Cover Story

Rethinking IBM

An Exclusive Account of Lou Gerstner's First Six Months

Judith H. Dobrzynski in New York
Business Week

October 4, 1993

On this sunny Sept. 9, the platform in the ballroom of the Falls Church (Va.) Marriott hotel is swathed in IBM's signature royal blue. Nearly a thousand talking, laughing IBMers -- bused and vanned in from an area stretching from Richmond to Baltimore -- fill row upon row of chairs. Just before 3:30 p.m., Louis V. Gerstner Jr. walks in. He shakes some hands and is introduced. Standing on the platform's edge, arms across his chest, he launches into what is, in essence, a stump speech.

IBM's new chairman is on the road again. He starts with a humorous anecdote. Next, he gives his analysis of IBM -- its strengths (world-class technology, extraordinary people, customers that want IBM to succeed), and weaknesses (towering costs, a preoccupation with internal processes, and slowness). "I didn't make it work out that there are three strengths and three weaknesses." Pause. "But on an IBM foil, it would have to work out that way." Foils -- overhead slides -- were used in every Big Blue presentation until he arrived.

Then he delivers the real punch line: "That is not a bad hand to play," he says. "The strengths are very, very hard to duplicate. And the weaknesses, we ought to be able to fix. Because it's us." With those words, Gerstner -- who left RJR Nabisco Holdings Corp. to take over from John F. Akers on Apr. 1 -- just about sums up his high-stakes campaign to rescue the ailing $ 62 billion computer company from itself. Few question the raw potential of the tarnished titan of Armonk, N.Y. But having watched IBM choke on its own culture, arrogance, and misguided strategies, having witnessed eight years of failed remedial efforts, many wonder whether Gerstner, or anyone, can fix Big Blue.

Gerstner has no time or tolerance for such doubts. He's busy rethinking IBM, top to bottom. The IBM he heads is a dysfunctional company: It has taken $ 28 billion in write-offs since 1986 and has lost money for the past three quarters. Its stock is off 75% from a peak of 176 in 1987. To succeed, the 51-year-old Gerstner must mount a cultural revolution at IBM: He must convince 250,000 employees in 140 countries to change the way they think and act. It's the toughest job in Corporate America today.

In a series of exclusive interviews with BUSINESS WEEK -- in Armonk, in Virginia, aboard an IBM Gulfstream IV jet, and preceding a two-day strategic-planning conference in Palisades, N.Y., on Sept. 19 -- Gerstner discussed his days at IBM. He let BUSINESS WEEK accompany him on a typical outing -- a tour of IBM's Falls Church offices and laboratory and a town meeting with the Mid-Atlantic sales region's employees. And he shared videotapes of other meetings he had with employees and customers. By his count, he has talked to "thousands" of customers and 20,000 workers -- a sign of the urgency he feels.

The pressure on him comes from far beyond Wall Street. Boeing Co. Chairman Frank A. Shrontz has warned Gerstner that the nation's No.1 exporter can't build airplanes without IBM computers. Ford Motor Co. Chairman Harold A. Poling told him that "we are so dependent on IBM that your success is almost as important as our own." And Nobel prize-winning geneticist Joshua Lederberg recently grabbed the IBM CEO at a meeting in New York, poked him in the chest, and said: "Do you know that you have one of the great technical treasures in the world at IBM? You better not blow it."

His task is all the more difficult because Gerstner -- unlike, say, Microsoft Corp.'s William H. Gates III or Intel Corp.'s Andrew S. Grove -- is no expert on technology. With a background as a consultant at McKinsey, president of American Express, and CEO of RJR, Gerstner sees the challenge as a management problem. The technology IBM develops is as good as any rival's, he insists. So he's out to cut costs, slash IBM's bureaucracy, get products to market fast, and turn Big Blue into a customer-driven enterprise. If he does that, he reasons, IBM's scientists and engineers will deliver high-quality products that will help customers run their businesses better.

Critics may doubt that -- and Gerstner knows that IBM can't rest on its technological laurels. He recently implemented what he calls "one of the most important things I've done." Instead of relying on a chief technologist, as IBM has done in recent years, Gerstner created a technology-assessment system. Now, teams of experts from throughout the corporation's empire will come together to review current and emerging technologies. Gerstner himself will chair these efforts.

