Incrementality is the share of conversions or revenue that happened because of a marketing activity and would not have happened without it. It is measured by comparing a group exposed to the activity with a randomly held-out control group; the difference is the incremental lift.
What is incrementality in marketing?
Google describes Conversion Lift as an incrementality tool that compares a treatment group that sees your ads with a control group that does not; the difference in conversions is the lift caused by the ads.1Source 1 · Google Ads HelpAbout Conversion Liftsupport.google.com Tests can be user-based (groups built from aggregated user attributes) or geo-based (groups of regions, which also supports offline sales).1Source 1 · Google Ads HelpAbout Conversion Liftsupport.google.com
Lift %=(Treatment conversion rate − Control conversion rate) ÷ Control conversion rate
- Incremental conversions
- Treatment conversions − Control rate × Treatment size
- Incremental CPA
- Spend ÷ Incremental conversions
- iROAS
- Incremental conversion value ÷ Spend
Worked example
| Metric | Treatment (90%) | Control (10%) |
|---|---|---|
| Users | 900,000 | 100,000 |
| Conversions | 21,600 | 2,000 |
| Conversion rate | 2.4% | 2.0% |
| Expected without ads (2.0% × 900,000) | 18,000 | — |
| Incremental conversions | 3,600 | — |
| Lift | 20% | — |
Why isn’t attribution enough?
Attribution counts the conversions of people who saw or clicked an ad, including people who would have bought anyway. In a study of 15 large US advertising experiments at Facebook, Gordon, Zettelmeyer, Bhargava and Chapsky found that observational methods often failed to produce the same effects as the randomized experiments, even after conditioning on extensive demographic and behavioral variables.2Source 2 · Gordon, Zettelmeyer, Bhargava & Chapsky, Marketing Science, 2019A Comparison of Approaches to Advertising Measurement: Evidence from Big Field Experiments at Facebookkellogg.northwestern.edu Randomized holdouts are the most reliable way to calibrate ROAS and media mix models.
Common mistakes
- Underpowered tests. Small control groups or short tests produce confidence intervals wider than the effect.
- Changing creative or budget mid-test, which muddies what was measured.
- Contaminated geos: neighboring regions share media markets and shoppers.
- Generalizing one result forever. Re-test when spend levels or audiences change.
How to track incrementality in Kimo
Kimo stores each lift test as a record (dates, channel, treatment and control, lift, interval) and applies the measured incrementality factor to platform-reported conversions from Google Ads and Meta. The Command center can then show attributed, incremental and blended (MER) views of the same spend, so budget debates start from causal numbers.
Frequently asked questions
What is a holdout group?
Do incrementality tests cost money?
How often should I run lift tests?
Sources
2 references- About Conversion Lift (opens in a new tab)Google Ads Helpsupport.google.com
Treatment vs control; user-based and geo-based tests; incremental CPA and iROAS.
- A Comparison of Approaches to Advertising Measurement: Evidence from Big Field Experiments at Facebook (opens in a new tab)Gordon, Zettelmeyer, Bhargava & Chapsky, Marketing Science2019kellogg.northwestern.edu
Observational methods often fail to match randomized experiments.
External sources were accessed at the time of writing. Kimo product details, customers and figures in examples are illustrative unless a source is cited.
