Key stat: 28% of US buy-side decision-makers plan to run 11 or more incrementality tests a year with AI-driven improvements, more than triple the 8% who do so today, according to a February report from IAB and BWG Strategy.

Beyond the chart:

• Regular testers have good reason to want more tests. Among US senior decision-makers, 60% trust incrementality testing more than any other measurement method, and 78% believe at least 10% of their marketing spend goes to waste due to insufficient measurement, according to Haus.
• But AI can't fix messy inputs, and more tests pay off only if the data feeding them holds up. Among US enterprise marketers, 55% say poor data quality is the biggest barrier to acting on AI-generated measurement insights, according to Winterberry Group.

Use this chart: Drop this in measurement planning meetings when you're pitching budget for AI testing tools. Show it to leadership when a few tests a year starts to look like enough. Benchmark your annual test count against peers.

Related EMARKETER reports:

• Marketing Data Trends 2026 (EMARKETER subscription required)
• Media Mix Modeling Trends 2026 (EMARKETER subscription required)