The Incrementality Gap Nobody Measures

Understanding the Gap

Attribution models assume that advertising caused conversions. They see a click, then a purchase, and draw a causal arrow between them. But correlation is not causation.

A customer already planning to buy might search for reviews, click an ad in the results, and complete their purchase. Attribution credits the ad. But the purchase was happening regardless. The ad was a detour, not a driver.

Incrementality measures what happens when you remove advertising entirely. The sales that disappear are truly incremental. Everything else was cannibalisation, coincidence, or halo effect.

Where the Gap Hides

Different campaign types have different incrementality profiles. Understanding where the gap is largest helps prioritise measurement efforts.

The Compound Problem

Because low-incrementality campaigns often show the best ROAS, budget naturally flows toward them. Branded search looks incredible. Retargeting looks efficient. The campaigns with genuine impact often look worse on paper.

Without incrementality measurement, optimisation actually works backward. You invest more in campaigns claiming credit for sales that would happen anyway, while starving the campaigns genuinely creating demand.

This is how sophisticated brands end up with efficient-looking accounts that are not actually driving growth. The numbers look right. The bank balance tells a different story.

Measuring the Gap

There are several approaches to incrementality measurement, each with tradeoffs:

The Finance Conversation

CFOs increasingly question marketing efficiency. When asked to justify Google Ads spend, attribution data is not convincing. Finance understands that the ad platform claiming credit is not an independent source.

Incrementality testing provides the evidence finance needs. It shows which spending genuinely drives growth and which is expensive rent-seeking. This transforms marketing from a cost centre that claims credit into a profit driver that proves value.

Closing the Gap

Once you measure incrementality, you can optimise for it. This often means counterintuitive moves: reducing branded search, shortening retargeting windows, investing more in prospecting despite lower apparent ROAS.

The goal is not to eliminate the incrementality gap entirely. Some level of cannibalisation is inevitable. The goal is to understand it, account for it, and make budget decisions based on true value rather than attributed value.