Every regulated iGaming operator has sat through the same meeting: a new affiliate or paid channel posts strong sign-ups, the dashboard shows a healthy CPA, and the budget gets renewed. Six months later, blended costs are up, retention is flat, and nobody can explain where the players came from.
The explanation is usually simple. In an affiliate-heavy mix , bonus-hungry depositors, casino messenger traffic, brand-search interception ,last-click attribution credits channels for players they never created.A channel that "works" on last-click can be adding zero incremental value , or worse, cannibalising demand you would have captured anyway.
Measuring incremental value requires a different discipline:incrementality testing.Here is how to design it, before you spend another budget cycle trusting the dashboard.
What incrementality actually means
Incrementality answers one question:how many of these players would we have acquired without this channel?If the answer is "most of them," the channel is not generating growth , it is generating cost.
The gap between attributed and incremental is widest where your mix is densest:
The mechanics are not exotic , randomised experiments on media spend. What separates good operators is discipline on design, duration, and decision thresholds.
Method 1: Holdout groups
The cleanest test is aclean holdout: a randomly selected slice of your target audience excluded from the channel while the rest is exposed. For partners targeting logged-out traffic, use an exclusion or suppression segment; for addressable audiences, a random split.
The incrementality read is simple:lift = (treated conversion − control conversion) ÷ control conversion.If lift is near zero, the channel is harvesting demand, not creating it.
Method 2: Geo-split tests
A holdout needs addressable audiences. When the channel is broadcast , programmatic, TV, out-of-home, sponsorships ,geo-split testingis the workhorse. Split matched regions into treatment and control geos, launch in treatment only, and compare new depositors or NGR *between* geos.
Watch the failure modes:
Method 3: Staggered launches
For channels that cannot be held out or geo-split , a league sponsorship going live nationally, a payment-provider panel, a new brand partnership ,staggered or switch-back designsare the pragmatic option. Run the channel in alternating windows: two weeks on, two weeks off, repeated, comparing conversion and NGR against prior-year periods and a control set of players.
Method 4: Promo codes and trackable links for partnerships
Partnership deals sit in a measurement grey zone: the partner controls the placement, traffic blends with organic, no tag survives every surface.Dedicated promo codes and trackable links are the minimum viable measurement layerfor sponsored-content and partnership deals.
Set the minimum detectable effect before spending
The most common mistake isdesigning the test after the budget conversation, instead of before it.The minimum detectable effect (MDE) , the smallest genuine uplift the test can reliably distinguish from noise , must be agreed before launch, as it determines test size and cost.
When a "working" channel fails the test
Plenty of channels that pass the last-click screen fail incrementality tests:
The response is not necessarily a kill. It is a renegotiation:move the commercial model to match true incrementality.A brand-terms affiliate whose lift is 10% should be paid for the marginal 10% , via rev-share with downside protection, or a hybrid CPA capped at incremental value.
The practical first step
Before your next budget meeting, write down three numbers: the channel's claimed CPA, itsincremental CPA from your last test of that channel type, and the MDE your next test is powered to detect. If you cannot produce the middle number, the conversation has not started.
Run one holdout or geo test on your single largest affiliate or paid channel this quarter. It costs a rounding error against the budget you free when you stop paying last-click prices for players you were going to get anyway.
About Digital Fuel
Digital Fuel is a performance marketing consultancy and commercial growth partner for the global iGaming, sports betting, and digital entertainment sectors. We help operators and B2B suppliers plan and execute market entries, from licensing-stage strategy to acquisition, retention, and partnership programmes that deliver measurable, sustainable growth.
To build incrementality testing into your acquisition strategy, explore our /services or contact the team at /contact to arrange a discussion.
Frequently asked questions
What is incrementality in the context of iGaming?
How do holdout groups work for measuring incrementality?
What are geo-split tests and when should they be used?
How can promo codes help in measuring the effectiveness of partnerships?
What is the minimum detectable effect and why is it important?
Ready to put this into practice?
Get in touch