Studies show two things about customer-relationship management: It will remain the hottest area of information-technology spending for years to come—and most of the billions of dollars spent thus far on customer-relationship-management systems, or CRM, have been wasted.

Just probe for examples of how CRM has been fumbled, though, and everyone quickly closes ranks. “Bad CRM? I’ve heard of it. Check down the street.” Of course, such failures at their companies are just urban legends.

“Companies absolutely hate to admit that they haven’t gotten their arms around CRM, especially if they’ve already tried mightily to do it and maybe spent millions of dollars in the effort,” says Rich Hebert, chief executive of ISky Corp. (www.isky.com), which operates call centers for large companies. After all, he says, “CRM has been deemed the Holy Grail these days.”

Even though these disasters often involve seven-figure investments, they are relatively easy for companies to bury. When a blown implementation of enterprise-resource-planning software kept Hershey Foods Corp. (www.hersheys.com) from putting candy on store shelves right before Halloween, that was front-page news in the Wall Street Journal. It is harder to find out, though, when a company’s CRM system causes call-center operators to miss a chance to cross-sell products and services or annoys customers by suggesting services they clearly don’t want. The company can blame weak sales on the uncertain economic environment, and who’s to know?

But the problems are out there: Meta Group Inc., a research firm (www.metagroup.com), says 55% to 75% of CRM implementations initially fail to meet objectives. So Context went in search of examples of the secret shame of CRM calamity and found some companies that were willing to ’fess up. Here are their stories:

(MIS)UNDERSTANDING THE CUSTOMER: GENERAL MOTORS ACCEPTANCE CORP.

General Motors Acceptance Corp.’s commercial-mortgage operation, known as GMACCM (www.gmaccm.com), has rocketed into a leadership position in business real-estate loans. Set up as a separate unit in 1994 with a portfolio of $5 billion of mortgages, GMACCM now has a portfolio of more than $100 billion. Much of the reason is that the business was built on a platform of the latest, most advanced information technology, such as the $3 million system installed three years ago to let employees digitize loan documents and zap them to each other around the world. The technology platform lets GMACCM be fast and efficient while providing good service.

“I don’t think any competitor on the commercial-mortgage side has made the kind of investment in technology that we have,” says Mike Lipson, an executive vice president of GMACCM.

Technology initiatives usually have worked smoothly for GMACCM—but CRM proved to be a much tougher nut to crack.

GMACCM Chief Information Officer Niraj Patel says he and other executives saw CRM as a way of holding customers’ hands better, while standing out from competitors. So, three years ago, GMACCM asked the consulting firm PricewaterhouseCoopers (www.pwcglobal.com) to redesign its customer-relationship apparatus so that it could operate well into the new century. Patel recalls that GMACCM had modest goals. It primarily wanted more efficiency and automation, plus ways for call-center operators to know a bit more about its borrowers. For instance, did they always call to make sure their payments were posted to the correct accounts?

The problems began, according to the GMACCM executives, when PwC stumbled into a fundamental misunderstand- ing of the complexity of the GMACCM customer base. PwC, which recently agreed to be acquired by computer hardware and consulting-services provider International Business Machines Corp. (www.ibm.com), declined to comment for this article. A company spokeswoman cited a policy of not commenting on client work.

Explaining the complexity, Lipson says, “Our average loan is about $3 million, but we have loans as small as $100,000 and as large as a $1.6 billion loan for Rockefeller Center. And we can have anyone from the chairman of the company down to someone in the accounts-payable department calling us with questions.”

The new system treated all customers pretty much the same way—badly. Patel says the consultants didn’t challenge GMACCM’s existing structure, which divided customer service along functional lines, differentiating, for example, between representatives who answered questions about originating loans and those who handled loan-servicing concerns. When PwC’s overhaul was unveiled two years ago, call-in customers were essentially expected to know GMACCM’s customer-relations structure well enough to know what kind of person could answer their question, and then request that department.

Even worse, PwC suggested that GMACCM get much more aggressive about trying to use cost-efficient automation to answer as many of customers’ questions as possible. The consultants installed a sophisticated voice-response system so that customers could call an 800 number, and then tap in their loan number and other information on the phone keypad to obtain the information they were seeking, such as loan balances.

