Introduction to Customer Churn Prediction

Customer churn prediction leverages historical sales trends to determine the likelihood that customers will stop placing new orders.  The model updates daily, generating churn risk scores for each customer to help businesses identify at-risk customers and take action to reduce churn.  Combine these attrition insights with customer ABCD scores to focus efforts where they make the most impact: retaining customers who are highly profitable to the business.

• Minimal Risk: 1-20% chance of churning

• Low Risk: 21-40% chance of churning

• Moderate Risk: 41-60% chance of churning

• Elevated Risk: 61-80% chance of churning

• High Risk: 81-99% chance of churning

Churn risk is calculated using RFM statistical analysis based on each customer's sales history, taking into account how recently they placed their last order, how frequently they typically place orders, and the monetary value trends for their orders.  The customer's overall invoice count, sales totals, and margin totals are also taken into consideration, among other factors. 

The model defaults to using the last year of data from the most recent invoice, and numbers update dynamically as customers churn or shift between risk categories.  Customers must have at least three historical invoices for accurate predictions, otherwise they get a simplified risk calculation.

For more information on churn risk calculations and where to see churn risk in the Profit Max Platform, refer to the in depth Customer Churn Prediction documentation.