Harvesting Information

Through data, giving customers what they want

By Don Hopper

Is it possible for a community bank to predict which products and services its customers are likely to need—and when? Using insights from predictive customer analytics to forecast behaviors and preferences, community banks are now better able to identify ready-to-buy customers.

Banks of all sizes have access to multiple, complex data sets—part of what has been termed “Big Data,” see related feature on page 54—that can provide timely, actionable information to help fine-tune their marketing and performance management strategies. The multichannel data available to them represents billions of customer transactions and millions of banking behavior observations distilled from account processing, debit and credit card processing, online and mobile banking, and electronic bill pay systems.

But it’s not enough to simply have this information in hand. Data analytics engines can turn these seemingly unrelated pieces of anonymous customer data into algorithms and predictive analytical models, which can help forecast customer needs and actions—much like a credit score provides insight into possible future behaviors.

It’s all about predicting what’s going to happen next with a customer. Making sense of complex data sets means mining demographics and basic online behavior tracking to build predictive models that provide timely, actionable insights. The result is a cost-effective, accessible approach to analytics that works for community banks of all sizes.

Recently, when one $2.4 billion-asset community bank wanted to grow its automobile loan business among its more than 55,000 customers, it used its in-house transactional data to identify those individuals most likely to use the bank’s installment loan products. Based on information gleaned from the statistical analysis, and criteria such as average banking balances, number of transactions, product preferences and usage patterns, and tenure at the bank, the bank launched a direct mail campaign targeted to individuals who were most likely to purchase autos in the near future.

The campaign’s response rate was more than three times what’s usually seen in the financial industry. Along with a threefold increase in the number of auto loan applications, the bank experienced a 20 percent rise in its auto loan approval rate due to the high number of qualified, creditworthy customers who received the mailing.

As another example of the power of data analytics and predictive modeling, one $2 billion-asset community bank was experiencing a troubling uptick in attrition after customers were charged for non-sufficient funds events. The bank used a predictive scores model to determine a customer’s likely response to an NSF fee and the individual’s potential profitability. Once the predictive modeling was completed, the bank had the information it needed to establish rules for imposing an NSF fee or waiving the charges in an effort to avoid customer defections.

Success stories like these point to the power of predictive customer analytics that can help community banks figure out the next, best action for each customer, including cross-selling and marketing opportunities. Accessing this type of information no longer requires a complex, in-house infrastructure, because technology providers can perform the analysis and deliver easy-to-use, highly targeted marketing lists.

Knowing what a customer wants and needs is important to building lasting, profitable banking relationships. Now technology is available to community banks to put previously untapped transaction data to use to drive revenue, compete more effectively and create a more valuable customer experience.