Google only launched the Topics API in January and began global testing in July. Topic API (short for Application Programming Interface) replaces an earlier proposal known as Federated Learning of Cohorts or FLoC, which was found to have privacy holes. Google has been very thoughtful about how we change data sharing on the Chrome web browser and Android devices. Apple rocked the advertising industry by ditching cookies from Safari starting in 2017, and last year it hit the advertising business by starting to prevent apps from collecting too much information from consumers on iPhones.

Unlike Apple, Google is the world’s largest internet advertising company, and withholding data in a way that harms a competitor’s advertising business would require more scrutiny. Google faces a pending antitrust lawsuit from the U.S. Department of Justice, revealing more details about how the search giant controls much of its online advertising on the publishing and advertising side. promised to be. Earlier this year, Google delayed cookie deprecation in Chrome for another year, he said, until 2024, arguing it would give the industry more time to test programs like Topic and other alternative IDs.

Google’s privacy sandbox proposal has many skeptics. Jason Kint, CEO of Digital Content Next, a trade association for the publishing industry, said: “Here’s the real problem.” Kint has been an outspoken critic of tech giants such as Google and Meta, demanding just compensation from publishers.

RELATED: Google’s cookie chaos is splitting the ad industry

According to Farrow, where the Google Topics API scored higher was in the area of ​​transparency. This proposal, in contrast to his previous FLoC proposal, gives consumers more details about why they were tagged with a particular interest based on their browsing. FLoC was less revealing about what categories people were grouped into as part of their assigned cohorts. While the topics were more contextual, FLoC aimed at anonymity by diluting individual identities with a larger audience pool. Still, FLoC raised concerns that nosy participants in the advertising market could reverse-engineer her identity and obtain personal information.

In our initial tests against Xandr’s Topic API, we found strength in the way Google organizes categories that can be used to target ads. “We found the topics broadly aligned with what we know and expect from our users,” he says. Xandr studied how Google analyzes the content of his website and how it labels sites for advertising targeting.

Google uses the context of websites such as the New York Times, Vogue Business, and Elle to generate interest categories. Instead of using third-party cookies that reveal more personal information about consumers, Google Chrome shares with advertisers a small sample of interests collected over a limited period of time. Publishers and brands are working on their own identity frameworks to continue personalized advertising based on obtaining consent from consumer and first-party data relationships. There are more of his Google ad management tools for publishers to help you use these alternative methods.

Xandr’s tests also found that Facebook.com was one of the domains that exhibited Chrome’s advertising characteristics. This website contributes to interests such as news, social media, sports, arts, literature, entertainment, home and garden. There are about 350 topics. Xandr found that scale was an issue, as 12.4% of US web domains were not assigned a topic. “The percentage of domains without topics is not negligible,” Xandr said in his July blog post, outlining his testing on topics.

Xandr is not ready to go all out to adopt Google’s privacy sandbox proposal, Farrow said. Advertisers and publishers are still cautious: “Is this something that actually provides value, or is it really unnecessary?” Farrow said.