Launching with AI from Day One: How a New Tier 2 Casino Gained 29% More Turnover by Personalizing the Lobby Before It Had History
Most casinos personalize later. This one started on launch day.
When launching a new iGaming product, personalization usually comes later. The typical sequence is: go live, acquire players, build up behavioral data over time, and then layer in recommendations once there's enough history to work with.
There's nothing wrong with that approach. It's practical. But it does mean that during the early months, when the product is forming its first impression with every new player, the lobby experience is the same for everyone.
A New Casino, a New Approach
A Tier 2 operator was launching a new casino brand in a competitive market. Rather than following the standard playbook (go live with a static lobby and optimize later), the team decided to ship personalization from the start.
The operator integrated The Playa's AI-powered game recommendations in the first 6 months of operation. From the first session of the first player, the "Recommended Games" block was powered by AI, not by a manually curated list.
The bet was straightforward: if new players get a lobby that adapts to them immediately, they'll find games they enjoy faster, play longer sessions, and come back more often. Rather than launching generic and adding personalization months down the road, the operator chose to have it working from the first days.
How it works with zero player history
The obvious question: how do you personalize players you've never seen before?
The Playa's model doesn't require months of behavioral data to start working. In the first sessions, it uses early signals, initial game choices, session length, betting patterns, deposit behavior, to begin building a player profile. As more data comes in, the recommendations sharpen.
For a new casino, this means the model and the player base grow together. Each player's lobby evolves from their very first interaction, and the system gets smarter across the entire platform as more players join.
The design and layout of the lobby stayed standard, no special UI, no extra complexity. The only difference from a non-personalized launch was what appeared inside the "Recommended Games" block.
The A/B test: 6 weeks, even split
To measure the impact, the operator ran a 6-week A/B test with a clean 50/50 split across the player base:
- Control group: Standard lobby with a default game selection, the same for all players.
- Test group: The Playa's AI recommendations, personalized per player from their first session.
Both groups saw the same lobby design, same game catalogue, same promotions. The A/B test was conducted only among active players in both groups. The only variable was the recommendation logic in the "Recommended Games" block.
The results
After 6 weeks, the personalized group showed clear gains across engagement and revenue metrics:
Players in the personalized group explored more of the catalogue, games per user increased by 7.2%, meaning they weren't just playing more, they were discovering more titles that matched their preferences. That broader engagement translated into 29.1% higher turnover per user and 5.7% higher ADPU.
The retention curve tells its own story. Both groups tracked closely through the first three weeks. By week 4, the personalized group pulled ahead. Retention grew by +1.1%pt compared to control, and the saved the gap +0.9%pt by week 5. For a brand-new casino with no prior player relationships, that separation in retention is significant: it compounds into meaningfully different LTV over the player lifecycle.
But what about...
"You can't personalize without data, the model has nothing to work with on day one."
The model needs around 2-3 months of a casino's operational history to produce solid recommendations, and it continues to improve as more player data comes in. This operator had already been running long enough for The Playa to train the model before the A/B test began. Personalization wasn't switched on blindly from zero. It was deployed once the model had enough history to work with. The 29.1% turnover gap after 6 weeks reflects a model that was ready, not one that was guessing.
New players don't know what they want yet, just show them popular games."
Popular games are popular on average, but individual players have preferences from their very first session. The +7.2% increase in games per user means players explored a wider variety of titles when recommendations were personalized. Instead of everyone clustering around the same popular games, players discovered titles that matched their individual taste. Personalization doesn't narrow the catalogue. It opens it up.
"Retention at a new casino is all about bonuses and acquisition quality, not the lobby."
The only difference during the test was the lobby logic. The retention gap at week 4-5 is entirely attributable to the player's product experience, specifically, whether the lobby adapted to them or not.
What this means for operators launching new products
Most operators treat personalization as a phase-two investment, something you add once you have enough data and enough players. This case shows it can be a first days advantage.
A new casino that personalizes from the start doesn't just perform better on metrics. It builds player habits around a personalized experience from the beginning, rather than training players in a generic lobby and then trying to shift their behavior later.
For operators planning new brand launches or market entries: the cost of launching without personalization isn't zero. It's the engagement and revenue gap that accumulates from the first days of operation until you eventually add it.
Launching a new casino or entering a new market?
Test AI lobby personalization as part of your launch stack, pilot in 2–4 weeks.
