Incrementality Testing — iGaming GlossaryIncrementality testing measures the true causal impact of an iGaming marketing channel or campaign by comparing treated audiences to randomised or geo-matched control groups.

Incrementality Testing

**TL;DR:** Incrementality testing measures the true causal impact of an iGaming marketing channel or campaign by comparing treated audiences to randomised or geo-matched control groups.

What it means

Where attribution models distribute credit between channels, incrementality answers a different question: "If I turned this channel off, how many FTDs would disappear?" Methods include geo holdouts (turn off a channel in 10 DMAs/regions for 4 weeks), Meta/Google native lift tests (ghost ads to control users), and synthetic-control causal models.

Incrementality is especially valuable for upper-funnel channels (TV, sponsorship, CTV, programmatic) where last-click is blind, and for evaluating affiliate value above the brand-search baseline.

Formula / How it's measured

Incremental conversions = Conversions(treated) − Conversions(control), often normalised by population.

Example: a US sportsbook runs a 4-week geo lift in 12 states with Meta enabled vs 6 matched control states with Meta paused. Treated states show 14,200 FTDs vs 11,600 expected from controls → 2,600 incremental FTDs. Meta spend in treated states $1.04M → true CAC $400 vs platform-reported CAC $190.

Why it matters for operators

Incrementality usually reveals that platform-reported conversions overstate true contribution by 20–60%, especially on retargeting and brand-bid campaigns. Operators who routinely run incrementality reallocate budget toward genuinely demand-creating channels and stop overpaying for cannibalised conversions.

Common benchmarks (2026)

Geo holdout duration: 3–8 weeks
Min markets per arm: 8–12 for reasonable power
Meta Lift / Google Conversion Lift: native tools available
Typical "true vs platform" CAC inflation: 25%–60%
Cadence: 1–2 incrementality tests per major channel per year