The deal that made sense in 2016 is losing you money in 2026

When I started in affiliate marketing in 2008, a deal was a simple thing. An affiliate sent players, an operator paid for them, and if the numbers roughly worked at the end of the quarter, everyone renewed. The math was forgiving because the market was forgiving. Acquisition was cheap, markets were new, and a player who stuck around for a few months covered a lot of sins.

 

Almost none of that is true anymore, and yet a surprising number of affiliate deals are still structured as if it were.

 

What changed underneath

 

Three things moved at once, and most operators have priced in one of them at best.

 

The first is cost. Industry media has been saying for a while now that the affiliate-dependent, bonus-driven acquisition model is under pressure from every direction: regulation, tax rises and rising media costs all at once. In my own conversations, the CPA an operator would have laughed at five years ago is now the opening position.

 

The second is retention, which is the part people prefer not to talk about. Optimove, which aggregates data across its operator base globally, puts the churned segment at 55% of a typical operator’s customer base on average. Churned meaning deposited once and then went quiet for thirty days or more. That is the single biggest group of players most operators have. More than half of everything they ever paid for.

 

The third is regulatory pressure on the product itself. Bonus restrictions, affordability checks, advertising limits. Each one lowers the value of an acquired player and lengthens the payback period, and each one lands hardest on exactly the kind of traffic that was only ever there for the welcome offer.

 

Put those together and the old deal breaks. Not dramatically, but inevitably, over a year, in a way that only shows up when finance runs the numbers and asks why affiliate spend went up and margin didn’t.

 

The math nobody runs

 

Let me walk through an example. The numbers are illustrative, but the shape is one I see constantly.

 

An operator pays an affiliate a flat rate CPA per first-time depositor. Say €200. The affiliate sends 500 FTDs in a quarter, so the invoice is € 100,000. On the operator’s dashboard, this is a good quarter: volume is up, cost per acquisition is on target, everybody’s happy.

 

Now look at it the way an accountant would. Half of those depositors put in the minimum, took the bonus and left within a month. They are worth, generously, €20 or €30 each in gross revenue before you even think about bonus cost. A third are ordinary players who will churn somewhere in months two to four. And a small slice, maybe fifty players, will still be active at the end of the year and will produce the vast majority of whatever this cohort ever returns.

 

The operator paid €200 for each of the fifty good players. It also paid €200 for each of the two hundred and fifty who were gone before the bonus wagering was complete. Same price. Same line on the invoice. Same “success” in the monthly report.

 

This is what I mean when I say the deal structure decides the margin. The affiliate did nothing wrong. They sent what they were paid to send. The operator got exactly what it priced for, which was volume.

 

Why it persists

 

If this is so obvious, why is it still how most of the industry works?

 

Partly because CPA is easy. One number, paid once, no arguments about revenue share six months later. Partly because the person negotiating the deal and the person who sees the twelve-month cohort data are rarely the same person, and by the time the second one has the numbers, the first one has renewed.

 

And partly because the alternative requires honesty on both sides. The operator has to tell the affiliate what its players are actually worth, which most operators treat as confidential. The affiliate has to accept being paid differently for different traffic, which most affiliates treat as a pay cut.

 

The market is starting to move anyway. More and more programs are shifting new agreements to hybrid structures, a lower CPA paired with a revenue share, and in Tier-1 markets it’s becoming the default rather than the exception. That’s the industry admitting, deal by deal, that a flat price for an unknown quantity was never a great idea. It’s a start. It is not the same as knowing what you’re buying.

 

I have spent seventeen years on the affiliate management side of this table, and I’d say the operators who do best by their affiliates are the ones who know precisely what a player is worth to them and are willing to say so out loud. Everybody else is still paying 2016 prices for 2026 players.

 

That’s the problem I’ll be picking apart at the Acceler8 Lab at SBC Summit Lisbon. Join me on September 29, 16:00, by registering at https://gr8.tech/acceler8-lab/