One of those carriers is Brussels-based P&C insurer Corona Direct (US$65 million in revenue), a subsidiary of DEXIA (Brussels). To acquire more clients from direct mail, the carrier implemented customer analytics. "We acquired a lot of customers by using direct mail, but it was an expensive way to acquire clients," says Philippe Neyt, commercial director for the carrier. "We were sending our 4 million letters at a price of half a euro each [approximately US$0.65]."
During the mid-90s through the early part of the current decade, the carrier structured its data into a data warehouse, relates Neyt. "After we structured our data, we asked ourselves how we were going to make money with it," he says.
According to Neyt, Corona Direct decided that it would use its data to better target clients on its prospect direct mail list. The carrier purchased SPSS' (Chicago) Predictive Analytics with the specific goal of decreasing its costs by 20 percent without losing more than 10 percent of sales. "We had a lot of client history in the data warehouse, and the software could take all of the fields together to tell us our best prospects to mail to depending on gender, age, type of client and location," describes Neyt.
Before using the tool, in 2000 Corona Direct had mailed more than 1.6 million letters and received 7,000 responses for a quote. Of those respondents, the carrier gained 2,626 new policyholders, according to Neyt. However, with the use of SPSS Predictive Analytics in 2003, the carrier conducted 1.1 million mail drops, received responses for quotes from 6,300 potential clients and gained 2,677 new policyholders. "We sold the same number of contracts with 25 percent less costs," says Neyt. "We knew that with this we could reach ROI within a year."
Because the model incorporates each year's data, it improves with time and the carrier sees an increase in response rates every year, Neyt notes. Currently, the carrier is considering integrating the SPSS product into its call center to help with cross-selling efforts by scoring the likelihood of products' appeal to specific customers. "We hope to be ready to test it in 2007, depending on the capability of our sales team," says Neyt.






