W3 Ceddl 201312

User Manual: W3

Open the PDF directly: View PDF PDF.
Page Count: 38

1
Customer Experience Digital Data Layer 1.0
Revision Date: December 10, 2013
Community Group Final Report
This version:
http://www.w3.org/2013/12/ceddl-201312
Authors:
Individual
Organization
Filip von Reiche Acceleration
Miles Fender
Murray Williams
Accenture
Reza Jalili Adobe
Jay Myers Best Buy
Jon Revill Blue Cross and Blue Shield of North Carolina
Jason Walsh BlueKai
Blane Sims
Jared Vestal
Eric Lunt
BrightTag
Paolo Margari British Council
Patrick Wyatt Criteo
Thomas Brune Cubite GmbH
Sam Law Conversion Foundry
Simon Rumble Data Republic
Anna Long
Eric Feinberg
Digital Analytics Association
Peter Loveday Digital Window.
Fares Aldik DIGITO Agency
2
Daniel Karpantschof Economist Digital, The
Mark Prince
Josh Goodwin
Joe Orlet
Ensighten
Mathieu Jondet Eulerian Technologies
Willem Paling Foxtel
Laura Holmes
Brian Kuhn
Justin Cutroni
Vishal Goenka
Lukas Bergstrom
Google
Brian Hendrixson HSN
Viswanath Srikanth
Aubrey Rupp
Hutch White
IBM
Casper Blicher Olsen IIH Nordic
Amin Shawki InfoTrust LLC
Andrew Thomas Intelligent Research
Martijn van Berkum GX Software
Lee Isensee Localytics
Jonathan Weber LunaMetrics
Gagan Kanwar
Emilie Laffray
William Hartley-Booth
Marin Software
Joseph Galarneau Mezzobit
Aurelie Pols Mind your Privacy S.L
Anton Gething nToklo
Nicolas Malo Optimal Ways
Roel Willems Orange Valley Consultancy
Matt Stannard 4Ps Marketing
Rod Jacka Panalysis
3
Harry Hurst
Ian McCaig
Stephen Elliott
QuBit
Keith Watkins Red Hat
Jonathan Conway Reevoo
Oliver Schiffers SapientNitro
Alexander Dean Snowplow Analytics
Adam Ware SwellPath
Toby Doig
Angus Glover Wilson
TagMan
Ali Behnam Tealium
David Henderson Triggered Messaging
Frederic Abrioux wemoTech
David Evans Whirlybird Consulting Limited
Evan Klein Zaelab
Copyright © 2013 the Contributors to the Customer Experience Digital Data 1.0 Specification,
published by the Customer Experience Digital Data Community Group under the W3C
Community Final Specification Agreement (FSA). A human-readable summary is available.
Abstract
This specification describes a method for surfacing customer experience digital data on a web
or other digital resource as a JavaScript Object which can be used for communicating this data
to digital analytics and reporting servers.
Status of this Document
This specification was published by the Customer Experience Digital Data Community Group. It
is not a W3C Standard nor is it on the W3C Standards Track. Please note that under the W3C
Community Final Specification Agreement (FSA) other conditions apply. Learn more about W3C
Community and Business Groups.
4
Table of Contents
Abstract
Status of this Document
1. Introduction
2. Design Rationale and Goals
3. How to Read this Document
4. Beginning with Examples
5. Privacy and Data Security Implications
5.1 Privacy Object Categorizes Vendors
5.2 Data Security property Categorizes Data
5.3 Implementation Example
6. The Customer Experience Digital Data Object
6.1 The Root JavaScript Object
6.2 Page Identifier Object
6.3 Page Object
6.4 Product Object
6.5 Cart Object
6.6 Transaction Object
6.7 Event Object
6.8 Component Object
6.9 User Object
6.10 Privacy Object
6.11 Security Object
6.12 Version Object
6.13 Extending the Specification
7. Industry Specific Examples of Using the Specification
7.1 Example from Air Travel Industry
7.2 Example from Healthcare Insurance Industry
8. Acknowledgements
9. References
A. Reserved Names and Identifiers
B. Privacy — Additional Considerations
B.1 Recommendations to Privacy and Security Solution Providers
B.1.1 Privacy
B.1.2 Data Security
B.2 Recommendations for Enhancing Privacy and Data Security
B.2.1 Privacy
B.2.2 Data Security
B.3 Privacy Regulations
B.3.1 Site Visitor Privacy Regulations
B.3.2 Data Security Protection Regulations
5
1. Introduction
Collection and analysis of visitor behavioral and demographic data has become an integral part
of web application design and website success, whether accessed through browsers on laptop,
mobile, kiosk, tablet or another device. This data is central to site performance analysis,
dynamically tailoring site content to visitor activity and interest and retargeting visitors based on
their behaviors.
Increasingly, multiple vendors are involved in the data collection process for a given digital
property, and each has a solution to be implemented on the page. As a result, page design has
become more complex and development cycles have lengthened as different requirements for
data surfacing and formatting are added to the implementation process. Further, changing or
adding vendors sometimes requires that the development team change designs to
accommodate vendor-specific requirements. Common data items must be continually surfaced
in different ways, and each design requirement is a custom effort. Companies are searching for
a simpler, more flexible, and standardized method to surface this common data across their
digital properties.
This document details the specification for a standard data layer that collects this valuable user
interaction information for subsequent use in analysis and reporting. The information in this
document will be relevant to digital analysts and website implementation engineers, as well as
to marketing professionals who need to understand user experience data that is being gathered.
6
2. Design Rationale and Goals
Customer experience digital data items that are tracked and captured by different vendors are
commonly understood elements used in digital analytics, but vendor-specific format
requirements and code assignments create design complexity and vendor-dependency in site
design.
