Evergent achieves 94% churn prediction accuracy for gaming giant

Faced with growing acquisition costs and a noticeable increase in churn of heavy gaming users, the unidentified gaming company partnered with Evergent to better predict and prevent player churn. Churn was defined as 14 consecutive days of inactivity.

Evergent’s proprietary AI model was deployed in eight weeks and began generating insights by analysing a dataset of over 2 million users, evaluating behavioural signals including login frequency, progression metrics, and transactional activity.

Key outcomes included 94% churn prediction accuracy using behavioural modelling and limited user data, and segmentation by risk and value to enable targeted retention strategies.

Vijay Sajja, founder and CEO of Evergent, said: “By achieving 94% churn prediction accuracy in this gaming sector, we’ve shown that businesses beyond our traditional sectors can benefit enormously from our proven ability to turn behavioural data into actionable, revenue-driving insights—with no heavy lifting required.”

Next steps for the companies include model retraining with expanded datasets, integration of campaign feedback for continuous improvement, and deployment of personalised re-engagement campaigns using real-time churn scores, alongside marketing automation platform integration to enable timely, targeted offers to at-risk user segments and help minimise voluntary churn.