The causal lift in conversions attributable to advertising, measured by comparing an exposed group against a matched holdout that did not see the ads.
Incrementality is the share of conversions that an ad campaign actually caused, as opposed to conversions that would have happened anyway. Haus describes incrementality testing as a controlled experiment that randomly assigns users to a treatment group that sees ads and a control or holdout group that does not; the difference in conversions is the incremental lift caused by the ads (haus.io).
This is the antidote to platform-reported attribution. A platform claims credit whenever its pixel touched a buyer, but many of those buyers would have converted regardless. The holdout reveals the true causal contribution. Eight X, citing Haus public case studies (Bombas, True Classic, Liquid Death), reports that platform-reported ROAS overstates measured incremental ROAS by roughly 1.5x to 3x (eightx.co).
Incrementality (a randomised geo or audience holdout), marketing mix modeling (aggregate statistical estimation), and multi-touch attribution (user-level credit assignment) form a triangulation. MTA and platform numbers describe correlation; incrementality and MMM estimate causation.
Blufire treats incrementality as the causal counterweight to the golden-rule problem: channel and CAC figures from Shopify last-touch or platform pixels are framed as reported, and incrementality is how an operator establishes what those channels are really worth before reallocating a dollar.