As a simple instance, Vendor A may capture some digital data for page details as a
concatenation of PageID + PAGENAME + PAGECATEGORY while Vendor B may capture the
same through distinct variables vendorB.page.pageId, vendorB.page.pageName,
vendorB.page.pageCategory. Further, the names of the variables, the data structure name
and hierarchy, and methods of extensibility all vary on a per-vendor basis. Frequently, custom
code must be written to capture data to meet a given vendor’s requirements.
Cumulatively, differences of this type between vendors permeate across all relevant customer
experience digital data objects including pages, orders, shopping carts, registrants, and more,
increasing the complexity of customer experience digital data management at the
site.
Figure 1: Vendor specific tags on a digital property
To this end, a standard data object that represents common data elements in a standardized
way will allow development teams to implement design structures that populate the standard
data object. Vendor code placed on the page would reference that standard structure,
simplifying the process of onboarding or changing vendors, and shaving off expensive
development cycles.
7
Figure 2: Common data object accelerates deployment, simplifies site management
The proposed standard data object is a JavaScript object because of ubiquitous support for
JavaScript in web browsers and web-based applications, as well as in other forms of digital
properties like mobile devices, kiosks, and other digital content.
The use of a JavaScript object means that the data is not embedded in the HTML markup and
would not affect rendering of a page or performance. Developers will need only to populate the
data fields in the object that are applicable to the page, keeping the size and complexity of the
object to a minimum. This coding will never change regardless of vendor additions or deletions
unless new data elements were required. Because the data is a standard object, vendors who
recognize the object would provide code that references that object, requiring no modification of
the page other than to drop the vendor code container onto the page.
This standard will thus yield a shorter development cycle using fewer resources and there will
be a savings to the client in time and money. The addition or deletion of vendors will be
independent of considerations of page redesign, allowing decisions to be made based on
vendor efficacy rather than cost of migration.
8
3. How to Read this Document
To understand the value of the specification, Sections 1–3 give a high-level understanding
behind the rationale and applicability of the specification. Section 4 shows examples of how
code will be incorporated into your digital property, and Section 5 discusses privacy
implications. Sections 6–7 dive into the specific objects and their properties.
Also note that none of the objects in the specification are designed to be required — use
of each object is at the discretion of the digital property owner. However, once an object has
been selected for use, the object must adhere to the structure as specified in this document to
remain conformant.
The term reserved is encountered in the specification when defining Object Literals. In this
context, names of certain values are reserved for specific use within an Object Literal, and these
values should be populated only for the intended purpose to maintain the integrity of the
specification.
The keywords MUST and REQUIRED, SHOULD and RECOMMENDED, MUST NOT, SHOULD
NOT and NOT RECOMMENDED are to be interpreted according to RFC 2119.
9
4. Beginning with Examples
The specification relies on a JavaScript Object (JSO) to collect customer experience data. User
interaction with a brand's digital property occurs in a wide variety of areas: customer service, e-
commerce transactions, information portals, B2B partner interaction, and more, and the JSO is
designed to accommodate these (as well as being extensible). Further sections of this
document will delve into detailed specifications for the JSO, but following are simple examples
of an instance of the JSO to illustrate its usage.
The JSO is designed to be contained within a root object called digitalData — this is a
matter of convenience and gives a common starting point. All other objects are sub-object from
this root object. There is a pageInstanceID that is used to identify the page being measured
within a unique environment — development, staging, or production, for example. Beyond that,
the specification includes sub-objects such as page, product, cart, transaction, event,
component, user, privacyAccessCategories, and version for collecting different types
of data in the JSO. (Additional objects can be added to digitalData as part of the extension
mechanism.) Within the sub-objects, the specification defines a number of standard names for
properties, while custom properties can also be added through an attributes object.
As an illustrative example, the digitalData object with the page object could be populated as
below
digitalData = {
pageInstanceID: "MyHomePage-Production",
page:{
pageInfo: {
pageID: "Home Page",
destinationURL: "http://mysite.com/index.html"},
category:{
primaryCategory: "FAQ Pages",
subCategory1: "ProductInfo",
pageType: "FAQ"},
attributes:{
country: "US",
language: "en-US"}
}
};
The specification allows the use of sub-objects and their properties as needed in any particular
implementation. While the presence of the sub-objects is optional, sub-objects that are
populated must adhere to the syntax and semantics defined in this document to be
conformant with the specification.
Another instance of populating digitalData with both page and product objects is shown
below:
10
digitalData = {
pageInstanceID: "ProductDetailPageNikonCamera-Staging",
page:{
pageInfo:{
pageID: "Nikon Camera",
destinationURL:
"http://mysite.com/products/NikonCamera.html"},
category:{
primaryCategory: "Cameras",
subCategory1: "Nikon",
pageType: "ProductDetail"},
attributes:{
Seasonal: "Christmas"}
},
product:[{
productInfo:{
productName: "Nikon SLR Camera",
sku: "sku12345",
manufacturer: "Nikon"},
category:{
primaryCategory: "Cameras"},
attributes:{
productType: "Special Offer"}
}]
};
The product sub-object is an array with additional details to identify the product further.
As is evident from these examples, different pages that incorporate different types of data may
populate different parts of the JSO to pass those data back to the server(s) for further analysis.
11
5. Privacy and Data Security Implications
Site visitor interactions with a website are collected as data, and could potentially include
personally identifiable information or sensitive information about site visitors. The owner of a
website is responsible for the legal consequences of the data collected, published in the site,
and shared with technology vendors. This regulation differs from jurisdiction to jurisdiction and
commonly takes the form of two classes of legislation:
● Site visitor privacy regulation (e.g., the EU Cookie Directive)
● Data security protection regulations (e.g., HIPAA)
Site visitor privacy regulations restrict site practices around whether or not to track a visitor’s
behavior. Data security protection regulations legislate whether a site owner can store and
share visitor data, as well as how and for what duration you can retain that data (for more
details on regulations, see Appendix B).
