InformationWeek.com

Welcome Guest. | Log In| Register | Membership Benefits

March 29, 1999

Profitable Customers

Businesses are using IT to identify high-yield clients and formulating new strategies for dealing with those that aren't

By Rick Whiting with Jeff Sweat

Related links:

Web Ties That Bind

The Front Line

Sidebar: Buy More, Save More

IT Management Resource Center

Enterprise Management Resource Center

growing trend in business--the use of IT to better manage customer relationships--is running headfirst into a financial necessity that drives all companies: the need to generate profits. The overlap of these two concerns is creating a fertile new IT environment in some companies that is being used to identify blue-chip customers and treat them accordingly, while developing aggressive new strategies for borderline and unprofitable customers. In the never-ending quest to gain and keep customers, the emphasis is increasingly on quality, not quantity.

The interest in tracking customer profitability is spreading across industries and encompasses all types of customers. It was a factor in Federated Department Stores Inc.'s decision last year to acquire Fingerhut Cos., a mail-order and direct-marketing company that built a $1.6 billion business selling products to lower-income households. Fingerhut, which manages a database of 75 million consumers who have purchased its products, may have something to teach Federated, which owns more upscale retailers such as Macy's and Bloomingdale's, about the art of using IT to understand and market to its customer base. "The modeling and technology we use to determine who we'll market to is the lifeblood of our company," says Bill Flach, Fingerhut's corporate research and analysis director. "We couldn't survive without the segmentation and targeting that we do."

Central to Fingerhut's success is its strategy of letting customers buy on credit. Using the transaction history of those customers--including purchases, product returns, and credit history--the company has created models for identifying customers with the highest potential for targeted marketing campaigns. To do it, Fingerhut is using data warehousing and data mining tools from IBM and SAS Institute Inc. Flach says Fingerhut will help Federated develop models for its mail-order operations. Federated and Fingerhut also plan to compare their customer databases and may share data.

The trend is made possible by the fact that many companies have completed projects to build data marts and data warehouses of customer information, and they're now using data mining algorithms to discover patterns of information within them. New "customer intelligence" applications from vendors such as DataSage, Epiphany, Exchange Applications, and Prime Response Group not only identify profitable customers but help managers act on that intelligence. "Technology is really driving this," says Martha Rogers, a partner with Peppers and Rogers Group, a consulting company that specializes in customer-relationship management.

Customer-profitability analysis presents a huge opportunity, but only a fraction of companies have pursued it in a serious way, say experts. "We've seen about 15% to 20% of our customers starting to do sophisticated profitability analysis," says Dan Harrington, marketing VP for NCR Corp.'s national accounts solutions group, which markets data warehouse systems and profitability modeling software for vertical industries.

Gauging customer profitability isn't easy. Some companies use activity-based costing and similar methods for assigning costs to individual customers when calculating their profitability. But it's an approach that doesn't take into account customers who aren't profitable now but might be in the future. "If you don't see this, you might not court your most valuable customers," says consultant Rogers.

Businesses have to be careful that these strategies don't backfire, warns Ken Robb, VP of marketing with Brodbeck Enterprises Inc., which operates eight supermarkets in Wisconsin and Illinois. Brodbeck is careful not to let its customers think they are being treated differently, even though it's really giving better deals to its best customers (see story, p. 52). "In the supermarket, it doesn't feel quite right that some people would get better treatment," he says.

Concentrated Profits
Some companies that have started down this path are surprised to discover that a relatively small percentage of their customers account for most or all of their profits, while the majority make no contribution to the bottom line--or worse, erode it. The First National Bank North Dakota in Grand Forks discovered that only 10% of its customers, both individual and commercial, accounted for virtually all of its profitability, according to Michael Winkel, the bank's CFO and chief operating officer. This experience isn't unique. Within the financial-services industry, only about 20% to 30% of customers are generally profitable, says senior research analyst Richard Bell at the Tower Group, a financial-services technology firm.

By many accounts, the banking industry, with its razor-thin profit margins, is ahead of others when it comes to identifying profitable customers. "The top 20% of our customers is our franchise," says Christopher Kelly, senior VP of database marketing at Bank of America in San Francisco. The bank uses an NCR Teradata data warehouse with SAS Institute's Enterprise Miner data mining software to identify its most profitable customers, and Exchange Applications' marketing automation software, Valex, to manage marketing campaigns based on those findings.

Bank of America's goals are to retain its best customers--the annual customer turnover rate in banking is 30%--while identifying opportunities to sell them additional services. The bank develops profiles of its top accounts, which relationship managers assigned to the top 10% of the bank's customers use to target services. It also creates probability models to identify customers who are in danger of taking their business to a rival.

The effort to identify profitable and unprofitable customers is often an outgrowth of attempts to create a "customer view" of a company's business. A major problem companies face when trying to identify profitable customers is that companies themselves are organized around product lines or services. While determining the profitability of a product is easy enough, doing the same for customers is trickier, especially for customers that buy multiple products or services from the company.

That was a problem faced by First National Bank North Dakota, which provides a range of services, including commercial and retail banking, brokerage, and trust management. In the past, each bank division ran on a separate legacy or IBM AS/400 system, so answering a question like "What services do we provide Mr. Smith?" was nearly impossible. So was identifying the bank's most profitable customers. Consequently, customers who contributed greatly to the bank's bottom line didn't always get the attention they deserved, says Winkel, who tells the story of how the bank once bounced a check that exceeded a customer's checking account balance--even though the customer had a trust fund with the bank worth several million dollars.