From One-Size-Fits-All to 1:1: How a Tier 3 Operator Grew ARPU 31.7% with AI-Powered Lobby Personalization
Personalization is the new battleground for player attention
iGaming operators are competing for the same players across the same game catalogues. The product difference increasingly comes down to how well a platform understands each player and adapts to their preferences. Operators that treat the lobby as a static storefront are leaving engagement and lifetime value on the table.
The "Top", "Popular" or just a "Casino" block sits at the top of the lobby, one of the highest-traffic spots on the platform. Most operators fill it with popular titles or manually curated picks that change periodically. It works, but it treats every player the same.
The starting point: a strong team that had hit a ceiling
A Tier 3 mono-geo operator with over three years of market presence had built a large, loyal player base. These weren't new sign-ups exploring the platform, these were habitual players with established routines, familiar with the lobby layout, the game catalogue, and how the platform works.
The operator had a "Casino" category at the top of the lobby, one of the most visible spots on the platform. The same static list of games went out to every player. No segmentation, no adjustment by player type or behavior. Everyone saw the same titles regardless of what they actually played.
They came to The Playa not to fix something broken, but to test a hypothesis: that even conservative, long-tenured players would respond to a lobby that actually adapts to them individually. That personalization could give these players something fresh, without disrupting what already worked.
The solution: AI that learns what each player wants
The operator plugged in The Playa's AI-powered lobby personalization to replace the static list inside the existing "Casino" category with dynamic, player-level recommendations. Same block, same position, same design, the only thing that changed was the logic deciding which games each player sees.
The Playa's model analyzes each player's gaming activity, session patterns, game preferences, betting dynamics, interaction history and generates a personalized game list in real time. Rather than surfacing what's popular platform-wide, it surfaces what a specific player is most likely to engage with.
The model updates as player behavior changes. No manual refreshes needed.
From pilot to production: measuring long-term impact
The operator first ran a short A/B test that confirmed positive results. Based on that, the team decided to roll out The Playa's personalization to the full player base while keeping a balanced control group to continuously measure the impact in production.
Over 14 weeks of live operation, the two groups were compared:
- Control group: The "Casino" block curated manually by the operator's team.
- Personalization group: The same block powered by The Playa's AI recommendations, personalized per player.
Same block design, same lobby position, same player base. The only variable was the recommendation logic, so the results reflect the real, sustained effect of personalization running in production, not a short-term test spike.
The results
After 14 weeks, the AI-personalized group beat manual curation across every metric that matters:
Players stayed active longer. Days per user increased by 5.8%, a direct sign that they were finding more relevant content in the lobby. That deeper engagement translated into stronger LTV across the board: ARPU grew by 31.7%, average turnover per group by 32.2%, average deposit by 29.4%, and average bet by 10.9%. All measured over 14 weeks in production with balanced groups.
Addressing the skeptics
"It's just a reshuffled popular games list."
A popularity-based list shows the same games to everyone. The Playa's model builds an individual ranking for each player based on his/her gaming, payment and bonus activity. Two players opening the lobby at the same time see entirely different recommendations. The +31.7% ARPU lift comes from relevance, not reordering.
"Our team already knows our players well enough."
No team can manually personalize recommendations for thousands of concurrent players across hundreds of games. Manual curation works at the segment level. The 32.2% turnover gap between AI and manual isn't marginal, it reflects a difference in approach, not effort.
"Changing the lobby is risky, players might not like it."
With a conservative, long-tenured player base, that's a fair concern. These players have routines, and there's real skepticism about whether they'll respond well to changes. That's exactly why we started with an A/B test because we wanted to try different approaches and find what resonates with a specific operator's audience, without risking the entire base. In this case, the test showed that players not only accepted the personalized lobby but came back more often, days per user grew by 5.8%. But the broader point isn't about any single metric. It's that you test, learn what works for your players, and scale from there.
What this means for operators
When players consistently find games that match their preferences, they play longer, return more often, and generate more value over their lifetime. The "Casino" block can drive that, without redesigning the lobby, buying more traffic, or adding headcount.
The more players interact with the platform, the sharper the model gets. It requires no ongoing manual work and starts producing uplift immediately after launch.
For operators still running manual or popularity-based recommendations: the 31.7% ARPU gap measured in this test exists in your data too. The question is whether you're capturing it.
Want to see what AI lobby personalization could do with your player base?
Run a pilot in 2-4 weeks and measure the uplift on your own platform.