As the Customer Experience Digital Data initiative represents centralizing data collection, the
JSO was designed with vendor-neutral objects that site owners can use to control which third-
party technologies to enable on a site, and which data can be shared with the enabled
technologies.
Leveraging the specification for privacy and security requires building the privacy object as
well as the security objects for those components of data objects that are meant to be
protected. Actual enforcement depends on processing that is external to the JSO. (See 6.10
and 6.11 for the specification of the privacy and security objects, and see Appendix B for
recommendations to solution providers).
5.1 Privacy Object Categorizes Vendors
The JSO allows for the creation of a privacy object that is meant to categorize the various
vendors with whom data may be shared into a small but meaningful set. The privacy object
can be used to create vendor categories relevant for the website. For example, a site may
create vendor categories such as "Analytics", "Personalization", and so forth. Categorizing the
vendors in this fashion makes it easier to indicate what type of data sharing is appropriate with
each.
5.2 Data Security property Categorizes Data
The JSO also allows for the creation of an optional security sub-object that is attached to any
object in the specification. This security object identifies the categories of vendors (defined in
the privacy object) with which that particular data property can be shared. For example,
productInfo.security = "Analytics" implies that the productInfo property can be
only shared with vendors categorized as "Analytics" vendors and vendors placed in the
"Default" category in the privacy object.
12
5.3 Implementation Example
Consider the following example, of an e-commerce company, www.nonexistent-store.com,
which uses tools from the vendors calc.com, adsRus.com, and audit.com, and uses the
digitalData object below:
Because www.nonexistent-store.com operates primarily in the Netherlands, they have a
requirement for privacy, and because they sell medication there are also concerns around the
sensitivity of the data they collect and store.
To use the data standard, first they will categorize their technology vendors into categories:
● Default: nonexistent-store.com, audit.com
● Analytics: calc.com
● Recommendations: adsRus.com
To implement those categories, they will code the
digitalData.privacy.accessCategories sub-object:
13
digitalData.privacy.accessCategories.categories[1] = {
categoryName: "Default",
domains: ["nonexistent-store.com", "audit.com"]};
digitalData.privacy.accessCategories.categories[2] = {
categoryName: "Analytics",
domains: ["calc.com"]};
digitalData.privacy.accessCategories.categories[3] = {
categoryName: "Recommendations",
domains: ["adsRus.com"]};
Second, they will identify any objects within the JSO that need data security, and add
security objects for those objects. (The Default category is always allowed, because it
represents the first party.) In this example, there is a need to protect parts of the product,
cart, and transaction data from Recommendations category of tools because one product
is sensitive. The following will restrict the data to only the “Analytics” and “Default” categories.
digitalData.product[2].productInfo.security =
digitalData.cart.item[4].productInfo.security =
digitalData.transaction.item[4].productInfo.security = {
productName: "Analytics",
description: "Analytics",
productUrl: "Analytics"};
digitalData.product[2].category.security =
digitalData.cart.item[4].category =
digitalData.transaction.item[4].category = {
productType: "Analytics"};
(Transaction data commonly contains information that could potentially identify a visitor; this
example does not enumerate that configuration for brevity.)
Finally, the site will deploy a privacy enforcement technology, presumably via script tag at the
top of the head in the page, which controls access to information in the digitalData object
based on these categories. Recommendations to security solution providers are provided in
Appendix B, as well as considerations for extending the specification in those specific
implementations.
14
6. The Customer Experience Digital Data Object
This section carries the core specification.
Use of this specification can be adapted to specific cases, since the objects in the specification
can be included or omitted as necessary or desired. None of the objects in this specification are
required, but where objects are included they MUST conform to the Object Names and Types
listed here.
6.1 The Root JavaScript Object
This section is normative
The root JavaScript Object (JSO) MUST be digitalData, and all data properties within this
specification MUST fall within the hierarchy of the digitalData object.
The following sub-objects are defined as children of the digitalData object.
● digitalData.pageInstanceID
● digitalData.page
● digitalData.product[n]
● digitalData.cart
● digitalData.transaction
● digitalData.event[n]
● digitalData.component[n]
● digitalData.user[n]
● digitalData.privacyAccessCategories
● digitalData.version
The following subsections detail each of these sub-objects.
6.2 Page Identifier Object
This section is normative
The Page Identifier is among the most widely used web analytics data properties, and is among
the top level digitalData objects. A Page Identifier, where included, MUST have the
following Object Name & Type.
Object Name
Type
digitalData.
pageInstanceID
String
A unique identifier for a page or other piece of content for which data is being collected.
This value SHOULD distinguish among environments, such as whether this page is in
development, staging, or production. (Contrast with digitalData.page.pageID below, which
may be unique only within a particular environment.)
15
6.3 Page Object
This section is normative
The Page object carries significant details about the page, and the most commonly used data
elements are captured by the specification below. The Page object and its children, where
included, MUST have the following Object Names & Types.
Object Name
Type
digitalData.
page
Object Literal
Describes the page.
digitalData.page.
pageInfo
Object Literal
Describes details about the page.
digitalData.page.pageInfo = {
pageID: "316",
pageName: "Rogaine Hair Regrowth Treatment",
destinationURL: "http://site.com/r.html",
referringURL: "http://www.google.com/url?&q=&esrc=s",
sysEnv: "mobile",
variant: "2",
version: "1.14",
breadCrumbs: ["home","Products","haircare"],
author: "J Smith",
issueDate: "2013-09-01",
effectiveDate: "2013-09-20",
expiryDate: "2014-09-20",
language: "en-US",
geoRegion: "US",
industryCodes: "SIC codes",
publisher: "BusDev"
};
Reserved: pageID (String), pageName (String), destinationURL (String), referringURL
(String), sysEnv (String), variant (String), version (String), breadcrumbs (Array::String),
author (String), issueDate (String or Date Object), effectiveDate (String or Date
Object), expiryDate (String or Date Object), language (String), industryCodes (String),
publisher (String)
For destinationURL, and referringURL, RECOMMENDED values are
document.location, and document.referrer, respectively.
