Mountain View, California-based giant Google is working on a new machine learning algorithm that aims to replace tracking cookies with privacy-focused alternatives. Google is calling the system Federated Learning of Cohorts (FLoC), and will allow businesses to send ads to groups of potential customers rather than specific individuals. The system uses on-device machine learning algorithms to create clusters of people with similar browsing habits. All the data analysed by the algorithms is kept private on the browser and not uploaded anywhere else.
Google says that this approach will effectively let individual users hide in the crowd. On Monday, January 25, Google released new data from simulations of the technique conducted by Google's ads team. The tech giant said that the results were almost as effective as cookie-based approaches. "Our tests of FLoC to reach in-market and affinity Google Audiences show that advertisers can expect to see at least 95 percent of the conversions per dollar spent when compared to cookie-based advertising," the company said in a blog post.
Google plans to make its FLoC available for public trials through a March update to Chrome. it will also test the cohorts with advertisers in the second quarter of this year. In October 2020, Google released early results from its FLoC tests indicated that interest-based cohorts generate big improvement in recall and precision over random clustering.
While the technology still remains under development, the tests suggest that FLoC could provide more privacy than many tracking cookies.