Your payout was declined by an issuer you have no contract with

Search for gambling payments and you will find a great deal about acceptance rates, routing, local payment methods, chargebacks and conversion. Almost all of it is about money coming in. The side where money goes out — the payout that reaches a player's card — is written about far less, and the reason is that most of the decisions on that side are not made by the operator or by its payment provider. They are made by scheme rules the operator cannot vary, by a country list it does not publish, and by an issuing bank it has no contract with and cannot call.

That is an uncomfortable position from which to run a retention-critical process, and it is worth understanding properly rather than treating each declined payout as a support ticket.

MCC 7995 follows you even when gambling is not your business

Visa's Merchant Data Standards Manual defines merchant category code 7995 as "Betting, including Lottery Tickets, Casino Gaming Chips, Off-Track Betting, Wagers at Race Tracks and games of chance to win prizes of monetary value". That much is unremarkable. The part that catches people is the card-absent rule, in the manual's own words: "If the Merchant conducts online gambling transactions, however, it must use MCC 7995 for all transactions, even if gambling is not the Merchant's primary business."

So a media company with a betting vertical, a sports platform with a wagering feature, or a rewards business that added a prize draw does not get to code the gambling part 7995 and the rest something cleaner. Card-absent, the whole thing is 7995.

There is one qualification, and it is worth stating properly because it cuts against the paragraph above. For US merchants, lottery and gambling transactions that qualify for MCC 7800 — government-owned lottery — or 7801 and 7802 must use those codes for all transactions, again regardless of primary business. And the manual adds that if such a merchant also sells other goods or services, it must use the appropriate MCC for those. So the one population that does get to separate its coding is the one operating under a US government-owned or government-licensed structure. Everyone else, card-absent, is 7995 across the board.

MCC 7995 also sits on Visa's high-integrity-risk MCC list for card-absent transactions, without the regional limits that qualify some of the other codes on that list. What follows from that classification downstream is our reading rather than something the manual states.

The manual does close one obvious workaround, in terms: "Acquirers must prohibit agents from utilizing dynamic Merchant descriptors for Gambling Merchants." Read precisely, that binds agents and payment facilitators using descriptors that rotate — not a gambling merchant's own fixed descriptor. It is still one of the few areas of payments where there is nothing to negotiate.

The same payout is legal in one country and unavailable in the next

Payouts to cards run as Original Credit Transactions, and gaming OCTs are not supported everywhere. One major payment provider's implementation documentation, last updated in October 2025, lists Visa gaming OCTs as restricted in fifty countries — Nigeria, Malaysia, India, Brazil, Hong Kong, the United States, Japan and Singapore among them. The same page lists forty-seven entries for Mastercard gaming payouts, almost entirely European states and former Soviet republics; note that it is forty-seven entries rather than forty-seven distinct markets, since the Channel Islands and the Isle of Man appear both individually and inside the United Kingdom entry.

A second provider's documentation describes the Mastercard Gaming and Gambling Payments Program as covering domestic and, where permitted, cross-border gaming payment transactions, and states that payments for purposes other than gambling are not permitted on that rail and are processed through the Mastercard MoneySend programme instead. Both providers put funds availability at typically thirty minutes — and the more precise of the two adds the condition that matters here: within thirty minutes after a participating issuer approves the authorisation. The issuer is in the loop even on the timing claim.

The operational consequence is that the same payout product, built once, works in one market and simply does not exist in the one next door. That is not a failure of the provider, and it is not something a better integration fixes. It is a map.

Now the uncomfortable part, and we went looking specifically to avoid saying it loosely. None of the material in the two paragraphs above is published by the card schemes. We searched the public Mastercard Rules for the Gaming and Gambling Payments Program, for gaming payment transactions, and for payout rules attaching to MCC 7995 and 7994. None of it is there. The country eligibility lists, the market counts, the thirty minutes and the MoneySend routing all come from payment providers' own integration documentation. Anyone outside a scheme member is reading a second-hand account of a rule they cannot audit, and there is no reason to assume two providers' lists agree. One further example of why that matters: the requirement to supply a cardholder name on Mastercard payouts for merchants under 7995 and 7994 is documented as that provider's own API requirement, and the provider does not attribute it to Mastercard. It may well be a scheme rule. We cannot show you that it is.

