Definition

What is advertising incrementality?

Advertising incrementality is the share of conversions that would not have happened without advertising. It is measured by comparing a group exposed to the ads against a comparable unexposed group, rather than by crediting observed touchpoints. It answers the executive's question: if I cut this budget, what do I actually lose in sales?

Why attribution is not enough

An attribution model distributes credit across the interactions it observed. It does not say whether the sale would have happened without the advertising. The clearest case is a search for your name: a customer who already knows you clicks on your brand ad, then buys. The report counts an advertising conversion even though that customer would probably have found you without the click.

Google makes the distinction itself: it describes its incrementality measurement tool as intentionally ignoring standard attribution rules.

Three ways to measure it

The controlled pause. You stop a campaign for a defined period and compare sales. The result remains indicative.
The geographic test. Comparable regions see the ads, others do not. It works with sales closed offline.
The randomized trial. The platform randomly splits exposure between users. It requires a high volume of conversions.

The conditions and pitfalls of each method are detailed in our analysis for executives.

How to read a result

An incremental result always comes with a margin of error. A lift (the difference in sales between the exposed group and the control group) presented without an uncertainty interval is not a measurement. A test also describes a situation, not a law: the incrementality of a brand campaign in January does not predict that of a product campaign in May.

Formula

Cost per incremental conversion = ad spend ÷ (conversions in the exposed group minus conversions in the control group, at comparable size)

Incremental conversions are the difference between the two groups. The same difference, expressed in revenue, gives the incremental return on ad spend.

Example

In the exposed regions, a campaign costs $9,000 and the report credits it with 100 conversions, or $90 per conversion. The control regions, comparable and without ads, produce 62 at equal population. Only 38 conversions are therefore incremental: the cost per incremental conversion is about $237 (fictional example).

How we read it

We treat incrementality as the question that decides a budget. A budget allocated on attributed conversions alone can overinvest in campaigns that capture existing demand. A weak result is therefore not a failure: that budget can go to campaigns that create demand.

A useful test starts with a decision named in advance: what you will cut or increase depending on the result. Below a certain volume of conversions per region, a business gains more by first measuring its cost per sale.

Not to be confused with

Data-driven attribution: Attribution distributes credit across observed touchpoints, with no comparison group. Incrementality compares an exposed group with an unexposed one to isolate the causal effect.
Lift: Lift is the difference in sales measured between the exposed group and the control group. It expresses incrementality and is only worth something with its margin of error.
ROAS: ROAS divides attributed revenue by ad spend. It includes sales that would have come without advertising; its incremental version keeps only those the advertising caused.

Related concepts

Related services

Frequently asked questions

Can incrementality be measured without cutting budget?

Not really: any causal measurement requires a group that does not see the advertising. The scale and duration are negotiable. One region in six for eight weeks costs less than a general shutdown.

Can incrementality be measured with sales closed offline?

Yes, with the geographic test. It compares regions rather than individuals, which makes it compatible with sales closed by quote, through a sales rep or in store.

← The full glossary