Interviews, insight & analysis on digital media & marketing

Turning obstacles into opportunities: how publishers can continue to thrive without traditional identifiers

By Jessica Werner, Senior Director, Publisher Development, ID5

With the recent announcement of Google limiting the functionality of GAIDs (Google Advertising IDs) on Android devices, as well as Google Analytics’ elimination of IP addresses, the result is the removal of yet more signals that contribute to identifying users. This, alongside the changes imposed by iOS15, and the upcoming deprecation of third-party cookies, portrays a future with fewer and fewer signals available to run personalised and measurable advertising campaigns.

The scope of the issue

The GAID announcement is certainly no new scenario for the digital media industry. The deprecation of traditional identifiers is something we have seen time and time again. With Mobile ad IDs (MAIDs), we have seen Identifier for Advertisers (IDFA) limited to being only available once the user has opted in to share it with the app. Recent news also confirms incoming changes to the Google Advertising IFA (GAID) set for the same opt-in basis in two years time.

In CTV, Google Analytics will no longer be logging or storing IP addresses. On the web, we know that third-party cookies have already been deprecated by Safari and Firefox browsers. Chrome is set to do the same in 2023.

All these signals that the industry has been relying on for so long are being restricted one by one in a privacy overhaul. The tightening of data privacy regulations and removal of traditional identifiers are initiatives that should guarantee higher protection of users’ privacy and publisher data. Cookie matching, for instance, allows for companies to access user data whether they have business relationships with the publisher or not. Nevertheless, the fact that there are fewer signals available for ad tech players to identify users on the web, mobile, and CTV results in additional limitations when it comes to addressability, which is likely to severely affect publishers’ monetisation efforts.

The role of alternative IDs

With traditional identifiers and signals being blocked, the industry must adapt and move towards a new, privacy-first and consumer-focused solution to thrive in the next era of digital advertising.

Several identity solutions have entered the market in recent years to tackle the disappearance of signals and the resulting identity crisis.

One of the most established identity resolutions is the first-party user ID, an alternative identifier enabling user-level recognition across websites and platforms without reliance on third-party cookies or MAIDs. First-party user IDs work by collecting a variety of publisher-provided signals to reconcile a user across various domains. They can collect these signals via deterministic and probabilistic methods. ID solution providers differentiate themselves from their competitors based on the signals that they leverage and the methodologies they apply.

Balancing accuracy and scale

Some ID solutions rely solely on deterministic methods to reconcile an ID across websites or apps. To do that, they need to use persistent signals such as hashed email addresses or login IDs. This method has the benefit of accuracy in identifying a user, yet falls short when it comes to scale. Today, only a small percentage of users log into websites, meaning only a limited number of users can be identified via deterministic IDs. On top of this, emails are being largely limited by Apple iOS, such as with its ‘Hide my Email’ initiative.

Relying on such signals alone means missing out on the opportunity to identify a huge portion of our potential audience. To provide a more scalable approach to identifying users, and enable publishers and brands to address a larger proportion of traffic, some identity solutions use probabilistic methodologies to reconcile IDs across different channels. Probabilistic IDs rely on the collection of signals, including IP address, page URL, user agent string and timestamp, and on algorithms that aggregate and process such signals to infer the identity of a user. Probabilistic IDs are, indeed, not as accurate as deterministic ones. On top of that, the removal of more signals can potentially make probabilistic IDs less reliable when it comes to accuracy.

The best approach for a future with fewer signals

Data privacy regulations are constantly evolving and browsers and operating systems are continuing to limit different signals at a rapid rate.

To keep up with evolving data privacy regulations and future-proof addressability in an increasingly signal-poor world, the combination of both probabilistic and deterministic methods is key to success. This allows ID solution providers to rely on a variety of signals to provide both scale and accuracy.

To ensure that users’ privacy preferences are respected, all types of signals, no matter if they are hashed emails, IP addresses or page URLs, need to be collected and processed only with users’ consent. There’s a common misconception that deterministic reconciliation methods are more privacy-compliant than probabilistic ones. The assumption is that if a user logs in to a website, then they are also comfortable with being identified across such websites. The reality is that users’ permission is needed, no matter the signal or reconciliation method, especially in highly regulated regions like Europe.

Another way to ensure that publisher’s identity providers receive the signals needed to identify users that have consented to be recognised is to integrate these solutions server-side. Server-to-server integration enables publishers to share signals with their ID solution provider via a direct encrypted connection between the two servers. Besides allowing publishers to share signals without being impacted by browser and OS restrictions, server-to-server integration also ensures that the transfer of data is much more secure, achieving an even greater level of privacy protection. This also allows them to receive and store the user IDs within their server databases and associate them with their client-side, first-party ID.

The time to act is now

The road ahead may seem daunting, but with more and more traditional signals exiting the equation, the adoption and integration of alternative identity solutions have never been more paramount. To stay ahead of the curve and continue to thrive, publishers need to ensure they select the right partners, offering the best approach to boost addressability and work closely with them to ensure identity partners receive the right signals via the most efficient and privacy-safe integration mechanisms. This industry shift marks a huge opportunity for the industry to move towards a more efficient, privacy-first way of operating – and it’s time to grab it with both hands.