If your expansion plan depends on card payouts in a specific market, the answer has to come from your sponsor rather than from anybody's blog, including this one.

The issuer declining you may be doing exactly what its customer asked

Here the analysis turns, because the operator's instinct — treat every decline as friction and engineer it away — is sometimes wrong.

Many consumer banks now offer gambling blocks, and the research on how they are built is clear that they work by filtering merchant category codes. Monzo, Starling, Nationwide, Revolut and Lloyds all describe such a feature on their own pages today, all of them card-level blocks operating in practice on debit cards, several of them explicit that they stop "most" gambling payments rather than all, and none of them reaching bank transfers.

It is worth noting how fast that landscape moved. When the University of Bristol's Personal Finance Research Centre surveyed the field in July 2020, it named Nationwide as one of four large banks that did not then offer a gambling block, and Revolut did not appear in the report at all. Both offer one now.

When one of these banks declines a transaction coded 7995, it is not a technical failure and it is not the bank being difficult. It is the bank doing precisely what its customer instructed it to do, using the only signal the card networks give it — which is the same MCC the operator is obliged to use.

So the mandatory coding described above is also what makes consumer self-exclusion at the payment layer possible. Those two things are the same mechanism seen from opposite ends, and an operator that lobbies for looser coding is, whether it intends to or not, lobbying against a tool a lot of its customers rely on.

The practical point for anyone building payout flows: a decline on a deposit and a decline on a payout are not the same event and should not be handled the same way. A blocked deposit may be a customer exercising a control. A blocked payout is money the player is owed, sitting somewhere it cannot land, and it needs an alternative route rather than a retry.

The real bottleneck is upstream of all of this

Acquiring for gambling is a mature, if expensive, market. There are specialist acquirers and a well-understood relationship between dispute and fraud ratios and penalties — Visa's acquirer monitoring thresholds tighten to 150 basis points for excessive merchants in Europe, the US, Canada and Asia-Pacific from 1 April 2026, down from 220, though those thresholds are scheme-wide rather than gambling-specific and the ratio they measure combines fraud and non-fraud disputes rather than chargebacks alone. A gambling operator that wants to accept cards can find someone to do it.

Issuing is a different shape. Putting a payout onto a card, or issuing a card to a player or an affiliate, requires a licensed institution willing to sponsor the BIN and carry gambling exposure on its own licence and its own regulatory relationships — unless you are a principal scheme member yourself, which almost nobody in this sector is. Licensed institutions are on public registers. Which of them will sponsor a BIN carrying gambling exposure is on no register, and in our experience it changes. The constraint is not technology and it is not pricing. It is how many institutions are prepared to have this sector on their book at all.

Which is also why the issuing side is written about so little. The people who understand it are the sponsors and the programme managers, and neither has much commercial reason to explain publicly how narrow the field is.

The regulators are working the other end, and not for the reason operators assume

In Great Britain, reverse withdrawals — the pending-withdrawal window in which a player could cancel a withdrawal and put the money back into play — were prohibited for all remote operators from 31 October 2021, under remote technical standard 14B: "Consumers must not be given the option to cancel their withdrawal request."

It is tempting for the industry to read that as a liquidity measure, and it is worth resisting. The Gambling Commission's rationale was consumer protection, on evidence from research, casework and people with lived experience that the function harms engaged and vulnerable gamblers. The Commission also noted the relatively low benefits the functionality brought to operators. And the causation runs the opposite way to the folk version: as the Commission put it, for reverse withdrawals to take place a delay in payment processing is required. The delay existed to enable the player-facing feature. Remove the feature and the delay loses its justification.

On cost, the Commission recorded that no significant increases in payment processing costs were reported during the earlier temporary ban, while being explicit that the feedback did not constitute a formal evaluation — and that most industry respondents did not share the cost concern, which had been raised by some operators.

