirtasks.Ineed
tokeepthescheduleuptodateandensurethatwe’renotslipping.Itypicallypublishmy
scheduleintheweeklystatusemails,somanagingscheduleriskiscritical,”saidLee.
“Thisisgreat!Givemeonemorebigone,”Terryreplied.
“Ithinktheotherhugeoneismanagingquality.Isetthequalityrequirementswiththe
customersearlyinprojectinitiation,andIconstantlyneedtomanagethetesterstoensure
theyfollowthoserequirementsandmeetthequalitybarexpectedbythecustomers,”said
Lee.
“Believeitornot,allthosetypesofriskmanagement—andmore—areaninherentpart
ofScrum.Theyarejustspreadoutamongthedifferentroles.Forexample,theteamis
responsibleforquality,andtheScrumMasterforensuringtheteamcollaboratessothat
qualityisbuiltintotheproject.Theteamisalsoresponsibleformeetingtheirsprint
commitments,whichhastodowithschedule,”explainedTerry.“I’lltellyouwhat:Ifyou
meetwithmelaterintheweek,IcanmapeveryriskyoucanthinkoftoaScrumteam
role.”
“Let’sdoit,”agreedLee.“Ineedtounderstandhowwholeteamriskmanagement
works.”
TheyagreedonatimelaterintheweekforTerrytoworkwithLee.Terryexplained
howhemanagedriskonprojectsandshowedLeehowtheScrumrolescollectively
managedriskandhowagile,inandofitself,wasonebigriskmanagementtechnique.
OnceLeerealizedhowmuchriskmanagementwasbuiltintotheScrumframework,he
feltbetteraboutgivingScrumatry.Beforelong,hehadtransitionedintooneofthemost
engagedproductownersinthecompany.
TheModel
LikeLee,manyprojectmanagershaveatoughtimemakingthetransitiontoScrum.They
struggletounderstandhowScrummanagesalloftherisksthattraditionallyhave
consumedprojectmanagers.Consider,forexample,thelistoftypicalsoftwareproject
risksillustratedinTable35-1.
TABLE35-1TypicalSoftwareProjectRisks
Theserisksdon’tjustgoawaywhenateamtransitionstoScrum.Scrumhandlesallof
theserisksandevensomethathistoricallyhaven’tmadethelist—suchasdysfunctional
teams,badhires,andmisalignedHRpolicies.Theseareallfactorsthatcandisrupta
softwareprojectbutthatmaynothavebeenconsideredpartoftraditionalrisk
management.Scrumdoesn’tignorerisk,itjustdividesitamongthedifferentScrumroles:
ScrumMaster,productowner,andteam.Ialsoliketocategorizeeachidentifiedrisktype
intooneofthreecategories:social,customer,andtechnical.Table35-2illustrateshow
softwareprojectrisksaretypicallydistributedamongthesethreerolesandcategories.
TABLE35-2ScrumRiskManagementbyRole
Thecustomerrisksarehandledbytheproductowner,thetechnicalrisksbytheteam,
andthesocialrisksbytheScrumMaster.
Forexample,ScrumMasterstakeonalloftheresponsibilitiesassociatedwithbuilding
andmaintaininghealthyteamsandremovingimpediments.Productowners,ontheother
hand,spendmuchoftheirtimewiththecustomer,whichmeansthemoretraditionalrisk
managementitemsofbudgetandscopefalltothatrole.Whentheyspendtimewiththe
team,productownersarerefiningtheproductbacklogtoensurestory,schedule,and
budgetrisksarebeingmanaged,whichfreestheteamtomanagethetechnicalrisks.
WhatthisclearlydemonstratesisthatScrumprojectsdon’tignorerisks;theyjustdivide
themupabitdifferently.Let’sdiveabitdeeperintothedifferentriskcategories.
CustomerRisk:ProductOwner
Customerriskincludesalloftherisksthathavetodowiththecustomer—specifically,
time,scope,andmoney.InScrum,productownersmanagethecustomerrisks.Tohelp
addressandmitigaterisks,productownersmeetwithstakeholdersoftentoreview
functionality,discussandupdatereleaseplans,reviewprojectbudgets,andmore.Product
ownersalsoworktounderstandcustomerneedsandidentifycustomerdelighters.(The
roleofproductowneriscoveredinmoredetailinChapter5,“ImplementingtheScrum
Roles.”)Productownersareconstantlyobservingthebehaviorofthestakeholdersand
customerstoensurethatallisgoingwell.Somecommonrisksthatproductownersneed
tokeepaneyeoutforincludethefollowing:
Customers/stakeholdersaren’tshowinguptothesprintreviewmeeting.
Customerssaythattheydon’tknowthestatusoftheproject.
Customersaresurprisedinthesprintreview.
Theproductbacklogcontainsmorestoriesthancanbeaccomplishedwiththe
currentprojectbudget.
Customersradicallychangeprojectdirectioneverysprint.
SocialRisk:ScrumMaster
Socialriskhastodowithpeople,especiallytheteamandhowitsmembersareworking
together.Inaddition,asChapter8,“TheCaseforaFull-TimeScrumMaster,”
demonstrates,socialriskalsoincludesissueswiththoseoutsidetheteamwhomightput
theprojectatrisk,includingproductowners,customersorstakeholders,otherteams,and
companymanagement.
Traditionalprojectmanagersmaynotconcernthemselveswithteamhealthorteam
happinessthewayaScrumMasterdoes.Yet,asshowninChapter23,“SustainablePace,”
teamsthatworkatasustainablepaceperformmuchbetterthanteamswhoareconsistently
workinginoverloadedconditions.OneoftheScrumMaster’smanyjobsistoprotectthe
socialinteractions,emotions,andoverallstabilityoftheteam.
ScrumMastersneedtobevigilantaboutlookingforbehaviorsthatcouldindicatea
developingsocialrisk,includingthefollowing:
Teammemberswhoshowuplatetothedailyscrum
Aproductownerwhodistractstheteamwithnewfeaturerequestsmidsprint
Teammemberswhoavoidpairing
Ateamthathastoomuchworkinprogress
Ateamthathands“alltheworktothetesters”lateinthesprint
TechnicalRisk:Development(Core)Team
Technicalriskisjustwhatitsoundslike—riskwithatechnicalfocus.Ifindthemost
commontechnicalrisksresidewiththeteammembersthemselves.Ifsomeorallofthe
teamislackingknowledgeofthesystemunderdevelopment,theprojectcanquicklygo
south.SomeofthebestwaystomanagetechnicalriskaredescribedinChapter9,“Why
EngineeringPracticesAreImportantinScrum,”andChapter19,“KeepingPeople
EngagedwithPairProgramming.”
Technicalriskfactorsusuallyincludearchitecture,design,systemandothertechnical
documentation,testing,devops,andmore.Signsoftroubleoftenincludethefollowing:
Toomuchmanualtesting
Lackofsystem,code,anddesigndocumentation
Integrationissues
Environmentinconsistency(e.g.,build,test,development)
Risinglevelsoftechnicaldebt
IntheEnd…
ThesampleriskindicatorsarejustafewofthemanybehaviorsthatallScrumteam
membersshouldconsiderwhenmeasuringthehealthoftheproject.Thegoalshouldbe
workingtowardapotentiallyreleasableproductincrementeverysprintthatmeetsthe
definitionofdonecriteria.Teamsthatachievethisgoalestablishapredictablerhythmthat
makeseveryoneinvolvedfeelmoreconfidentabouttheproject’sultimateoutcome.
KeystoSuccess
Scrumdoesn’tignoretraditionalprojectrisks;instead,riskmanagementisdividedamong
theScrumroles.Thisideacanbechallengingtograspforsomewhohistoricallyhave
beenincontrolofeverythingthroughouttheprojectlifecycle.Ihaveseenproject
managers,developers,testers,andevensoftwarearchitectsstrugglewiththisconcept.
TohelpmakethetransitiontoScrumeasier,Ioffertwokeystosuccess:First,letgo,
thenletagile.
LetGo
Iwastrainedasatraditionalprojectandprogrammanager.Itwasaprimarythrustofmy
universitystudiesandlaterIearnedvariousprojectmanagementcertifications.Iwas
expectedtomanageteamsfromdesigntorelease,whichincludedmanagingalltherisks
associatedwiththeirprojects;theprojectplan,theworkbreakdownstructure,the
estimation,thedesign,everything.Iwasaccountablewithoutauthority.
IstruggledwithlettinggowhenIfirstadoptedScrum.IbelievedthatifIdidn’t
personallymanagetherisks,noonewould.IquicklyrealizedthattosucceedwithScrum,
Ineededtochangemythinking.Thisfeltliketheequivalentoflearningtorideabicycle
alloveragain.IhadtoletgoofcontrolandtotrustthepeopleandteamsIworkedwith.It
wasdifficultandtooktime.SomedaysIstillstrugglewithmycontrolissues.
