Rethinking the Casino Lobby with AI

Earlier this year, The Playa surveyed 151 senior iGaming decision-makers from operators and platform providers to examine how artificial intelligence (AI) and machine learning (ML) are being applied across the player journey, including acquisition, segmentation, predictive analytics, bonus management, lobby personalisation and customer support.

Rather than focusing solely on adoption rates, the research explored the outcomes operators are achieving and the operational approaches underpinning them. Lobby personalisation emerged as one of the most notable findings. While adoption remains relatively low, operators that have implemented AI-driven personalisation report consistently positive results.

Adoption Remains Limited

Just 17 per cent of operators surveyed have fully deployed AI or ML-powered lobby personalisation. A further 25 per cent are actively testing or planning deployments, while 58 per cent continue to operate static lobbies with little or no personalisation. The research also highlights significant differences in what operators mean by “personalisation”.

Among operators using AI-driven lobby management:

• 24 per cent personalise content at an individual player level.
• 35 per cent personalise by player segment.

Among operators not using AI:

• 48 per cent display the same fixed lobby to every player.
• Just five per cent provide any meaningful player-level personalisation.

The distinction lies in how game ordering decisions are made, whether based on a single static lobby, predefined business rules or models responding dynamically to individual player behaviour.

Refresh Frequency Matters

One of the clearest differences between AI adopters and non-adopters is how frequently lobbies are updated. Nearly half (47 per cent) of operators using AI-driven lobby personalisation refresh their game selections daily or more frequently, compared with just 27 per cent of operators without AI.

This is significant because refresh cadence directly affects relevance. A lobby updated every few weeks reflects historical averages, while a daily refresh responds to recent behaviour, such as the games played during a player’s latest sessions, preferred categories and changing engagement patterns. The survey suggests AI-powered operators are therefore better positioned to keep their lobby content timely and engaging.

Measurable Commercial Impact

Operators already measuring the performance of lobby personalisation report encouraging commercial outcomes. Among respondents tracking gross gaming revenue (GGR):

• 63 per cent reported GGR improvements of between three and nine per cent.

Among those measuring retention:

• Every respondent recorded monthly retention improvements of between two and six percentage points.

The strongest performance gains were reported among active players, with 46 per cent of operators identifying this group as delivering the greatest uplift. This is consistent with AI models having access to richer and more recent behavioural data for highly engaged customers. Benefits for new and returning players were also reported, although with less consistency due to more limited player histories.

Why Adoption Remains Low

Despite these results, adoption remains relatively modest. The survey suggests this is less about confidence in AI and more about implementation complexity.

Unlike internal automation projects, lobby personalisation requires integrated player data, trained machine learning models and deployment within the live product before measurable benefits emerge. As a result, implementation timelines are longer and operators often need more time before performance can be properly assessed.

The findings indicate this should not be mistaken for scepticism. Overall, 76 per cent of operators surveyed plan to expand their use of AI and machine learning. Lobby personalisation also tends to follow other AI initiatives. Many operators first focus on segmentation and predictive analytics, building the data foundations needed before introducing personalised lobby experiences.

What Successful Operators Have in Common

The survey identified several common characteristics among operators reporting the strongest results.

Create a unified player profile. Successful personalisation relies on bringing together session data, betting activity, game preferences and transactions into a single customer view.

Start early. Even limited behavioural data from a player’s first session can support meaningful recommendations, with models becoming more accurate as additional data is collected.

Model behaviour, not simply game popularity. The most effective systems analyse exploration patterns, session length, betting behaviour and preferred mechanics, while maintaining enough variety to encourage continued discovery.

Refresh frequently. Daily updates allow recommendations to reflect current player interests rather than historical averages.

Define success before launch. Operators reporting the clearest results established measurement frameworks in advance, identifying key metrics such as GGR, retention or turnover and maintaining control groups throughout testing.

A Significant Opportunity

The survey paints a consistent picture. A relatively small group of operators has already deployed AI-driven lobby personalisation and is reporting measurable improvements in revenue and retention, while the majority of the market continues to rely on static game lobbies. With only 17 per cent of operators having fully implemented the technology, lobby personalisation remains an emerging use case rather than an industry standard.

Whether it represents the right investment for an individual operator will depend on factors including data maturity, product priorities and technical infrastructure, but the findings suggest it remains one of the more underutilised applications of AI currently available to the industry.