What is incrementality?

The sales that happened because of a campaign and wouldn’t have happened without it.

Incrementality asks the question that matters most for a budget: what would have happened without this? The sales, tickets or sign-ups you’d have lost if the campaign hadn’t run are incremental. The ones that would have come anyway aren’t, even if an ad was clicked first.

You can’t see incrementality in a single report, because every report only shows what happened with the ads on. To see it, you need a comparison: a group that saw the campaign and a similar group that didn’t. For most brand teams the cleanest version is a holdout test by region. It takes a few weeks and some patience, and it’s the only number in this glossary that speaks to cause.

Why it matters

If people searching your name would have bought anyway, those sales aren’t incremental, even if an ad was clicked first.

An example

Example (made-up numbers): a campaign runs in 10 regions and is held back in 10 similar ones for four weeks. The regions with ads sell 1,100 cases a week, and the regions without sell 1,000. The campaign’s incremental effect is about 100 cases a week, even if the platform’s report credits it with 600.

Where it misleads

Assuming attributed results are incremental. People who search your name, open your email or use a code they found online were often going to buy anyway.

How Arcus handles it

When you need proof of cause, we design and read a regions-on, regions-off test, using your own numbers as the scoreboard. If a test is too small or too short to call, we tell you before it starts.

More in our guide: Incrementality, explained without the math