Lettinggoforatraditionalprojectmanagerwillbeachallenge.Ifyou’restruggling
withchangingyourmindset,readbooksbyauthorssuchasCarolDweck,whowrotethe
bestsellingbookMindset:TheNewPsychologyofSuccess.
Butaboveallelse,trustyourteam.You’reallinthistogether.Thewholeteammight
havetotakealeapoffaith,butbelieveme,thesooneryoucanletgoofthepast,thefaster
youcanembraceyourfuture.
LetAgile
Oddlyenough,eventhoughthischapterdiscussesriskmanagementandhowtodivide
risksupamongrolesandbreakthemdownintocategories,agiledevelopment,inandof
itself,isentirelyariskmanagementtechnique.
Frequentreleaseshelpcustomersdeterminewhattheymeanwhentheyaskfor
something.Workingincross-functionalteamsdrivesknowledgesharingand
accountability,removingtheriskofasinglepersonretainingalltribalknowledgeandthen
leavingthecompany.Reviewingtheproductbacklogwithstakeholdershelpsrefinethe
releaseplaninrealtime.Sprintreviewsandretrospectiveshelpidentifyopportunitiesfor
improvement.Andmostimportant,theconversations—betweenteammembers,
stakeholders,customers,andothers—helpresolverisksandissuesandensurethat
everyoneisonthesamepage.
Intheend,agile—andbyassociation,Scrum—isallaboutmanagingriskthroughthe
useoffrequentinspectandadaptcycles,fromsprintplanningtothedailyscrumtothe
sprintretrospective.Letgoandletagileworkthewayitwasdesignedto.Youwillsoon
findthat,likeme,youaremuchbetterofftrustingyourteamsthanyouweretryingto
controleverythingsinglehandedly.
WorksConsulted
Dweck,Carol.2007.Mindset:TheNewPsychologyofSuccess.NewYork:Ballantine
Books.
Rubin,Kenneth.2013.EssentialScrum.UpperSaddleRiver,NJ:Addison-Wesley.
Appendix.ScrumFramework
Scrumisaframeworkformanagingcomplexprojects(seeFigureA-1),suchassoftware
development.Scrumusesaprescribedsetofroles,meetings,andartifactsanddividesa
largeamountofworkintoone-weektoone-monthchunkscalledsprints.Itisbasedon
andemploysagileprinciples(seeChapter1,“Scrum:Simple,NotEasy,”formoreon
Scrum’sunderlyingprinciples).
FIGUREA-1TheScrumframework(downloadacopyathttp://mitchlacey.com)
Iliketothinkofthebalancebetweentheroles,themeetings,andtheartifactsthatmake
upthesprintcycleaslikethatofafinelytunedracecar.Tiny,incrementaladjustmentscan
leadtoexceptionalwins.Atthesametime,aslightknockorissuewiththecarorits
enginecanleadtolessoutput,slowerspeeds,andrapidenginewear.Howdoesthis
analogyrelatetoScrum?Thefollowingsectionsexploreeachcomponentindetail.
TheRoles
Scrumconsistsofthreeroles:theScrumMaster,theproductowner,andthedevelopment
team(Ialsorefertothedevelopmentteamasteamorthecoreteamthroughoutthebook).
Togethertheyworkintheinterestsofthecustomersandstakeholderstoturnthevision
intoaworkingproductorservice.Iftheywerepartsofaracecar,thedevelopmentteam
wouldbetheengine,theproductownerthedriver,andtheScrumMasterallthelubricants
andsensors.
ScrumMaster
Aracecarhasgaugesandsensorstohelpmonitortheengineandusesoiltohelplubricate
it.Withoutoil,theenginewouldgrindtoahalt,destroyingitselfintheprocess.Theoilis
everywhere,keepingthepartsoftheengineworkingsmoothly,coolingthem,andensuring
theyperformwellunderstress.
TheScrumMasterislikethat.Hehasembeddedsensorsandgaugesthatallowhimto
identifywhentheteamisnotperformingtoitsability,andhehastheskills(lubrication)
neededtoassistincorrectingtheissues.BeingtheScrumMasterisahardjob.Agood
ScrumMasterissomeonewhonoticesnonverbalcues,iscomfortablewithconflict,isan
effectivecommunicator,canbuildtrustandearnrespect,andunderstandsteamdynamics.
AgoodScrumMastercan“evolution-ize”development.Hecanbuildtrustnotonlyinthe
teambutwithcustomersaswell.FormoreonchoosingtherightScrumMaster,see
Chapter5,“ImplementingtheScrumRoles.”
ProductOwner
IftheScrumMasteristheoilandsensorsintheengine,theproductowneristhedriver.
Theproductownerpointsthecarinthecorrectdirectionandmakestheminute
adjustmentsnecessarytostayoncourseanddeliverresults.Sherepresentscustomerand
stakeholderinterests.Herjobistoestablish,nurture,andcommunicatetheproductvision
totheteamandtheotherstakeholders,managetheprojectreturnoninvestmentand
financials,andmakedecisionsaboutwhentocreateanofficialrelease(withcustomerand
stakeholderinput).Theproductowneristheonepersonultimatelyresponsibleforthe
successorfailureoftheproject.Shedecideswhatisdeveloped,whenitisdeveloped,and
whetherthatwhichhasbeendevelopedmeetsexpectations.
DevelopmentTeam
Thedevelopmentteamistheengineoftheracecar.Allthedrivingandlubricationinthe
worldareofnousewithoutanengine.Thedevelopmentteamexecutestheproduct
owner’svisionwiththehelpoftheScrumMaster.Theteamiscomposedofthepeople
neededtodeliverthework—developers,testers,architects,designers—anyonewhois
needed.Adevelopmentteamisideallymadeupofpeoplededicatedfull-timetothe
project.(Realitysometimesmakesthisimpossible.SeeChapter3,“UsingTeam
ConsultantstoOptimizeTeamPerformance,”forwaystocreateahigh-performingteam
underless-than-idealcircumstances.)Theteamisresponsibleformanagingitswork,its
commitments,andtheexecutionofthosecommitments.
MostScrummaterialsaysthattheidealteamsizeisseven,plusorminustwopeople.I
preferevennumbers,asitfacilitatesbetterXPengineeringpracticeintegration,somyrule
wouldbesix,plusorminustwo.Theteamisjustthat,ateam—rolesandtitlesshouldbe
removed,whichhelpsbuildthecamaraderiearoundtheteam.Thegoalistoremovethe
mindsetof“I’madeveloperandIonlywritecode”andshifttheattentionto“I’mateam
memberwhoisresponsiblefordeliveringthiswork,andIcannotdoitalone.”Ona
Scrumteam,testersmaywritecodeanddevelopersmaywritetests—cross-functionalityis
agoodthing.
TheArtifacts
Scrumisnotanartifact-heavydevelopmentprocess,butthreeartifactsarecrucialto
success:theproductbacklog,thesprintbacklog,andtheburndown.
TheProductBacklog
Theproductbacklogisthemasterlist,prioritizedbybusinessvalueandrisk,ofallthe
featuresandfunctionalityneededtoimplementthevision,andintheend,theproductor
serviceunderdevelopmenttoachievethatvision.Theproductownerisresponsiblefor
managingtheproductbacklog,keepingituptodate,prioritizedorordered,andclear.
Unlikearequirementsdocumentorproductspecification,theproductbacklogisnever
complete.Instead,itisalivingdocumentthatexpandsorcontractsinresponsetochanges
inpriority,value,orrisk.Itemscanbeaddedordeletedfromtheproductbacklogatany
time.Onceadded,theyareprioritized(orordered)bytheproductownerinrelationtothe
otheritems:bugs,features,enhancements,nonfunctionalrequirements,andsoon.
Priorityisdeterminedbybusinessvalueandrisk,orwhateverorderingmakessensefor
yourcompanyandproject.Thehighestpriorityitemsareslatedfordevelopmentfirst
whilethelowestarecompletedlast(seeFigureA-2).Theteamworkswiththeproduct
ownertoestimatethesizeoftheproductbacklogitems(PBIs),relativetooneanother.
Thatsizecanbeexpressedinpoints,T-shirtsizes,oranyothernon-time-specificunit.The
keyistostayawayfromtime-basedestimationforPBIs.Thosespecificswillcomelater
(seeChapter11,“ReleasePlanning,”formoreonestimation).
FIGUREA-2Keeptheproductbacklogorderedandprioritized.
High-prioritystoriesshouldbesmallandclearsothattheycanbebroughtintothe
sprint.Lower-priorityitemscanbelargeandfuzzy.Ashigh-priorityitemsarecompleted
andlowerpriorityitemsbegintorisetothetopofthebacklog,thebiggerstoriesneedto
bebrokendownintosmallerchunks,aprocesscalleddecomposition(seeChapter12,
“DecomposingStoriesandTasks”).
