industry
Revenue share, CPA and negative carryover explained with worked figures, what each payment model does to a comparison site's incentives, and the four signals that separate a site trying to be honest from one that is not.
Almost every casino comparison site you will find is paid by the casinos it compares. That includes this one. Our own arrangements are set out on the affiliate disclosure page and the scoring we apply is on the methodology page.
Knowing that a site is paid is not very useful on its own. Knowing how it is paid is, because the payment structure shapes which operators get recommended, which facts get printed, and which get quietly left out. This guide explains the mechanics with worked numbers, then sets out what an honest site looks like from the outside.
The plumbing is simple. A comparison site links to an operator with a tracking identifier attached. The operator's software records that identifier against any account opened from that click, usually with a cookie window of some weeks. From that point the account is tagged to the affiliate for as long as it exists.
The affiliate then earns on that account under one of three commercial models. The model is negotiated per operator and is almost never disclosed publicly, which is the first thing worth understanding about the industry.
Cost per acquisition pays a fixed fee for each player who meets a qualifying condition. The condition is typically a first deposit above a minimum, sometimes with a requirement that the deposit is wagered once.
A representative structure is a fixed fee of a few hundred units of currency per qualifying player. Payment happens once. What the player does afterwards is commercially irrelevant to the affiliate.
The incentive that creates is straightforward. Under CPA, a site is rewarded for the number of people who deposit, and for nothing else. Whether they enjoy the site, get paid on withdrawal, or come back next month does not change the payment. A site running purely on CPA has no financial reason to care about withdrawal reliability, which is the single most consequential thing about a casino.
Revenue share pays a percentage of the net gaming revenue generated by tagged players, month after month, usually somewhere between a quarter and a half.
Net gaming revenue is not the same as player losses. Deductions are applied before the share is calculated, and the deduction list is where the details hide. Bonus costs are commonly deducted. Payment processing fees often are. Some contracts deduct an administrative or platform fee as a flat percentage. Two contracts quoting the same headline percentage can pay materially differently once the deductions are applied.
The incentive here is longer term than CPA. The affiliate earns only while the player keeps playing, which at least aligns the site with operators that do not alienate their customers. It also aligns the site with player losses continuing indefinitely, which is worth saying plainly rather than dressing up.
This is the mechanism most readers have never heard of, and it is the one that most distorts behaviour.
Under revenue share, a month in which tagged players win more than they lose produces negative net gaming revenue. The affiliate's share of a negative number is negative.
Negative carryover determines what happens next. With carryover, that negative balance is carried into the following month and must be worked off from future revenue before any payment resumes. Without carryover, the balance resets to zero at month end and the affiliate simply receives nothing for that month.
Take one cohort of players at a 40% revenue share over three months.
| Month | Net gaming revenue | Affiliate share at 40% |
|---|---|---|
| 1 | −$5,000 | −$2,000 |
| 2 | $3,000 | $1,200 |
| 3 | $4,000 | $1,600 |
Now settle it both ways.
| Month 1 paid | Month 2 paid | Month 3 paid | Three-month total | |
|---|---|---|---|---|
| With negative carryover | $0, balance −$2,000 | $0, balance −$800 | $800 | $800 |
| Without negative carryover | $0, balance reset | $1,200 | $1,600 | $2,800 |
Same players, same results, same headline percentage. The payment differs by a factor of three and a half, entirely because of one clause.
The behavioural consequence follows directly. Under negative carryover, a single large winner can put an affiliate into a hole that takes months to clear. That creates a quiet preference for operators without the clause, for CPA deals that cannot go negative, and, uncomfortably, for traffic that loses steadily rather than traffic that occasionally wins big. No site will ever tell you this is happening. It is simply what the contract structure rewards.
| Model | Paid when | Rewards | Blind to |
|---|---|---|---|
| CPA | Once, on qualifying deposit | Volume of new depositors | Withdrawal reliability, player retention, anything after day one |
| Revenue share, no carryover | Monthly, on positive months | Operators players stay with | Downside of a bad month, absorbed by the operator |
| Revenue share, with carryover | Monthly, after clearing a deficit | Steady losing cohorts | Nothing, which is the problem |
| Hybrid | Reduced CPA plus reduced share | A mix of both | A mix of both |
There is a further layer. Commercial rates differ between operators, sometimes substantially. A site with no fixed, published scoring method will drift towards ranking the best-paying operator first, not by conspiracy but by ordinary commercial gravity. Placement in a top position is also sold directly at some sites, and is not always labelled.
Networks add another step. Many affiliates work through aggregators that contract with operators and pass on a portion, which means the site writing the review may not know the underlying terms at all.
You cannot audit a site's contracts. You can audit its output, and the output carries four signals that are difficult to fake.
Sourced facts. Every load-bearing claim should be traceable to a primary source. An RTP figure should come from the game's own information screen or the provider's published specification, not from another comparison site. A licence claim should be checkable against the regulator's public register. A fee or a withdrawal minimum should come from the operator's own terms page. A site that cites nothing has either done the work and hidden it or not done it.
Dated verification. Casino terms change without notice. Wagering multipliers move, withdrawal limits move, games get redeployed at different RTP configurations. A claim without a date attached is a claim about an unknown point in the past. Look for a checked-on date against specific facts, not a single site-wide timestamp that updates automatically.
Published scoring. If a site ranks operators, the weights should be published in advance, applied uniformly, and detailed enough that you could recompute a score yourself and get the same answer. Scoring that exists only as a number next to a logo is a marketing device. Ours is on the methodology page, including the weights and what each one is measured from.
Unflattering details left in. This is the strongest signal available, because it is the one that costs the site money. A review that finds nothing to criticise about a commercial partner is not a review. Real examples of the kind of detail an honest site keeps in: naming a licence tier accurately, including the fact that Anjouan licences are among the least demanding available and that Curaçao has moved to issuing licences directly; flagging that a title such as Aviator runs at 97% but is configurable by the operator down to 96% or 94%; recording documented withdrawal delays; and stating where an operator cannot legally serve a reader at all.
That last point is a live test. In France the ANJ never authorises online casinos, so any site presenting offshore casinos to French readers as a legitimate option is misrepresenting the legal position, and the loi du 9 juin 2023 bans influencers from promoting unauthorised gambling. In the United States, regulated online casino play exists in Delaware, Michigan, New Jersey, Pennsylvania, Rhode Island, West Virginia, Connecticut and Maine. A site that never mentions the regulated option while promoting offshore alternatives has made a commercial choice and presented it as advice.
A fifth signal is quieter: what is absent. Operators that run no affiliate programme, or that decline to work with a given site, may simply not appear in any listing. Absence from a comparison table is rarely a verdict. It is usually a commercial fact, and few sites say so.
We take affiliate commission. It is how the site is funded. That is a conflict of interest and it does not stop existing because we have described it.
What we can do is make the conflict checkable. The scoring weights are published before they are applied. Facts carry sources and dates. Criticism of operators we are paid by stays in the text. None of that makes us disinterested. It makes us auditable, which is a lower and more honest claim.
Treat us the way you would treat any other interested party. Verify one thing yourself. Pick a claim on this site that matters to your decision, such as a licence, a withdrawal limit or an RTP figure, and check it against the primary source. If it holds, that tells you something real about the rest. If it does not, you have learned something more useful still.
The commercial model behind this industry is not a scandal. It is ordinary advertising, and it funds work that is often genuinely useful. The failure is not being paid. The failure is being paid and pretending otherwise.