A client has extreme ebbs and flows to their user base depending on the season; certain events throughout the year result insubstantial changes to the volume of active users month-to-month. The challenge was to analyze these patterns and create a solution for reducing the seasonal drops.

Our team analyzed the client’s user base, and was able to identify the key metrics that allowed us to predict which users were at a high risk of churn. Armed with this research, we designed, developed and put forward a predictive model that flagged these customers before they actually churned.