Incrementality

Incrementality is the extra revenue, orders or signups that exist only because you ran a campaign. You find it by comparing a group that saw the campaign with a comparable group that did not, then treating the gap between them as the true effect. Everything the untouched group bought is business you would have kept without spending anything.

It is not the revenue a platform reports back to you. Ad platforms and email tools credit themselves for orders from people who were already going to buy, which is why reported ROAS and incremental ROAS can differ sharply. Attribution answers which touchpoint gets the credit; incrementality answers whether the sale needed the touchpoint at all.

Split your audience, withhold the campaign from one slice, then compare revenue per person in each group. Incremental revenue equals (revenue per exposed person minus revenue per held-out person) multiplied by the number exposed. Divide that by spend for incremental ROAS. The honest denominator is total campaign spend, including creative and discount cost. The flattering version divides only by media spend, or compares total exposed revenue to holdout revenue without normalising group sizes, which inflates the result whenever the exposed group is larger.

ROAS counts every sale the platform claims; incrementality counts only the ones that would not have happened. A Holdout Group is the mechanism you use to measure it. Statistical Significance tells you whether the gap you saw is real or noise. MER looks at all revenue against all spend without separating caused from captured, so it moves slowly but lies less.