TheSprintBacklog
Thesprintbacklogistheoutputoftheplanningmeeting.Itisessentiallythelistoftasks
thattheteamneedstocompleteduringthesprinttoturnaselectedsetofproductbacklog
itemsintoadeliverableincrementoffunctionality.UnlikePBIs,sprintbacklogtaskshave
atime-based(hourly)estimate.Sincetheteamisdoingthework,itisresponsiblefor
keepingthesprintbackloguptodate.
Duringasprint,newtasksmightbediscoveredandtasksthathavealreadybeen
identifiedmightbeadjusted.Thisisperfectlynormalbehavior.Theteamsimplyaddsthe
newtasks(andanestimateoftheworkremainingontasks)tothesprintbacklogoradjusts
thewording(and,ifnecessary,theestimatedworkremaining)fortasksinprogress.Asthe
teamcompletestasks,thetasksshouldbemarkedasdoneonthesprintbacklog.Thesprint
backlogshowsteammemberswhatworkiscompleteandwhattasksremain.Thisdata
helpsteammembersrunaneffectivedailyscrum(dailystandup).
Whilechangestothesprintbacklogareexpected,theScrumMastershouldbewatching
forpatternsaboutthetypeoramountofworkthatisaddedoradjusted.Ifapattern
emerges,itcanteachateamquiteabitaboutthesystemasitisbuiltandpossiblyabout
theteamitself.
TheBurndown
WhileTheScrumGuidepublishedbySchwaberandSutherlandinOctober2013saysthat
burndownsarenolongerarequiredartifactinScrum,Ibelievetheyarestillessentialtools
inateam’sarsenal.Scrumrequiresthatteamscommunicatehowmuchworkisremaining
andhowtaskcompletionistrendingthroughoutthesprint.ThebesttoolI’vefoundto
trackthisprogressisthesprintburndown,agraphical,real-timepictureofwhatworkis
remaininginthesprint.Itiscalculatedbygraphingthenumberofhoursremaining(y-
axis)againstthenumberofdaysremaining(x-axis).Burndownscanbehandwrittenor
createdwithatoolsuchasMicrosoftExcel.Toupdatethesprintburndown,the
ScrumMasterortheteamplotsthenumberofhoursremainingonthechartattheendof
eachday.Connectingthosepointsshowsagraphoftheworkstilloutstandingasthedays
ofthesprintgoby.Youcanvisualizewhethertheteamisontracktocompleteallthework
remainingbyextrapolatingthelinetothelastdayofthesprint.
TheMeetings
Scrumhasfourmeetings:theplanningmeeting,thedailyscrum(ordailystandup),the
sprintreview,andtheteamretrospective.
PlanningMeeting
Eachsprintbeginswithatwo-partsprintplanningmeetingattendedbytheteam,
ScrumMaster,andproductowner.Foraone-monthorfour-weeksprint,thistwo-part
meetingshouldlasteighthours.Foratwo-weeksprint,planforaboutfourhours.Asa
generalruleofthumb,multiplythenumberofweeksinyoursprintbytwohourstoget
yourtotalsprintplanningmeetinglength.
Partoneofthesprintplanningmeetingisareviewofpotentialproductbacklogitems
forthesprint.Thisisthetimefortheproductownertodescribewhatshewantstosee
builtforthenextsprint.Duringthispartofthemeeting,itisnotuncommonfortheteam
tobanterbackandforthwiththeproductowner,askingclarifyingquestionsanddriving
awayambiguity.Bytheendofsprintplanningpartone,theteamselectsasprintgoal,
whichshouldbeaone-sentencedescriptionoftheoveralloutcomeofthesprint.This
helpslaterwhenquestionsaboutdepthandbreadthcomeup,inthatiftheworkdoesnot
directlytietothesprintgoal,thenitisnotdoneduringthesprint.
Duringparttwoofthesprintplanningmeeting,theteamdecideshowtheworkwillbe
built.Inthismeeting,theteambeginsdecomposingtheproductbacklogitemsintowork
tasksandestimatingthesetasksinhours.Theproductownermustbeavailableduringthis
meetingbutdoesnothavetobeintheroom.Infact,manyteamsfindithelpfultowork
withouttheproductownerduringthisdetailedpartofthemeeting.Manyteamsfindthey
enjoydiscussingmanyimplementationpossibilitieswithoutworryingthattheproduct
ownerwillpanicormisunderstand.Iftheproductownerdoesremainintheroom,the
ScrumMasterneedstotakechargeofthispartofthemeeting,keepingtheteamfocused
andfreetoexplorepossibilitieswithoutbeinglimitedbytheproductowner’sownideasor
opinions.
DailyScrum
ThemostfrequentmeetingthatoccursinScrumisthedailymeeting,otherwiseknownas
thedailyscrumordailystandup.Regardlessofwhatitiscalled,thepurposeremainsthe
same:Givetheteamtheopportunitytosyncdaily,atthesametimeandatthesameplace.
Thedailyscrumisadailyplanningmeetingwheretheteamanswersastandardsetof
questions.Thethreequestionsmostoftenusedareasfollows:
1.Whathaveyouaccomplishedsincethelastmeeting?
2.Whatwillyouaccomplishtoday?
3.Whatimpedimentsorobstaclesareinyourway?
Someteamsaddafourthquestion.Formoreonthefourthquestion,seeChapter18,
“TheFourthQuestioninScrum.”
Thedailyscrumisnotadeep-diveproblem-solvingmeeting;itismeanttoidentify
whatpeopledid,whattheywilldo,andtheirissues—nothingmore.Anyotherissuesneed
tobetakenofflineasactionitemsfortheScrumMastertoresolveortheteamtodiscuss
further,dependingontheissue.Atthesametime,thedailyscrumisnotastatusreporting
meeting.Peopleshouldbetalkingtoeachother,notreportingtheiractivitiestothe
ScrumMaster.Itspurposeistokeeppeoplefocusedonwhattheyhavedoneandwilldo.
TheScrumMaster’sonlyfunctionistofacilitatethemeeting,notactasadetectiveor
taskmaster.
Theteamdetermineswhenthemeetingwilltakeplace,butitshouldbeeverydayatthe
sametimeandplace.TheScrumMasterandalltheteammembersarerequiredtoattend
everyday.Theproductownerandotherinterestedpartiesareinvitedtoattend,butitis
importanttounderstandthatthismeetingisfortheteam,notforexecutives,managers,
devleads,oranyoneelse.Ifnon–teammembersdoattend,theyarein“read-only”or
“listen-only”mode.Theyarenotallowedtoactivelyparticipate,butcanwatch,listen,and
learn.
Donewell,dailyscrumsimproveteamcommunication,identifyissuesand
impediments,andbuildasenseofcamaraderie.Readmoreabouthowtorunaneffective
dailyscrummeetinginChapter17,“FacilitatingaProductiveDailyScrum.”
SprintReview
Onthelastdayofthesprint,theteamholdsasprintreview.Thismeeting(theduration,
likethesprintplanningmeeting,dependsonsprintlength)allowsthecustomersan
opportunitytoreviewtheprogressmadeontheprojecttodateandprovidesthemavenue
togivedirectionorguidancetotheteamwithregardtotheproject.TheScrumMaster,
team,andproductownermustallattendthesprintreview.Additionally,theproductowner
ofteninvitescustomers,keystakeholders,andexecutivestothismeeting.
Duringthereview,theteamrecapsthegoalofthesprintandpresentstheworkthatis
done(seeChapter7,“HowDoYouKnowYou’reDone?,”tounderstandwhatthis
means.).Itisnotuncommonforcustomerstopraisetheteaminthismeeting,norisit
uncommonforthecustomerstoaskforchanges.Afterall,thereasonteamsdoScrumisto
providequickfeedbackloopsandmultipleopportunitiestoinspectandadapt;thisreview
isoneofthoseloops.Anynewrequestsshouldbeputontheproductbacklogand
prioritized.Inthismeeting,theteamshouldaskforcustomeracceptance.Sometimes,
customerssaythatsomefunctionalitywasnotdeliveredorwasnotdeliveredasexpected.
Inthosecases,workmaybeputbackontheproductbacklog.Ifthisisarecurringpattern,
itshouldbenotedbytheScrumMasterandaddressedwiththeteam.
Formoreoneffectivesprintreviews,seeChapter15,“TheSprintReview.”
SprintRetrospective
Thesprintretrospectivefollowsthesprintreview.Itprovidestheteamanopportunityto
identifywaysinwhichitcanimproveitsworkprocessandexecutionoftheScrum
framework.ThemeetingismandatoryforallScrumteammembers.