“When we fired it up, we found that 99% of our customers—literally—were hitting zero so they could talk directly to a live operator,” Lipson says. “While a personal-account customer might be willing to punch through a whole bunch of numbers, like for an American Express Gold Card, when he was calling about a commercial loan, he wasn’t willing to do the same.”

During the few months in 1999 that most of GMACCM’s 20,000 customers were calling up GMACCM and furiously “zeroing out,” Lipson says the competition was “using this against us as a marketing tool.” While he declines to estimate the cost of the debacle, he says it was “a big deal.”

“We were irritating people, and they didn’t want to do business with us going forward,” Lipson says. Internal complaints welled up too, from loan officers who were losing deals because customers “were dissatisfied with the service from us,” Lipson recalls.

By the time, about 20 months ago, that Lipson was promoted to executive vice president in charge of client relations, putting him atop the company’s CRM apparatus, GMACCM’s new system was doing way more harm than good. He had to overhaul the overhaul.

Right off the bat, Lipson junked the automated responses in favor of live operators. He also reorganized the client-service operation along product lines, and, when a loan is originated, the company gives a customer the 800 number of a specific representative for any further contact, rather than “sending a letter to the borrower saying, ‘Contact GMAC Commercial Mortgage.’” In addition, Lipson says he has “upgraded personnel,” especially seeking phone reps with real-estate experience.

Only recently has Lipson become convinced that he has reversed the problems. “We’re beginning to see more compliments coming from our customer base, and internally,” he says. “And I’m spending less and less time on complaints.”

MAYBE, MAYBE NOT: OWENS CORNING

Owens Corning (www.owenscorning.com) is an archetypal Fortune 500 industrial. Its headquarters lobby on the banks of the Detroit River, in Toledo, Ohio, includes a 10-foot-long map of the world with color-coded lights representing the company’s dozens of installations around the globe. But in the early 1990s, because of its formidable size, the company’s dealings with customers were a mess.

Owens Corning was acquiring a number of smaller companies to expand its product portfolio beyond insulation, a strategy that was creating more points of contact with customers. At around the same time, senior executives were dabbling in another bit of strategic broadening, using the company’s well-known Pink Panther icon as the wedge for a new positioning of the company as a home-improvement expert. Unfortunately, all this expansion was creating widely varying levels of service experiences for customers, and often a confusing duplicating of information about a single account in several databases, each of which had been organized around a different product line.

There were different electronic pockets of customer responses that had been collected, says Richard Kaverman, a marketing staff member at Owens Corning through most of the ’90s. One was from people who’d called in for roofing products, another for insulation products, and yet another for windows. In some places, Kaverman says, the data storage was as crude as having actual warranty cards that were stuffed in shoeboxes.

The chaotic systems caused problems for some customers that escalated all the way to the chairman, one of Owens Corning’s marketing executives from the early ’90s said.

At 84 Lumber Co. (www.84lumber.com), a big customer for insulation and roofing products, Owens Corning’s problems “definitely led to some confusion on our end,” says Alex Eadie, director of contract purchasing for 84 Lumber.

“In terms of marketing philosophy,” Kaverman says, “[Owens Corning] just wasn’t there yet. Rather than having a [sophisticated, data-intensive] approach to marketing, they had this belly-to-belly concept: If you could just get me down in front of this person across a table, that’s how I’m going to market.”

Owens Corning did possess a strategic advantage that was unusual for an old-line supplier of such mundane goods as shingles and insulation: It had a cadre of young managers who understood the promise of CRM and was pressing to exploit it.

A guerrilla corps of CRM-oriented operatives inside the marketing and information-technology departments had met marketing consultant Martha Rogers at nearby Bowling Green State University and got fellow Owens Corning managers to gather with her and hear her ideas on one-to-one customer relationships and how to manage them. The managers pushed for funding and won approval for some pilot projects, such as one to improve the call center for 1-800-GETPINK, a general direct-response number that ran in the company’s television ads.

