Data Layer

2 Kudos

The goal of this article is to introduce the data layer: what it is, how it works, and its many benefits.

In this article:

Table of Contents Placeholder

What is a data layer?

A data layer is a specification of the vast amounts of customer interaction data that flow from your digital properties. The data will come from many sources, including your desktop/mobile websites, mobile apps, connected devices, and offline sources to name a few. This data powers your third-party vendor solutions and becomes the foundation of your data-driven initiatives.

The definition of your data layer starts with identifying the most important activity of your visitors, called events. Events encompass everything from page views, site searches, cart adds, and purchases to social interactions, button clicks, form submissions, and logins. These are all trackable events that will inform your critical business decisions. Each event will contain contextual information, called event attributes. The data layer comprises a list of these events and their corresponding attributes.

Tealium reserves a data layer variable named tealium_event to identify the events you track. You can define your own event names, but we have a standard set of events defined in the Tealium Events extension.

Here are a few examples of events that are common to many data layer definitions:

Site Search

Event Attribute	Description	Example Value
tealium_event	The event being tracked.	"search"
page_type	The type of page being viewed.	"search"
search_keyword	The term entered into the search.	"red boots"
search_results	The number of search results returned.	"42"

Newsletter Signup

Event Attribute	Description	Example Value
tealium_event	The event being tracked.	"newsletter_signup"
newsletter_name	The name of the newsletter.	"Weekly Digest"

Add to Cart

Event Attribute	Description	Example Value
tealium_event	The event being tracked.	"cart_add"
product_id	The ID of the product added to the cart.	["B51378"]
product_name	The name of the product added to the cart.	["Widget XYZ"]
product_quantity	The quantity of the product added to the cart.	["2"]

Once these events and attributes are defined for your business, your data layer will serve as the one true definition of your data across all digital assets and customer interactions.

Why do I need a data layer?

The data layer helps you take ownership of your most valuable asset--your data. You rely on many third-party vendors to power your business, each with their own data collection needs, and the data layer is a centralized definition of that data. A well-defined and managed data layer brings the following benefits:

Governance
Take control of your data by establishing a governance policy around your data layer.
Standardization
Use terminology that makes sense to your organization.
Marketing Agility
A well implemented data layer allows your marketing strategies to adjust quickly to changing conditions by eliminating your dependency on development resources.
Time/Cost Savings
Save time and money for developers by eliminating constant requests for incremental changes to the website, mobile applications, etc.

How does it work?

Once your data layer is defined you will install it across your digital properties. Your developers will follow the installation guides for each platform to ensure the specifications of the data layer are followed consistently. This data will then be used by in your tag vendor configurations and campaign strategies.

We recommend a naming convention that is simple and easy to comprehend. Some examples include:

Event Attribute	Description	Example Value
order_id	The unique identifier of a completed order.	"1234"
order_total	The total amount paid for the order.	"123.45"
customer_id	A unique identfier for the visitor.	"0123456789"
customer_zip	The visitor's zip code.	"92101"

This data can then be passed along to your third-party vendors in the format they require. This process is made simple with iQ Tag Management and Universal Data Hub.

As an example, the attribute named order_id is referenced across vendors using several different names, such as s.purchase_id, transid, pid, and oid to name a few. Since the data layer has defined order_id as the single source of the Order ID, it can be reused to send that value to the other vendors in the format they expect.

Event Attribute	Value	Vendor Attribute	Vendor Name
order_id	"1234"	s.purchase_id	Adobe Analytics
		transid	Rocket Fuel
		pid	Twitter Conversions
		oid	Conversant