Consumer Identity Graphs are Not the Same as Marketing Databases

These days everyone says they have a consumer ID graph. In this article, the author suggests 26 of the best graphs available. If it feels like everyone is jumping in on the identity graph game, it’s because they are!

The great number of consumer ID graphs is in direct response to marketer demand. Clearly, ID graphs are seen as a reliable way to target consumers once the third-party cookie finally crumbles. But can we assume that all ID graphs are created equal?

If you’re not sure what an ID graph is, or how to distinguish it from a marketing database, read on!

High-Level Purpose of Consumer ID Graphs

The purpose of an identity graph is to uncover an identity of a consumer and to confirm that identity using various individual data points. Full stop.

Let’s say you want to know about the users who engage with your website, mobile app or other digital property, including how to reach them in the future. A consumer ID graph lets you:

Resolve an IP address, email address, MAID or any other identity to a single individual

Associate that individual to a specific household

Associate other data points, such as demo, geo, lifestyle info, in order to get a richer understanding of the people who are interested in your brand

Enable you to reach them wherever they are, across any channel or device

Consumer ID Graphs vs. Marketing Databases

Think of a consumer ID graph as an elaborate jigsaw puzzle, where each interconnecting piece contains an element of information that helps you accurately identify the customer behind any screen or device. To succeed, the provider’s goal is to capture as many individual data signals as possible in order to resolve and validate the individual’s identity.

Marketing databases, on the other hand, provide a list of consumers to target. This is very different from a consumer ID graph, and it solves a very different use case.

Here’s the key difference between the two: marketing databases tell you who to target, while consumer ID graphs allow you to recognize your customers and prospects as they switch from screen to screen and channel to channel, enabling you to deliver a seamless brand experience.

The Challenges of Building a Consumer ID Graph

It’s not easy to build a true ID graph. They take years of maturity to build up a proper ID history, and they require many data partners to contribute various ID data so that the provider can validate and confirm all the data points that go into the graph. To ensure accuracy, they also require an offline anchor (e.g. a physical address) that is connected to online behavior.

And given that the American population continuously starts new jobs, activates new devices and moves households, ID graphs require multiple real-time online data feeds.

Finally, they require hundreds, if not thousands, of proprietary scripts to determine the accuracy and confirmation of data. A reputable ID graph provider, like Throtle, never assumes data is accurate. We test, validate and test again against rigorous criteria.

The Throtle Advantage

Marketers across all industry sectors tell us the Throtle user ID graph delivers amazing results. We’re not surprised, as we designed it from the ground up with accuracy in mind. To begin, we use deterministic, not probabilistic, data. Do we think we know this user’s identity? Or do we have ample deterministic data points to say with certainty who a user is? If we can’t validate the data, it’s not included in our graph.

We also start at the individual level, not the household level. Household level requires clustering, which is inexact, and can lead to marketers spending their media budgets on people who aren’t the right audience for your ads. It’s much better to start at an individual level and ladder up to the household.

We also put a premium on accuracy. For instance, we built and utilize a proprietary and industry leading validated identity graph, and ensure the highest level of accuracy. We use a wide array of personal and persistent data points, including email addresses, postal addresses, mobile numbers, device IDs, MAIDs, and so on to validate identities consistently and in real time.

Before we label an identity as accurate, we go through extraordinary effort to validate it. For instance, each data point in our ID Graph is validated by confirming it’s the same across multiple and disparate sources that we’ve hand-picked and tested rigorously. We also have an external process that incorporates five levels of validation.

Throtle is the only identity and onboarding firm that has a rigorous data hygiene process to ensure deterministic and individual accuracy. We perform data quality checks, validate, standardize and parse customer data, as well as enhance customer records.

We’re also firm believers in transparency. Transparent results are a necessity for any marketer seeking to make the right strategic marketing decisions. That’s why we provide clients with transparency in their customer data and the reporting to prove it.

Our identity graph is built to allow the extension of records with emails and MAIDs – a process that is unique to Throtle, which increases match rates and delivers stronger campaign performance.

Finally, as our customers will attest, our match rates are much higher than the industry standard, thanks to our data-centric approach and ability to append multiple individual identifiers, including email addresses, physical addresses or phone numbers. We also offer insist on a complimentary match test prior to working together to prove the Throtle difference.

The Data Behind our Verification

If verification is key to accurate consumer ID graphs, what data does Throtle use to validate our data? A good (and fair!) question.

First, we create Individual Throtle IDs that are connected to any and all identifiers for each individual, including his or her name, address, email, phone, MAIDs, IP Addresses, etc. These identities are also connected to cookies for online browsers, cookieless IDs and mobile devices (MAIDs) to target individuals when browsing the web or mobile web. When possible, we authenticate identity with log-in data (e.g. if a consumer logs in to the New York Times from a home computer, mobile phone and work laptop, we can tie those devices back to the same individual).

Finally, we verify accuracy with multi-source corroboration and external reviews, as stated above.