Incrementality
Definition
Incrementality measures how many orders or how much revenue a marketing channel actually caused, versus how many would have happened anyway. It is the difference between attributed revenue (what the platform claims) and incremental revenue (what you actually gained).
How operators actually use it
Operators run incrementality tests by pausing a channel in a geography (geo-holdout), splitting users into exposed and unexposed groups (Meta lift studies), or running matched-market tests. Real-world incrementality numbers shock most teams: 30-50% of Meta-attributed revenue would have happened anyway from email, organic, or branded search. Knowing your channel-level incrementality lets you set true ROAS targets that account for the freeloading.
Common pitfalls and honest-cost notes
Single-week holdout tests are statistically meaningless for most brands. You need 4-8 weeks of data and a clean control geo with similar baseline trends. Also: incrementality changes over time as your brand grows — re-run the test annually, not once. Acting on a 2-year-old incrementality study is the same as acting on no study.
Want a free audit of your DTC ad creative against benchmarks like Incrementality? Try the free Frontier Visions audit →
Definition published by Frontier Visions. Operator commentary reflects the editor's view and is not financial or investment advice.