With Google confirming it will phase out cookies in 2024, brands are urgently looking for a new, future-proof solution for post-cookie digital advertising. They know success can no longer depend on the easy convenience of privacy-invasive personalised tracking. And consumers frown upon user tracking, so achieving fair consent at scale, as regulations increasingly demand, has become impossible.
Multiple emerging cookieless solutions are trying to convince brand marketers that they can deliver the same results as when individual tracking was the norm, ranging from unified IDs to contextual, semantic, and cohort-based targeting. As brands set out to find the best path forward in their transition to a post-cookie landscape, here are three tips that will offer a future-proof road map.
Forget cohorts, focus on personas
Cohort-based advertising, like Google Topics, which substitutes the original Federated Learning of Cohorts (FLoC), has claimed superiority as the next best thing to cookies. By collecting users’ browsing history, as opposed to data on the individual, this technique leverages user behaviour to segment general topics at an aggregate level (cohorts).
But cohort-based targeting still relies on gathering user information, without users being fully conscious or aware of it, and relies on tracking their digital behaviour.
Think about scaling in the long term
Multiple unified ID solutions already exist, but they run into a common problem: they are siloed and not interoperable. What’s more, they also still require consent, which makes it more difficult as users reject tracking en masse. In addition, they depend on a specific network of publishers, who do not want to share their users’ data either, which makes their reach extremely limited.
The Private Relay setting on Apple devices and its Safari browser for example encrypts users’ IP addresses and browsing data, making it impossible to reconcile IP addresses to unique IDs.
Demand more than contextual and semantic targeting
Despite the AI-driven advances of contextual and semantic targeting, these methods continue to be insufficient on their own. This targeting technique tries to predict who is looking at a page or app based on the context instead of specific user interests. In this way, it practically makes wild guesses, and, when it’s the only targeting method deployed for a campaign, it falls short of offering real understanding and audience engagement to brands.
Let’s take a common example: If user A is looking at the Sports page of a general news website, they will be categorised as a sports fan, triggering sports-related ads. This might well be the case: user A could indeed be a fan of one or several sports, but what about those other interests they no doubt have, which could also provide rich insights? What if User A also has a dog, but never visits pet-focused websites? When contextual and semantic techniques focus only on a tiny number of people gleaned from focusing on very specific topics, they miss these “hidden enthusiasts” and the advertising opportunities their various interests bring.