Much else has happened in the six months since Gerstner replaced Akers, who threw in the towel after eight years in IBM's executive suite. Gerstner has reversed Akers' plans to split the company into autonomous units and has set up task forces to explore growth opportunities. He has ordered more layoffs and cost-cutting. He restructured senior management and changed the way some executives are paid. He has begun reengineering IBM's vast, bureaucratic maze. And he has promised that IBM will execute its strategies better.

But the world isn't impressed. IBM's stock price remains stubbornly in the 40s -- partly because Gerstner has kept his activities under wraps. Critics, seeing cost-cutting as Gerstner's primary tangible achievement, want more.

Gerstner's plans, in reality, are anything but one dimensional. As he plots IBM's future, he's relying on a broad but simple conviction: Since information-technology products are coming out with increasing frequency, even knowledgeable customers need help sorting out the confusion. That trend, he believes, will favor IBM, with its broad product line and worldwide reach.

"We must become the premier solution-provider to our customers," he says. That's why Gerstner will keep Big Blue intact, offering hardware, software, systems, and services. It's why he plans to develop a cadre of people with expertise in specific industries who will, when necessary, recommend rival products if IBM's don't measure up. But Gerstner isn't excluding customers who want less from IBM. He says he'll let customers decide the relationship -- down to the use of a toll-free phone line for ordering virtually any IBM product.

Gerstner's "solutions" sell is hardly a new notion to computer marketers -- or to IBM. And even if the new CEO does better than his predecessors in instilling this approach, skeptics question whether customers will trust IBM or any other hardware maker to be truly impartial when suggesting the best solutions (page 96).

In response, Gerstner tells the skeptics to wait and see. "What I'm trying to do is deliver results not promises, results not vision, results not concepts," he says. "The world is cynical about IBM's promises." Specifics will come later. Still, as he prepared to lead the strategy session on Sept. 19, he continued to talk freely about his tenure at IBM.

If the new job weighs heavily on Gerstner, he barely shows it. Ever intense, he seems energized by the challenge. He regularly strides down the tangerine-carpeted halls of IBM headquarters in shirtsleeves to find someone he must see at that very moment -- to the surprise of staffers used to being summoned for scheduled sessions with CEOs who rarely left the corner office. "Before, there was no impromptu anything," says a veteran IBMer.

Only the fishbowl aspect of the job annoys Gerstner. Describing a recent visit to his father in the hospital, he says his dad complained about being stuck frequently with needles. "'Why do they keep doing this?' he asked me," Gerstner remembers. "And I said I feel the same way."

So might IBM -- with Gerstner the master prober. Upon his arrival, Gerstner asked each of IBM's 13 division chiefs to write a 10-page report detailing the business environment and critical issues the unit faced. He then spent a half day going through each business unit with its senior staff. He flew to Europe, visiting four countries in five days. Also on his agenda: lunch with former CEO Thomas J. Watson Jr., the founder's son, and talks with former CEOs Akers, John R. Opel, and Frank T. Cary. He put in time reading letters from customers and E-mail in which employees poured out their thoughts and, sometimes, their hearts.

"For the first month, I listened, and I tried very hard not to draw conclusions," he says. Even so, Gerstner's mind tended to race ahead. He woke up nearly every night at 3 a.m. "with an answer that was clear. By 6:30, it was murky again." Before long, Gerstner was making initial judgments: "I knew I had about 30 days before I had to start saying something, and that's when I came out with the four priorities."

On Apr. 21, Gerstner told his top 24 managers that the company was going to concentrate on "right-sizing," developing strategic directions, improving employee morale, and "bear-hugging" customers. The last served a second purpose, aiding Gerstner's fact-finding: In each of the next three months, he ordered everybody in the room to meet with five customers, both large and small. Each of their subordinates was told to do the same. He instructed everybody to send him a report on each of these visits -- which he took home to read each night.

Gerstner revealed his priorities publicly on Apr. 26 at IBM's annual meeting in Tampa. The confab, packed with 2,300 shareholders, was his official debut. It was no party. Angry attendees lashed out at IBM's board. For two-and-a-half hours, Gerstner took questions. Answering one, he revealed that he would not split IBM into freestanding Baby Blues -- his first break with Akers' plans.

By mid-May, Gerstner had finished his initial internal fact-finding. Much of what he found he didn't like: Personnel, compensation, data-processing, and other administrative systems were a mess. IBM was hamstrung by internal wrangling among units. Oversight by headquarters strangled action. Some dealings required 18 signatures. It was "bureaucracy run amok," he later said.