Separately, since 31 January 2024, licence condition 5.1.2 requires remote casino, bingo and betting licensees — with narrow exclusions for ancillary, host and remote betting intermediary licences — to accept payment from customers using their Great Britain gambling facilities only through regulated payment services. That is a rule about money coming in rather than going out, but it belongs to the same narrowing of what is permitted.

The Netherlands went further on enforcement. On 13 August 2024 the Kansspelautoriteit reported action against thirteen operators over payout obstruction — nine ordered to stop and four more under suspicion — on two specific practices: minimum withdrawal amounts, and wagering requirements imposed on players' own deposited funds before payout. All thirteen were required to confirm in writing that they use neither. Note the scope: this is about playing through money the player deposited, not about wagering requirements on bonus funds, which is a different question.

And the position has moved since, in a direction that cuts against the tidy version of this argument. On 2 April 2026 the KSA returned to the subject, restating that credits must be paid without undue delay — but reporting that it had found no indications that operators were still applying payout conditions, and confirming that clearly disclosed payout fees remain permitted. The enforcement worked, or the practice was never as widespread as the 2024 action implied. Either way, "regulators are cracking down on slow payouts" is a weaker story in the Netherlands in 2026 than it was in 2024.

The fair summary is narrower than the one the industry tells itself. Regulatory tolerance for holding a payout for commercial reasons is going to zero. Holds for anti-money-laundering source-of-funds checks and for reasonable suspicion of fraud or false registration remain expressly permitted, and those are the two most common lawful holds in practice.

What we do not know, and what the industry keeps quoting anyway

Three things, and the first is the most repeated number in this subject.

There is no current evidence on payout speed and retention. The one methodologically sound dataset we could find is a TrueLayer and YouGov survey of more than three thousand players who gamble at least once every couple of months, across six European markets — the UK, Germany, France, Spain, Italy and the Nordics — with at least five hundred respondents per territory. Its fieldwork was conducted in October 2020. It found that 82% rated fast payout of winnings important when choosing a provider, that 55% were likely to switch to a service offering instant withdrawals, and that a good sign-up offer never made it into any country's top five priorities. Those are useful findings. They are also nearly six years old, vendor-sponsored, and predate the current payments landscape entirely. Figures circulating in 2026 claiming instant withdrawals lift retention by a specific percentage trace back to processor and affiliate marketing with no stated sample or methodology — in one case to a 35% figure attributed only to "recent studies" — and we are not going to repeat them.

Whether faster is straightforwardly better is genuinely contested. Every commercial instinct in this industry says reduce the time to funds. But a payout that lands in thirty seconds also removes a pause, and there is a real argument — made by people whose job is player protection rather than conversion — that some friction on the way back into play is protective. We do not have a settled view on where that line sits, and we are suspicious of anyone in payments who claims to.

The scheme material is second-hand and we cannot fix that. Stated above and repeated here because it is the single biggest caveat on this page: we could not find any of the payout-side gaming rules in the schemes' own published documents. Treat every specific claim in that section as a prompt to ask your sponsor, not as an answer.

What this means in practice

If you are an operator, the useful exercise is to separate your payout failures into the categories above, because they have nothing in common except the symptom. A scheme-rule failure will never be fixed by a retry or a different provider. A country-eligibility failure needs a different instrument entirely, not a better card integration. An issuer-side block is somebody else's customer relationship and is frequently working correctly. Most payout dashboards we have seen collapse all three into one decline rate, which is why the number never improves.

If you are a payment provider or a service provider selling into operators, the honest version of your product story probably involves saying which of these you can affect and which you cannot. The gap between what an integration can fix and what a sponsor relationship can fix is not obvious to the buyer, and being clear about it early is worth more than the deal it costs you.

We build on multiple licensed institutions offering the same service rather than on one, which helps with the sponsorship bottleneck and with concentration risk. It does not help at all with a scheme-level country restriction — no amount of redundancy makes a gaming OCT work where the network does not support one, and anyone who tells you otherwise is selling something. If you have data on payout speed and retention that is more recent than 2020, we would genuinely like to see it.