Three delivery models dominate the market. Each comes with trade-offs that are genuinely context-dependent, so anyone telling you one is universally better than the others is selling something.

Turnkey and white-label solutions get you to market fast. A white-label provider supplies the platform, the licence (often operated under their umbrella), the games, payments, and sometimes even the CRM. Your role is brand, marketing, and player acquisition. Time-to-market can be weeks. Initial capex is low.

The costs show up later. Revenue share models (typically 30% to 50% of GGR, sometimes higher) compound as you scale. Customisation is constrained to what the provider’s configuration layer allows. Your front-end differentiation is limited. And you’re entirely dependent on the vendor’s product roadmap. When UKGC tightens affordability check requirements, you wait for the provider to implement the change. If they prioritise differently, you have limited recourse.

Migration off a white-label platform while maintaining live operations is one of the most painful technical exercises in iGaming. We’ve seen operators trapped for years because the switching cost (data migration, regulatory re-certification, player communication, commercial renegotiation with payment providers and game studios) exceeds the annual cost of staying, even when staying is demonstrably suboptimal.

Turnkey solutions sit in a middle ground. You get a pre-integrated platform stack but hold your own licence and have more control over operations. The technical debt risk is lower than white-label, but you’re still constrained by the vendor’s architectural decisions. If their PAM uses a monolithic backend, you inherit that limitation regardless of what your own engineering team would choose.

API-first and headless architectures flip the model. You own the front-end entirely. The platform exposes its functionality through well-documented APIs. You can integrate best-of-breed components for different parts of the stack: one vendor for PAM, another aggregator for game content, a third for payments. The upfront investment in integration engineering is higher. The operational complexity is greater. But you own the player experience, you can swap components without rebuilding everything, and you’re not locked into a single vendor’s commercial terms or release schedule.

The honest framing: API-first is the right architecture for operators with sufficient engineering capability and scale to justify the investment. For a new entrant targeting a single market with limited technical resource, a turnkey solution with a clear contractual exit path might be the pragmatic choice. The mistake is treating the initial deployment model as permanent rather than planning the migration path from day one.

Game count is a vanity metric. An operator with 3,000 well-curated titles, strong lobby merchandising, and fast game loads will outperform one with 12,000 games that players never find.

When evaluating game aggregators and content portfolios, focus on these dimensions:

Portfolio quality and exclusivity. Which tier-one studios are available through the aggregator? Do they have direct integration with Evolution for live dealer, or are they routing through a sub-aggregator (adding latency and cost)? Can you get early access to new releases from studios like Pragmatic Play, or are you in the standard release queue? Exclusive content windows, even short ones, have measurable impact on player engagement.

RGS performance. The Remote Game Server is where the game logic executes. Latency matters. A 200ms round-trip time on a slot spin might seem acceptable until you compare it with a competitor offering 80ms. Players notice. Measure P95 and P99 latency, not averages. Ask about the aggregator’s uptime per studio. Some aggregators mask provider-side instability poorly, resulting in broken game sessions that your support team has to handle.

RNG certification. Every game in a regulated market requires certified Random Number Generation. This is non-negotiable. But the certification body matters. UKGC-licensed operators need games tested by approved test houses. Verify that the aggregator’s content is certified for every jurisdiction you plan to operate in. Content that’s MGA-certified but not UKGC-approved is useless if you’re launching in the UK.

Commercial terms. Aggregator revenue share typically ranges from 5% to 15% of the studio’s GGR share, depending on volume and exclusivity. Some aggregators charge flat monthly fees plus a lower percentage. Understand the full cost stack: studio GGR share plus aggregator margin plus any platform fees. On a popular title with a 15% studio share, an aggregator adding 10% on top means 25% of GGR leaving your P&L before you’ve paid for anything else.

Technical integration. How does the aggregator handle wallet calls? Synchronous or asynchronous? What happens when a game round completes but the wallet callback fails? Idempotency handling, reconciliation tooling, and the quality of their back-office reporting all matter for operational efficiency.

Vendor pricing in iGaming is deliberately complex. The headline number rarely reflects the actual cost of operation.

Revenue share models (common in white-label and some turnkey arrangements) typically range from 30% to 50% of GGR for white-label, dropping to 10% to 25% for turnkey platforms with less bundled service. These percentages look manageable at low volumes but become eye-watering as you scale. An operator generating €5M monthly GGR on a 35% revenue share is paying €1.75M per month for platform services. At that point, the capex for a custom build pays for itself within a year.

Fixed monthly licensing fees range from €15,000 to €100,000+ depending on the platform’s scope and the operator’s volume tier. These are more predictable but often come with minimum commitment periods (two to three years) and escalation clauses.

Hidden costs are where the real damage occurs. Setup fees (€50,000 to €500,000). Customisation charges billed at day rates 30% to 50% above market. Per-player or per-transaction fees that only become material at scale. Charges for additional jurisdictions. Penalty fees for falling below minimum GGR thresholds.

To calculate real TCO over a three to five year horizon, model the following: licensing or revenue share at projected volume growth, integration costs for payments/KYC/games, customisation and ongoing development, compliance change requests (assume at least two major regulatory changes per year requiring platform updates), hosting and infrastructure (if not included), and the cost of the engineering team required to manage the vendor relationship. Include a realistic estimate for the cost of eventual migration, because the probability of remaining on the same platform for more than five years is lower than most operators assume at signing.