iGaming Revshare vs CPA: Choosing the Right Deal

Most affiliates settle the iGaming revshare vs CPA question once, early, based on temperament rather than arithmetic. The cautious ones take CPA and never revisit it. The patient ones take revshare and tell themselves they are building an asset. Both are picking a default and applying it to traffic sources with completely different economics.

The decision is per campaign, and it turns on two numbers. The first is the player-level break-even: how many months a surviving player must stay before revshare overtakes the CPA. Nearly everyone calculates this one. The second is the cohort-level break-even: how many months before a hundred first-time depositors, most of whom churn, collectively overtake a hundred CPA payments. Almost nobody calculates this one, and it is typically three to four times longer.

What follows is both calculations worked out in full, the case for hybrid deals, why paid buyers behave the way they do, and how to move a deal in your favour once you have real data. The math is simple. The consequences of skipping it are not.

The three deal shapes, precisely defined

CPA. A fixed payment for each qualifying first-time deposit. The qualification rules matter enormously — minimum deposit, minimum wagering, time window between registration and deposit — and are covered in the full breakdown of how casino affiliate programs work. Payment usually lands in the next billing cycle. Your exposure to the player ends there.

Revenue share. A percentage of the net gaming revenue that player generates, paid monthly, usually for the life of the account. Common headline rates run from 25% to 50%, often tiered on monthly FTD volume. The percentage applies to NGR, not to what the player loses, which is a distinction worth several thousand words on its own.

Hybrid. A reduced CPA plus a reduced revenue share on the same player. Typical shapes are around half the standard CPA plus half the standard percentage, though the split is negotiable and the two halves rarely add to a whole.

There is a fourth shape you will meet: tiered revshare with a monthly reset, where your percentage depends on how many FTDs you delivered that month. It looks generous and behaves badly, because a slow month drops your rate on your entire existing player base, not just on new players.

iGaming revshare vs CPA at the player level

Take the standard scenario. A player generates $60 of NGR per month. Your revenue share is 40%. The alternative offer is a flat $120 CPA.

• Monthly commission under revshare: $60 × 0.40 = $24
• Months to match the CPA: $120 ÷ $24 = 5 months

If that player stays active for five months, the two deals pay identically: $120. Month six onward is pure upside on revshare. If the player quits in month three you collected $72 and the CPA would have paid $120.

The general formula is simply CPA divided by (monthly NGR × your rate). Keep it on a sticky note. It gives you an instant read on any offer an affiliate manager puts in front of you, and it makes obvious what most rate negotiations ignore: raising your revshare from 40% to 45% moves break-even from 5 months to 4.4 months, which is a smaller improvement than it sounds.

The trap in this calculation

This number describes a player who survives. It says nothing about the probability that any given player does. A five-month break-even sounds comfortable because five months feels short. It is not short in casino retention terms — a large majority of first-time depositors never see month five.

That is the gap between the calculation everyone does and the one that actually determines your income.

The cohort-level break-even

Model a hundred FTDs and let them decay. Here is a reasonably healthy retention curve for a mid-market casino brand, expressed as players still active in each month:

Add the first twelve months: 100 + 58 + 42 + 34 + 29 + 26 + 24 + 22 + 21 + 20 + 19 + 18 = 413 active player-months. The second twelve months add another 168, for 581 player-months over two full years.

Hold NGR at $60 per active player-month and your 40% share pays $24 per player-month.

• Twelve-month revshare total: 413 × $24 = $9,912
• Twenty-four-month revshare total: 581 × $24 = $13,944
• CPA alternative: 100 × $120 = $12,000, paid within about 45 days

To match the CPA you need $12,000 ÷ $24 = 500 player-months. Running the cumulative total: 413 at month 12, 430, 446, 462, 477, 492, then 506 at month 18.

So the cohort-level break-even is month eighteen. Not month five. Same players, same rate, same operator — and the honest payback period is more than three times longer than the player-level math suggests.

Over the full two years, revshare returns $13,944 against the CPA's $12,000. A 16% premium, earned by waiting a year and a half and carrying all the risk in between: term changes, brand failure, negative carryover eating months you thought you had banked, and the operator's retention quietly deteriorating.

Why the two numbers diverge so hard

The player-level calculation implicitly assumes survival is certain. The cohort calculation prices in the fact that roughly 40% of your depositors are gone by month two and most of your long-run revenue comes from a small tail of players who stay for years.

That tail is real and it is valuable, which is why revshare is not a bad deal. But it means revshare is a bet on the shape of a distribution, and the shape varies enormously between operators. A brand with a strong VIP program and disciplined reactivation has a fat tail. A brand that acquires on aggressive bonuses and churns has almost none, and its revshare is close to worthless past month four regardless of the headline percentage.

Hybrid deals: what they actually solve

Run the same cohort on a hybrid of $60 CPA plus 20% revshare.

• Upfront: 100 × $60 = $6,000
• Revshare over 24 months: 581 player-months × $60 × 0.20 = 581 × $12 = $6,972
• Total: $12,972

Now note something that surprises most people. The hybrid overtakes the pure $120 CPA at exactly the same moment pure revshare does — month eighteen. That is not a coincidence. A 50/50 hybrid is a linear blend of the two deals, so half of each side cancels and the crossover point is unchanged.

Which means hybrid does not buy you a better long-run total or a faster payback against CPA. What it buys is variance reduction and cashflow. Half your return is locked in immediately and cannot be taken away by a bad clause, a churned brand, or a jackpot winner. The other half preserves your exposure to the tail. If you are uncertain about the operator's retention — which you should be, until you have a cohort of your own data — hybrid is the rational position while you find out.

The mistake is treating hybrid as a compromise you take when you cannot decide. It is a specific instrument for a specific condition: high uncertainty about the revenue curve, combined with a need for working capital.

