Build skills-based multiplayer mobile gaming applications with cash prizes
One Goal.
Improve price sensitivity to maximize in-game store monetization.
Sometimes the best way to solve a problem is to redefine what’s being asked. For us and this example, this meant interviewing relevant prominent researchers and companies, for a deep understanding of what’s ‘in the box’.
Then, based on our findings and existing gaming industry knowledge, we could deliver a radical ‘out of the box’ solution.
We combined advanced large language models (LLMs) with Multi-Armed Bandit strategies. The real-time algorithms could then be deployed to optimize in-game store configurations based on individual customers.
This allowed us to overcome the challenges listed above, removing Game Story’s previous reliance on manual and tedious programming to optimize four key areas of value.
Intelligent optimization
LLMab uses AI to continuously analyze player behavior, spending patterns, and game-specific metrics.
This dynamic approach allows Game Story to more accurately refine store offerings based on clean, objective data.
Personalization at scale
The new system automatically creates and evaluates multiple store configurations for different player segments.
For Game Story customers, this brings more personalized experiences that lead to higher engagement and revenue.
Long-term revenue growth
Unlike traditional A/B testing, LLMab intelligently optimizes offerings in real-time rather than simply offering fixed variations.
Game Story gains greater long-term player value, balancing their immediate sales with higher player retention and engagement.
Adaptive learning
The system uses a multi-armed bandit approach, adapting to changing player behaviors over time.
As a result, Game Story can explore new configurations while maximizing the benefits from what’s already driving positive results.
Revenue boost
Achieved within this timeframe using a ‘minimum request, maximum impact’ approach
We created a best-of-both-worlds approach using the power of LLMs and Multi-Armed Bandit frameworks for Game Story, a mobile gaming company. This included using AI for continuous monitoring and optimization of in-store offerings, improving personalization of experiences, balancing immediate revenue needs with longer-term retention, and deploying adaptive learning for customized configurations.
The resulting solution, LLM Multi-Armed Bandit (LLMab), led to more nuanced analysis and human-like recommendations, a more holistic approach around player demographics and financial balance plus in-game events, efficiency with automation, and data-enriched insights. LLMab was built to interact with standard database systems and offer a well-defined API for integration, ensuring long-term scalability and sustainability. Within 1 month this achieved a 2.5% revenue boost for Game Story.
