Segment Twice, Mail Once

“TONIGHT SHOW” host Jay Leno and Hell’s Angel maximum leader Sonny Barger are both Harley-Davidson enthusiasts. But a marketer using the same pitch to sell motorcycles to both likely would risk being publicly mocked by one — or stomped by the other. Without additional information, recency, frequency and monetary analysis won’t reveal key differences between customers.

Last January, Hibernia National Bank put this fanciful example to a real-world test. Using past purchase behavior and attitudinal data, the bank modified a direct mail campaign to existing customers and improved response by 33%.

Furthermore, deposit-product sales to bank customers receiving the tailored efforts were 44% higher than those who received generic solicitations. The New Orleans-based bank also found that revenue generated from these customers was 47% higher than those not receiving the new creative.

The offer for Tower Gold, a loan and deposit sales program that targets current customers, was the same in all the letters. The incentive — a good rate on a certificate of deposit — remained constant as well. What changed was the tone of the letter.

“I can blow the whole process if I don’t manage the sales approach,” says Hibernia’s senior vice president and segment manager Stan Demarest. “If the sales approach is bad I can lose 75% of my leads.”

This came as little surprise to Craig Wood, president of Monitor MindBase, a unit of Yankelovich Inc. in Chapel Hill, NC that supplied the attitudinal segmentation. Wood, who uses the Leno/Barger example to illustrate RFM’s limitations, says “People don’t buy products, they buy solutions to problems.”

Monitor MindBase separates consumers into distinct preconfigured attitudinal groups, similar to the process used by Claritas, which relies on demographic information. For its test, Hibernia pulled a list of roughly 32,300 customers with strong potential (as indicated by their product use history) to sign up for Tower Gold.

Half were further analyzed according to their likely attitudinal segmenting, based on data Hibernia gathered through customer interaction.

Cover Story

Armed with this information, Hibernia’s advertising agency created 11 cover letters, each with a headline, opening message, benefits and response channel targeted to the specific group.

The text for each segment could not have been more varied, even though the offer was the same. “As your children grow older, perhaps you’ll find more time to poke around the garden or watch the news at night. You’ll probably be spending more time thinking about your savings…” started one. Another began, “You work hard to provide for the ones you love, so when you are considering professional advice about your household finances…” while a third assured recipients that “You are at a point where you deserve to kick back and watch the world go by…”

Up-and-coming technophiles were steered to Hibernia’s Web site (www.hiberniabank.com), while older, more staid customers received letters that didn’t even mention online channels.

The mailing also tracked four clusters of likely Tower Gold customers who received letters tailored only by the product segmenting process, without the additional attitudinal data. This allowed Hibernia to determine whether using Monitor MindBase created enough lift to justify it.

The mailing also included a control group of 8,000 customers that received a generic letter with the same incentive.

Not only did the Monitor MindBase segments significantly outperform the generic letter in the January test, but among the product-based segmented efforts the attitudinal pieces drew between 18% and 22% higher responses depending on the segment and the product sold.

But why bother? Why not assume that any profitable customer is a likely target for additional purchases? As it turns out, even the most profitable clients are not all created equal.

For example, high-value Hibernia customers include fixed-income households with many accounts. These prospects are saturated with retail bank products and potential to buy more is not there. “[Insufficient funds] or late fee customers are high-value relationships,” Demarest says. “I don’t want to waste dollars marketing to them — [they probably would be turned down for most credit products] — but I want them to have all the checks they want.”

The program has been so successful in generating revenue that the bank plans to reduce its prospect mailing to consumer clients from 2.1 million pieces in 2001 to 1.7 million in 2002.

But use of the Monitor MindBase segments has also influenced Hibernia’s customer service. The entire database has now been scored with it, and all customers have been assigned a segment. When any of them contact the bank, sales representatives can bring up five important pieces of information.

These include the next financial product likely to be purchased, whether to retain the customer or allow attrition, the customer’s value, general demographic information, and whether this segment generally prefers to have products offered proactively or doesn’t need an aggressive sales effort.

This information makes service reps’ jobs easier and more effective, Demarest says.