The licence is the start of the obligation
The most consistent failure in market entry is treating the grant of a licence as the finish line. Everything that has been built so far is a precondition; the obligation begins now, and in a newly regulated market the regulator is likely to look at new licensees early.
The first year has a characteristic shape: a launch period where acquisition dominates, an early compliance review, a first reporting cycle where the data problems surface, and a point six to twelve months in where the original business case meets reality. Planning for all four beats planning for the first.
Launch
Soft launch first where permitted. A limited release exposes payment failures, verification friction, reporting defects and support gaps at a volume the team can absorb. Some regulators require a supervised trial before full launch: Maryland, for example, requires each sports wagering licensee to complete a controlled demonstration, live wagering on two scheduled dates under the observation of regulator staff, before it is permitted to launch. Where no such step is required, doing it anyway is sound.
Instrument the funnel before you buy traffic. Registration completion, verification pass rate, first deposit conversion by payment method, and time to first bet. In a new market the verification pass rate is the number most likely to surprise you, because the local identity data sources behave differently from the ones you know.
Expect the payment mix to be wrong. The methods you predicted would dominate frequently do not. Watch deposit success rate by method in the first fortnight and be prepared to reprioritise the integration roadmap.
Do not launch marketing into an unresolved compliance question. Advertising rules in new markets are frequently stricter and more specific than operators expect, covering channel, timing, content, the use of sports figures, affiliate conduct and the wording of bonus offers. Ontario, for example, prohibited the use of active or retired athletes in igaming advertising from 28 February 2024, except to promote responsible gambling. An early advertising breach in a market where the regulator is forming its view of you is disproportionately expensive.
Own your affiliates from day one. In many frameworks the licensee is responsible for the conduct of its affiliates (in Britain, licensees are responsible for the actions of third parties they contract with and must be able to terminate an affiliate that breaches an advertising code), including their advertising, their claims and in some markets their targeting. A new market with an aggressive affiliate programme and no monitoring is a predictable enforcement case.
The first compliance review
Newly licensed operators are often reviewed early. What is examined is consistent enough to prepare for.
Whether the policies submitted with the application are actually operating, rather than filed. Whether responsible gambling tools function as certified. Whether AML thresholds and processes are those described. Whether reporting submissions are complete, on time and accurate. Whether key persons are in post and exercising the authority attributed to them. Whether marketing has complied. Whether customer complaints have been handled to the local standard, including escalation to the local dispute resolution body: in Britain, a customer whose complaint is not resolved within eight weeks must be able to refer the dispute to an ADR entity, free of charge.
A common finding, and one of the most damaging, is the gap between the application and the operation: an applicant described a control in detail, was licensed on that basis, and did not implement it. That finding is worse than an absent control, because it goes to candour.
Prepare by auditing yourself against your own application before the regulator does. Take the submitted policies, and for each material commitment, evidence that it is happening.
Reporting, and the data problems that surface
The first reporting cycles are where data model mismatches appear.
Typical problems: a figure the regulator defines differently from your platform, for example whether bonus stakes count in turnover or whether voided bets are included; a timing boundary where the regulator's day does not match yours; a customer categorisation that does not exist in your data; and reconciliation differences between what was reported and what the finance team files for duty.
Two disciplines prevent these becoming findings. Reconcile the regulatory submission to the financial records every period, and investigate differences rather than noting them. And where a definition is ambiguous, ask the regulator in writing and keep the answer, because a documented interpretation agreed in advance is a defence and an assumption is not.
Late or amended submissions are noticed. In Britain, for example, licensees must submit an accurate regulatory return within 28 days of the end of each quarter. A pattern of resubmissions in the first year shapes the supervisory relationship for years afterwards.
When the business case meets reality
Six to twelve months in, the numbers will differ from the model. What matters is which assumption broke, which is why the second lesson insisted on recording them separately.
Acquisition cost higher than modelled is a common outcome, particularly in markets that opened to many operators at once. The question is whether it is a launch-cohort effect that will normalise or a structural feature of a market with restricted advertising.
Revenue per customer lower than modelled usually means the product or payment mix is wrong, or that the local customer behaves differently from the comparable market used in the model.
Channelisation worse than expected means licensed operators are competing with an unlicensed sector that faces none of your costs. This is a market-level problem no single operator solves, and it is a legitimate reason to exit.
Regulatory change is among the risks most likely to invalidate a case. Tax increases, advertising restrictions and product limits in the first years of a framework are common. The Netherlands, whose legal online market opened on 1 October 2021, raised its gaming tax from 30.5% to 34.2% on 1 January 2025 and to 37.8% in 2026; Ontario, which opened in April 2022, introduced its athlete advertising ban after reviewing the market's first year. That is why the test set in advance was whether the case survives a materially worse tax rate and an advertising ban.
Compliance cost higher than modelled is very common, because the ongoing obligation is consistently underestimated at the application stage.
Set a review point in advance, with the criteria written down, and hold to it. The alternative is the pattern that recurs across the sector: a market that never quite works, a business case revised upward each year, and an eventual exit several years and a great deal of money later than the evidence supported.
Exit, and doing it properly
Exiting a licensed market is a regulated process, not a switch.
Expect obligations to notify the regulator in advance, to communicate with customers on a defined timeline, to return customer balances in full, to honour open bets or settle them fairly, to retain records for the statutory period after the relationship ends (in Britain, casino operators must keep customer due diligence records for five years after the business relationship ends under the Money Laundering Regulations 2017), to complete final reporting and duty payments, and to surrender the licence formally rather than letting it lapse.
The Gambling Commission's guidance on closing a licensed business shows what this means in practice for an online operator leaving Britain: contact the Commission's licensing team in advance with a closure plan covering customer numbers, customer funds held and open liabilities; give customers at least two weeks' notice before the website closes, with four weeks preferred; keep a route for refunds open for around four weeks after closure; keep access to ADR for complaints; and surrender the licence only once the Commission has assurance that bets have been settled and customers told.
The reputational dimension is real and underweighted. An operator that exits a market messily, leaves customers unable to withdraw, or surrenders under regulatory pressure carries that into every future application, because suitability assessment is retrospective and regulators talk to each other.
The corollary is worth stating plainly: plan the exit at entry. Knowing what an orderly withdrawal requires, and what it costs, is part of knowing whether the entry is worth making.
Doing the second market
The first market costs the most and teaches the most. The operators who scale well convert that learning into reusable structure.
What generalises: the localisation architecture, so language, currency, payment method and content are configuration rather than code; the compliance policy framework, with a core plus market-specific overlays rather than a fresh document per market; the reporting abstraction, so a new regulator interface is a mapping rather than a rebuild; the certification process and testing house relationship; and the market assessment template with its assumption log.
What does not generalise: the regulator relationship, the payment mix, the customer behaviour, the competitive position, and the specific rules. Assuming these transfer is the second-market error, and it is expensive in a different way from the first-market one.
A candid final point. Every mature operator has entered a market it should not have, usually because the size of the prize was assessed before the cost of entry, and usually because nobody wrote down in advance what would constitute failure. The discipline this course teaches is mostly about writing things down before you want a particular answer.