Media measurement is at a turning point. Publishers, advertisers, and anyone who depends on solid data to make money are facing tough choices. At the CIMM Summit in New York City, Ben Hovaness from Omnicom Media gave a clear warning. He said artificial intelligence could either lock in the industry’s worst habits or finally bring real transparency and quality. But that will only happen if the industry steps up and makes real changes.
In Google Consent Mode V2, conversion modeling only activates after a threshold of 700 clicks within 7 days per country and domain group is reached, highlighting the dependence of modeling accuracy on sufficient traffic volume.
Hovaness warned that AI could make this confusion even worse. He described a future where vendors’ performance numbers go straight into buyers’ AI agents-tools that are already catching on and could be everywhere by 2031. These agents might shift budgets automatically to vendors who report the highest conversion rates, even if those numbers are pumped up by modeling and not real results. This kind of feedback loop, he said, could keep bad data in play and push both buyers and sellers to spend more on ads that don’t actually work.
IAB Europe has opened public consultation on new proposals for measuring in-store and travel media, emphasizing that in the absence of sensor technology, a more transparent approach to modeled assumptions is required.
But the summit wasn’t all bad news. Hovaness also laid out a better path, where AI helps make high-quality measurement faster and more accessible. AI could cut down the time and effort needed to run solid studies, check results, and update models. This would make strong measurement normal and affordable. Publishers and advertisers could keep up with fast-changing markets and adjust their strategies using data that actually matches what’s happening now.
This shift wouldn’t just make measurement easier and cheaper. It could also help rebuild trust in the numbers that drive billions in ad spending. Hovaness said the industry has a clear choice: keep hiding behind unclear data and risk a crisis of trust, or invest in transparency and start a new era of accountability. This urgent decision matches what’s been seen in earlier coverage about editorial power shifts and AI’s effect on marketing trust.
Recent moves in the industry show just how tangled measurement has become. AGF’s video measurement system now models data from over 1.7 million devices with direct measurement, and the goal is to reach about 3.4 million devices, according to an AGF system upgrade announcement. At the same time, Google Ads and GA4 use different ways to model conversions, which leads to mismatched numbers-especially when users say no to cookies and direct tracking isn’t possible, as explained in industry analysis.
For publishers and media operators, the message is clear. The tools are changing, but so are the risks. Those who focus on clarity and make sure to separate real results from modeled ones will be in the best spot as AI changes how measurement works. Letting synthetic data go unchecked could put not just single campaigns, but the trust in all digital advertising, at risk.
