IAB drafts AI advertising measurement framework as agents replace human clicks

Ads are starting to reach machines instead of people. The advertising industry is still figuring out how to count what happens next.

The Interactive Advertising Bureau plans to publish a framework Nov. 12 for measuring and crediting conversions influenced by artificial intelligence, according to Digiday’s reporting on the trade group’s work.

Caroline Giegerich, the IAB’s vice president of AI, is drafting the framework with a working group of tech companies, publishers, agencies, measurement vendors and brands. She declined to identify the participants.

The goal, Giegerich said, is to “create a shared framework for measuring and crediting AI’s role in conversions, especially when traditional signals are minimized.”

Those signals include UTM tracking parameters and referral data. But an AI-mediated purchase can make them less useful. An agent might read a product page, compare options and help a user buy something without producing the trail that follows a direct click.

That leaves a basic question: What counts as evidence that AI influenced a sale?

Giegerich said the framework will likely separate AI’s role into two stages. One covers situations where an AI system surfaces something to a user, creating awareness or intent. The other covers cases where AI helps make a decision.

Assigning credit inside those stages has been harder.

When asked what was hardest for the working group to agree on, Giegerich said “everything.”

Publishers have a particular stake. Giegerich told Digiday they argue their content may have informed an AI response and that they should not be excluded when credit is assigned.

The project follows IAB work this month on measuring visibility in the AI era. Separately, the trade group has pursued a proposal for federal legislation addressing uncompensated AI scraping. That policy effort is distinct from the attribution framework Digiday described.

Jaime Schultheis, head of global data partnerships at Bombora, a business-to-business data company, said the effort could help create a more reciprocal relationship between publishers and technology companies. She told Digiday that, over time, large technology companies have built audiences with help from publisher content while offering publishers little in return.

Michael Bishop, co-founder of AI-native ad platform OpenAds, raised another problem. Even with an industry standard, AI platforms may still provide limited visibility into how they influenced a purchase. If platforms control those signals, outside measurement firms may have to rely on platform-provided data.

Bishop compared that possibility to Facebook’s relationship with third-party measurement vendors, where the platform retained control over the underlying systems while outside firms acted more like auditors.

Giegerich acknowledged a related problem. Some of the evidence the industry would want for attribution, she said, does not currently exist. The IAB wants the framework to identify what is missing and who might provide it.

The issue reaches beyond advertising. Apple is reportedly considering usage-based payments to publishers whose reporting is used by its revamped Siri, an arrangement that raises its own questions about how individual uses are counted and verified.

Publishers are also competing with brands for AI referral traffic. Recent data cited by Media Copilot found a growing share of ChatGPT referral visits going to brand homepages.

The IAB framework will not settle all of those questions. But it is an attempt to create a common way to talk about one of them: when AI sits between a publisher, a customer and a sale, who gets credit — and what evidence proves it?