Google gives post-cookie ad tech update

Internet giant shows how ads could perform through a combination of new ad targeting and privacy measures.

Google gives post-cookie ad tech update

Google released new data about what the post-cookie future may look like in advertising, revealing a small dip in performance. Although Google said it was still encouraged by the results, the tech giant outlined some potential drawbacks in its research paper.

Google still intends to remove third-party cookies from Chrome at the end of 2024, ditching the online tracking tools in a bid to catch up to privacy measures pushed by other platforms, such as Apple. Google released an initial study today showing how ads perform within its ecosystem when they don’t use cookies for ad targeting. Instead of cookies, Google has been developing what it calls the Privacy Sandbox, which is a set of tools that tap into more anonymous, aggregated internet data to target the ads. In this early limited test, Google claimed that the ads performed almost on par with ads that used cookies.

Also read: Apple's privacy policies under fire from ad tech industry

Google measured the amount of advertising spend that went through the privacy-enhanced pathway, using advertising spend as a “proxy” to understand the potential scale. Spend was down 2% to 7%, while ad effectiveness dropped between 1% and 3% as measured through a metric called “conversions per dollar,” or how much money the advertiser spent to drive an action by the consumer, said Dan Taylor, Google's VP of global ads. The ads maintained a relevance score based on how often consumers clicked on the ads within 90% of cookie-powered ads, Taylor said.

The push to eliminate cookies comes as “80% of people are now concerned about the state of their online privacy,” Taylor said in a conference call, “and almost half are turning away from services due to privacy concerns. That’s scary.”

Google’s tests are being watched closely by the ad industry, but the trials have been slow going. Advertisers have already moved off cookies in other ecosystems, such as within Apple. Meanwhile, ad tech companies such as The Trade Desk, LiveRamp and Neustar, among others, are developing advertising IDs that replicate the utility of cookies. Supply-side ad platforms that work with publishers to sell internet ads are coming up with their own post-cookie products.

Google needs to adopt new privacy measures or risk consumers migrating to devices and browsers that prioritize data security. Apple has made privacy a centerpiece of its marketing, even as it builds its own ads business. Google already has a powerful position in internet ad markets and any changes it makes to data-sharing must take into account its many competitors in ad tech. Google is being investigated in the U.S. and Europe over its competitive practices, and regulators are watching if its privacy changes unfairly affect rivals.

Also read: Google asks judge to dismiss DOJ's digital ad antitrust case

Google tweaked another portion of Privacy Sandbox this week, renaming FLEDGE to “Protected Audience API.” FLEDGE was an unwieldy acronym that stood for “first locally executed decision over groups experiment.” Google said the name change reflected the fact that FLEDGE is entering the next stage of development and will be more widely available for testing. FLEDGE is a way to run ad auctions without revealing personal data about consumers.

In the recent Privacy Sandbox test, Google tried the Topics API, which is a core component of its post-cookie system. Instead of tracking a person’s entire web history like cookies do, the Topics API collects a handful of contextual interests about the consumer, and those interests go into ad targeting. The Topics API, however, cannot work on its own to replace cookies, Taylor said. “Topics API is not being designed as a standalone offering or ad tech in a box,” Taylor said.

For that reason, the tests combined Topics API data with other contextual advertising tools and with signals from publishers, known as publisher provided IDs.

Google also used AI to see if that would improve ad efficiency. AI is being used more often in post-cookie ad tech because machine learning can improve targeting and bidding in the ad auctions. “Especially without a cross-site identifier like [third-party cookies], Google’s AI can help advertisers derive more value out of otherwise fragmented privacy-preserving signals,” Google said in its research.

Google’s Privacy Sandbox has faced some skepticism, and there are some who think Google will have a hard time meeting its deadline to eliminate cookies next year. For its part, Google said these early tests are ongoing, and that they should not be taken as the final word on how ads will perform once cookies do go away.

“The results, while they’re encouraging, shouldn’t be considered an unequivocal indicator of Google’s performance after the third-party cookie deprecation,” Taylor said.

Google released the full results of the tests in a white paper on GitHub on Tuesday.