Incrementality

Incrementality measures the actual lift in sales, signups, or other outcomes that a marketing investment causes — versus what would have happened without it. It is the answer to the question "if I turned this channel off tomorrow, how much revenue would I actually lose?" For e-commerce brands that spend heavily on paid acquisition, incrementality testing is the most accurate way to separate ads that drive new revenue from ads that simply take credit for sales that were going to happen anyway.

The need for incrementality grew alongside the limitations of last-click and multi-touch attribution. iOS privacy changes, cookie deprecation, and the rise of branded search ads created a measurement environment where conventional attribution platforms over-report on paid channels — particularly bottom-funnel and branded search — because they capture conversions that would have occurred organically. A brand running incrementality tests often discovers that 20-40% of "paid" conversions are not actually incremental.

The standard incrementality test methodology is geo-holdout testing: temporarily turning off advertising in a representative subset of geographies and comparing conversion rates to matched control geos that still receive advertising. The difference reveals true causal lift. Other approaches include ghost ad testing (showing PSA ads to a control group), conversion lift studies in Meta and Google's own platforms, and synthetic control modeling.

Incrementality is most valuable when used to make budget reallocation decisions. If brand search has 95% conversion attribution but only 30% incremental lift, much of that spend is paying for traffic that would have arrived organically — and the freed budget can shift to a top-funnel channel with 60% incremental lift even if its attribution number looks worse. The brands that win at media efficiency in 2026 are the ones who allocate by incrementality, not by last-click attribution.