The properties onsiteSearchTerm and onsiteSearchResults are RECOMMENDED for
measuring the query term and number of results returned for onsite search functions.
For fields containing dates, JavaScript Date Objects or Strings using ISO 8601 are
RECOMMENDED.
For fields containing language or locale, ISO 3166 and 639 are RECOMMENDED.
16
digitalData.page
.
category
Object Literal
Because of the wide range of methods for categorization, an object literal is provided for page
categories.
digitalData.page.category = {
primaryCategory: "FAQ Pages",
subCategory1: "ProductInfo",
pageType: "FAQ"
};
Reserved: primaryCategory (String)
The name primaryCategory is RECOMMENDED if you included only one set of categories
for pages, or for your primary set of categories. All other names are optional and should fit the
individual implementation needs in both naming and values passed.
digitalData.page.
attributes
Object Literal
This object provides extensibility to the Page object. All names are optional and should fit the
individual implementation needs in both naming and values passed.
6.4 Product Object
This section is normative
The Product object carries details about a particular product with frequently used properties
listed below. This is intended for data about products displayed on pages or other content. For
products added to a shopping cart or ordered in a transaction, see the Cart and Transaction
objects below.
The Product object and its children, where included, MUST have the following Object Names &
Types.
Object Name
Type
digitalData.
product[n]
Array::Object Literal
List of products.
digitalData.product[n].
productInfo
Object Literal
This object describes the product.
digitalData.product[n].productInfo = {
productID: "rog3000",
productName: "Rogaine",
description: "Hair Regrowth",
productURL: "http://site.com/r.html",
productImage: "http://site.com/rog300_large.png",
productThumbnail: "http://site.com/rog300_thumb.png",
manufacturer: "Pharma",
size: "300ml" };
17
Reserved: productID (String), productName (String), description (String), productURL
(String), productImage (String), productThumbnail (String), manufacturer (String), sku
(String), color (String), size (String)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.product[n].
category
Object Literal
Because of the wide range of methods for categorization, an object literal is provided for
product categories.
digitalData.product[n].category = {
primaryCategory: "Haircare",
subCategory1: "Men's",
productType: "Thinning Hair Treatments"
};
Reserved: primaryCategory (String)
The name primaryCategory is RECOMMENDED if you included only one set of categories
for products, or for your primary set of categories. All other names are optional and should fit
the individual implementation needs in both naming and values passed.
digitalData.product[n].
linkedProdu
ct[n]
Array::Object Literal
List of products linked to this product.
digitalData.product[n].linkedProduct[n].
productInfo
Object Literal
As for for each linkedProduct.
Reserved: productID (String), productName (String), description (String), productURL
(String), productImage (String), productThumbnail (String), manufacturer (String), sku
(string), color (String), size (String)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.product[n].
attributes
Object Literal
This object provides extensibility to the Product object. Any additional dimensions related to the
product can be provided. All names are optional and should fit the individual implementation
needs in both naming and values passed.
6.5 Cart Object
This section is normative
The Cart object carries details about a shopping cart or basket and the products that have been
added to it. The Cart object is intended for a purchase that has not yet been completed. See the
Transaction object below for completed orders.
18
The Cart object and its children, where included, MUST have the following Object Names &
Types.
Object Name
Type
digitalData.cart.
cartID
String
An identifier for a particular shopping cart.
digitalData.cart.
price
Object Literal
This object provides details of the cart price. The basePrice SHOULD be the price of the
items before applicable discounts, shipping charges, and tax. The cartTotal SHOULD be
the total price inclusive of all discounts, charges, and tax.
digitalData.cart.price = {
basePrice: 200.00,
voucherCode: "Alpha",
voucherDiscount: 0.50,
currency: "EUR",
taxRate: 0.20,
shipping: 5.00,
shippingMethod: "UPS",
priceWithTax: 120,
cartTotal: 125
};
Reserved: basePrice (Number), voucherCode (String), voucherDiscount (Number),
currency (String), taxRate (Number), shipping (Number), shippingMethod (String),
priceWithTax (Number), cartTotal (Number)
For currency values, ISO 4217 is RECOMMENDED.
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.cart.
attributes
Object Literal
This object provides extensibility to the cart as a whole. Any additional dimensions related to
the cart can be provided. All names are optional and should fit the individual implementation
needs in both naming and values passed.
digitalData.cart.
item[n]
Array::Object Literal
List of items in the cart.
digitalData.cart.item[n].
productInfo
Object Literal
As for digitalData.product[n].productInfo, for each item.
19
Reserved: productID (String), productName (String), description (String), productURL
(String), productImage (String), productThumbnail (String), manufacturer (String), sku
(string), color (String), size (String)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.cart.item[n].
category
Object Literal
As for digitalData.product[n].category, for each item.
Reserved: primaryCategory (String)
The name primaryCategory is RECOMMENDED if you included only one set of categories
for products, or for your primary set of categories. All other names are optional and should fit
the individual implementation needs in both naming and values passed.
digitalData.cart.item[n].
quantity
Number
Quantity of this particular item in the cart.
digitalData.cart.item[n].
price
Object Literal
As for the main cart object digitalData.cart.price, where tracking on each individual
item is desired.
Reserved: basePrice (Number), voucherCode (String), voucherDiscount (Number),
currency (String), taxRate (Number), shipping (Number), shippingMethod (String),
priceWithTax (Number)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.cart.item[n].
linkedProduct[n]
Array::Object Literal
List of products linked to this item in the cart, as for
digitalData.product[n].linkedProduct[n], for each of the item.
digitalData.cart.item[n].linkedProduct[n].
productInfo
Object Literal
As for digitalData.productInfo, for each linkedProduct.
