Scope
- Churn is a huge problem for the client. Example – Two-cycle churn is between 60 & 75%:
- Cannot optimize wholesale cost as it is too risky to place on bundles
- Loss of revenue
- Reduced predictability of business
- Churn problem compounded by pre-paid model as there is no way to contact the customer after he has churned and attempt a win back
Solution
Adopted a phased-approach for delivery of churn management solution
- Phase 1 involved data collection, aggregation & analysis, development of statistical models to predict churn & manual generations of campaign lists
- Phase 2 was focused on automation
- Phase 3 is about global roll out (work-in-progress)
Impact
- Added competitive advantage by predicting the possibility of churn in advance; possible to take measures to prevent churn
- Churn prediction & management will also help ensure effective spend of marketing in the context of base management