An insurance provider was spending significant resources in trying to retain unprofitable customers calling to cancel existing policies.
Lack of understanding of customer value resulted in a non-differentiated approach where every call was answered by a customer service representative.
Developed and deployed predictive models that routed low-value customers to an IVR and focused expensive customer service resources on valuable customers
- Internal and external data assets were combined to profile and value customers, and to evaluate alternative scenarios
- A predictive model was designed to predict retention probability on a cancellation request call, based on past data on call history and a host of known customer characteristics
- Predictive elements of the Retention model
Improved profitability by reducing customer service costs by 25% and still retaining over 90% of at-risk revenues.
- The analysis revealed significant effort was invested in unprofitable customers that were very unlikely to be retained
- Based on subsequent profit maximization analysis, 35% of lowest expected value calls were routed to an IVR system. Retained at-risk revenue diminished from 40% to only 37%, overall program profitability improved over 15%