Incrementality expresses how much of the result was actually generated by the campaign and how much would have happened anyway, without it. It is the most important and most ignored question in marketing measurement. Advertising systems credit themselves with the conversions they managed to touch, not the ones they caused – a customer who has already decided and is searching for your name clicks on the ad, and the system counts the sale, even though they would have bought anyway. The gap between reported and actual return tends to be very large for brand campaigns and remarketing. It can only be measured through an experiment: part of the audience or territory does not see the advert, and the result of both groups is compared. Without such a test, the effect of a channel can only be discussed in estimates, not in figures.
See also: Geo experiment, Media mix modelling, Statistical significance of a test.
