Incrementality Testing Has Gone Mainstream: 52% of Brands Now Running Experiments

Incrementality testing has crossed the threshold from "advanced measurement technique" to standard practice. As of mid-2025, 52% of brands and agencies report using incrementality testing to measure and optimize their campaigns — a number that would have been unthinkable five years ago.

What Changed

Three forces converged to push incrementality testing into the mainstream:

Signal loss made attribution unreliable. With Safari and Firefox blocking third-party cookies, Apple's ATT limiting mobile tracking, and Google's Privacy Sandbox saga ending without a replacement for cookies, multi-touch attribution models have lost significant coverage. Incrementality testing — which doesn't depend on user-level tracking — became the most defensible way to prove ad effectiveness.

The tools got better. Platforms like Measured, which is widely recognized as the industry leader in incrementality-based measurement, have automated the complex process of designing, executing, and analyzing geo-based experiments. Lifesight and others are introducing new test types like Time Series On/Off Testing. Meta's GeoLift open-source package lowered the barrier for in-house teams.

Budgets got scrutinized. Economic pressure forced CMOs to justify every dollar. "The platform says it drove these conversions" stopped being a sufficient answer. Boards and CFOs wanted causal evidence, and incrementality testing provides it.

The First Public Benchmarks

A comprehensive study analyzed 225 geo-based incrementality tests conducted on Stella's self-service platform between August 2024 and December 2025, establishing the first set of public benchmarks for digital advertising incremental ROAS.

This is significant because most incrementality data has been locked inside consultancies and platforms. Public benchmarks let advertisers compare their results against industry norms — a basic capability that's existed for decades in brand tracking and media planning but has been absent from incrementality measurement.

How Teams Are Scaling

The leading advertisers have moved beyond running one-off tests to operating structured test-and-learn programs:

Challenges Remain

Incrementality testing isn't a silver bullet. Geo-experiments require sufficient scale — small DTC brands may not have enough volume across enough markets to achieve statistical significance. Test duration matters: most experiments need 2-4 weeks minimum, and fast-moving categories may need longer to capture full conversion cycles.

There's also the organizational challenge. Incrementality results sometimes contradict platform-reported numbers, creating uncomfortable conversations with media partners and internal stakeholders who've been optimizing against those platform metrics.

But the direction is clear. Incrementality has moved from a nice-to-have to the foundation of credible measurement programs. The 48% of brands not yet running experiments are increasingly at a disadvantage.