Feb 23, 2001
SUMMARY: ESL Federal Credit Union, a full-service credit union serving Eastman Kodak employees and their families, represents $1.8 billion in assets and has 165,000 members, making it one of the largest credit unions in the nation. However, ESL was having difficulty accurately identifying likeliest prospects for products such as Internet banking. So in last year, the credit union decided it was time to begin leveraging more customer data in its marketing efforts to find out which online ad click throughs had the highest propensity to actually open an online account. || |
ESL Federal Credit Union, a full-service credit union serving Eastman Kodak employees and their families, represents $1.8 billion in assets and has 165,000 members, making it one of the largest credit unions in the nation. Historically, ESL used its in-house Marketing Customer Information File to define a target audience based on age, income, psychographic coding, and approximately 20 other customer-specific data elements.
However, ESL was having difficulty accurately identifying likeliest prospects for products such as Internet banking. So in January 2000, the credit union decided it was time to begin leveraging more customer data in its marketing efforts. “The goal for us was to find out which of our customers had the highest propensity to purchase which product,” explains Roger Rassman, Vice President, Marketing Manager. “With that intelligence, we would be better able to target our product and service offerings to those most likely to qualify and purchase.”
With the help of an eFS solutions enabler, Digital Insight, ESL set out to uncover detailed customer information by applying advanced targeted marketing techniques. The project consisted of collecting and massaging data that consisted of more than 1,000 attributes per household.
In modeling that data, ESL’s portfolio was scored at the customer-level, and the entities were rank-ordered into twenty segments —- or “twentiles” —- with a segment number of 20 indicating the highest propensity to respond/purchase, and 1 indicating the lowest propensity. The scores and the segment numbers from this model were then appended to ESL’s database, so the credit union could use these scores and segments in its direct mail and telemarketing programs for Internet banking, auto loans and home equity loans.
ESL also employed the results of propensity modeling into its Internet channel in an effort to speed and streamline the loan application and approval process and improve the online customer experience for the institution’s more than 23,000 Internet banking users.
Soon after ESL began using more intricate propensity models based on some 1,100 data elements per household, the credit union's average campaign response rates increased by 50%.
In addition, because of propensity modeling, each customer interaction became more profitable. Rassman says, “We now know from the start whether or not a customer is high propensity —- whether or not they are likely to buy, are credit worthy, and so on —- which allows us to close the loan deal within hours, not the days it used to take. We are currently averaging about 400-500 loan applications a month this way.”
ESL also counts as an intangible benefit the fact that propensity models have helped prevent customer attrition by saving ESL from sending out inappropriate marketing messages to customers.
NEXT: ESL now wants to take the targeted approach to the front lines —- to provide front-line staff with customer intelligence instantly so that they know when a member standing before them (literally, online at a particular given moment) has a very high propensity, for example, to take out a home equity loan.