Define Incrementality in Marketing

Incrementality is the additional outcome, such as sales, sign-ups or revenue, that a marketing activity causes and that would not have happened without it. It separates the conversions you merely observed from the conversions you actually created. This page defines the term, shows how it differs from attribution, and explains how marketers measure it.

A plain definition of incrementality

Incrementality asks one question: what would have happened if we had not run this campaign? The gap between the outcome with the marketing and the outcome without it is the incremental effect. If a customer would have bought anyway, that sale is not incremental, even if an ad was shown to them shortly before.

This is why incrementality is a causal idea rather than a counting idea. Counting tells you how many conversions sit next to your ads. Incrementality tells you how many of them your ads were responsible for. The difference matters most for channels that reach people who were already likely to buy, such as branded search or retargeting.

How incrementality differs from attribution

Attribution assigns credit for a conversion to the touchpoints a customer passed through, using a rule such as last click or a weighted model. It describes a path. It does not prove that any step on the path changed the customer’s decision, so it can over-credit channels that simply sit close to the point of purchase.

Incrementality tests the effect itself. It compares what happened with the marketing against what would have happened without it, which makes it a better guide to where extra budget will actually add results. The two approaches are not rivals. Attribution is fast and detailed, while incrementality is the check on whether the credit it hands out is deserved.

Ways to measure incrementality

The most direct method is a controlled experiment. A group is held out from seeing the marketing, a similar group is exposed to it, and the difference in outcomes is the incremental result. Geography-based tests work the same way, switching spend on or off in some regions and comparing them with others. The key is that the groups are comparable and that the test runs long enough to be reliable.

Experiments are not always practical for every channel or budget, which is where modelling helps. Marketing mix modelling uses historical sales and spend data to estimate what each channel contributes, and it can be calibrated against test results. Seeda’s platform includes marketing mix modelling and incrementality testing, so both views of true incremental performance can sit together.

Using incrementality to make budget decisions

Once you know which spend is incremental, budget decisions become clearer. A channel with strong reported results but low incrementality may be collecting credit for sales that were coming anyway, and moving some of that money elsewhere may cost very little. A channel with modest reported results but high incrementality may deserve more.

Incrementality also changes how you read marketing ROI. Return calculated on all attributed revenue can flatter a channel, while return calculated on incremental revenue shows what the spend really added. Treat each result as a current estimate, and retest as your mix, seasons and audiences change.

Frequently asked questions

What is the simplest way to define incrementality?

Incrementality is the extra result your marketing causes that would not have happened without it. It is the difference between running a campaign and not running it.

Is incrementality the same as attribution?

No. Attribution shares out credit across the touchpoints on a customer’s path, while incrementality measures whether the marketing changed the outcome at all. Many teams use attribution for detail and incrementality to check it.

How do you measure incrementality?

The usual approach is a controlled test with an exposed group and a held-out group, compared on the same outcome. Marketing mix modelling can estimate it across channels where testing is difficult.