Continuing on his information-gathering mission, Gerstner now went outside IBM. He consulted with the heads of most major U.S. information-technology companies -- Gates and Grove among them. He talked with other top executives. The longest session -- six hours -- was with John F. Welch Jr., CEO of $ 57 billion General Electric Co. Gerstner wanted to know how Welch leveraged GE's size without being stymied by it. Welch calls it "making 'big' pay."

As these thoughts swirled in Gerstner's mind, he convened his first planning forum, a two-day affair in Chantilly, Va. There, Gerstner set up task forces to study 12 areas of strategic concern (table). Covering product areas, such as multimedia and software, and other opportunities, such as customer service, the titles show how Gerstner first viewed IBM's future. Now, he says the groupings "weren't perfect... But I didn't want to wait six months to define the issues exactly right." Gerstner says two of the 12 -- he won't say which -- were resolved during the Sept. 20-21 strategy session. Two more replaced them.

Gerstner stayed a day longer in Chantilly. U.S. sales chief Robert J. LaBant was hosting a 200-customer forum on May 19, and Gerstner kicked if off with these words: "One of the most important things I can say to you is there is now a customer running IBM." He promised to focus on delivering value and quality. Attendees, having heard much of this before, asked tough questions. They didn't hide their disappointment with IBM's past rescue plans, either. Louis Valente from SmithKline Beecham PLC asked Gerstner when customers could expect change. "Now. We're going to put a blowtorch to these things," Gerstner vowed.

June was filled with more meetings, more trips. On the Fourth of July, Gerstner flew to Tokyo to meet customers. When he got back, he faced several major decisions, as IBM prepared to report second-quarter earnings on July 27. The largest: a write-off to cover restructuring. Gerstner spent a week looking at IBM's financial outlook through 1995. He huddled often with Jerome B. York, the new chief financial officer.

They had plenty to chew on. IBM's gross margins had sunk from 55% in 1990 to 40%, or about industry average, in 1993. Meanwhile, its operating expenses hovered around 30% of sales. Throw in research outlays, and IBM was spending 42% of sales on expenses in 1992 vs. 31% spent by rivals. Unable to charge its once standard premium prices, IBM started posting operating losses in late 1992.

One day, York walked into Gerstner's office with some surprising numbers. While IBM was slashing its work force by 120,000, other outlays had skyrocketed. Using numbers from the first half of 1993, annualized, York had found that while personnel costs dropped by $ 3 billion from 1990, other expenses had jumped by $ 7 billion. Too much inventory was being written off. Benefits, warranty, and purchasing expenses were rising. Data processing was costing IBM 7% of sales, vs. an average of about 2% at other U.S. companies. IBM's CEO nearly fell off his chair.

Worse, York couldn't determine why it had happened. The costs were spread around 13 operating units. And when IBM was decentralized in late 1991, management didn't adapt its data-gathering system. York and Gerstner established nine task forces to study IBM's information systems, inventory management, purchasing, transportation, and other administrative areas. Reports are due at the end of September.

Meanwhile, the two worked backward from the numbers IBM collects about its rivals. To get costs in line with or better than theirs, they figured IBM has to cut $ 7 billion from its expenses by 1995. About $ 5.25 billion will come from already announced work-force cuts and operating efficiencies. The task forces will find the remaining $ 1.75 billion in overhead. For now, IBM plans no more layoffs or write-offs.

Gerstner released the $ 7 billion cost-cutting goal on July 27, as well as second-quarter results and an $ 8.9 billion write-off. That's when he made his first mistake. Speaking at New York's St. Regis hotel, Gerstner told the press: "There's been a lot of speculation as to when I'm going to deliver a vision. The last thing IBM needs right now is a vision. What IBM needs right now is a series of very tough-minded, market-driven, highly effective strategies in each of its businesses." He was hammered. Some news reports printed only the middle sentence. Critics, including many customers, had a field day.

Gerstner has his reasons: "I could spend a lot of time focusing this company on a mission statement, and employees would love it. There would be debate instead of action, when the most important thing we had to do was implement better and sooner. I want to stay away from abstractions and deal with the marketplace."

That won't satisfy the doubters, though. And Gerstner keeps getting nagged. When one employee asked about it at a mid-August town meeting, Gerstner blurted out: "Our mission is to be the most successful information-technology company in the world. O.K., you want a vision statement. Fine, we got it, now let's go back to work."