TheScrumMastertypicallyfacilitatesthemeeting,helpingtheteambuildalistof
prioritizedimprovementitemsitcanimplementinfuturesprints.Theseitemsshouldbe
reviewedatsubsequentretrospectivessotheteamknowshowitisdoingandwhether
improvementsarehappening.Onceagain,thedurationhingesonthesprintlength.
Inthismeeting,theteamshouldanswer(atleast)thefollowingtwoquestions:
1.Whatwentwellduringthissprint?
2.Whatcouldbeimprovedinthenextsprint?
Everyoneshouldparticipate.TheScrumMastershouldnotbegoingaroundtheroom
askingpeoplefortheiranswers.Instead,theteamshouldbeactivelywritingitsanswers
onawhiteboard.Oncetheteammembershavefinishedanswering,theyshouldwork
togethertoprioritizetheitemsandbegindecidingwhethertheitemswillbeaddressed.
Retrospectivesthatdonothavealotofactivityorthatdonotdrivechangearefrustrating
foreveryoneinvolvedandoftenleadtotheteamquestioningtheretrospectives’value.The
valuecomesthroughtheopportunitiesforchangeidentifiedintheretrospective.Ifthe
opportunitiesarenotactedonandtheteamdoesnotimprove,thereisalargerissue.For
moreonrunningeffectiveretrospectives,seeChapter16,“Retrospectives.”
PuttingItAllTogether
Thepartsandrulesbythemselvesareeasytounderstand.Puttingthemtogetherinthe
mosteffectiveway,however,poseschallenges.Movingbacktotheracecaranalogy,each
partmustbefinelytunedforoptimumperformance.Thechaptersinthebookdivedeeply
intoimplementationdetailsandreal-lifestrategiesforoptimizingperformance;moving
fasterwhileremainingincontrol,anddealingwiththeknocks,blowouts,andbackfires
thatcomefromtryingtoraceinthecompetitiveworldofsoftwaredevelopment.
Index
Acceptancetests,inTDD,129–131
Acceptancewindow,contractualagreement,362
Accountability,teamconsultants,47–48
Activetasklimit(ATL),incollaboration,373–377
Agenda,dailyScrummeeting,209–210
Agileteams,successfuloutsourcing,332
Aligningincentiveswithoutcomes
aligningaroundcustomersatisfaction,412–413
earlyinvolvementofdevelopmentteams,413
end-to-enddeliverymodel,412–415
prioritizingprojectportfolios,414–415
settingthefocus,412–415
shiftingresources,414–415
sidebenefits,415
astory,409–412
Aligningincentiveswithoutcomes,keystosuccess
dedicatedteamsversusrandompeople,416
executivesupport,417
integratingsalesanddevelopment,416
organization-widecoordination,417
prioritizingprojectportfolios,416
Allocatingwork,successfuloutsourcing,333
Anarchicprojects,technologyandrequirements,14
Artifacts.SeeScrum,artifacts;specificartifacts.
ATL(activetasklimit),incollaboration,373–377
Automatedintegration,inTDD,129–131
Automatingdocumentation,323
Backlog.SeeProductbacklog.
Beginner’smind,225–226,230–231
Belshee,Arlo,223,225–227,230
TheBigWalltechnique
description,342
example,342–346
astory,340–341
Blindestimationofteamvelocity.SeealsoEstimatingteamvelocity.
decomposingthereferencestory,59
estimatingvelocity,60–61
versusothertechniques,66
overview,58–61
points-to-hoursapproximation,59
productbacklog,58
teamcapacity,59–60
Blockingissues,dailyScrummeeting,207–208
Boehm,Barry,171,357
Brainstorming,96–97
Breakingupfights,roleoftheScrumMaster,113
Brooks,Fred,45
Brooks’law(addingmanpowertolateprojects),45,233,398–399
Budgets,truecostsofoutsourcing,330–332
Bugs.SeeDefectmanagement.
Burndowncharts
description,431
sustainablepace,273–274
Burnout,269–275
Busfactor,214
“BuyaFeature”game,199
Cadence,dailyScrummeeting,213
Cancelingthesprint,259–260
Candidatescreening.SeeImmersiveinterviewing,candidatescreening.
Cards,collectinguserstories,306–307
CareerBuilder.com,399–400
Carnegie,Dale,252–253
Carnegieprinciples,teamculture,253
Categorizingissues,definitionofdone,97–98
Chairsinmeetings.SeeStandingversussitting.
Change,roleoftheScrumMaster,115
Change,stagesof
chaos,16
foreignelements,16
Kotter’seight-stepmodel,30
latestatusquo,15–16
newstatusquo,17
practiceandintegration,16–17
Satir’sStagesofChange,15–17
Changemanagement,contractualagreement,353–356,361
Chaos,stageofchange,16
Clark,Kim,374
Cleland-Huang,Jane,308
Clientrole,combiningwithotherroles,75–77
Codedebt.SeeTechnicaldebt.
Codereviews,pairprogrammingasreal-timereviews,129
Codesmells,125–126
Codingstandards,definition,13
Cohn,Mike
collaboration,371
consolidation,98
datacollection,64
outsourcing,329
oncustomers,362
Collaboration.SeealsoTeams.
ATL(activetasklimit),373–377
keystosuccess,380–382
pairprogramming,373
relativesizingandtimeestimation,372
shorteningsprints,377–378
astory,367–371
taskboards,378–380
two-weeksprints,377–378
workitems,limiting,373–377
Collaboration,estimatingtasks.SeealsoTaskPoker.
ATL(activetasklimit),373–377
incollaboration,371–373
openitems,limiting,373–377
astory,367–371
TaskPoker,371–373
Collectivecodeownership,12
Commitment,Scrumvalue,8
Communication
emergencyprocedures,260
enlistingScrumsupport,31–32
releaseplanning,155
retrospectivemeetings,197
Scrumvision,31–32
successfuloutsourcing,335
Competencies.SeeSkillsandcompetencies.
Completingaproject.SeeDeliveringworkingsoftware;Done,defining.
Complexprojects,technologyandrequirements,14
Complicatedprojects,technologyandrequirements,14
ConeofUncertainty,357
Confirmation,collectinguserstories,307
Conflictavoidance,dailyScrummeetings,221
Conformity,teamculture,247–249
Consolidating
improvements,33
issues,98–100
Contingencyplans.SeeEmergencyprocedures.
Continuousintegration
definition,13
successfuloutsourcing,335
inTDD,126–128
Continuouslearning
dailyScrummeetings,221
implementingScrum,18
Contracts
acceptancewindow,362
customeravailability,361–362
keystosuccess,361–363
prioritization,362–363
ROIcutoff,363
astory,349–353
trustfactor,363
Contracts,projectmodel
changemanagement,361
deliverables,360–361
Contracts,rangesandchangesmodel
changemanagement,361
costestimation,359–360
costpersprint,estimating,360
discoverycontract,358–360
overview,358
paymentoptions,specifying,360
projectcontracts,360–361
releaseplanning,360
teamvelocity,determining,360
timeline,determining,359–360
versustraditionalcontracts,358
userstories,creatingandestimating,359
usertypes(personas),identifying,359
Contracts,traditionalmodel
changemanagement,353–356
ConeofUncertainty,357
overview,353–356
timing,356–358
Conversations,collectinguserstories,307
Conway’slaw(organizationalstructureinthecode),xxi
Corehours
colocatedteams,138–140
distributedteams,140–141
keystosuccess,142
part-timeteammembers,141–142
astory,135–138
Coreteams.SeealsoDevelopmentteams;Teams.
earlyinvolvement.SeeAligningincentiveswithoutcomes.
memberresponsibilities,74
optimalsize,45–46
riskmanagement,423–424
skillsandcompetencies,42–43
astory,35–39
versusteamconsultants,43–44,46,49.SeealsoTeamconsultants.
Costs
ofabadhire,399–400
documentation,318–321
ofimmersiveinterviewing,406
projects.SeeEstimatingprojectcost.
releaseplanning,152–153
Courage,Scrumvalue,8
Criticalpaths,implementingScrum,9–11
Culturalchallenges,truecostsofoutsourcing,329–330,332
Culture,team
addingnewmembers,238
Carnegieprinciples,253
conformity,247–249
culturalgoals,247–250
empowerment,252–253
innovation,247–249
institutionalmeans,248
keystosuccess,251–254
Merton’sstraintheory,246–251
Merton’stopologyofdeviantbehavior,246–251
rebellion,247–249
retreatism,247–249
ritualism,247–249
roleinoutsourcing,329,332
socialdeviance,246
sprintlength,86
astory,241–246
Cunningham,Ward,227–228
Customers
availability,contractualagreement,361–362
environment,sprintlength,86–87
estimatingteamvelocity,57
riskmanagement,421,423
sprintlength,86–87
viewofproductbacklog,343–346
Cycletime,burnout,272
DailyScrummeetings
agenda,209–210
blockingissues,207–208
busfactor,214
cadence,213
commonobstacles,205,208
conflictavoidance,221
continuouslearning,221
deepdives,210–211
description,432–433
fourthquestion,432
glossingoverproblems,212
handsignals,211
hiddenimpediments,212–213
interruptions,210
keystosuccess,212–215
layout,209–210
legacysystems,180
nonverbalcommunication,221
punctuality,209
rambling,210–211
rhythm,209–210
scheduling,208–209
standardthreequestions,432
standingversussitting,214
astory,205–208
successfuloutsourcing,333–334
teamconsultants,47
teamwork,214
vagueness,212
Dailystandupmeetings.SeeDailyScrummeetings.
