A large regional retail bank sought to better target cross-sell marketing to its existing customer base. Insights about the drivers of cross-sell opportunities were critical to shaping the specific offer and messaging. Chris Cahill led the effort to devise and deploy a suite of predictive models that drove specific, successful cross sell activity.
Approach
- Analyzed an extensive sample of customer deposit and loan product ownership, balances, and transactions to provide a comprehensive view of the key customer behaviors
- Identified the customer product affinities and migration patterns to determine the most common cross-sell product combinations
- Developed a suite of eleven cross-sell models to predict the likelihood of the customer taking relevant deposit and loan products
- Produced a specific product recommendation based on model scores, product profitability, and other business rules
Results
The bank incorporated these cross-sell scores and recommendations into a year-long direct mail campaign:
- In each quarter, customers received better timed, more relevant cross-sell offers
- The design of the program allowed for significant cost-efficiencies in the direct mail production and mailing costs
- The program produced response rates nearly double earlier mail efforts and significantly higher total balances
