The cheap licence is usually the expensive one

Operators send us two quotes side by side and ask why one jurisdiction costs several times the other. It is a fair question. It is also the wrong comparison, because a licence is not a product with a price, it is a set of permissions, and the fee you pay to obtain it is the smallest number attached to it. The rest arrive later, monthly, and they decide whether the business works.

What the fee actually buys

An application fee buys you an assessment and a grant. That is all it buys.

It does not buy a market, a bank account, a game library or an affiliate deal. Those come from the same document, but they are not priced into it, and they are where the money is.

Three parties read your licence after the regulator has finished with it. Payment providers read it to set your risk category and your rate. Game studios and aggregators read it to decide whether their content may appear on your platform and on what terms. Affiliates read it to decide whether sending you traffic is worth the chargeback exposure and the reputational risk of being associated with you. None of them owe you an explanation, and none of them will improve their terms because you found the licence cheaply.

That is the whole mechanism. The fee is a one-off; the commercial terms it produces are permanent until you change the licence, and changing the licence is the expensive part.

The markets a cheap licence removes

Every gambling licence carries a restricted-territory list. That list is a revenue forecast, written by somebody who has never seen your business plan.

Anjouan is the clearest case, and we say this as a firm that files Anjouan licences and recommends them regularly. The regime is fast, light and genuinely well matched to a crypto-facing operator launching a new brand outside Europe. Its geo-block list is also binding, and it covers the United States, France, Australia, the Netherlands, Singapore and other restricted territories, with no EU market access attached to the grant.

So if your traffic plan runs on European search and European affiliates, the licence you obtained in three weeks has just deleted most of your addressable market. No amount of compliance work puts it back.

Curaçao carries its own list and its own conditions. The Isle of Man carries a heavier substance and oversight burden and, in exchange, opens payment relationships and game-studio contracts that lighter regimes do not reach. Kahnawake and Tobique sit between the two. Not one of the five is the cheap option or the expensive one in any absolute sense: each is cheap for one operator and ruinous for another, and which of the two you are is settled entirely by your market list.

Which is why the first thing to price is not the licence. It is your top five markets by expected revenue, checked against the restricted list of every jurisdiction on the shortlist, before anybody quotes you anything.

We do that at the start of an engagement rather than the end. It is the only step capable of invalidating the whole plan, so it goes first.

The price of a timeline nobody believes

The second cost hides inside the promise of speed.

Regulators in the lighter jurisdictions really are fast. Anjouan grants in weeks from a complete file, and that figure is accurate. It is also beside the point, because the regulator was never the constraint. The payment layer is, and the payment layer runs on its own calendar no matter how quickly your certificate arrived.

We have watched the same sequence often enough to call it a pattern rather than bad luck. An operator buys the fast licence and holds the grant within a month. Then spends four more months trying to find a payment relationship that will accept it. Throughout those four months the platform licence is running, the team is on payroll, the content deals have minimum guarantees attached, and the marketing budget is either burning or frozen. The licence was quick. The launch was not, and the burn was modelled as though it would be.

The opposite error costs just as much. An operator commits to a heavier European regime on a quoted timeline, then finds that the source-of-funds file for one shareholder needs six weeks of document recovery from a bank in another country, and the whole schedule slides behind it. That is not a slow regulator. It is a preparation problem nobody costed.

Both errors have one fix. Open the payment and banking conversation on day one, in parallel with the application, and treat the answer you get there as your real launch date.

A licence you cannot bank is a certificate, not a business.

What re-licensing actually costs

This is the number that ends the argument, and Curaçao has just demonstrated it at scale.

Under the previous ordinance, most operators in Curaçao did not hold a licence at all. They held a sub-licence issued by one of a handful of master licence holders. It was among the least expensive recognised gambling authorisations available anywhere and it was sold on exactly that basis for two decades.

The National Ordinance on Games of Chance took effect on 24 December 2024 and abolished the structure. The Curaçao Gaming Authority now licenses operators directly, sub-licence certificates of operations are no longer valid and must not be displayed, and existing licences were converted into provisional ones pending full review against the current standard.

Read that back as a cost line. Every operator on a sub-licence had to file a complete application, to a materially heavier standard, on somebody else’s deadline, while already trading.

What that involves in practice, none of it optional: a new application built to a standard the original file was never written to meet; a corporate entity satisfying the current requirements, which under the reformed regime means real local substance, resident key personnel, an actual office rather than a virtual one, records kept on the island; and a compliance suite rewritten rather than refreshed, because the AML, responsible-gaming and technical standards all moved.

Then the part operators consistently underestimate.

Every payment relationship is re-underwritten against the new entity and the new licence. Every supplier and aggregator contract is re-papered. The player database is migrated and, in many cases, re-verified. Domains, seals and terms of service change under the players’ feet at the same time, and a proportion of those players do not come back through the door.

The re-licensing exercise costs more than the original licence did. It costs more than the gap between the cheap jurisdiction and the correct one ever was. And it lands at the worst possible moment, because regimes reform when they are under external pressure, which is precisely when your payment providers are already reviewing their exposure to your sector.

How to count it properly

Four questions, asked before anything is signed.

Which of my markets does this licence permit? Not whether it is recognised. Not whether it is respected. Which specific territories can I lawfully take money from, and what share of my forecast sits in them. If the honest answer removes a third of the plan, the licence is not cheap and never was.

What does the payment layer cost under this licence? Ask for the rate and the rolling reserve, not the setup fee. A few points of difference on processing, applied to every transaction for the life of the business, dwarfs any licensing fee ever quoted. It is the largest recurring number in the entire comparison and it is almost never in the spreadsheet.

What recurs, and who performs it? Renewals, annual clearances, reporting calendars, audits, substance that has to stay real rather than merely stated. Licences lapse quietly. The first sign is usually a payment provider noticing before you do.

What happens if this regime reforms? Not hypothetical. Curaçao reformed. Costa Rica has been proposing a framework for years without enacting one, which is why the heavily advertised “Costa Rica gambling licence” remains a municipal business permit with a marketing name attached rather than a gaming authorisation at all. Ask what your migration path looks like, and carry it as a liability from day one instead of a problem for a future version of the company.

Count those four properly and the comparison usually inverts. The jurisdiction that quoted higher was the cheaper one, and it was not close.

Where we stand

We file gambling licences in Anjouan, Curaçao, the Isle of Man, Kahnawake and Tobique, and we file across the crypto, payments, forex and fund regimes in Europe and beyond. We are not neutral about jurisdictions and we do not pretend to be. We are neutral about which one you end up in, because we deliver in all of them, and we are paid to be right rather than quick.

In practice that means two things. We will tell you when the cheaper jurisdiction is genuinely the correct answer, which happens more often than the market assumes. And we will tell you when the one you have already settled on is going to cost you a market you were counting on, which is the conversation nobody selling a licence wants to have with you.

We map your activity to the regulator that supervises it, choose the jurisdiction that fits your commercial model, and file the application ourselves, alongside the company formation and the payments work that has to run in parallel for the licence to be worth holding at all.

If you are comparing quotes right now, bring us both of them and the market list they are meant to serve. We will tell you which one is actually cheaper, before you pay anybody. Including us.