Datacollection
historicaldata,391–393
retrospectivemeetings,198–199
sampletable,392
instorypoints,391–393
Datagatheredovertimemodel,sustainedengineering,179
Dates.SeePlanning;Scheduling.
Debt.SeeTechnicaldebt.
Decomposingstories
estimatingteamvelocity,59
example,161–164
granularity,167
astory,157–159
Decomposingtasks
estimatingtasksizes,164–167
example,164–167
granularity,167
sprintlength,87
astory,157–159
Decomposingthemes
example,162–163
granularity,167
Dedicatedteammodel,sustainedengineering,179–182
Dedicatedteams.SeeCoreteams.
Dedicatedtimemodel,sustainedengineering,178
Deepdives,dailyScrummeeting,210–211
Defectmanagement
frequenttesting,171
keystosuccess,172–173
onlegacysystems,173
overview,170–172
pairprogramming,129
settingpriorities,171–172
astory,169–170
value,optimizingandmeasuring,295
Definitionofdone.SeeDeliveringworkingsoftware;Done,defining.
Degreeofconfidence,releaseplanning,148–151
Deliveringworkingsoftware.SeealsoDone,defining.
end-to-endscenarios,286–287
expansion,283–284
identifyingacorestory,281–282
keystosuccess,284–287
limitinguseraccess,282–283
numberofusers,282–283
prioritizingrisk,283
rework,285–286
astory,277–281
validation,283–284
windowofopportunity,283
DeMarco,Tom,274
Denne,Mark,308
Derby,Esther,198,200–202
Designconceptcards,227
Developmentpractices,truecostsofoutsourcing,330
Developmentteams,definition,428–429.SeealsoCoreteams.
Developmentalstages,teamgrowth,235–237
Discoverycontract,contractualagreement,358–360
Documentation
inagileprojects,321–322
automating,323
committingto,323
commondocuments,listof,317–318
cost,318–321
explainingyourprocess,323
keystosuccess,323–324
listoffeaturesandfunctions.SeeSprintbacklog.
planningfor,322–323
purposeof,317
sprintreviewmeetingdecisions,189–190
stabilityversusvolatility,322
startingprojectswithout,322–323
astory,313–316
versusworkingsoftware,316–317
Documentation,approachesto
early,318–319
late,319–320
asyougo,320–321
Dollardemonstration,285
Done,defining.SeealsoDeliveringworkingsoftware.
brainstorming,96–97
categorization,97–98
consolidation,98–100
creationandpublishing,100–101
exercise,95–101
keystosuccess,101
participants,96
purposeof,101
sorting,98–100
astory,93–95
inTDD,132
teamdefinitionofdone,94
undonework,101
Duration,sprintreviewmeetings,187
Duvall,PaulM.,128
Educating
individuals,TDD,132–133
organizations,roleoftheScrumMaster,115
stakeholders,298
Efficiencyversuseffectiveness,274–275
Emergencyprocedures,teamoptions
cancelingthesprint,259–260
communication,260
don’tpanic,260–261
gettinghelp,258
keystosuccess,260–261
maintainingfocus,261
overview,257–258
reducingscope,258–259
removingimpediments,258
astory,255–257
Employeecosts
estimatingprojectcosts,308–309
outsourcing,329
roleoftheScrumMaster,109–112
Empowerment
enlistingsupportforScrum,32–33
teamculture,252–253
Endgame,releaseplanning,153–154
Endingaproject.SeeDeliveringworkingsoftware;Done,defining.
End-to-enddeliverymodel,412–415
End-to-endscenarios,286–287
Engineeringpractices.SeeSustainedengineering;TDD(Test-DrivenDevelopment).
Environment
physical,retrospectivemeetings,197–198
political,estimatingteamvelocity,57
productowner,sprintlength,86–87
Epics,definition,160
Erdogmus,Hakan,129
Estimates
becomingcommitments,387
ascommitments,54
inranges,385–386,388–391
Estimates,relative
incostestimation,307–308
Fibonaccisequence,59,303
Estimating
hoursfromstorypoints.SeeStorypoints.
productbacklog.SeeProductbacklog,prioritizingandestimating.
projectresources.SeeEstimatingprojectcost.
remainingworkload.SeeBurndown.
tasks.SeeTasks,estimating.
trendsintaskcompletion.SeeBurndown.
Estimatingprojectcost
contractualagreement,359–360
costpersprint,360
employeecosts,309
functionalspecifications,307
keystosuccess,310–311
MMF(minimalmarketablefeature)set,308
outsourcing,truecosts,330–332
planningpokertechnique,307
releaseplanning,309–310
roughlyrightversuspreciselywrong,307–308
astory,301–305
teamcosts,309
teamvelocity,308–309
techniquesfor,307–308
Estimatingprojectcost,userstories
cards,306–307
confirmation,307
conversations,307
creating,307–308
prioritizing,308
sizing,301–305,307–308
threeC’s,307
Estimatingteamvelocity
communicatingestimatesasranges,62–64
comparisonoftechniques,65–66
estimatesascommitments,54
fromhistoricaldata,56–57,66
keystosuccess,65–66
multipliers,64–65
politicalenvironment,57
productownerandcustomer,57
forprojectcost,308
projectsizeandcomplexity,57
astory,51–56
teamnewness,57
truncateddatacollection,64–65
variables,57
waitandsee(realdata)method,61–64
Estimatingteamvelocity,byblindestimation
decomposingthereferencestory,59
estimatingvelocity,60–61
versusothertechniques,66
overview,58–61
points-to-hoursapproximation,59
productbacklog,58
teamcapacity,59–60
Expansion,deliveringworkingsoftware,283–284
Expendabilityofteammembers,214
Extendingsprintlength,92
Externalfocus,pairprogramming,225
ExtremeProgramming(XP).SeeXP(ExtremeProgramming).
Facilitation,roleoftheScrumMaster,114
Feathers,Michael,182
Featurelist.SeeProductbacklog;Sprintbacklog.
Featurework,292–294
Fibonaccisequence,59,303
Finishingaproject.SeeDeliveringworkingsoftware;Done,defining.
Focus,Scrumvalue,8
Forecastingwithimmersiveinterviewing,398
Foreignelements,stageofchange,16
Forming,stageofteamdevelopment,235–237
Fourthquestion,dailyScrummeetings,217–221,432
Fowler,Martin,126,182
Functionlist.SeeProductbacklog;Sprintbacklog.
Functionalspecifications,estimatingprojectcost,307
Gabrieli,John,225
Geographicdistance,costsofoutsourcing,332
Gettingstarted.SeeImplementingScrum.
Glossingoverproblems,dailyScrummeeting,212
Granularity,decomposingstories,167
Grenning,James,371
Groupcohesion,costsofoutsourcing,329
Guidingcoalition,enlistingsupportforScrum,31
Handsignals,dailyScrummeeting,211
Hedden,Trey,225
Help,emergencyprocedures,258
Helpingout,roleoftheScrumMaster,114
Hiring
interviewingcandidates.SeeImmersiveinterviewing.
outsourcing.SeeOutsourcing.
Historicaldata
collecting,391–393
estimatingteamvelocity,56–57,66
estimatingworkload,393
Hittingthewall,267–269
Hofstede,Geert,329
Hohmann,Luke,199
Humphrey’slaw(gatheringuserrequirements),xxi
IBM
keydimensionsofculturalvariety,329
TDD,benefitinteams,124–125
Immersiveinterviewing
costofabadhire,399–400
definition,398,400–401
asaforecast,398
hiringfortherightreasons,398–399
howtohire,400–401
long-termfitversusshort-termexpertise,400
organizationalevaluation,398
skillsandcompetencies,400
astory,395–397
Immersiveinterviewing,candidatescreening
managersandnontechnicalpeople,404
overview,401–402
preparationandsetup,402–403
scoringcandidates,403–404
senior-levelpeopleversusjunior-level,402–403
Immersiveinterviewing,keystosuccess
focusoncompetencies,405
hiringstrongpeople,406
investingintheprocess,406
repeatablehiringprocess,405
skillsandcompetencies,405–406
understandingcosts,406
ImplementingScrum.SeealsoPeople,enlistingsupportof.
combiningwithExtremeProgramming,12–13
continuouslearning,18
exposingissues,12
identifyingcriticalpaths,9–11
keystosuccess,17–18
learningbasemechanics,17
inmidstream,18
patience,17–18
potentiallyshippablecode,13
Scrumplanningversustraditionalmethods,9–11
shiftingmindsets,9
astory,1–6
timeframe,18
underlyingvalues,7–8
understandingtherules,17
Improvingexistingcode,125–126.SeealsoRefactoring.
