Skip to content
Thursday 19 November 2026 marketing · daily

Martech · AI attribution

IAB’s AI attribution blueprint: 45% struggle to measure

IAB's new framework classifies conversions as human-owned, AI-assisted or AI-executed, as 45% of ad buyers struggle to compare AI and traditional journeys.

IAB's AI attribution blueprint: 45% can't measure AI's role
In this story
  1. Why marketers need this now
  2. Three conversion pathways
  3. Claims are not credit
  4. What to do now

The Interactive Advertising Bureau has unveiled a measurement framework for a problem every performance marketer is beginning to feel: when an AI assistant sparks a purchase, there may be no click to credit.

Called The AI-Era Attribution Blueprint, the framework gives advertisers, agencies, publishers and AI platforms a shared way to determine who can claim influence when a sale is shaped or completed by AI. It does not replace existing multi-touch or last-click systems; it sits alongside them to classify the extra claims AI creates.

Why marketers need this now

AI is compressing purchase journeys. When a shopper asks an assistant for a recommendation and buys through it, the familiar click, referral or session trail may simply not exist. That absence gets filled with proxies.

According to IAB’s 2026 Outlook Study September update, 45% of ad buyers say comparing AI-driven journeys with traditional ones is their top measurement challenge, while 44% lean on branded search and direct traffic as a stand-in for AI influence they cannot directly observe.

Caroline Giegerich, IAB’s vice president of AI, put the risk bluntly: “the clicks and referrals that attribution depends on are thinning out or disappearing.” The consequence is misattribution, and budget moving toward the wrong channel.

Three conversion pathways

The blueprint classifies every conversion into one of three buckets:

  • Human-owned: a person created the intent, made the decision and completed the purchase.
  • AI-assisted: AI helped shape or inform the journey, but a person still made the decision.
  • AI-executed: AI completed the transaction, with or without human decision-making.

For each conversion, the framework asks who created the intent, who made the decision and who completed the purchase. Those answers determine which parties are eligible to make a claim and what evidence they need.

Claims are not credit

The framework separates being eligible for influence from actually receiving attribution. A decision claim is for the party that influenced or made the purchase decision. A contributory claim is for parties that shaped the outcome without making the final call.

The unresolved problem is evidence. Much of the data that would prove AI’s role is either not collected or sits inside AI platforms. So the blueprint hands responsibilities across the ecosystem: brands estimate how many conversions fall into each path, AI platforms label paid versus organic recommendations and disclose sources, agencies apply a common claim-and-evidence standard, and publishers identify where their content contributed.

What to do now

Marketers should not wait for a perfect data feed. A practical first step is to audit current conversion data and estimate the share that is human-owned, AI-assisted and AI-executed, then compare that with what last-click reporting assigns to search and direct. That gap is the size of the attribution risk.

The framework also asks agencies to separate claims reported by platforms from claims independently verified, a discipline worth adopting immediately.

IAB plans to bring brands, agencies, AI platforms, retailers and publishers together to share findings and refine the measurement as more data surfaces. For now, the blueprint offers a common language, not a complete solution.

Written by

Marketing Junkies Desk

Marketing Junkies covers agency moves, campaigns, martech and adtech launches with an Indian and global lens. Every story is reported in our own words.