Gamelight's AI Driven Algorithm

In a strategic partnership, Gamelight managed to lead an impressive campaign utilising their AI Algorithm. To fulfil the partner's objectives, Gamelight employed their AI algorithm to generate user profiles, identifying intersections between highly engaged existing users and potential new ones. This strategic approach aimed to introduce the game to users with a high likelihood of sustained engagement. Leveraging exclusive access to valuable datasets, Gamelight continually refined its AI algorithm through advanced machine learning techniques.



The AI algorithm ensures effective targeting of users likely to download the title. It operates with a focus on optimizing key metrics such as ROAS growth, ARPU and User activation, striving for peak performance.



In a large campaign with mobile game developer SuperPlay, 2 million users' data was analysed across 11 geos with an allocated budget of $500,000 daily ad spend. Under these conditions, Gamelight’s platform achieved impressive results:



Gamelight showed a 38% higher D30 ARPU than all other media sources.

Superplay has reached 100%+ ROAS profitability on fully matured cohorts.

Scaling Gamelight increased the total average ARPU on app stores by 15% across the whole game.

improved store rankings by 48 positions in the Free Games US store charts.



Impressive long-term retention and ROAS growth were observed, with a D30 of +212%, a D90 of +387%, and +508% for D180.



During special events Halloween Special, X-Mas Boost and Valentine’s Day Special. Superplay achieved an impressive +32% boost in revenues and +24% user activation.



The numbers showcase the algorithm's effectiveness in achieving outstanding outcomes. Manual optimisation could take weeks without showing any results, but the self-optimisation of the algorithm is created to be able to deliver the best results and game recommendations to the users, without bias. Making it possible to deliver impressive numbers while creating more enjoyable and meaningful experiences for users.