Incentives,aligningwithoutcomes.SeeAligningincentiveswithoutcomes.
Innovation,teamculture,247–249
Institutionalmeans,teamculture,248
Institutionalizingnewapproaches,33
Internalfocus,pairprogramming,225
Interruptions,dailyScrummeeting,210
Interviewing.SeeImmersiveinterviewing.
James,Michael,113
Jansen,Dan,225
Junior-levelpeople,interviewing,402–403
Kerth,Norman,202
Kessler,Robert,129
Kotter,John,30
Kotter’smodelforenlistingsupportforScrum,30
Larsen,Diana,198,200–202
Latestatusquo,stageofchange,15–16
Lawsofsoftwaredevelopment
Brooks’law(addingmanpowertolateprojects),45,233,398–399
Conway’slaw(organizationalstructureinthecode),xxi
Humphrey’slaw(gatheringuserrequirements),xxi
Ziv’slaw(predictability),xx–xxi
Layout,dailyScrummeeting,209–210
Learningorganizations,35
Legacysystems.SeealsoSustainedengineering.
dailyreleasesandstandups,180
defectmanagement,173
goalplanning,180
keystosuccess,181–182
retiring,182
retrofitting,182
stakeholdermeetings,180–181
astory,175–177
stranglerapplications,182
tribalknowledge,176
Legalagreements.SeeContracts.
Maintaining
oldcode.SeeLegacysystems;Sustainedengineering.
thereleaseplan,151–152
Managementsupportforteamconsultants,48
Managers,interviewing,404
Managingpeople,roleoftheScrumMaster,113
Martin,Robert,126
Masterlist.SeeProductbacklog;Sprintbacklog.
McConnell,Steve,45,357,388
Meetings.SeealsoPlanning.
chairs.SeeStandingversussitting.
daily.SeeDailyScrummeetings.
standingversussitting,197,214
astory,1–6
teamconsultants,46–47
typesof,431–434.Seealsospecificmeetings.
Menlo,ExtremeInterviewingtechnique,401
Merton,RobertK.,246–251
Merton’sstraintheory,teamculture,246–251
Merton’stopologyofdeviantbehavior,246–251
Micro-pairing,pairprogramming,224,227–230,231
Microsoft,124
Miller,Ade,128
MMF(minimalmarketablefeature)set,308
MMR(minimalmarketablerelease),347
Multipliers,estimatingteamvelocity,64–65
Multitasking,negativeeffectsof,374
Myers,Ware,45
Nass,Clifford,374
Newstatusquo,stageofchange,17
Nielsen,Dave,354–355
Noisereduction,pairprogramming,128
Nontechnicalpeople,interviewing,404
Nonverbalcommunication,dailyScrummeetings,221
Norming,stageofteamdevelopment,235–237
NorthCarolinaStateUniversity,124
Offshoring.SeeOutsourcing;Teammembers,adding.
Oldcode
maintaining.SeeLegacysystems;Sustainedengineering.
refactoring,13,125–126
Openness,Scrumvalue,8
Ophira,Eyal,374
Osborn,AlexF.,96
Outcomes,aligningincentiveswith.SeeAligningincentiveswithoutcomes.
Outsourcing,keystosuccess
agileteams,332
allocatingthework,333
continuousintegration,335
contraindications,336–337
dailystandups,333–334
hiringnorth/southversuseast/west,333
maintainingtheScrumframework,333–334
pairedprogramming,334–335
projectmanagement,335–336
real-timecommunication,335
retrospectives,334
sprintreviews,334
teambuilding,332–333
travelrequirements,335–336
workpackages,330,333,335
Outsourcing,astory,325–328.SeealsoTeammembers,adding.
Outsourcing,truecosts
culturalchallenges,329–330,332
developmentpractices,330
estimatingbudgets,330–332
geographicdistance,332
groupcohesion,329–330
increasedoverhead,329
long-termretention,329
projectmanagement,329–330
transitioncosts,328–329
workingacrosstimezones,332
Overloadingteamconsultants,49
Pacing.SeeSustainablepace.
Pairchurn,226–227
Paircycletime,226
Pairprogramming
beginner’smind,225–226,230–231
benefitsof,128–129,373
bugreduction,128
incollaboration,373
definition,12–13
designconceptcards,227
distractions,225,230
externalfocus,225
integratingnewteammembers,234
internalfocus,225
keystosuccess,230–231
micro-pairing,227–231
noisereduction,128
outsourcing,334–335
pairchurn,226–227
paircycletime,226
Ping-Pongpattern,227–228
promiscuouspairing,225–227
asreal-timecodereviews,128
astory,223–225
inTDD,128–129
PairProgrammingPing-PongPattern,227–228
Parkingunresolvabledisagreements,347–348
Patience
enlistingsupportofpeople,34
implementingScrum,17–18
Patterns,determining,298–299
Patton,Jeff,285
Paymentoptions,contractualagreement,360
People,enlistingsupportof.SeealsoManagement;Teams.
communicatingavision,31–32
consolidatingimprovements,33
creatingavision,31
creatingshort-termwins,33
empoweringparticipants,32–33
establishingasenseofurgency,30–31
formingaguidingcoalition,31
institutionalizingnewapproaches,33
keystosuccess,33–34
Kotter’seight-stepmodel,30
patience,34
providinginformation,34
sponsors,31
astory,23–29
Performing,stageofteamdevelopment,235–237
Personas(usertypes),identifying,359
Personnel.SeeManagement;People;Teams.
Physicalenvironment,retrospectivemeetings,197–198
Ping-Pongpattern,pairprogramming,227–228
Planning.SeealsoEstimating;Meetings;Scheduling.
goalsforlegacysystems,180
listoffeaturesandfunctions.SeeProductbacklog;Sprintbacklog.
prioritizingandestimatingproductbacklog,346
releases.SeeReleaseplanning.
retrospectivemeetings,196–198
Scrumversustraditionalmethods,9–11
sprintreviewmeetings,189
astory,1–6
forteamconsultantdowntime,49
Planningmeetings
description,431–432
teamconsultants,47
PlanningPoker,371–372.SeealsoTaskPoker.
Planningpokertechnique,estimatingprojectcost,307
Points-to-hoursapproximation,59
Politicalenvironment,estimatingteamvelocity,57
Potentiallyshippablecode
implementingScrum,13
astory,277–281.SeealsoDeliveringworkingsoftware.
PowerPointslides
astory,184–186
templatefor,187–188
Practiceandintegration,stageofchange,16–17
Preciselywrongversusroughlyright,307
Preconditionsforsprints,295–296
Preplanning,prioritizingandestimatingproductbacklog,346
Principlesofclassdesign,126
Prioritizing
bybusinessvalueandrisk,429–430
contractualagreement,362–363
defectmanagement,171–172
issuesinretrospectivemeetings,194–195,199–200
itemsforreleaseplanning,155
productbacklog.SeeProductbacklog,prioritizingandestimating.
projectportfolios,414–416
risks,deliveringworkingsoftware,283
userstories,308
Problemresolution,roleoftheScrumMaster,113
Productbacklog.SeealsoSprintbacklog.
definition,429–430
estimatingteamvelocity,58
prioritizingandestimating,429–430
Productbacklog,prioritizingandestimating
TheBigWalltechnique,342–346
customerview,343–346
emulatingtheteam,342–343
focusingdiscussion,347–348
keystosuccess,346–348
meetingsupplies,348
MMR(minimalmarketablerelease),347
parkingunresolvabledisagreements,347–348
preplanning,346
settingtimelimits,347
shiftingestimates,348
stakeholderview,343–346
astory,339–341
Productownerrole
cancelingthesprint,259–260
combiningwithotherroles,75–77
definition,428
estimatingteamvelocity,57
responsibilities,73
riskmanagement,421,423
Progressreporting.SeeDailyScrummeetings;Retrospectivemeetings;Sprintreview
meetings;Value,optimizingandmeasuring.
