As of yesterday, Google unveiled a new targeting proposal for advertisers that is not reliant on third-party cookies. Known as ‘Topics’, this new targeting proposal will replace FLoC (Federated Learning of Cohorts).
How Does FLoC Work?
Federated Learning of Cohorts worked by effectively segmenting audiences into groups based on their browsing activity. For example, PC fans might be placed in one specific cohort, while avid gym goers might be put in another, enabling advertisers to target these users based on their interests as opposed to individually tracking them.
How Does Topics Work?
Using your browsing history, Google will assign users three topics that advertisers can use to target. These topics could range from music and travel to fitness and cooking.
According to Google, topics will be “curated to exclude sensitive categories” such as gender and race.
Google also says that is has about 300 topics for testing, which is expected to grow into the thousand figures. Once every three weeks, older topics are automatically deleted.
The tech giant also speaks about how it’s working on allowing users to turn off the feature entirely as well as remove specific topics.
What Makes Topics Different to FLoC?
In a nutshell, Topics is just a simplified version of FLoC with certain measurements put in place to prevent something known as ‘fingerprinting’.
Fingerprinting is the process whereby advertisers attempt to target individuals despite the efforts put in place to anonymize users.
With Topics, sites will only be able to see a user’s three topics, triggering the mechanisms to serve them the most relevant ads based on that that topic.
When Will Google Get Rid of 3rd Party Cookies?
Google says that it will get rid of third-party cookes in Chrome by 2023. Although the proposal of Topics is still in the very early stages of development, this trial period will give advertisers a sense of how Topics campares to third-party cookies.
As said by Ben Galbraith, senior product director at Google:
“With Topics, we’re looking for a sweet spot that is a meaningful improvement for user privacy but also provides a useful signal. We’re looking forward to feedback from the ecosystem and from the trials to assess other people’s views and how effective this ultimately is.”