Still, CRM work was stymied every time it seemed to be building momentum. For example, there was a pilot project to test a CRM model in connection with a new line of roofing products being test-marketed in Portland, Ore., but the project was scotched when Owens Corning quickly decided to roll out the products nationwide. For three years in a row, beginning in 1997, the CRM enthusiasts were told to put together budgets outlining their plans for seven-figure investments in various expansions of the company’s customer-relations capabilities. Then the funding would disappear.

“We’d turn in our budgets in July, during the budgeting cycle, and those would get approved. By October, they’d be cut significantly. By January, they’d be cut back so far that we only could operate what already was in existence at the beginning of the [prior] year,” Kaverman remembers. “So, by 1999, we’d arrived at the same place with CRM that we were in 1992.”

Much of the money that might have gone to CRM was siphoned off into a massive enterprise-resource-planning project that was designed to replace 200 disparate computer systems with a single platform that could support the operations of multiple business units across multiple countries. The seven-year effort cost $280 million.

“You have to have your internal transactions, your base transactions and processes, in place and operating appropriately, effectively, and reliably before you automate anything to the end customer,” says David Johns, currently the chief information officer with Owens Corning.

Johns acknowledges that Owens Corning had purchased a CRM system from German software producer SAP AG (www.sap.com) as part of the enterprise-resource planning package it bought from the software company and could have kicked CRM into action anytime. But “everything was being sunk into core business systems,” says Steve Smoot, a former Owens Corning marketing executive. CRM “didn’t sink in as a priority at the time.”

Johns thinks the CRM advocates “had good intentions” but “got the cart ahead of the horse”—in particular, in wanting to use CRM software to deal with customers over the Internet. The advocates’ approach “just would have provided inaccurate data, faster,” Johns says. The CRM boosters “thought that the Internet was some sort of magic technology platform that made everything in the back office work perfectly,” he says.

Still, Johns concedes that Owens Corning should have focused more on what customers wanted, rather than working only on getting its internal processes right. If he had it to do all over again, “We would have spent more time working with the customer and gone backward from there,” Johns says.

John Sviokla, vice chairman of management-consulting firm DiamondCluster International Inc. (the publisher of this magazine, www.diamondcluster.com), says Johns is right. “These CRM journeys need to start out by working deeply with customers to discover what is critical to them. This work goes well beyond call-center operations to the design of the whole business. Great firms understand this.”

The project to integrate its internal computer systems was a success, and, in the middle of 2001, Owens Corning began a drive to implement a program for dealing with customers on the Web, using SAP software and starting with a new home-handyman service called HomeExperts. But Owens Corning is now operating under the huge handicap of supervision by a U.S. bankruptcy court, where it sought protection from its creditors two years ago amid asbestos litigation. The company also is operating without its would-be CRM pioneers, including Kaverman and Smoot, who became frustrated and left in 1999 and 2000.

“There is no way of measuring what it might have cost Owens Corning by not doing [CRM] for six or seven years,” says Rogers, the marketing consultant who encouraged Owens Corning’s CRM advocates.

TECHNOLOGY TROUBLE: PERSEUS DEVELOPMENT CORP.

If there is any company that should have grasped the importance of CRM early on, it is Perseus Development Corp. (www.perseus.com). After all, the company was founded nine years ago based on the value of communicating with customers: Its business is formulating and conducting customer surveys for other companies.

The company started out mainly providing survey-writing software for $179, and it quickly acquired a long list of customers, ranging from huge corporations to mom-and-pop outfits such as ophthalmologists’ offices and college professors who polled students on their performance at the end of the semester.

Yet everybody at Perseus was so busy designing the next version of the software and expanding sales that nobody even transferred customer-registration information into something as basic as a contact-management program, admits Jeff Henning, co-founder and chief operating officer. “We were getting data spit back to us, but we weren’t doing anything with them,” recalls Scott Washburn, former head of marketing who left closely held Perseus in early 2002 to become a marketing consultant for Tusk Creative Group.

By 1999, Henning and others realized that Perseus simply had to keep better track of its customers and that, by allowing them virtually no way to interact except with its technical-support department, the company was missing out on many growth opportunities, such as the chance to sell extra software licenses to people calling in with unrelated questions. So Henning and Perseus co-founder and Chief Executive Richard Nadler decided to purchase a Clientele CRM module from Epicor Software Corp. (www.epicor.com).