Projectcontracts,360–361
Projectmanagement
dutiesmappedtoroles,74–75
successfuloutsourcing,335–336
truecostsofoutsourcing,329–330
Projectportfolios,prioritizing,414–416
Projects
anarchic,14
complex,14
complicated,14
costestimation.SeeEstimatingprojectcost.
duration,sprintlength,85–86
rankingcomplexity,14
simple,14
sizeandcomplexity,estimatingteamvelocity,57
technologyandrequirements,14
Promiscuouspairing,225–227
Provost,Peter,224,227–228
Punctuality,dailyScrummeeting,209
Putnam,Lawrence,45
Quality.SeeValue.
Questions,dailyScrummeetings
fourthquestion,217–221,432
keystosuccess,220–221
standardthreequestions,432
astory,217–220
Questions,sprintretrospectivemeetings,433–434
Quizfordeterminingsprintlength,88–90
Rambling,indailyScrummeetings,210–211
Range,Estimate,Focus,Learn,Effort,Communication,Team(REFLECT),393–394
Rangesandchangesmodelforcontracts.SeeContracts,rangesandchangesmodel.
Rangesinestimates.SeeStorypoints.
Rants,retrospectivemeetings,194
Rate-limitingpaths.SeeCriticalpaths.
Ratingthesprint,retrospectivemeetings,200
Realdata,estimatingteamvelocity,61–64
Rebellion,teamculture,247–249
Refactoringoldcode,13,125–126
Referencestories,303
REFLECT(Range,Estimate,Focus,Learn,Effort,Communication,Team),393–394
Relativeestimates
incostestimation,307–308
Fibonaccisequence,59,303
Releaseplanning
contractualagreement,360
degreeofconfidence,148–151
deliveringworkingsoftware,155–156
determiningtheendgame,153–154
estimatingprojectcosts,309–310
inputs,147–148
keystosuccess,154–155
maintainingtheplan,151–152
outcomes,153–154
apreliminaryreleaseplan,147–148
prioritizingworkitems,155
projectcost,152–153
refiningestimates,155
Scrumplanning,155–156
astory,143–146
updatingtheplan,155
Removingimpediments,roleoftheScrumMaster,113
Reporting
progress.SeeDailyScrummeetings;Retrospectivemeetings;Sprintreviewmeetings;
Value,optimizingandmeasuring.
teamperformance,roleoftheScrumMaster,113
Resolvingproblems,roleoftheScrumMaster,113
Respect,Scrumvalue,8
Retreatism,teamculture,247–249
Retrofittinglegacysystems,182
Retrospectivemeetings
attendance,198
basicprinciples,201
benefitsof,196–197
“BuyaFeature”game,199
communication,197
datacollection,198–199
description,433–434
duediligence,196–197
groundrules,198
importanceof,200–201
keystosuccess,200–201
physicalenvironment,197–198
planning,197–198
prioritizingissues,194–195,199–200
purposeof,200–202
rants,194
ratingthesprint,200
roleoftheScrumMaster,198–199
running,198–200
scheduling,201–202
standardtwoquestions,433–434
standingversussitting,197
astory,193–196
successfuloutsourcing,334
teamconsultants,47
timing,197
Reviewmeetings.SeeSprintreviewmeetings.
Rework,deliveringworkingsoftware,285–286
Rhythm,dailyScrummeeting,209–210
Riskmanagement
addingteammembers,239
coreteam,423–424
customerrisk,421,423
keystosuccess,424–425
prioritizing,283,429–430
productowner,421,423
byrole,422
ScrumMaster,423
socialrisk,423
astory,419–420
technicalrisk,423–424
typicalsoftwareprojectrisks,420–421
Ritualism,teamculture,247–249
ROIcutoff,363
Roles
choosing,74–75
descriptions,428–429.Seealsospecificroles.
keycompetencies,74–75
keystosuccess,78–79
mappedtoprojectmanagerduties,74–75
mixing,75–78
astory,69–72
Rothman,Johanna,171
Roughlyrightversuspreciselywrong,307
Satir,Virginia,15
Satir’sStagesofChange,15–17
Scenarios.SeeStories.
Scheduling.SeealsoDone,defining.
dailyScrummeeting,208–209
retrospectivemeetings,201–202
undonework,101
Schwaber,Ken,7,23,92,272,431
Screeningcandidates.SeeImmersiveinterviewing,candidatescreening.
Scrum
artifacts,typesof,429–431.Seealsospecificartifacts.
definition,6–7
evaluatingyourneedfor,13–14
gettingstarted.SeeImplementingScrum.
planning,155–156.SeealsoReleaseplanning.
ScrumEmergencyProcedures,257
Scrumframework,successfuloutsourcing,333–334
Scrummeetings,typesof,431–434.Seealsospecificmeetings.
Scrumroles.SeeRoles.
Scrumvalues
commitment,8
courage,8
focus,8
openness,8
respect,8
ScrumMaster
combiningwithotherroles,75–77
definition,428
responsibilities,73
inretrospectivemeetings,198–199
riskmanagement,423
rotatingamongteammembers,79
ScrumMaster,asfull-timejob
breakingupfights,113
day-to-daytasks,112–114
drivingorganizationalchange,115
educatingtheorganization,115
employeecosts,109–112
facilitatingteamactivities,114
helpingout,114
impactontheteam,106–112
keyfunctions,104
managingpeople,113
removingimpediments,113
reportingteamperformance,113–114
resolvingproblems,113
servantleadership,114
astory,103–106
Senior-levelpeople,interviewing,402–403
Senseofurgency,enlistingsupportforScrum,30–31
Servantleadership,roleoftheScrumMaster,114
Shippablecode.SeePotentiallyshippablecode.
Shore,James,129
Shorteningsprints,377–378
Short-termwins,enlistingsupportforScrum,33
Simpleprojects,technologyandrequirements,14
Size
coreteams,45–46
teamconsultantpools,45–46
userstories,301–305,307–308
Skillsandcompetencies
coreteams,42–43
focusinginterviewsoncompetencies,405
immersiveinterviewing,400,405–406
keycompetenciesforroles,74–75
teamconsultants,42–43
testingcompetenciesinnewteammembers,234,237–238
Slides
astory,184–186
templatefor,187–188
Sliger,Michele,372
Socialdeviance,teamculture,246
Socialrisk,423
Softwaredevelopment.SeeProjects.
SOLIDclassdesignprinciples,126
Sortingissues,definitionofdone,98–100
Spikes,293–294
Sponsors,enlistingsupportforScrum,31
Sprintbacklog,430–431.SeealsoProductbacklog.
Sprintlength
choosing,84–85,88–90
companyculture,86
criteriafor,84–85
customergroup,86–87
decomposingtasks,87
inexcessofonemonth,92
extending,92
guidelinesfor,88–90
keystosuccess,91–92
productownerenvironment,86–87
projectduration,85–86
quizfordetermining,88–90
Scrumteam,87
shortening,astory,370
stakeholdergroup,86–87
astory,81–84
twoweeks,astory,370
warningsigns,90
Sprintretrospectivemeetings.SeeRetrospectivemeetings.
Sprintreviewmeetings.SeealsoDailyScrummeetings.
description,433
documentingdecisions,189–190
duration,187
encouragingparticipants,190
keystosuccess,188–190
overview,186–187
planning,189
preparingfor,187–188
running,188
stories,customeracceptance,190
astory,183–186
successfuloutsourcing,334
teamconsultants,47
Sprintreviewmeetings,PowerPointslides
astory,184–186
templatefor,187–188
Sprints
canceling,259–260
cost.SeeEstimatingprojectcost.
preconditionsfor,295–296
ratingduringretrospectivemeetings,200
reducingscope,258–259
removingimpediments,258
shortening,377–378
two-week,377–378
Stabilityversusvolatility,documentation,322
Stacey,Ralph,14
Stakeholders
educating,298
meetings,legacysystems,180–181
prioritizingandestimatingproductbacklog,343–346
sprintlength,86–87
Standingversussitting
dailyScrummeeting,214
retrospectivemeetings,197
Standupmeetings.SeeDailyScrummeetings.
