Like Fifth Third, Wachovia also sees the implementation of a predictive analytics approach to customer service as an ongoing process. According to Wachovia's Thorpe, however, the bank's new head of insight and innovation (which is part of the marketing division), Ramin Eivaz, who comes from the consumer packaged goods industry and oversees the statistics and modeling group, has challenged Thorpe and his team to look for best practices in analytics and retention from other industries.
"Our [retention] strategy now can best be described as having three tiers," Thorpe explains. "At the bottom is 'hindsight.' In the middle is 'insight.' At the top is 'foresight.' We're in the 'insight' phase right now. We have a very large data warehouse and many legacy systems that we're reorganizing to get the information more quickly and model more broadly."
The ultimate goal, Thorpe continues, is to achieve foresight so the bank can predict attrition before it happens. "We've always been good at attrition analysis," he says. "Now we're building life models -- how the customers evolve over the lifetime of their relationships with us -- to anticipate their needs and make product recommendations that are more relevant so we don't see the attrition." Thorpe notes that Wachovia takes a hybrid approach to building its analytics capabilities -- part in-house and part vendor-built -- but he declines to name the vendors with which the bank is working.
Toronto-based Royal Bank of Canada (US$605 billion in assets) is taking a more targeted approach to analytics, deploying the technology specifically in its lending business, another area, in addition to cards, where analytics are more widely deployed across the industry. According to Neil McLaughlin, the bank's vice president of personal lending, analyzing the right price points for its customers is one way to keep them from leaving.
"By using analytics tools to dovetail our pricing and personal marketing strategies, we were able to tackle problems we had around our relationship strategy," McLaughlin explains, adding that although RBC built its own predictive models, the bank is employing price optimization software from Nomis Solutions.
Bringing the Nomis solution on board involved a very large, cross-functional effort to bring together the data and feed it into the Nomis product, McLaughlin notes. "We rolled it out for originations initially," he says. "We tested it geographically and did a pre/post analysis to look at the spreads within those geographies. We saw positive results from selecting a better price for loans. We were able to go underneath and understand where the customer sees value in the product."
McLaughlin stresses that the bank gives employees discretion in terms of augmenting the suggested pricing models based on their own market knowledge. "Understanding the price of convenience and rate sensitivity to customers is important, especially as rates on personal loans increase," he says. "There are many benefits to pricing the loan right in the first place."