Reserved: productID (String), productName (String), description (String), productURL
(String), productImage (String), productThumbnail (String), manufacturer (String), sku
(string), color (String), size (String)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.cart.item[n].
attributes
Object Literal
This object provides extensibility to each item within the cart. Any additional dimensions related
to the item can be provided. All names are optional and should fit the individual implementation
needs in both naming and values passed.
20
6.6 Transaction Object
This section is normative
The Transaction object is similar to the Cart object, but represents a completed order. The
Transaction object contains analogous sub-objects to the Cart object as well as additional sub-
objects specific to completed orders.
The Transaction object and its children, where included, MUST have the following Object
Names & Types.
Object Name
Type
digitalData.transaction.
transactionID
String
A unique identifier for a particular transaction; usually an existing order identifier.
digitalData.transaction.
p
rofile
Object
A profile for information about the purchaser, typically associated with a registered user.
digitalData.transaction.profile.
profileInfo
Object Literal
An extensible object for providing information about the purchaser.
digitalData.transaction.profile.profileInfo = {
profileID: "humanbeing12345",
userName: "me"
};
Reserved: profileID (String), userName (String), email (String)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.transaction.profile.
address
Object Literal
An extensible object for providing (billing) address information for the purchaser.
digitalData.transaction.profile.address = {
line1: "673 My Street",
line2: "Apt 1",
city: "Austin",
stateProvince: "TX",
postalCode: "78610",
country: "USA"
};
Reserved: line1 (String), line2 (String), city (String), stateProvince (String),
postalCode (String), country (String),
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
21
digitalData.transaction.profile.
shippingAddress
Object Literal
As for address, but for the shipping address.
Reserved: line1 (String), line2 (String), city (String), stateProvince (String),
postalCode (String), country (String),
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.transaction.
total
Object Literal
This object provides details of the final price the purchaser has to pay. The basePrice
SHOULD be the price of the items before applicable discounts, shipping charges, and tax.
The transactionTotal SHOULD be the total price inclusive of all discounts, charges, and
tax.
digitalData.transaction.total = {
basePrice: 200.00,
voucherCode: "Alpha",
voucherDiscount: 0.50,
currency: "EUR",
taxRate: 0.20,
shipping: 5,
shippingMethod: "UPS",
priceWithTax: 120,
transactionTotal: 125
};
Reserved: totalPrice (Number), voucherCode (String), voucherDiscount (Number),
currency (String), taxRate (Number), shipping (Number), shippingMethod (String),
priceWithTax (Number), transactionTotal (Number)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.transaction.
attributes
Object Literal
This object provides extensibility to the transaction as a whole. Any additional dimensions
related to the cart can be provided. All names are optional and should fit the individual
implementation needs in both naming and values passed.
digitalData.transaction.
item[n]
Array
List of items in the transaction.
digitalData.transaction.item[n].
productInfo
Object Literal
As for digitalData.product[n].productInfo, for each item.
The productID property is RECOMMENDED.
22
Reserved: productID (String), productName (String), description (String), productURL
(String), productImage (String), productThumbnail (String), manufacturer (String), sku
(string), color (String), size (String)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.transaction.items[n].
category
Object Literal
As for digitalData.product[n].category, for each item.
Reserved: primaryCategory (String)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.transaction.item[n].
qua
ntity
Number
Quantity of this particular item ordered.
digitalData.transaction.item[n].
price
Object Literal
As for the main transaction object digitalData.transaction.price, where tracking on
each individual item is desired.
Reserved: basePrice (Number), voucherCode (String), voucherDiscount (Number),
currency (String), taxRate (Number), shipping (Number), shippingMethod (String),
priceWithTax (Number)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.transaction.item[n].
linkedProduct[n]
Array::Object Literal
List of products linked to this item in the order, as for
digitalData.product[n].linkedProduct[n], for each item.
digitalData.transaction.item[n].linkedProduct[n].
productInfo
Object Literal
As for digitalData.productInfo, for each linkedProduct.
Reserved: productID (String), productName (String), description (String), productURL
(String), productImage (String), productThumbnail (String), manufacturer (String), sku
(string), color (String), size (String)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.cart.items[n].
attributes
Object Literal
This object provides extensibility to each item within the transaction. Any additional dimensions
related to the item can be provided. All names are optional and should fit the individual
implementation needs in both naming and values passed.
23
6.7 Event Object
This section is normative
The Event object collects information about an interaction event by the user. An event might be
a button click, the addition of a portal widget, playing a video, adding a product to the shopping
cart, etc. Any action on the page could be captured by an Event object.
The Event object and its children, where included, MUST have the following Object Names &
Types.
Object Name
Type
digitalData.
event[n]
Array
List of events that were measured.
digitalData.event[n].
eventInfo
Object Literal
This object describes the event.
digitalData.event[n].eventInfo = {
eventName: "Add News Portal",
eventAction: "addportal",
eventPoints: 200,
type: "contentModifier",
timeStamp: new Date(),
effect: "include portal 1234"
};
The eventInfo object MUST include eventName.
Reserved: eventName (String), eventAction (String), eventPoints (Number), type
(String), timeStamp (String or Date Object), cause (String), effect (String)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.event[n]
.
category
Object Literal
Because of the wide range of methods for categorization, an object literal is provided for page
categories.
digitalData.event[n].category = {
primaryCategory: "Portal",
subCategory1: "dashboard"
};
Reserved: primaryCategory (String)
The name primaryCategory is RECOMMENDED if you included only one set of categories
for events, or for your primary set of categories. All other names are optional and should fit the
individual implementation needs in both naming and values passed.
24
digitalData.event[n].
attributes
Object Literal
This object provides extensibility to each event. Any additional dimensions related to the event
can be provided. All names are optional and should fit the individual implementation needs in
both naming and values passed.
6.8 Component Object
This section is normative
The Component object is intended to capture information about a content component included
as part of a page, such as a video. Interactions with the component — such as playing the video
— would be an Event, captured by the Event object above.
The Component object and its children, where included, MUST have the following Object
Names & Types.
