Incrementality
Incrementality is the portion of revenue that would not have occurred without a specific marketing action. It measures causation, not correlation. Unlike attribution models that assign credit based on user journeys, incrementality asks a counterfactual question: what would revenue have been if this channel did not exist?
The Short Version
Incrementality separates the revenue you *caused* from the revenue you just *claimed*.
Visual Explanation
What Is Incremental ROAS?
Your ROAS might look great. That doesn’t mean it worked.
The Attribution Illusion
Modern ad platforms are designed to claim credit. If a user sees an ad and buys, the platform claims the sale.
But many of those users were going to buy anyway. Without measuring incrementality, you pay for sales that were already happening (subsidizing organic demand) instead of fueling new growth.
How it works
Establish a baseline of 'organic' performance
Introduce marketing spend as a variable
Measure the lift above the baseline specifically caused by that spend
Common Misconceptions
Confusing 'Lift' with 'Incrementality' (Lift can be temporary; Incrementality is structural)
Assuming 100% of attributed sales are incremental (They almost never are)
Ignoring Baseline Revenue (You must subtract what would have happened anyway)
Frequently Asked Questions
QHow do you measure incrementality without turning off ads?
We use Ridge Regression and counterfactual modeling on historical data to estimate impact without needing destructive 'blackout' tests.
QIs incrementality just for large brands?
No, but it requires enough signal (spend/revenue variation) to measure. Usually, brands spending $50k+/month see the clearest results.