Subscription service
Churn prediction for a subscription service
Each week the CRM receives a list of customers likely to cancel, with the likely reason.
The challenge
The retention team contacts customers after they cancel, when winning them back is hardest. Usage and payment data exist, but nobody turns them into early warnings.
The solution
- 01
Combine usage, payments and support history into signals for each customer.
- 02
Score the risk that each customer cancels, along with the main reasons.
- 03
Send the ranked list to the CRM, so the team can act before renewal.
Related examples
Demand forecasting for a retail chain
a forecast for each product in each store
Every week, each store gets a forecast per product that allows for promotions, holidays and the season.
The timeline depends on
how complete the history is, the number of products and stores, and the ordering system
Support assistant for an online store
answers drawn only from approved store content
The assistant answers from the store's approved policies and checks order status. Conversations it cannot resolve go to a person.
The timeline depends on
the state of the knowledge base, the number of integrations and the rules for verifying customers
Document processing for a logistics operator
documents checked before they reach the ERP
Waybills and invoices arriving by email and from scanners are read, checked and loaded into the ERP.
The timeline depends on
the number and variety of forms, the quality of the scans, and what the ERP allows