Object Name
Type
digitalData.
component[n]
Array::Object Literal
List of page components.
digitalData.component[n].
componentInfo
Object Literal
This object describes the component.
digitalData.component[n].componentInfo = {
componentID: "rog300v",
componentName: "How to Use Rogaine",
description: "Hair Treatment Video"
};
Reserved: componentID (String)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.component[n]
.
category
Object Literal
Because of the wide range of methods for categorization, an object literal is provided for page
categories.
digitalData.component[n].category = {
primaryCategory: "Haircare",
subCategory1: "Videos",
componentType: "Flash Movie"
};
Reserved: primaryCategory (String)
25
The name primaryCategory is RECOMMENDED if you included only one set of categories
for components, or for your primary set of categories. All other names are optional and should
fit the individual implementation needs in both naming and values passed.
digitalData.component[n].
attributes
Object Literal
This object provides extensibility to each component. Any additional dimensions related to the
component can be provided. All names are optional and should fit the individual
implementation needs in both naming and values passed.
6.9 User Object
This section is normative
The User object captures the profile of a user who is interacting with the website.
The User object and its children, where included, MUST have the following Object Names &
Types.
Object Name
Type
digitalData.
user[n]
Array::Object Literal
List of user(s) interacting with the page. (Although typically web data has a single user per
recorded interaction, this object is an array and can capture multiple users.)
digitalData.user[n].
segment
Object Literal
This object provides population segmentation information for the user, such as premium versus
basic membership, or any other forms of segmentation that are desirable. Any additional
dimensions related to the user can be provided. All names are optional and should fit the
individual implementation needs in both naming and values passed.
digitalData.user[n].
profile[n]
Array::Object Literal
A profile for information about the user, typically associated with a registered user. (Although
typically a user might have only a single profile, this object is an array and can capture multiple
profiles per user.)
digitalData.user[n].profile[n].
profileInfo
Object Literal
An extensible object for providing information about the user.
digitalData.transaction.profile.profileInfo = {
profileID: "humanbeing12345",
userName: "me"
};
Reserved: profileID (String), userName (String), email (String), language (String),
returningStatus (String), type (String)
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
26
digitalData.user[n].profile[n].
address
Object Literal
An extensible object for providing address information for the user.
digitalData.transaction.profile.address = {
line1: "673 My Street",
line2: "Apt 1",
city: "Austin",
stateProvince: "TX",
postalCode: "78610",
country: "USA"
};
Reserved: line1 (String), line2 (String), city (String), stateProvince (String),
postalCode (String), country (String),
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
digitalData.user[n].profile[n].
social
Object Literal
An extensible object for providing social information for the user profile.
digitalData.user[n].profile[n].social = {
twitter: "somebody",
twitterInfo: "stuff",
facebook: "somebody1234",
facebookInfo: "morestuff"
};
All names are optional and should fit the individual implementation needs in both naming and
values passed.
digitalData.user[n].profile[n].
attributes
Object Literal
This object provides extensibility to the profile.
All other names are optional and should fit the individual implementation needs in both naming
and values passed.
6.10 Privacy Object
This section is normative
The Privacy object holds the privacy policy settings that could be used to:
1. Capture and enforce site visitor consent to use tracking technologies on the site.
2. Together with Security objects described below, secure access to individual objects
within the JSO by categories of tracking technologies.
The Privacy object and its children, where included, MUST have the following Object Names &
Types.
27
Object Name
Type
digitalData.privacy.
accessCategories[n]
Array::Object Literal
A list of privacy categories.
digitalData.privacy.accessCategories[n].
categoryName
String
Definition of category names to be associated with security objects with the matching name.
A categoryName of Default MUST be included for privacy enforcement. All other categories
are custom. Common categories include Analytics, Recommendations, and Personalization.
digitalData.privacy.accessCategories[n].
domains
Array
Particular vendors associated with the particular category. This SHOULD be an array listing
domains for vendors associated with the category.
The Default category SHOULD contain all technologies which must be treated the same as first
party for privacy purposes (e.g., a tag management technology).
6.11 Security Object
This section is normative
The Security object is an optional sub-object of each object in the JSO, which can be used to
enforce data access control over that object. If a Security object is defined for an object, the
value should be a comma-separated list of categories defined as categoryName in the Privacy
object. If a Security object is not defined for an object, no data access controls will be enforced.
The Default category is required for security enforcement. The Default category has access to
all children of the digitalData element and does not need to be explicitly listed in Security
objects.
For instance, the security object associated with cartID would be written as:
Object being secured: digitalData.cart.cartID
Security object: digitalData.cart.cartID.security: "Analytics"
For object literals, the same structure used in the the object literal is repeated in the Security
object with values for the security levels. For instance:
Object being secured:
digitalData.transaction.profile.shippingAddress = {
line1: "673 Mystreet",
line2: "Apt 1",
city: "Austin",
stateProvince: "TX",
postalCode: "78610",
country: "USA",
shippingMethod: "UPS",
specialInstruction: "Beware of Dog"
};
28
Security object:
digitalData.transaction.profile.shippingAddress.security = {
line1: "Personalization",
line2: "Personalization",
city: ["Analytics","Personalization"],
stateProvince: "Analytics",
postalCode: "Analytics"
};
Because shippingMethod and specialInstruction do not have a security object defined,
there is no access control on those elements.
Object Name
Type
*.
security
String or Object Literal, depending on the parent object being secured.
Security sub-object that captures the specific setting for its parent object. The value of security
properties MUST be a string or array with values drawn from the values of
digitalData.privacy.accessCategories[n].categoryName.
An empty string for a Security object indicates that only the Default category is allowed to
access the data. No Security object denotes that no data security protection is necessary for a
piece of data.
Security object with empty string: digitalData.cart.cartID.security = "";
Only the Default category can access the cartID property. Synonymous with
digitalData.cart.cartID.security = "Default"; .