Incredibly, however, Nadler and Henning didn’t take Epicor up on its proposal to fully integrate Clientele with Perseus’s existing information systems at a cost of no more than another $20,000—despite Washburn’s strong recommendation to do so. “We thought their product was easy enough that it wouldn’t be difficult for us to integrate, because we didn’t have a pre-existing CRM system that we had to work around—and we were a software-development shop,” Henning remembers. “We felt at the day rate they would be charging to do it, we could get it done ourselves.”

Right away, the technical-support manager Perseus assigned to integrate Clientele “began hitting walls,” Henning says. For example, in allowing the Perseus sales force to place orders electronically, he couldn’t reconcile Perseus’s existing forms with the new forms called for by Clientele. Several more months went by, and Perseus’s employees still were no more able to interact knowledgeably with customers than they had been before Clientele was installed. In the meantime, some of Perseus’s competitors were raising the bar by completing their own, more successful CRM projects.

Each week, frustrated marketing staffers ended up manually pulling names in alphabetically organized batches out of Perseus’s database of several thousand customers to e-mail them messages, such as reminders that their 30-day free trials were expiring. When customers reached Perseus, technical-support staffers didn’t even have simple data points in front of them, such as the fact that the client would need to be upgrading its software soon. “Or it would take us a long time to answer a simple question such as how many licenses did they have or what was the status of an order,” Henning says. “Sometimes it took a couple of days.”

The recession in 2001 cooled activity at Perseus just enough so that Henning could collect his thoughts and hire an internal data manager, John Bowens, who is teaching Perseus to utilize the CRM system installed two years ago. This year, Perseus just missed Inc. magazine’s list of America’s 500 fastest-growing companies, at No. 520. “If we’d had our act together on CRM,” Henning concludes, “we would have made that list a long time ago.”

Washburn, the former marketing executive, says: “Revenue could have been 30% higher each year for the last two years.”

Buss is a free-lance author, journalist, and editorial consultant in Rochester Hills, Mich. As someone who spends a lot of time on the phone gathering information from all sorts of far-flung places, his personal dream about CRM is for the return of phone-company directory-assistance operators who can reliably dispense the correct phone number. He can be reached at DaleDBuss@aol.com.

DOING IT RIGHT

In a classic Saturday Night Live skit, a woman hears burglars breaking into her apartment and dials the police. She reaches an automated voice response system that goes something like this: “If you’re reporting an attempted murder, hit 1. If this is a robbery, hit 2.” The options drone on as two masked men bind and gag her, then loot the place.

The skit made me wince because it rang true. Too many companies get carried away with the technology that is now available for customer-relationship management, or CRM. Instead of starting with the gadgetry, you first need to think about just what you hope to accomplish for, or with, customers.

In my experience, the best approach includes three steps.

First, you find a “strategic nerve,” a pressing concern that you can help customers handle.

This requires lots of careful testing, because what customers care about isn’t always apparent. For instance, 3M Corp. (www.3m.com) found that its hospital-administration customers were most interested in reducing their backlog of bills, even though 3M could do other things that provide much more economic benefit.

Second, you rethink your organization and incentive structure, to make sure it corresponds with your new way of dealing with the customer.

Owens Corning (www.owenscorning.com) paid the price for missing this step. Although trying to provide a unified, companywide view of customers, Owens Corning didn’t provide incentives for the different parts of its sales organization to work together. As a result, improvements have been limited. Owens Corning can now send out marketing material the same day it’s requested, rather than days or weeks later, but that’s hardly enough improvement to justify the tens of millions of dollars that Owens Corning spent.

Third, you figure out what technology to use.

That likely will mean buying a CRM package but may require that you build key components yourself or do extensive customization. Harrah’s Entertainment Inc. (www.harrahs.com), for instance, had to do considerable custom work so it could institute a companywide loyalty program that would increase its share of the business from frequent gamblers. The work was hard, but Harrah’s chief operating officer Gary Loveman says he welcomed the difficulty, because it would make it tougher for competitors to copy him.

Because Harrah’s understood its customers and executed its idea superbly, its stock price has tripled since 1998.

In other words, doing CRM right can let you hit the jackpot.

—by John Sviokla