- March 4, 2021

What happens to ad targeting if user tracking goes away?

Rachel Gantz
Rachel Gantz
Managing Director, Proximic

On March 3rd, 2021 Google announced, much to most of the ad tech industry’s chagrin, that “once third-party cookies are phased out, [Google] will not build alternative identifiers to track individuals as they browse across the web, nor will [they] use them in [their] products,” according to David Temkin, Director of Product Management, Ads Privacy and Trust, Google. In practical terms, this means that hashed email addresses and their associated IDs will likely not be a scalable third-party cookie replacement. 

Google’s announcement last year that Chrome would phase out third-party cookies by 2022 was the catalyst for large-scale privacy-centric initiatives and called into question the future of user-level targeting. With consumer privacy rights at the core, it wasn’t long before other IDs that track individual user behaviors needed to be examined. March 3rd’s announcement from Google creates challenges for the many initiatives that aim to maintain user tracking through hashed emails. This all points to the painful questions that many in ad tech have secretly been contemplating: if all the persistent identifiers go away, how will advertisers reach their target audiences? Without these persistent identifiers, is programmatic advertising and more specifically programmatic audience targeting as we know it dead?

The short answer is yes. The programmatic ecosystem today is largely transacted on ID-based data. Marketers can go into any DMP and DSP and see hundreds of data providers with thousands of segments available that literally enable a brand to target an ad campaign to Millennial pet owners if they so choose. With third-party cookies still on track to be deprecated in 2022, and mobile ad IDs and connected TV identifiers likely to follow, it is unlikely that the exact tools, techniques, and data sources that advertisers use today will be available after third-party identifiers are deprecated.

However, before we all go and drown our sorrows in an Ina Garten Cosmopolitan, the more important question to ask is: can this type of granular audience targeting exist without scalable user-level identifiers? And the good news is that the answer to this question is YES! In January 2021 Comscore released a targeting solution called Predictive Audiences that doesn’t require the type of invasive user tracking or targeting frameworks that make consumers and regulators nervous. This solution instead translates granular consumer behaviors from our opt-in panels (yes, the same best-in-class panels Comscore has been cultivating for the past 20-years) to content-level signals. Without tapping into (or even accessing) their potential customer’s data, advertisers can leverage these segments to target content (web, mobile, CTV, and audio) that is likely to be a predictor of a defined consumer behavior. Not only is Comscore offering this for its own demographics and media consumption data, but is also partnering with data providers such as Eyeota, PlaceIQ, Polk by IHS Markit, and TransUnion to bring B2B, location-based, automotive, and non-FCRA financial segments to advertisers as well.

Programmatic advertising as we know it will be forever changed, that part is clear. Comscore’s approach to a cookieless future is based on the principle of interoperability fit for purpose. Comscore will take multiple approaches and create intentional redundancies in methodologies to provide both accuracy and robustness among the changes in the landscape. This includes partnering with the many emerging unified ID solutions, initiatives like the Privacy Sandbox, as well as creating our own solutions leveraging consumer panels as a source of truth. At Comscore, we’ve been preparing for this new programmatic reality for a long time to ensure that advertisers can still achieve their KPIs at scale in a privacy-focused world.