All the while, Gerstner was calling on customers and chatting by phone with at least two a day. Inside IBM, too, he gradually has been working his way around, striving to be "visible, showing what does this new guy look like and sound like."

Gerstner usually begins a site visit with a management meeting, often over a meal. He goes around the table asking each person what exactly he or she does. Then, he probes. "He uses a Columbo-style," says Bond Isaacson, director of marketing for IBM's Mid-Atlantic sales region. "He'll say 'Help me with this -- it just doesn't make sense that...' It's effective because it gets people talking."

To get to the rank and file, he sets up separate meetings without managers and tours plants or offices. Finally, he takes to the hustings in mass gatherings, leaving about 90 minutes for questions. He uses these exercises "to test-market my ideas."

Back in Armonk, Gerstner kept noodling over organizational issues. Outsiders suggested more decentralization, which worked well with IBM's PC company. But that, he worried, might worsen IBM's paralyzing turf battles.

For a few days in mid-August, he decided to go home at midday, taking papers to read in his library. "I was wrestling with decentralization because at heart I'm a decentralizer," he recounts. "But as I was looking at mail and customer reports, it became increasingly clear to me that the real issue of effectiveness, of winning in the marketplace, was finding ways to make the company work horizontally."

To thrive, he believes, IBM must spread its technology across product lines -- for example by using the PowerPC chip throughout its computer line. It has to design products and systems that will meet customer needs, not just sell technology for technology's sake -- and not become merely a services company. Finally, it must leverage its global size. Says Gerstner: "When a manufacturing company in Spain looks to IBM for a solution to a problem, they expect us to bring the best of IBM worldwide to it, not just the experience of IBM Spain." It would be foolhardy to break into Baby Blues. It would also be an acknowledgement of management failure.

Since this breakthrough, Gerstner has been designing ways to decentralize what he calls "the things that matter in running a business" but reinforce the things that benefit from size. That means decentralizing some things and centralizing others. "We're going to be complicated -- there is no simple solution," he says. So while unit managers can expect to define their customers, design their own products, manage most of their costs, and set prices, they'll be expected to cooperate more on such issues as technology and product announcements.

To make IBM more cohesive, Gerstner on Sept. 13 announced two changes in management structure. He created an 11-member executive panel, composed largely of operating chiefs, to make them find ways to work together better. He had already chaired its first meeting. Over lunch on Sept. 10, they talked through several issues, such as the sharing of intellectual property. Gerstner also changed the way panel members are paid. Pay will now depend on companywide results, not unit numbers. By 1995, IBM will alter compensation for another 100 to 200 top managers.

Next, he borrowed from Welch. Gerstner had stopped convening IBM's powerful three-member management committee, which set broad company policy and made many operating decisions. In its place, copying GE's corporate executive council, he instituted a "worldwide management council." Its 34 members -- operating chiefs, geographic managers, and staff heads -- will meet four or five times a year to announce operating results, compare practices, problems, and solutions, and thrash out companywide initiatives.

In August, Gerstner tinkered with one more bit of IBM's structure: He set up the new technology-assessment system. When former IBM President and Chief Technologist Jack D. Kuehler retired in August, Gerstner decided not to replace him. Instead, field experts -- not staff in Armonk -- will now guide technology investments. To enhance his role, Gerstner has already visited IBM's main lab in Yorktown Heights, N.Y. three times.

Gerstner doesn't believe these new structures are enough to prompt change in IBM's ranks. For that, he wants to manage "by principle rather than procedure." It's a hard-to-define concept. "It means when a situation arises, you don't go to a manual. You know in your heart and head what to do." To ignite this cultural revolution, Gerstner will change IBM's training and management-development programs. But he struggles to define what else he'll do. Finally, he says: "There is no formal way to change attitudes. It takes communication and reinforcement."

"It's just getting out and being me," he adds -- without a trace of self-consciousness.

Lou Gerstner is not an immodest man. He can even be unassuming. Nearly everyone calls him Lou, including IBMers who still refer to "Mr. Akers." But Gerstner does speak his mind. When an employee asked, "how are you going to compete with Microsoft?" he bristled and said: "I would have preferred if you had phrased your question, 'how are we going to compete with Microsoft?'" His friend, Vernon R. Jordan, recounts how, on a fly-fishing trip to Alaska in 1992, Gerstner was his roommate. "He would wake me up because he said I was snoring and it was disturbing him," the Washington lawyer says. "What I didn't tell him was that his snoring disturbed me."

