Random Forest Machine Learning Model to Predict Customer Cancellations

JKA created a Random Forest machine learning model to predict customer cancellations for a client in the retail space. The model was populated with only six months of data and successfully predicted with 95% accuracy the next month’s customer cancellations. When applied to 10,000 customer data, the churn rate was 7% (~700 customers) and the model correctly identified the 600+ customers who will cancel.

Our experience and knowledge in Predictive Analytics are applicable to other areas such as customer targeting, customer churn prevention, risk mitigation, forecasting and financial modeling.