Security object not specified: digitalData.cart.cartID.security = null;
Any category can access the cartID property (i.e., no data access controls are enforced).
6.12 Version Object
This section is normative
The version of this specification used by the digitalData object.
Object Name
Type
digitalData.
version
String
The version number corresponding to the W3C Community Specification for Customer
Experience Digital Data followed by the JSO. For digitalData objects conforming to this
specification document, the value SHOULD be digitalData.version = "1.0".
29
6.13 Extending the Specification
Extending this specification is straightforward: implementers can add appropriate sub-objects or
properties as needed.
Adding a new object. A new object can be added to the digitalData object to represent a
completely new type of data. This could be added as a first-level object, such as the following:
digitalData.newObject = { };
It could also take the form of adding a new sub-object to an existing object:
digitalData.transaction.newObject = { };
Adding a new properties. An existing object can be extended by adding additional name-value
pairs:
digitalData.user[n].profile[n] = { newValue: "value" };
If reserved names are employed, they MUST use the types specified. If other names are
employed, the values may take on any type. All of the first-level objects and some of their
children contain an attributes object for extensibility, which does not reserve any names.
When extending the data layer, names SHOULD NOT use JavaScript reserved words,
predefined names, or event handlers for the names of objects or properties. These are laid out
in Appendix A.
30
7. Industry Specific Examples of Using the Specification
The specification, as shown above, is easily extensible by adding new objects, sub-objects, or
additional name-value properties to object literals. Thus the specification can be used across
industries by either adding entirely new objects, or extending existing objects in the
specification.
7.1 Example from Air Travel Industry
Following are a number of possible examples of employing and extending this specification for
data for an air travel booking website.
● Core flight reservation related data can be used to populate and extend the
productInfo field as shown below:
digitalData.product[n].productInfo = {
productID: "734565538989889110",
description: "Business Class One-Way Ticket",
originAirportCode: "RDU",
originCity: "Raleigh",
originState: "North Carolina",
originCounty: "USA",
destinationAirportCode: "BOM",
destinationState: "Maharashtra",
destinationCountry: "India",
departureDate: new Date("December 15, 2013 14:20:00"),
arrivalDate: new Date("December 16, 2013 21:40:00"),
numberOfTravellers: 1
};
● As a travel product is moved into the cart details such as price, fees, and currency can
be used to populate or extend the digitalData.cart.price object literal. Additional
fields such as a confirmation number might extend the digitalData.transaction
objects.
digitalData.cart.price = {
basePrice: 1000.00,
currency: "USD",
fees: 200,
taxRate: 0.08,
cartTotal: 1296.00
};
● If a visitor carries out an on-site search for available flights, the search and its results
can be captured using the existing
digitalData.page.pageInfo.onsiteSearchTerm and
digitalData.page.pageInfo.onsiteSearchResult objects.
● Details regarding the Customer making the purchase can be captured in the User object.
Existing objects in the specification could be used, such as:
○ digitalData.user[n].profile[n].address, for the customer’s postal
address.
31
○ digitalData.user[n].profile[n].profileInfo for any other customer
identifiers.
○ A "frequent flyer club" property could be added to the
digitalData.user[n].segment object to capture the customer’s loyalty
level:
digitalData.user[n].segment = {
frequentFlyerClub: "Silver Elite"
};
7.2 Example from Healthcare Insurance Industry
As another example, to collect data in the healthcare insurance industry, it might be more
convenient to model a few entirely new objects to capture relevant information.
One such bespoke object could be digitalData.member, used to capture a member’ s
account details with the institution, which may frequently be insurance related. This could
include sub-objects as detailed below:
Example Member object
digitalData.member[n] = {
memberID: "2723 49202388 01",
age: "41",
groupRelationship: "436378",
groupName: "Employer\, Inc.",
relation: "Spouse",
gender: "M",
originalJoinDate: "2011-01-21",
postalCode: "15214"
};
Other useful objects may include an Application object and a Plan object, with additional sub-
objects under each.
Example Application object
digitalData.application = {
appID: "7565-2373-0086-8937",
source: "Telephone",
status: "Pending",
creationDate: new Date("December 15, 2013 14:20:02"),
completionDate: new Date("December 15, 2013 16:05:16")
};
Example Plan object
digitalData.plan = {
name: "Family Advantage 250",
type: "EPO",
policyStatus: "Current",
premium: 454.25,
effectiveDate: new Date("December 15, 2013 16:05:16") };
32
8. Acknowledgements
The editors would like to acknowledge, in addition to all the mentioned authors, the contributions
of Eliot Towb, Michael Niemann, Jim Colson, Christopher B Ferris, Tom Glover, Karen Myers for
giving this initiative the initial momentum and contributions to the specification.
9. References
RFC 2119: Key words for use in RFCs to Indicate Requirement Levels , S. Bradner, Author.
Internet Engineering Task Force, March 1997. Available at http://www.ietf.org/rfc/rfc2119.txt.
RFC 4627: The application/json Media Type for JavaScript Object Notation (JSON), D.
Crockford, Author. Internet Engineering Task Force, July 2006. Available at
http://www.ietf.org/rfc/rfc4627.txt.
33
Appendices
A. Reserved Names and Identifiers
Avoid the following Reserved Words, Predefined Names, and Event Handlers for the object or
property names when extending the specification.
