E

ver since Google Chrome announced that it would be shutting off support for third-party cookies, the programmatic advertising industry has been searching for alternatives.  One of the main alternatives being considered are universal IDs.

In this article, I’ll explain what universal IDs are and how they work. 

Key Points About Universal IDs

• • A universal ID is a unique user ID that allows AdTech companies to identify users across different websites and devices.
  • Some universal IDs will use either probabilistic data (e.g. IP address, browser type and model, and user-agent string) or deterministic data (e.g. an email address or phone number), or both, to produce an ID.
  • Although universal IDs operate in a very similar way to IDs stored in third-party cookies, they don’t scale as well.
  • The main universal ID solutions are The Trade Desk’s Unified ID 2.0 (UID 2.0), ID5, LiveRamp, Tapad, Neustar, Epsilon, Zeotap, Flashtalking, and LiveIntent.

What Are Universal IDs?

A universal ID is a unique user ID that allows AdTech companies to identify users across different websites and devices. Universal IDs are created using probabilistic and deterministic data. 

Some universal IDs operate within one environment, such as web browsers, while others aim to identify users across different environments, such web browsers and mobile devices.

For identifying users across different environments, device graphs are used to match together the IDs generated in web browsers with the ones generated in other devices, e.g. mobile IDs in smartphones.

Universal IDs have emerged in response to the end of third-party cookies in major web browsers like Safari and Firefox, and the planned end of third-party cookies in Google Chrome in 2023. These universal IDs perform the same functions as third-party cookies, but the difference is in how they are created.

How Do Universal IDs Work?

Each universal ID will work in a slightly different way, but the general flow looks like this:

1. An Internet user visits a webpage. 

2. As the page loads, a request is sent from the web browser to the universal ID service. This request can contain both probabilistic data (e.g. IP address, browser type and model, and user-agent string) and deterministic data (e.g. an email address or phone number). Some universal IDs will use either probabilistic or deterministic data, or both, to produce an ID.

3. The universal ID service will either create a new ID or read the one it created previously.

4. The universal ID service will send back a new ID or the one that it created previously to the web browser. 

5. The website can then pass that ID on to its AdTech partners, e.g. a supply-side platform (SSP). The SSP will then pass the ID on to other AdTech platforms, e.g. ad exchanges and demand-side platforms (DSPs). If the DSPs can recognise the ID, then they’re able to tell whether the person visiting the website is a member of their target audience and then bid accordingly.

With this ID, AdTech companies can also run measurement, frequency capping and attribution. 

How Do Universal IDs Differ from Third-Party Cookies?

Note: When we talk about “third-party cookies”, we’re essentially talking about IDs stored in third-party cookies.

Although universal IDs operate in a very similar way to IDs stored in third-party cookies, the key difference is how the IDs are created and used.

With IDs in third-party cookies, pretty much any AdTech company can create a third-party cookie and store a unique ID inside it, provided they’re able to get their code to load on the page, e.g. by partnering with the publisher or via piggybacking.

With universal IDs, it’s up to the publisher to pass on the signals (e.g. probabilistic and deterministic data) to the universal ID service.

Also, it’s much easier for AdTech companies to create and recognize their third-party cookies across different websites than it is for universal IDs. The reason for this, again, has to do with how the IDs are created. IDs in third-party cookies can be created and recognized across multiple websites, provided the AdTech company’s code is loaded on the website.

With universal IDs, especially ones that use deterministic data to produce IDs, it’s much harder because the user would have to use the same signal (e.g. an email address) across multiple websites in order for the ID solution to recognize it. 

For universal IDs created using probabilistic data, the match rate increases, but it’s not as accurate as IDs based on deterministic data.

What’s an ID Graph?

An ID graph is essentially a centralized repository of IDs that are created using probabilistic and deterministic data collected from multiple sources (e.g. websites and mobile apps) and devices (e.g. laptops and smartphones).

Are Universal IDs a Long-Term Option?

Even though the universal IDs are one of the most popular alternatives to third-party cookies, they face one main challenge — they still revolve around identification and rely on some type of ID.

The reason this is a problem is because walled gardens like Google and Apple are constantly strengthening their products to make them more privacy friendly.

For this reason, many folks in the industry believe that these ID solutions are rather short-term solutions.

Many say that the future of digital advertising and marketing won’t be done on an individual basis, but rather done in a privacy-friendly way where individuals are not identified.