Sterling,Chris,13
Stories
collaboration,estimatingtasks,367–371
contracts,349–353
corehours,135–138
coreteams,35–39
creatingandestimating,contractualagreement,359
dailyScrummeetings,205–208
decomposing.SeeDecomposingstories.
decomposingtasks.SeeDecomposingtasks.
defectmanagement,169–170
definingdone,93–95
definition,160,292
deliveringworkingsoftware,277–281
documentation,313–316
emergencyprocedures,teamoptions,255–257
enlistingsupportofpeople,23–29
gainingcustomeracceptance,190
hierarchyof,160
implementingScrum,1–6
legacysystems,175–177
meetings,1–6
pairprogramming,223–225
planning,1–6
potentiallyshippablecode,277–281
PowerPointslides,184–186
productbacklog,prioritizingandestimating,339–341
projectcost,estimating,301–305
questions,dailyScrummeetings,217–220
reference,303
reference,decomposing,59
releaseplanning,143–146
retrospectivemeetings,193–196
roles,69–72
ScrumMaster,asfull-timejob,103–106
slides,184–186
sprintlength,81–84,370
sprintreviewmeetings,183–186
sustainablepace,265–269
sustainedengineering,175–177
taskboards,370
TaskPoker,369–370
tasks,estimating,367–371
TDD(Test-DrivenDevelopment),119–123
teamconsultants,35–39
teamculture,241–246
teammembers,adding,233–235
teamvelocity,estimating,51–56
value,optimizingandmeasuring,289–291
Stories,estimatingprojectcost
cards,306–307
confirmation,307
Stories,estimatingprojectcost,continued
conversations,307
creating,307–308
prioritizing,308
sizing,301–305,307–308
teamvelocity,308–309
threeC’s,307
Storming,stageofteamdevelopment,235–237
Storypoints
advantagesof,387
datacollection,391
data-basedimprovements,391–393
estimatesbecomingcommitments,387
estimatesinranges,385–386,388–391
keystosuccess,391–393
REFLECT(Range,Estimate,Focus,Learn,Effort,Communication,Team),393–394
relationtohours,393
astory,383–386
Straintheory,teamculture,246–251
Stranglerapplications,legacysystems,182
Sustainablepace
burndowncharts,273–274
burnout,269–275
cycletime,272
definition,12
efficiencyversuseffectiveness,274–275
hittingthewall,267–269
increasingteamtime,274
keystosuccess,274–275
monitoringprogress,273–274
shorteningiterations,272
astory,265–269
Sustainedengineering.SeealsoLegacysystems.
dailyreleasesandstandups,180
goalplanning,180
keystosuccess,181–182
retiringthelegacysystem,182
retrofittinglegacycode,182
stakeholdermeetings,180–181
astory,175–177
stranglerapplications,182
tribalknowledge,176
Sustainedengineeringmodels
datagatheredovertime,179
dedicatedteam,179–182
dedicatedtime,178
Sutherland,Jeff
burndown,273,431
contractualdeliverables,361
estimatingteamvelocity,57
onlegacycode,181
contracts,358
Scrumdescription,7
ScrumEmergencyProcedures,257
terminationclauses,363
Tabaka,Jean,114
Taskboards
incollaboration,378–380
astory,370
TaskPoker
commonmistakes,380
versusPlanningPoker,371
astory,369–370
fortaskestimation,380
team-levelbenefits,380
TaskPoker,keystosuccess
averagingtaskestimates,381
hearingeveryvoice,380
investmentintheoutcome,381
sharedunderstanding,381
taskestimation,380
taskgranularityestimates,381
teamwork,382
Taskswitching,374
Tasks,decomposing
estimatingtasksizes,164–167
example,164–167
granularity,167
sprintlength,87
astory,157–159
Tasks,definition,160
Tasks,estimating.SeealsoTaskPoker.
ATL(activetasklimit),373–377
incollaboration.SeeCollaboration,estimatingtasks.
granularity,381
openitems,limiting,373–377
pairprogramming,373
relativesizingandtimeestimation,372
shorteningsprints,377–378
astory,367–371
taskboards,378–380
taskswitching,374
two-weeksprints,377–378
WIP(workinprogress),limiting,374–377
workitems,limiting,373–377
Taxesonteamperformance,292–293
TDD(Test-DrivenDevelopment)
acceptancetests,129–131
automatedintegration,129–131
benefitinteams,124–125
benefitsof,133
buildingintotheproductbacklog,132
codesmells,125–126
continuousintegration,126–128
definition,13
definitionofdone,132
educationin,132
gettingstarted,132
implementing,124–125
improvingexistingcode,125–126
keypractices,123
keystosuccess,131–133
limitationsof,131–132
pairprogramming,128–129
principlesofclassdesign,126
refactoring,125–126
astory,119–123
teambuyin,132
teamstatus,127–128
testautomationpyramid,129
trainingandcoaching,132–133
Teamconsultants
accountability,47–48
buildingyourteam,42–47
versuscoreteams,43–44,46,49.SeealsoCoreteams.
establishingapool,40–42
keystosuccess,47–49
managementsupport,48
meetings,46–47
optimalpoolsize,45–46
overloading,49
overview,39–40
planningfordowntime,49
skillsandcompetencies,42–43
small-scaleexperiments,48
astory,35–39
timemanagement,37–39
transitionplans,40–41
Teammembers
busfactor,214
combiningwithotherroles,75–77
expendability,214
rotatingtheScrumMasterrole,79
Teammembers,adding.SeealsoOutsourcing.
Brooks’law(addingmanpowertolateprojects),45,233
consideringteamculture,238
developmentalstages,235–237
dropinvelocity,238
forming,235–237
groupcohesion,329
integratingnewmembers,234,237–238
keystosuccess,238–239
norming,235–237
pairprogramming,234
performing,235–237
risks,239
storming,235–237
astory,233–235
testingcompetencies,234,237–238
Teamvelocity
contractualagreement,360
definition,51
estimating.SeeEstimatingteamvelocity.
estimatingprojectcost,308–309
penaltyforaddingteammembers,238
astory,51–56
Teams.SeealsoCollaboration;Coreteams;People.
auxiliary.SeeTeamconsultants.
building,successfuloutsourcing,332–333
buyintoTDD,132
capacity,estimatingteamvelocity,59–60
colocated,138–140
dedicated.SeeCoreteams.
definition,428–429
distributed,140–141
long-termretention,truecostsofoutsourcing,329–330
newness,estimatingteamvelocity,57
optimalsize,35
part-time,141–142
prioritizingandestimatingproductbacklog,342–343
reportingperformance,roleoftheScrumMaster,113–114
sprintlength,87
statusreporting,127–128
taxesonperformance,292
workschedules.SeeCorehours.
Teamwork,dailyScrummeeting,214–215
Technicaldebt,294–295
Technicalrisk,423–424
Testfirst.SeeTDD(Test-DrivenDevelopment).
Test-DrivenDevelopment(TDD).SeeTDD(Test-DrivenDevelopment).
Testing.SeealsoTDD(Test-DrivenDevelopment).
automationpyramid,TDD,129
competenciesofnewteammembers,234,237–238
frequent,effectsondefects,171
Themes
decomposing,162–163,167
definition,160
ThreeC’sofuserstories,307
TimeframeforimplementingScrum,18
Timelimits,prioritizingandestimatingproductbacklog,347
Timemanagement
relativesizingandtimeestimation,372
teamconsultants,37–39
Timezones,truecostsofoutsourcing,332
Timeline,contractualagreement,359–360
Timing
contractualagreement,356–358
retrospectivemeetings,197
Traditionalcontracts.SeeContracts,traditionalmodel.
Trainingandcoaching.SeeEducating.
Transitionplans
teamconsultants,40–41
truecostsofoutsourcing,328–329
Transparency,292–294
Travelrequirements,successfuloutsourcing,335–336
Trends,determining,298–299
Tribalknowledge,176
Truncateddatacollection,estimatingteamvelocity,64–65
Trustfactorincontracts,363
Tuckman,Bruce,235–238
Two-weeksprints,incollaboration,377–378
Ullman,Ellen,317
Undonework,rescheduling,101
Userstories.SeeStories.
Usertypes(personas),identifying,359
Vagueness,dailyScrummeeting,212
Validation,deliveringworkingsoftware,283–284
Value,optimizingandmeasuring
defectmanagement,295
determiningtrendsandpatterns,298–299
educatingstakeholders,298
featurework,292–294
keystosuccess,297–299
preconditions,295–296
presentingdata,296–297
spikes,293–294
stories,definition,292
astory,289–291
structuringdata,296
taxesonteamperformance,292
technicaldebt,294–295
transparency,292–294
Values,Scrum
commitment,8
courage,8
focus,8
implementingScrum,7–8
openness,8
respect,8
ValveSoftware,hiringpractices,404–405,406
Velocity.SeeTeamvelocity.
Vision,enlistingsupportforScrum,31
Wagner,AnthonyD.,374
Waitandsee(realdata)method,estimatingteamvelocity,61–64
Wall,hitting,267–269
Wheelwright,Steven,374
Williams,Laurie,128,129,223
Wilson,Brad,227
Wilson,Peter,227
Windowofopportunity,deliveringworkingsoftware,283
WIP(workinprogress),limitingduringtaskestimation,374–377
Workitems,limitingincollaboration,373–377
Workpackages,330,333,335
Workloadestimation.SeeBurndown.
XP(ExtremeProgramming)
implementingScrum,12–13
practicesrequiredforScrum,12–13
Ziv’slaw(predictingsoftwaredevelopment),xx–xxi

Navigation menu