Reserved Words:
break enum instanceof super yield
case export interface switch
catch extends let this
class finally new throw
continue for package try
debugger function private typeof
default if protected var
delete implements public void
do import return while
else in static with
Predefined Names:
alert element Infinity Object scroll
all elements isFinite offscreenBuf
fering
secure
anchor embed isNaN open select
anchors embeds isPrototypeO
f
opener self
area encodeURI java option setInterval
34
encodeURICom
ponent
JavaArray outerHeight setTimeout
assign escape JavaClass outerWidth status
blur eval JavaObject packages String
button event JavaPackage pageXOffset submit
checkbox fileUpload innerHeight pageYOffset taint
clearInterva
l
focus innerWidth parent text
clearTimeout form layer parseFloat textarea
clientInform
ation
forms layers parseInt top
close frame length password toString
closed frames link pkcs11 undefined
confirm frameRate location plugin unescape
constructor function Math prompt untaint
crypto getClass mimeTypes propertyIsEn
um
valueOf
Date hasOwnProper
ty
name prototype window
decodeURI hidden NaN radio
decodeURICom
ponent
history navigate reset
defaultStatu
s
image navigator screenX
document images Number screenY
35
Event Handlers:
onbeforeunlo
ad
oncontextmen
u
onkeydown onmousedown onmouseup
onblur ondragdrop onkeypress onmousemove onreset
ondragdrop onerror onkeyup onmouseout onsubmit
onclick onfocus onload onmouseover onunload
Examples of identifiers to avoid include:
digitalData.user.public
digitalData.form
digitalData.transaction.status
36
B. Privacy — Additional Considerations
B.1 Recommendations to Privacy and Security Solution Providers
The Privacy and Security objects were designed to be vendor and technology neutral, while
providing all of the necessary data for functional enforcement without vendor lock-in. The
following sections provide detail on how to enforce, and allow for open source and home-grown
solutions, as needed.
B.1.1 Privacy
Technology providers can build support for this standard by enforcing the configured
preference. A technology which supports this standard must:
1. Provide interface to gather visitor consent for categories of technologies.
2. Enforce the visitor consent preference, and thereby prevent the execution of sharing
technologies or prevent the sharing of data by technologies
As of September 2013, there are at least three current approaches to address this concern:
1. Conditional Logic -- These technologies use a helper function to wrap every 3rd party
JavaScript in conditional logic, governed by interface controls. For example:
github.com/creativeaura/eu-cookie-opt-in
2. Firewall -- This technology deploys JavaScript to the browser which blocks JavaScript
from being added if the site visitor has opted out. For example:
developers.google.com/analytics/devguides/collection/analyticsjs/advanced#optout
3. Uncookieing or Unpixeling -- These technologies send a request to individual or all
technology networks to opt-out from data collection. The technology networks set a third
party cookie (pixel) indicating that the site visitor should not be tracked. Subsequent
page loads are ignored by technology networks. For example: aboutads.info/choices/
B.1.2 Data Security
Technology providers can support this standard by providing the functionality to execute
controlled data sharing as defined by the website administrator. A technology which supports
this standard must:
1. Provide an interface for website owners to create privacy objects and security sub-
objects. This interface must include the flexibility to define privacy categories and
names since these can vary greatly across industries and organizations.
2. Some method of supporting de-identification or anonymization algorithms, either built
into the tool, through API-type call, or some external access.
37
B.2 Recommendations for Enhancing Privacy and Data Security
B.2.1 Privacy
A fellow subgroup of the W3C is working on an open standard by which web users can express
their desire to opt in or out of tracking technologies (Tracking Preference Expression, under
development by the W3C Tracking Protection Working Group). As that standard develops and
gains adoption, we hope and expect Privacy technologies will seek to be compatible with that
specification.
B.2.2 Data Security
Data can be protected by restricting which vendors it is shared with, as intended by this
standard, and by sharing it in appropriate privacy-protecting formats (add note for reference with
case studies). There is a clear place to build additional specification for data handling policies to
govern this sharing -- anonymization, data-retention, de-identification, deletion, encryption, to
extend the data access controls enabled by this specification. The W3C Tracking Protection
Working Group is also developing a standard to define compliance to users’ tracking
preferences (Tracking Compliance and Scope). With adoption of that specification as a
standard, we hope and expect privacy technologies will support it as well.
B.3 Privacy Regulations
B.3.1 Site Visitor Privacy Regulations
Countries have enacted legislation requiring that site owners offer site visitors the ability to opt-
out from or opt-in to having their interactions tracked. This class of regulation is commonly
referred to as "cookie laws", due to the general emphasis on cookies, one method of sharing
data across organizations. As a result of these laws, site owners have implemented in-house or
commercial solutions to address these concerns. As of September 2013, many European Union
countries have enacted legislation based on a 2011 European Union directive. Other regions
outside the EU are also considering or enacting legislation in this area. For instance, Mexico
and New Zealand already have legislation in this regard.
You need to identify the legislation impacting your site, prior to implementing a Privacy solution.
The local legislation will indicate whether visitors must have the ability to opt-out, or whether
visitors must consent to opt-in for tracking. This determines the interface options you can use,
and whether a site owner can track initial page views.
You should also understand which technologies deployed on your site share data, as these are
the technologies that regulations are enacted to restrict.
B.3.2 Data Security Protection Regulations
When you understand the data your site is collecting and the regulations that apply to it, you
need to develop your security approach. The following questions address several common
dimensions of security:
● How will you protect personally identifiable information (which may itself be defined
differently based on regulations) or other sensitive data such as bank account or
insurance subscriber numbers?
38
Common policies are to restrict data access, delete the data, encrypt the data, or de-
identify the data. At a minimum you will want to restrict who can access the data. As
such, this standard provides a way to configure access controls over the data. Other
options are to delete, encrypt the data, or even de-identify or anonymize the data -- alter
it in some way so that the data owners cannot be easily identified.
● How long and under what circumstances is this data retained?
Retention periods for individual data should be set as well as decisions about
aggregation and archiving.
● Is an oversight program a good fit for your website? Several organizations, such as
TRUSTe, the Digital Advertising Alliance, and the Network Advertising Initiative, have
programs providing privacy oversight for website administrators. These organizations
develop tracking and privacy practices that they expect their members to adhere to.
Most of these programs are tailored to online advertising, where there is the greatest
threat of sharing.
You should consider these aspects of security as you develop your plan for securing your data.

Navigation menu