Identify potential lost customers on time to allow a company take actions to avoid loosing customers. Use historical data and customer’s characteristics to learn who, why and when a customer will churn.


The team built a classification model to predict the probability of losing a customer. The trained model considers trends and statistics of the historical revenue as well as customer’s demographics like country and industry among others.


  • Technology

Tools Used

  • Data Science Experience (DSX)

Data Science Techniques

  • Binary Classification


This company was able to identify the customers that are at a high risk of being lost with 6 months of anticipation. In addition, the model outputs the key factors explaining why each customer will potentially churn. As a result, the company has 6 months to take action on high risk customers and re-engage with them.

Need help with a similar case?

Contact us to find out how we can help!