What Google’s “Topics” Means for the Future of Targeted Advertising
Here's why Google is making the shift, what it portends for advertisers and publishers, and the questions marketers and publishers should be asking as they assess Topics’ long-term implications.
By Mario Diez, CEO of Peer39
Google just announced “Topics,” the digital ad giant’s latest model for replacing third-party-cookie-based ad targeting. Google had previously proposed Federated Learning of Cohorts, or FLoCs, a means of targeting groups of users, instead of individuals, based on their interests. But with FLoCs facing criticism for being opaque and privacy-unsafe, Google retreated to “Topics.”
While FLoCs did not depart enough from the behavioral targeting methods that inspired legislation like The Ban Surveillance Advertising Act, Topics represents a more drastic shift. Google would monitor users’ browsing history over a rolling three-week period, determine what Topics are of interest to them, and allow advertisers to target audiences based on a few high-interest Topics. Topics’ value proposition is essentially that of contextual targeting — matching ads to the content of consumers’ digital experiences — with a hint, or three weeks, of data collection.
Topics is a sign that the leading digital ad company recognizes the future of advertising is primarily contextual, not behavioral. Increasingly unable to track users across the internet and develop granular behavioral and demographic profiles on them, advertisers will instead have to hone messaging by serving users ads that match the content of their digital experiences.
Here’s why Google is making the shift, what it portends for advertisers and publishers, and the questions marketers and publishers should be asking as they assess Topics’ long-term implications.
Why Google is shifting to Topics
The shift from FLoCs to Topics makes sense in the context of the yearslong movement in digital advertising from unfettered tracking to privacy. FLoCs were basically cookies by another name: Google was still monitoring the user’s web behavior, much in the same way cookies track users across sites. Despite some improvements, FLoCs landed Google in hot water in Europe, where the methodology did not clear the high bar of the General Data Protection Regulation.
While Google planned to immediately allocate users to interest-based cohorts under FLoCs, Topics will allow users to opt out of tracking. In addition, the methodology will only take advantage of user data for three weeks at a time. As a result, Topics brings ad targeting on the leading browser, Chrome, into line with advertising’s broader shift away from granular, consent-free tracking.
Finally, the shift to Topics is not surprising because Topics is hardly novel. Topics looks a lot like contextual targeting. In fact, in setting up its approximately 300 Topics, Google appears to be largely copying the IAB’s content taxonomy, which shows the ad giant is relying on existing contextual infrastructure to design its new targeting model.
What the shift to context means for advertisers and publishers
Topics signals that advertisers need to start learning now what contextual signals will allow them to better identify, measure, and target ideal audiences. Advertisers can no longer depend on massive, probabilistic third-party data sets on user behavior and demographics. Instead, they need to develop the first-party data and targeting infrastructure required to leverage the power of context, perhaps complemented by limited behavioral data.
Publishers should set up the infrastructure for advertisers to reach their audiences with contextually appropriate messaging. It would not be surprising to see Topics banners, similar to current cookie banners, asking users for consent to track their interests. Opt-in rates could be low. Publishers will have to make the case for tracking to users in order to create robust first-party data sets that boost contextual targeting and drive premium supply-side value.
Questions the ad industry should ask as it tests Topics
As the advertising industry scrutinizes Topics, there are potential shortcomings that advertisers and publishers would do well to consider.
First, Google said Topics will be based on the domains and subdomains of the sites a user visits and that it will not parse article text to determine Topics. This means that, while Topics resembles contextual targeting, it falls short of the latter’s more advanced methodologies, which combine on-page semantic analysis, sentiment analysis, and other contextual signals to deliver the most relevant possible ads. Advertisers may find the precision of Topics insufficient, a gap Google and other DSPs may seek to remedy.
Another key concern for advertisers is fragmentation. Topics will be limited to Chrome. So, while Topics may go a long way toward filling holes in targeting capacities on Google’s browser, the solution will not apply to user activity on Safari, Firefox, or others. This means advertisers may struggle to measure effectiveness across browsers and develop a unified view of users.
Lastly, Google’s move raises the question of choice for both advertisers and publishers. The privacy era emphasizes consumer choice over tracking and targeting, which is commendable. But what about publishers and advertisers’ choice over the tools at their disposal? The industry should encourage Google to be transparent and cooperate as all parties aim to develop solutions that respect end user consent while shoring up ad effectiveness, which finances so much of digital media.
