Incrementality measures the additional outcome caused by a marketing activity that would not have happened without it — the true causal effect, isolated from what would have occurred anyway.
This is the sharpest distinction in modern marketing measurement: attribution asks who gets credit; incrementality asks what actually changed because of the ad. A retargeting campaign can show a spectacular attributed ROAS by claiming credit for purchases from people who were already about to buy — incrementality testing is the method that catches this by comparing an exposed group against a genuinely comparable holdout group that saw no ads at all.
The standard method: split a comparable audience into a test group (exposed to ads) and a control group (held out), run the campaign, and measure the difference in conversion rate between the two groups. That difference is the incremental lift — the portion of results genuinely caused by the advertising, not just correlated with it.
Incrementality testing costs real, near-term revenue (the holdout group deliberately doesn't get ads that might have converted them), which is why most teams run it periodically (quarterly checks on major channels) rather than continuously on every campaign.
Incremental Lift
Incremental Lift = (Test Group Conversion Rate − Control Group Conversion Rate) ÷ Control Group Conversion RateExample
An app sends push notifications to 50,000 users (test group) and holds out 50,000 comparable users (control). The test group converts at 2.4%, the control group at 1.9%.
Incremental lift = (2.4% − 1.9%) ÷ 1.9% ≈ 26%. Roughly a quarter of the test group's conversions are genuinely attributable to the push campaign; the rest would likely have converted regardless.
Go deeper