Incrementality Testing.

Incrementality testing is a causal measurement methodology that determines how much of a marketing outcome — conversions, revenue, awareness — would not have occurred without a specific marketing activity. It uses controlled experiments (A/B holdout tests, geo holdout tests) to compare a group exposed to the marketing activity against a control group unexposed to it. The difference is the incremental impact attributable to the activity.

Why Incrementality Matters

Attribution models — last-click, first-click, data-driven — allocate credit across touchpoints but cannot prove causality. They reveal correlation, not incrementality. A retargeting campaign may appear to drive many conversions because it targets users who were already highly intent; but if those users would have converted anyway, the campaign's true incremental lift is near zero. Incrementality tests answer the causal question: would this conversion have happened without us spending money?

Incrementality Testing for SEO

SEO is notoriously difficult to measure incrementally because there is no obvious 'control' — you can't easily withhold SEO from a group of users. Geo holdout tests (comparing markets where SEO investment was paused against markets where it continued) are the closest approximation. Some enterprise SEO platforms use SEO split testing — enabling or disabling optimisations on statistically matched page groups — as a form of controlled incrementality measurement for on-page changes.