Operator Guide · 2026

Deciding how to launch a sportsbook is really three decisions pretending to be one: who owns the odds, who owns the platform, and who carries the risk when a market is priced wrong. Get those separated and the build-versus-buy argument resolves itself quickly. This guide walks through the three routes, what each actually costs you in attention, and the questions that decide it.

The three things you are actually choosing

Odds and trading. Someone has to price thousands of markets, move them as money arrives, and decide when to cut a limit. This is a specialist function with real risk attached, and it is the part almost every operator rents rather than builds. Even large brands run supplied odds with their own traders layered on top.

The platform. Account, wallet, bonusing, reporting, responsible gaming, the bet slip itself. This is the part where «build» is genuinely tempting, because it is the part players touch.

Liability. Who eats a mispriced market or a settlement error. In white label the licence holder does; in turnkey and own-platform models it is yours. That single line changes the risk profile of the whole business, and it is worth being explicit about before comparing prices.

Three routes side by side

Why duty rates changed the answer

The build case always rested on the same argument: high fixed cost now, better unit economics later. That argument survives only if the margin at the end is big enough to repay it. When the UK moved remote gaming duty from 21% to 40% in April 2026, and set remote betting duty to rise from 15% to 25% in 2027, it took a large bite out of exactly the margin the build case was borrowing against.

The visible result is operators who could plausibly have built choosing not to. A new UK licensee taking sportsbook, racing, virtuals, casino, CRM and payments as a supplied stack is not a small operator cutting corners — it is a rational reading of what a point of GGR is now worth.

What actually decides it

1. How many markets you intend to run

One market rarely justifies a build. Four or five, each with its own compliance profile, starts to — because you are then amortising the platform across several licences and the per-market configuration cost dominates.

2. Whether your edge is the product or the audience

If you win because you already have an audience — media brand, affiliate network, retail estate — buy the platform and spend everything on converting that audience. If you win because your product does something nobody else’s does, you eventually need to own the thing you are differentiating.

3. What happens on day 400

Launch is the easy part to model. Ask what adding a market, a payment method or a new vertical costs in month fourteen. Supplied stacks that make launch trivial and change expensive are a common and expensive trap.

The part most sportsbook plans get wrong

A sportsbook without a casino behind it leaves most of its revenue on the table. A football match is ninety minutes of which a handful are eventful, and a bettor with nothing to do between markets closes the app. Casino content is what fills that gap — and it only works if it sits on the same wallet, with one balance and one bonus system.

This is why the wallet architecture question belongs in the sportsbook decision rather than after it. If casino arrives later as a separate integration with its own balance, the cross-sell that justified having both never materialises. Our guide to wallet models covers why, and the casino API page describes carrying sport and casino on one connection.

A sensible sequence

Decide the licence model first, because it determines who carries liability and that is the hardest thing to change later. Pick odds supply second — this is a specialist choice and worth taking seriously. Choose the platform third, judging it on the cost of change rather than the cost of launch. Wire casino in from the start on the same wallet, even if you open with a small lobby.

If you already run a casino and are adding betting, the order reverses: your wallet and player base exist, so the question is only which sportsbook software plugs into them with the least disruption.

FAQ