Setting Up Customer Churn Prediction

Churn prediction analyses customer purchase patterns to identify who is likely to stop buying. It compares each customer's recent behaviour against their historical pattern. A customer who bought weekly for 6 months and has not purchased in 3 weeks scores as high risk. The system ranks all customers by churn probability so you can prioritise retention efforts.

AskBiz uses purchase frequency, recency, monetary value, and category breadth to score each customer. The model adapts to your business — it learns what normal looks like for your customers and flags deviations. A hardware store where customers buy monthly has different thresholds than a coffee shop where daily visits are normal. No manual configuration needed.

Go to Intelligence > Churn. The dashboard shows at-risk customers ranked by probability. Each entry shows the customer name, last purchase date, predicted churn score (0-100), estimated revenue at risk, and the suggested action — send a message, offer a discount, or make a personal call. The total revenue at risk across all at-risk customers is shown at the top.

For each at-risk customer, choose an action. If you have WhatsApp messaging enabled, send a personalised message directly from the dashboard. If you run a loyalty programme, offer bonus points. Track which actions work — AskBiz measures whether re-engaged customers return, helping you refine your retention strategy over time.