The Gold Mine Of Data In Customer Service

Last summer, a computer system at Whirlpool Corp. raised a warning flag: There was trouble brewing in a brand-new washing-machine model. It seemed that after just a few washloads, the machines were springing bad leaks. As soon as Whirlpool engineers determined the cause--a faulty hose clamp--manufacturing was immediately halted. More important, though, Whirlpool's computer helped identify each of the few hundred customers who had purchased the machines so that mechanics could be sent out to replace the offending part. "Imagine the property-damage liability if there had been a leak in a fifth-floor apartment," reflects Gary Lockwood, Whirlpool's director of consumer assistance.

Whirlpool wouldn't have been so lucky if it hadn't been so aggressive and imaginative in applying computers to its customer-service process. Like so many companies operating in highly competitive markets, Whirlpool has come to view customer service as vital to its success. For years, the appliance giant and many other companies have used computers to keep customers happy by giving the telephone representatives on service lines the information they needed to solve problems quickly. Now, the most forward-thinking companies are pushing even further: They're focusing on the information systems in their customer-service operations as a key source of marketing, sales, and design data.

"We're taking customer service from the must-have necessary evil it was in the past and turning it into a competitive advantage," says Sally Price, general manager of the nationwide customer-service center PepsiCo Inc. is setting up in Winston-Salem, N.C. "We want to identify problems and fix them after just 10 calls, not 10,000." Collected and sorted the right way, information gleaned from customer-service transactions can help every part of an enterprise. It can aid in holding down field-service costs, develop stronger relationships with customers, and, in many cases, actually help generate substantial new revenues.

That's why Corporate America will spend more than $1 billion this year on computers and related technology for customer-service departments. In a recent survey by consultants Computer Sciences Corp., some 70% of 782 large U.S. and European companies said customer service is now the main focus of their investments in technology. Virtually every one of the big management-consulting companies, along with hardware suppliers such as Sun Micro-systems, Hewlett-Packard, and IBM and a host of specialist software companies, are in the game, too.

At Whirlpool, customer information not only helps pinpoint faulty washing machines, it helps the Benton Harbor (Mich.)-based manufacturer lower its spare-parts inventories and drive harder bargains with suppliers of motors and other components. If a particular motor shows up in the computer as breaking too often, Whirlpool can employ that data to renegotiate the supplier's warranty on the part.

Similarly, the information gathered from customer-service reps is helping American Telephone & Telegraph Co. dispatch field maintenance people more efficiently and Consolidated Rail Corp. to better understand its customers' freight-shipping needs. And for just about any conceivable business, customer information can help identify prospects for follow-on sales efforts. That can mean a lot for the bottom line, says Mike Korchinsky, president of Axiom Management Consulting Inc. in San Francisco: It takes just 20% the cost of landing a new customer to sell to an existing one.

NO PROBLEM. Until recently, though, "know thy customer" was usually easier said than done. Computer technology just wasn't up to the job: The data existed but were too voluminous, too widely scattered throughout the organization, and recorded too inconsistently to make use of. But now, with high-powered workstations, extensive networks, specialized software packages, and extra-powerful data-base computers, technology is no longer a problem.

And with the new technology, applying computers to customer service is no longer just for giants: A $145 million, PC-communications software company called Attachmate Corp. uses data collected from the 165,000 support calls it receives each year to plan enhancements for its established products and give shape to entirely new ones. The data, managed with software from Vantive Corp., "give us a very clear idea of what's going on in the marketplace," says Mary Harwood, vice-president in charge of customer support.

In most cases, competitive pressure forces companies to probe the gold mine of customer-service data that they've been gathering for so long. At Otis Elevator Co., 360 U.S. field offices once had a local answering service take their customers' calls for repairs. But as independent service companies arose and Japanese-made elevators challenged Otis' reliability record, management decided to centralize and reengineer its customer service.

Today, the company's OtisLine center handles 1.2 million calls a year, 600,000 of which are requests for unscheduled repairs--a broken escalator, perhaps, or, worst case, people trapped in an elevator. When answering such a call, an Otis service rep starts by punching in the code that identifies the customer's building. Immediately, a record of the equipment there and a history of previous repairs shows up on his or her screen. A series of canned questions helps solicit new information, and within minutes, a radio-dispatch message is beaming its way to the appropriate Otis repairman's wireless data terminal.

IN THE BANK. But that's only the beginning of the process. After the repairman fixes the problem, he reports back to headquarters a description of what he did. Then, a full-time, 20-member engineering team reviews that and any similar cases in the computer to see if there's a pattern to the failures. Such reviews may lead to a redesign of an elevator or perhaps a change in maintenance procedures. "We've seen dramatic reductions in callbacks and entrapments," says Maria Gallo, manager of Otis' service-center operations. The same customer data also will be used in a coming initiative to plan the weekly maintenance rounds by Otis mechanics better--a process that's now done on paper. With today's office-building glut holding back new elevator installations, Otis makes most of its money from maintenance services, and every improvement in that department is money in the bank, says Gallo.

Not surprisingly, high-tech companies have been quick to reengineer customer service. AT&T Network Services' Columbia (Md.) facility manages the installation and repair of complex computer networks at customer sites from Maine to Virginia. But competition in that business is fierce, attracting numerous computer companies and independent contractors. So, after analyzing many of its business practices as part of a corporate drive to improve quality, the AT&T unit has built a customer-service data base using a package from Aurum Software Inc.

The new system has given AT&T a badly needed extra advantage. It keeps track of customers' hardware and software inventories, spare parts and their repairs, and a wealth of other information. Michael Crabb, information systems manager, says the data help AT&T develop a series of measures that describe for internal managers and customers alike just how well the company is living up to its contracts. Choosing a packaged solution instead of writing its own software "has put [AT&T] six to nine months ahead of the competition," Crabb suggests.

For many companies, the amount of data they're able to collect from and about their customers can be staggering--and difficult to make sense of. Whirlpool maintains records on close to 15 million customers and more than 20 million installed appliances, some of them dating back to the 1960s. To help plow through so much data, companies such as Whirlpool and Conrail have begun using specialized data-base computers that get masses of powerful microprocessors to scan volumes of records in parallel and seek out faint but significant patterns. Whirlpool and Conrail each use parallel processors from AT&T's former Teradata Corp. unit. IBM is expected this spring to come out with a similar processor aimed at the same kind of applications.

At Conrail, as at other railroads, the challenge has been to win back shipping business that has gone to trucking companies. Many trucks now can be tracked by satellite; only lately have railroad companies begun using technology to keep constant tabs on their widely scattered railcars.

And now, with its AT&T computer to sift through a massive data base of railcar and customer information, Conrail can address customers' shipping needs much more effectively, says Robert O. Wagner, the railroad's vice-president for information systems. It can set up accurate long-distance train schedules that can even take into account handoffs of railcars to railroads other than Conrail. Then, if bad weather arises or there's an equipment failure, the AT&T machine can calculate the best alternate route. "We can continually optimize our schedules," Wagner says, "because we understand the customer's requirements on an end-to-end basis."

That's what all companies need to do --understand and manage customer relationships from the first sale onward. And with computers keeping track of things at every stage, the chances are better that the customer relationships will never end.

Before it's here, it's on the Bloomberg Terminal.