Crumbs! Part 2 – Audience Targeting in a Cookieless World

Welcome back to part 2 of our series of articles covering cookieless preparedness. In our first article, found here , we delved into cookieless data collection and how marketers can set themselves up for success by following the 4 Cs of data collection, making sure to deliver Compliant, Customer-Centric and Connected data assets to power downstream personalisation and analytics.

In this article we’re going to turn our attention to audience targeting, before closing out the series with our final piece on measurement.

?

Same, Same, But Different.

The audience targeting landscape has been through rapid and continuous evolution since its inception. This evolution has been punctuated by certain critical “events” over the past decade. Whether it was the rise and fall (and rise!) of the ad network, the creation of the agency trading desk, the emergence of cross-device, the proliferation of audience data providers, and the rise of clean rooms.

One could certainly be forgiven for believing that the excitement and innovation surrounding audience targeting follows Einstein’s observation of energy, namely that it cannot be destroyed, but merely changes form. The removal of 3rd party cookies, marks yet another event in the long-running history of audience targeting, and although significant, will be survived, as audience targeting morphs into a new state.

A Chaotic Free-for-All?

?

In early 2024, we still anticipate that a significant volume of desktop and mobile web advertising will be traded and measured using 3rd party cookies. For a long time, cookies have supported some of digital marketing’s most tried and trusted use cases, whether that’s web retargeting, look-a-like modelling or audience suppression. But with 3rd party cookies set to go away, the reality is that many of these use cases will change or disappear altogether in the next 12-months.

?

So, what type of audience targeting capabilities are we able to use, in a world without 3rd party cookies.

?

1. The Walled Gardens – First we should acknowledge that most large advertisers invest significantly in walled garden media platforms like Amazon, Google and Meta. Although these platforms won’t be completely unaffected, they’re in a much stronger position, with most operating on one domain, where 1st party cookies can still be leveraged, and almost all having high volumes of logged-in customer data, negating the need for 3rd party cookies.
2. Contextual Targeting – When we move to the open web, we need to leverage existing and new signals, and in the hunt for predictive signals, contextual targeting has undoubtedly seen a renaissance. Using contextual signals to identify relevant content and other attributes, such as time, device, and general location information, can help brands to deliver a more personalised advertising experience.
3. Publishers, PPIDs & Universal IDs – A range of solutions on the market seek to replicate more closely the audience targeting brands have come to expect from digital marketing – namely historical interests / behaviours that are independent of the in-the-moment content consumption. Whether that is publishers offering 1st party audiences through private marketplaces or publisher provided IDs, or probabilistic solutions looking to re-imagine cross-domain data curation through statistical matching of 1st party cookies or other identifiers.
4. Propensity Modelling – Machine learning and AI could play a significant role in manifesting 1st party insights in cookie-free audience targeting. Propensity modelling harnesses 1st party analytics and user-agent signals to identify characteristics that demonstrate a high probability to drive conversions. These characteristics are then used to inform how we value inventory opportunities through custom bid strategies.
5. Customer Data and #PII – Customer records and the use of authenticated data sets, mapped to personally identifiable information such as email address or telephone number, has been an understandably popular option when addressing the removal of 3rd party cookies, as brands look to build strategies around more persistent forms of identity.
6. Clean Rooms – Data syndication through clean rooms, which act as technology-based mediators, allow for data sets to be uploaded into secure, neutral, privacy-focussed environments. Data can then be syndicated, or securely matched, for the purpose of insights generation and targeting. Importantly, this all happens without the physical movement of data between parties.
7. The Targeting API Revolution – Since Google announced its plans to remove 3rd party cookies in 2020, and simultaneously announced it was launching its Privacy Sandbox initiative, there has been a slow rolling launch of a range of solutions that focus on in-browser data processing, such as Google’s Topics API and Protected Audience API. Leveraging a new suite of browser-centric capabilities, these solutions allow for personalisation but limit data transfer by concentrating storage and decisioning to the browser.

?

Putting the Puzzle Together

If we were to plot these targeting solutions on a graph with scale as the Y axis and accuracy as the X axis, we’d quickly see that not all solutions are the same.

?

This ?highlights the reality that we’re very likely moving into a future where advertisers will be implementing a portfolio-based approach to targeting to satisfy the varied ways in which they seek to target audiences at different stages of the consumer journey, with contextual, probabilistic and propensity used to solve for upper-funnel, high reach targeting strategies, whilst publisher provided audiences offer a sense of normality with high quality audiences and inventory combines, and customer data / clean rooms offering specific tactical advantages at high accuracy on a known customer base.