Measurement leads can score an AI conversion claim before any budget moves by classifying the pathway, separating a claim from credit, and rejecting influence that is really branded search or direct. IAB released its AI-Era Attribution Blueprint on November 12, 2026, as buyers named comparing AI-driven and traditional journeys their top measurement challenge. Mis-scoring can send the next dollar to the channel that captured the visit, not the one that shaped the sale.
Updated November 12, 2026
Key Takeaways
- Classify each sale as Human-Owned, AI-Assisted, or AI-Executed.
- A claim is the right to be evaluated, not credit.
- Only 8.8% of seven-day visits come directly from AI referrals.
- Forty-four percent of ad buyers proxy AI with branded search and direct.
How do you classify the pathway behind an AI conversion claim?
The blueprint asks three questions at conversion: who or what created the intent, made the decision, and completed the purchase. Those answers classify the sale as Human-Owned, AI-Assisted, or AI-Executed, and that pathway decides who is eligible for a decision claim or a contributory claim. Mediapost reported that the blueprint distinguishes AI-influenced conversions, where AI shapes intent and a human decides, from AI-recommended conversions, where AI makes the choice. If AI does not contribute to or shape intent or the final choice, there is no new AI claim.
What should you check before scoring an AI conversion claim?
The blueprint works alongside existing attribution systems. It defines the additional credit claims AI creates and the evidence each one requires. A claim is not credit. It is only the right to be evaluated for it.
- Classify the pathway with the three questions before anyone names a claimant.
- Record no new AI claim if AI did not shape intent or the final choice.
- Match the party to a decision claim or a contributory claim, and list the evidence required.
- Treat referral links, seller records, and shopper surveys as partial proof only. They cannot show how AI shaped a decision before the click.
- Separate platform-reported claims from independently verified ones.
- Have the brand estimate each pathway’s share of conversions. Do not split one sale into credit percentages. The framework does not.
- Hold the budget move if the only support is undisclosed evidence inside an AI platform.
Which reported AI influence is really search or direct?
Caroline Giegerich, VP of AI at IAB, said AI-influenced sales are sometimes reported as search or direct traffic, “which can send the next marketing investment to the wrong channel.” IAB’s release put numbers on the habit: 44% of ad buyers rely on branded search and direct traffic as proxies for AI influence they cannot track, and 45% name comparing AI-driven and traditional journeys as their top measurement challenge, according to IAB’s 2026 Outlook Study: September Update.
Digiday reported Similarweb data cited in IAB’s report. Consumers who received a brand recommendation from an AI assistant were 2.5 times more likely to visit that brand’s website within seven days, but only 8.8% of those visits come directly from AI referrals.
| Path | Share of those seven-day visits |
|---|---|
| Directly from AI referrals | 8.8% |
| Attributed to search | 55.9% |
| Attributed to direct traffic | 19.9% |
Reject the AI conversion claim when the only observable path is branded search or direct. Measurement teams facing the same capture-versus-influence problem can apply that skepticism to on-device conversion reporting and reported ROAS.
What should you demand, and what still cannot be proven?
Before budget moves, ask platforms to disclose whether a recommendation was paid or organic, which sources informed it, and to label AI-originated traffic consistently. Ask publishers to name the contributory claim their content creates. Keep platform-reported claims apart from third-party checks. Buyers already know the visibility limit, including what ChatGPT ad proof lets buyers see.
The clearest gap is a publisher review inside an AI answer. The consumer may not click the citation, and the sale is reported as search or direct by the retailer. Digiday quoted IAB’s report: “The review shaped the sale, and the publisher is eligible for a contributory claim.” Training content “is generally not traceable from outside the model.” Grounding and citation are more observable, “although citation alone does not establish how much that content influenced the answer.” Until those signals connect to conversions, the contributory claim cannot be fully substantiated.
David Cohen, IAB’s CEO, said the blueprint “is honest about what the industry can’t yet substantiate.”
Frequently asked questions
What is an AI conversion claim under IAB’s blueprint?
It is not credit. IAB defines a claim as the right to be evaluated for credit, as a decision claim or a contributory claim, only with the evidence that pathway requires.
How does the blueprint classify a conversion?
Ask who or what created the intent, made the decision, and completed the purchase. The sale is then Human-Owned, AI-Assisted, or AI-Executed.
Why reject branded search as AI influence?
IAB’s 2026 Outlook Study found 44% of ad buyers rely on branded search and direct traffic as proxies for AI influence they cannot track. Similarweb data cited in the report says only 8.8% of visits within seven days after an AI brand recommendation come directly from AI referrals, while 55.9% are attributed to search.
What still cannot substantiate a publisher’s contributory claim?
Training content is generally not traceable from outside the model, and citation alone does not establish how much it influenced the answer. Until sourcing signals connect to conversions, the claim cannot be fully substantiated.
