FLEDGE and Topics: the post-cookie adtech ecosystem remains a work in progress

There are some positive noises in Google’s recent announcements, but there’s plenty of work still to be done on their cookie replacements.

As the Origin Trials prepare to start in earnest, Lukasz Wlodarczyk, VP of Programmatic Ecosystem Growth & Innovation at RTB House, digs deeper into how FLEDGE and Topics are shaping up…

Given Google’s key position in the global digital advertising ecosystem, the industry has been eagerly awaiting further news from the search giant about its plans for ad bidding, targeting and measurement in the privacy-first, post-third-party cookie era.

Late last month, Google shared more details of two new proposed post-cookie solutions for advertisers, known as FLEDGE and Topics, which have been developed in collaboration with a number of other ad tech companies (including RTB House). These new solutions are expected to be made available for testing by ad tech firms later this year.

The first of these proposed solutions, FLEDGE, stands for ‘First Locally-Executed Decision over Groups Experiment’. The basic idea behind this approach is that ad bid and targeting decisions happen at the device level. In other words, the user's browser on their device – rather than the adtech platform – stores information about what that person is interested in. This significantly reduces the volume of personal data being shared across the adtech ecosystem, while giving advertisers and publishers a non-intrusive, privacy-compliant mechanism with which to serve personalised ads. 

Google’s latest announcement provides some details of the upcoming Origin Trials of the FLEDGE API [application programming interface]. Many of the elements included in the full FLEDGE specification will not be available within the first Trials, including: Aggregated Reporting API, serving as a tool for understanding campaign outcomes from the business perspective; and k-anonymity limitation, for ensuring privacy protection. The initial Origin Trials will not be available on mobile devices. 

Positive moves

This latest announcement is positive overall, improving the industry’s understanding of the scope of tests that will be possible in the months ahead. Two of the most exciting inclusions in the specification are Product-level TURTLEDOVE and Outcome-based TURTLEDOVE, created and championed by RTB House. We’re pleased to see these contributions be made available as tools, designed to provide a measurable impact on the efficacy of advertising, and at the same time, protect the privacy of users. 

Google’s decision to make a publisher opt-in mechanism the default (with both the FLEDGE and Topics APIs) gives publishers more control over who will take part in the auction through auction configuration. It improves the transparency of the buying channels compared to the current state, which is also beneficial for publishers. This move is also in line with IAB Tech Lab proposal on buyers.json.

However, there are still several key questions that remain unanswered. For example, we are yet to find out when exactly the trials start or what percentage of Chrome traffic will be included in the Trials. It is also still unclear how browser-side resource management will work. What will be the split of computational power between the buyers? What will the browser do if there are too many interest groups for evaluation? Ensuring a level playing field for all entities regardless of their size will be a challenge for Google and is likely to be a hot topic of debate as more about the FLEDGE API becomes known.

Topics under discussion

Google has also recently offered an overview of its Topics API system, which will categorise a website’s hostname from an overall list of 350 topics to help with ad targeting. As yet, it’s unclear whether publishers will have control over the topics their sites represent or whether an in-browser algorithm will assign topics for them.

One clear privacy benefit gained from using generalised topics is the negation of third-party data based on tracking users’ cross-site behaviour and the fact that users become anonymous within groups. When privacy is considered, there’s no question that this is a move in the right direction.

In Topics API, individual users are assigned five topics within the set of 350, meaning brands should theoretically be able to reach new audiences using topics relating to users’ interests. This approach could be used in combination with contextual targeting, providing a greater level of precision over when to serve relevant ads. 

However, concerns have already been voiced that the Topics API, as proposed, appears to have a skewed focus towards larger entities, rather than SMEs. The more websites a company appears on, the more likely it is to get topics to target, disproportionately favouring large social networks, multi-category big publishers, supply-side platforms and ad networks.

A further problem identified is that larger websites containing countless sub-topics, such as any major news publication, will only match topics linked to the overall website domain. 

The global adtech system remains a work-in-progress; however, the latest announcements from Google and the general direction of the ad tech industry should be seen as positive developments. This is just the beginning of what will be a constructive discussion on how we, as an industry, can successfully adapt to maintain user privacy, advertising efficacy and true competitiveness in the advertising market field.