Your iGaming platform determines how fast you can enter a new jurisdiction, how cleanly you can respond to a UKGC licence condition change, whether your data infrastructure can support the ML models your personalisation vendor is promising, and how much it costs you every time a regulator updates their technical standards.

That’s the reality. The frustration CTOs and Heads of Platform consistently describe is the same: the platform they selected three or four years ago, whether white-label or a legacy in-house build, now constrains more decisions than it enables. Revenue share models that seemed tolerable at £2m GGR look punitive at £20m. Monolithic backends that worked for a single brand in one jurisdiction become a liability when the business wants to launch a second brand or enter a new market. Compliance changes that should take weeks take months, because the architecture wasn’t designed for regulatory modularity.

The pattern is consistent. Operators who treat the platform as an IT cost centre make procurement decisions optimised for speed and initial price. Operators who treat it as a strategic asset make architectural decisions optimised for five-year TCO, regulatory adaptability, and competitive differentiation.

Every section that follows is written for the second group.

Platform pricing models in iGaming broadly fall into three categories. Each has costs that don’t appear in the initial proposal.

Revenue share (white-label and turnkey). Typically 10-20% of GGR, sometimes with a minimum monthly fee. The attraction is low upfront cost and fast time to market. The hidden costs: this model becomes a significant drain as GGR grows; you’re paying for the vendor’s entire platform overhead whether you use it or not; switching costs are high because you don’t own the player database architecture; and your product roadmap is the vendor’s roadmap, not yours. At £10m GGR, a 15% revenue share is £1.5m annually. Over a five-year licence period, that’s £7.5m before you account for any customisation fees.

Fixed licence fee. Annual or monthly licence for platform access, typically £300k-£800k per year for a full-featured enterprise platform, plus implementation costs of £200k-£1m+ depending on complexity. Lower long-term cost than revenue share for growing operators. Hidden costs: customisation beyond the standard configuration is usually billed at the vendor’s day rates (£800-£1,500/day); you’re still dependent on the vendor’s release cycle for core platform updates; and integration work with your preferred suppliers may not be on their priority list.

Bespoke build (with a specialist development partner). Higher upfront investment, typically £1m-£3m+ for a full platform build over 12-18 months, with ongoing maintenance and development costs of £300k-£600k annually for a dedicated engineering team. The return: you own the IP, control the roadmap, and can architect specifically for your regulatory and commercial requirements. The risk: you need a partner who has actually built and shipped platforms in regulated markets, not a generalist software consultancy learning iGaming on your budget.

The honest comparison: for operators below £5m GGR with a single brand in a single jurisdiction, a white-label or turnkey platform is usually the right answer. Speed to market matters more than architectural purity. For operators above £10m GGR, operating multiple brands or jurisdictions, or with a strategic need for product differentiation, the economics of bespoke development become compelling within two to three years.

The range between those thresholds is where the decision is genuinely difficult and context-dependent.

AI in iGaming is real, but the gap between vendor demos and production deployment is wider than most operators realise. A personalisation engine that works on clean, structured, well-labelled historical data in a sandbox is different from one that needs to process real-time behavioural events from a production platform with inconsistent data taxonomies across products.

The prerequisite for any AI/ML capability is the data infrastructure. Event streaming, a unified player data model, feature stores for ML pipelines, and clean integration between your platform’s transactional data and your analytics environment. If your current platform exports daily CSV dumps to an SFTP server, you’re not AI-ready regardless of what any vendor tells you.

Practical AI applications that are delivering value today in iGaming: real-time transaction monitoring for AML (pattern detection across deposits, withdrawals, and gameplay that’s faster and more accurate than rule-based systems); responsible gaming intervention triggers (behavioural markers that predict harm before traditional thresholds are breached); and dynamic bonus optimisation (adjusting offer parameters based on player segment response data). All of these require the data infrastructure first.

Omnichannel and headless architecture. A headless, API-first platform decouples the front-end presentation layer from the back-end services. This enables native mobile apps, responsive web, retail terminal integration, and emerging channels (wearables, voice) from a single platform backend. For operators with both online and retail estates (Rank Group’s Mecca and Grosvenor brands are a clear example), this architecture allows a unified player experience and shared wallet across channels without maintaining separate platform instances.

The headless approach also accelerates front-end iteration. Your marketing and UX teams can redesign the player experience, run A/B tests, and deploy front-end changes without touching the platform backend. In a market where mobile performance directly affects player retention, the ability to iterate on your app independently of your platform release cycle is a measurable competitive advantage.