The short definition
Incrementality is the additional outcome caused by a marketing activity, over and above what would have happened anyway.
Attribution asks which touchpoint to credit. Incrementality asks whether the spend changed anything at all. They are different questions and the second is the one that decides whether money is well spent.
The classic example
A business runs ads on its own brand name. The reports look excellent: high click-through, low cost per click, strong conversion rate, superb return on ad spend.
But many of those people were searching the business by name. They had already decided. If the ad had not been there, most would have clicked the organic result immediately below it — free.
The attributed return is real. The incremental return may be close to zero.
Where the question matters most
Branded search advertising. As above. There are legitimate reasons to defend your name, particularly if competitors bid on it, but that is a defensive decision rather than a growth one.
Remarketing. Following people who already visited your site. Some were returning regardless.
Broad awareness spending. Genuinely hard to evaluate, in both directions — often written off unfairly, sometimes indulged unfairly.
How to test it honestly
The rigorous method is a holdout: withhold the activity from part of the audience or the market and compare.
For a small local business, a practical version:
Pause and watch. Turn a campaign off for two to four weeks and track total enquiries, not attributed ones. If overall volume barely moves, the channel was largely capturing demand that existed anyway.
This is uncomfortable, and it is the only way to find out. Choose a period that is not your peak season, and change nothing else at the same time.
The caveats
Search and content have long lag times, so a short pause understates their loss. Seasonality can swamp the signal entirely. And there is a real risk in pausing something that turns out to have been working.
Which is why this is a test to run deliberately, on one channel, in a quiet period — not a routine practice.
Why it is worth knowing about
Because a great deal of small business marketing spend is capturing demand rather than creating it, while being reported as though it created it.
Knowing which is which changes where the next dollar goes.
Applying it
If one channel reports results that look too good, it is usually the one worth testing.
Frequently asked questions
How is incrementality different from attribution?
Attribution asks which touchpoint to credit for a sale. Incrementality asks whether the spend caused the sale at all, or whether it would have happened anyway.
Why are branded search ads the classic example?
They report excellent numbers, but most of those people were searching your business name and had already decided. Without the ad, most would have clicked the free organic result just below.
How can a small business test it?
Pause one channel for two to four weeks in a quiet period and watch total enquiries rather than attributed ones. If overall volume barely moves, that channel was capturing existing demand.
What are the risks of testing?
Search and content have long lag times, so a short pause understates their loss, and seasonality can swamp the signal. Test one channel deliberately, off-peak, changing nothing else.