Incrementality testing measures the true causal effect of your advertising — how many conversions happened because of the campaign that would not have happened without it. It is the antidote to attribution inflation: last-touch and even multi-touch models routinely credit ads for conversions that would have occurred organically. Incrementality isolates the campaign contribution by comparing outcomes between a group that saw the ads and a matched group that did not.
Two main approaches for India CTV: (1) Geo holdout test — run the CTV campaign in some cities or states (test markets) and hold back spend from matched markets (control markets). Compare conversion rates between test and control geographies. The difference is the incremental effect. This is the most practical approach for India CTV because it does not require individual-level identity matching across platforms. (2) User holdout test (ghost ads) — within a single publisher, randomly assign some users to see the ad (exposed) and suppress for a matched holdout group. Requires publisher cooperation — feasible on JioCinema and YouTube CTV. Incrementality tests require minimum scale (INR 50 lakh+, 4+ weeks) to produce statistically significant results. Worth running for any brand spending meaningfully on CTV who wants to know if the channel is actually working.
Full guide
For a complete explanation, read: Incrementality testing for CTV advertising: how to measure what your campaign actually caused