Identity graphs help marketers do all the things marketers set out to do: personalize and target their ads and measure the success of those messages and campaigns. They ensure your brand can continue marketing on a 1-to-1 level across ad products once increased privacy measures are the norm across the web. 
 Working with third-party ID graphs brings marketers scale, but at the same time, a reliance on their partners’ capability to provide insights based on those graphs, which could be limited. This situation will only be exacerbated when third-party cookies go away since they will have even less access to customer information.
There are two types of methodologies that can be utilized in creating an ID graph: deterministic or probabilistic. 

Unique Identifiers: A Complex Situation for Marketers

Probabilistic matching uses unauthenticated data and is therefore not as accurate. This method, which unlike deterministic is highly scalable, assesses the probability of a match using AI and other pattern-seeking algorithms but provides much less certainty.
An ID graph offers a holistic picture of a customer based on all the identifying and behavioral data compiled across all their touchpoints. This allows marketers to link each customer across devices and channels and to their offline behavior in order to identify the person at an individual level.
Because brands have access to first-party data, and because first-party data is king as the web becomes more privacy-protected, brands are well-advised to begin owning their own ID graphs if they are not already doing so. 
One of those solutions is Identity Graphs, more commonly called ‘ID graphs.’

The Impact of Privacy

id graph
If you want control over your own ID graphs instead of relying on those of social platforms, you will need to find a vendor to partner with. You will also need to figure out which identifiers you already have captured from places like Google Analytics, Salesforce or other CRM, Data Management Platforms (DMPs) or Customer Data Platforms (CDPs)and which you may still want to go after. 

What Is an Identity Graph? [And How It Can Help]

In the real world, they may have a home address and a work address (where they sometimes send their packages). On any of those devices, there might be shared family logins for accounts on sites like Amazon or Netflix. On top of all of this, various cookies are picked up when they browse. 

Brands can create and own their own ID graphs using their data or they can leverage ID graphs from outside parties that interact with their customers, such as social media platforms or advertising networks. 

The waters have further been muddied by the move to increased data privacy on the web.  As third-party cookies are expected to be phased out soon, many alternative solutions are being developed to make up for the signal loss.
The Identity Graph can solve some of these problems by collecting disparate customer identifiers used both on- and offline and tying them together into one customer profile. Some ID graphs are more accurate than others, depending on the types of data they use.

What Are the Types of ID Graphs?

In the simplest terms, brands can have their own ID graphs or they can use someone else’s.

Create your own ID graph

This all creates a complex situation for marketers who seek to understand consumer behavior well enough to place the most personalized advertising possible in front of potential customers. 

Work with a third party ID graph

The push to privacy will impact all advertisers. Find out what you need to know about new restrictions, cookies, IDFA, first-party data, and all things privacy from our Tinuiti experts.

Identity Graphs can provide two types of customer profiles: authenticated and unauthenticated.

What Types of Customer Profiles Do Identity Graphs Provide? 

In the following article, we deep dive into why ID Graphs may be powerful solutions for marketers as they seek ways to continue personalized, targeted advertising in the face of ever-tightening privacy protection across the web.

Authenticated Profiles

Your identity partner will then work with you to take your data and apply their own special sauce to it to enrich and activate it into usable ID Graphs. One of the leading identity partner companies currently is LiveRamp. See this article for a complete list broken down by subset. 

Unauthenticated profiles

Unauthenticated profiles are less reliable because they are obtained not through logging in, but on short-term identification methods such as cookies (which are going away) and/or the ID of the particular device they are on. 

What Type of Methodologies Do Identity Graphs Provide?

People use a variety of different devices and email addresses when they are on the internet. In one day, a single person could jump from their laptop to mobile to tablet. They may be logged to a site with their personal email or their work email address (which may or may not be hashed). They may have two computers, one personal and one issued by their employer, which has two separate IP addresses. It’s likely they have a whole bunch of customer loyalty numbers as well as usernames for their various accounts.


The ID graph is a collection of disparate customer identifiers that can be tied together into one comprehensive and dynamic customer profile. An ID graph is created that synthesizes a consumer’s actions and behaviors on the web, on mobile, in-app, or in-store. 


You can think of an identity graph like one giant database that not only creates user profiles, but that is constantly at work matching records of the user across all the different places where they are consuming content, shopping, and living their lives. This ensures that each user’s record is as accurate, robust, and up-to-date as possible.

What are the Benefits of an ID Graph?

As you might guess, the first is privacy forward and the second is not. Authenticated profiles are those obtained through authentication measures such as logging in or purchasing something with a credit card. This user ID or email address is then used to discover and identify the same person across all their different channels and devices. 

How Do I Get Started with ID Graphs? 

Every time a consumer interacts with a brand, one (or more) of these unique identifiers is at play, making it impossible for marketers to accurately associate each user with their actions across various devices and channels.
Deterministic matching uses the types of authenticated data mentioned above. Because the data is certain, the matching is 100 percent certain.
Today, consumers use various devices and identifying information when they browse the web, which can make it difficult for marketers to tie individuals to their actions across devices and channels. Enhanced privacy measures across the web are only going to make this situation more difficult for marketers, who are going to lose signals. 

Similar Posts