Cashflow versus asset-building

Two affiliates with identical traffic can rationally choose opposite deals, because the constraint that binds them is different.

If your traffic is bought, cash is your input. Every month you wait for revshare is a month that cash is not buying more traffic. If your traffic is owned — a ranking site, an email list, a Telegram channel, a YouTube back catalogue — the marginal cost of the next player is near zero, so waiting costs you nothing except risk.

The reinvestment test

This is the calculation that decides it. Take the $12,000 of CPA cash and ask what it earns if redeployed.

Suppose your media buying returns $1.15 for every $1 spent, net, over a 60-day cycle. Redeploy three times and you have 12,000 × 1.15³ = 12,000 × 1.520875 ≈ $18,250, reached in roughly six months. That is a $6,250 gain, against revshare's $1,944 premium earned over twenty-four.

CPA wins that comparison decisively — but only if the assumption holds. The assumption fails in two common ways. First, capacity: most affiliates cannot triple spend at constant ROI, because the cheap inventory runs out and the marginal impression is worse than the average one. Second, volatility: a 15% net margin is an average across campaigns, and one bad month of testing can wipe a cycle's gain.

So the honest version of the test is: can you actually absorb three times your current spend at similar efficiency? If yes, take CPA and compound. If no, the cash has nowhere productive to go and revshare's tail is worth more than idle capital.

Why paid media buyers almost always need CPA

Beyond compounding, three structural reasons push buyers toward CPA.

Attribution decay. A media buyer optimises campaigns against a signal. If the signal arrives eighteen months later, it cannot inform bidding. CPA produces a clean, fast, per-click value that feeds directly into your bid logic and your sub-ID level optimisation. Revshare gives you a number you will only trust after a year.

Platform payment terms. Ad platforms bill on their schedule, not yours. Running revshare while paying media invoices net-7 means financing an eighteen-month payback out of working capital, which is a lending business, not an affiliate business.

Traffic quality mismatch. Bought traffic frequently converts at high FTD rates and low retention — exactly the profile where CPA extracts maximum value and revshare extracts almost none. Selling a low-retention cohort at a fixed price is not exploitation, it is correct pricing of what you have.

The mirror case is equally true. If you rank for high-intent casino review terms and your players arrive already researched and self-selected, you are producing an unusually retentive cohort and CPA underpays you badly. That is the population where the revshare tail earns its reputation.

An iGaming revshare vs CPA framework you can apply per campaign

The last row is not a joke. Plenty of affiliates have taken theoretically correct revshare deals into insolvency while the theory was still correct.

Negotiating a deal change once you have data

Rate conversations go badly when they are opinion against opinion. They go well when you arrive with reconciled numbers.

Gather one full cohort, minimum thirty to fifty FTDs from a single identifiable source, tracked with unique sub-IDs. Then present four things: NGR per FTD by geo, the month-by-month retention of your cohort against whatever benchmark the manager will share, your bonus-abuse and chargeback rate, and the specific structure you want.

Frame it around what the operator cares about, which is not fairness but predictability and cost per acquired value. If your players retain better than the program average, say so with the curve. If your chargeback rate is negligible, that is worth real money to them and worth mentioning.

Then ask for something bounded. A scoped trial — one geo, one traffic source, ninety days — gets approved by an affiliate manager without escalation. A blanket rate increase across the account needs sign-off from someone who has never heard of you. Practical asks that tend to succeed:

1. Removal of the administrative fee rather than an increase in the percentage. It costs the operator less to concede and is worth roughly the same to you.
2. A hybrid trial on one source, rather than switching the whole account.
3. Grandfathering: any rate change applies to new players only, so your existing cohort is not repriced.
4. A written carryover reset, monthly, in the terms rather than in an email.
5. A CPA bump tied to a volume commitment you are confident you can hit, with no penalty clause if you miss.

Bring a comparison too. Knowing what a competing brand offers for the same geo changes the conversation, and the broader mechanics of that trade-off are covered in the general treatment of revenue share versus one-time CPA.

Where this goes wrong

Comparing headline numbers across programs. A 45% share with six NGR deductions and a 10% admin fee pays less than a 35% share with three deductions. Compare the effective percentage of gross gaming revenue, not the number on the landing page.

Switching deal type mid-cohort. Some programs move you from revshare to CPA prospectively but stop paying on the existing base. Confirm in writing that historical players continue to earn before you agree to anything.

Using account-wide averages. Your Germany search traffic and your push traffic are different businesses with different retention curves. Averaging them produces a deal that is wrong for both. Segment before you decide — the reason per-source KPI tracking matters is precisely that it lets you price each source correctly.

Ignoring the discount rate on your own money. Revenue two years out is not worth revenue today, especially in a vertical where brands disappear and terms change unilaterally. If a revshare deal only wins by 16% over 24 months, it is not really winning.

Taking a tiered revshare without modelling a bad month. Check what happens to your rate on the entire player base if FTD volume drops below the tier threshold. On some programs the answer is brutal.

Working with Shazam on iGaming deal structures

The deal you are offered depends heavily on the volume standing behind the request, which is the main reason individual affiliates get worse terms than their traffic deserves. We negotiate on aggregate across our partner base, which gets access to higher caps, custom payout bumps and structures — hybrid splits, carryover resets, grandfathering — that are hard to obtain on your own. Partners take 50% revenue share, with tracking infrastructure and sub-ID structures built and maintained on our side so the cohort data you need for any renegotiation actually exists, and sites and landing pages built at no cost.

We will also tell you when CPA is the right answer for a source, because a partner who runs out of working capital is not useful to anyone. If you want to work through the numbers on a specific brand or geo, start with the Telegram bot or sit in the community chat and see how other partners are structuring theirs. More on the verticals we cover is on the main site.