Incrementality
What This Page Answers
Incrementality measures whether advertising caused outcomes that would not have happened otherwise. It is the difference between attributed performance and causal impact.
Attribution vs Incrementality
Attribution asks:
Which touchpoint gets credit?
Incrementality asks:
What would have happened without the ad?
A campaign can have strong attributed ROAS and weak incrementality if it mostly reaches users who would have converted anyway.
Why It Matters
Incrementality helps answer:
-
Is this channel creating new demand or just claiming existing demand?
-
Is retargeting actually increasing conversions?
-
Does paid social drive more brand Search?
-
Did Performance Max add new sales or cannibalize Search/Shopping?
-
Should we increase or reduce budget?
Common Test Methods
| Method | Use Case |
| Holdout test | Withhold ads from a control group |
| Geo experiment | Compare treated and untreated regions |
| Conversion lift test | Platform-managed lift measurement |
| Matched market test | Compare similar markets over time |
| Budget cut test | Observe effect of reducing spend carefully |
When To Prioritize Incrementality
Use incrementality thinking when:
-
Budget is large.
-
Retargeting looks too good.
-
Platform reports disagree with MER.
-
A channel claims many view-through conversions.
-
PMax or broad automation may cannibalize existing demand.
-
You are making a major budget reallocation.
Practical Constraints
Incrementality tests need discipline:
-
Clear hypothesis
-
Stable test period
-
Sufficient volume
-
Clean geography or audience split
-
No major promotions during the test if possible
-
Predefined success metric
-
Enough time for conversion lag
Practical Rule
Attribution helps operate campaigns. Incrementality helps decide whether the spend deserves to exist.