Google will use AI to optimize how often users see ads

Google explains that this new tool, which it plans to bring to display offerings in Google Ads in the future, uses traffic patterns where a third-party cookie is available and analyzes them at an aggregated level across Google Ad Manager publishers to create predictive traffic models. This allows the ad platform to estimate how likely it is for users to visit different publishers the same ads through Google Ad Manager, and  to optimize how often those ads should be shown to users whose PCs lack a third-party cookie present.

Google’s already using machine learning in Google Ads, albeit mostly to generate ad suggestions, better match users’ searches, and adjust video spot bids. But Google Ads Privacy product manager Rahul Srinivasan asserts this new AI-driven frequency management approach is more “privacy safe” than workarounds like fingerprinting, which rely on user-level signals like IP address. If all goes according to plan, it should result in fewer instances of users repeatedly encountering the same ads as a result of blocked or restricted cookies, he says.

“Since we aggregate all user data before applying our machine learning models, no user-level information is shared across websites. Instead, this feature relies on a publisher’s first-party data to inform the ad experience for its own site visitor,” said Google Ads Privacy product manager Rahul Srinivasan. “This is a step in the right direction as we work across Google to raise the bar for how our products deliver better user experiences while also respecting user privacy.”

Today’s announcement, which coincided with Google’s discussions this week with advertising and publishing partners in Europe at a series of events in London, comes after the company announced it would introduce protections in Chrome to protect users from cross-site cookies and fingerprinting.