When he takes control, Gerstner sometimes riles people by pulling a little too hard on the reins. At RJR, sitting down to dinner with four grocery-chain CEOs, he amazed his guests -- all customers -- by assigning each a topic to discuss. Others tell of times when he's so brisk that he seems rude. He takes his competitive drive everywhere. On the golf course, where he's a mediocre player, he has soured some outings by pressuring partners to win.

Yet by all accounts, Gerstner can be quite charming. He will banter with anyone who can engage him. Colleagues also admire his devotion to his family -- his wife Robin and his children, Louis, a college junior, and Elizabeth, a freshman. When they were young, Gerstner always left work early on report-card day. Even now, he tries to depart around 6:30 p.m., usually with briefcases of papers to review in his traditional brick home in Greenwich, Conn. On weekends, when he's not golfing, he likes to putter around in a vegetable garden. He regrets not having more time at his second home, in Hobe Sound, Fla.

But Gerstner never lets up -- and never has. He grew up in a middle-class Catholic home in Mineola, N.Y., the second of four sons. Both parents worked. Education was the key to success, and Gerstner excelled at Chaminade High School, Dartmouth College, where he earned a degree in engineering magna cum laude, and Harvard business school. He did just as well when he joined the work force, winning accolades for his marketing and strategic skills.

His credentials impressed IBM's search committee. So did his reputation as a quick study whose mind rarely strays far from business. RJR President Lawrence R. Ricciardi remembers their first trip together, when both worked at AmEx. The pair arrived in Brussels and went for a walk. While Ricciardi was taking in the scenery, he noticed that "Lou was walking slowly and looking at all the storefront windows. I soon realized that he was counting American Express decals, to see what percentage of stores we'd penetrated."

Ricciardi says Gerstner wanted this job: "It will be the business story of the decade, so he's got an opportunity now that he didn't have at American Express or RJR. They're not of the same magnitude as redirecting an institution like IBM. So I think he's fired up."

Gerstner is working six-and-a-half-day weeks, 16- to 18-hour days. "There have been some days when I said, 'this is really tough,'" he admits. "I would spend hours absorbing the technology, racing from one issue to another. Just getting my arms around the dimensions of this company can burn you out at the end of the day." Still, he adds, "by the next day, I'm ready to go again."

One sign of his drive: Employees of RJR Nabisco, accustomed to the serious side of Gerstner, have been stunned to learn that he's funny and self-effacing at IBM. But RJR -- having enjoyed the free-spending reign of Gerstner's predecessor, F. Ross Johnson -- hardly needed loosening up. IBM does. The change says something about Gerstner's management abilities: He will do whatever it takes to accomplish a job, even adapting his personal style. Says one RJR manager: "He must be working very hard at this. That reinforces what I've always said about Lou -- he has a tremendously strong will, maybe the strongest I've ever seen."

Like many people, Gerstner sees Sept. 1, as much as Jan. 1, as the start of a new year. So he spent August thinking about his next phase at IBM. York's numbers provided one foundation: He wants to end the company's preoccupation with head count as a focal point of productivity increases. Instead, he will shift the spotlight to IBM's processes and procedures. His new management committees should start producing results, too. And the strategy debate will point him in new directions.

For now, Gerstner has more skeptics than fans. For every Tom Watson, who says he's "impressed" and "very optimistic" about the company under Gerstner, there are dozens of doubters. But he's not concerned. "Outsiders are looking for a single silver-bullet solution for IBM's future, and there isn't one," he says.

Gerstner, in fact, is upbeat. In August, he spent $ 772,250 on 17,500 shares of stock. In June, despite job pressures, he shot his first-ever hole in one, the seventh hole at Monterey's Cypress Point course. And he's settling in at Armonk. He recently rummaged through IBM's basement and found a raft of antique calculators. He's having them placed around the headquarters building, which he likens to an insurance company's. And he's moving some of the machines into his own office.

One CEO who briefly chatted with IBM's search committee calls the task "a 10-year" job. Gerstner wouldn't agree. He would say it was never going to end.

STRATEGY Plans to keep IBM together, offering hardware, software, and services to provide "solutions" for customers, big and small. Started boosting industry-specific expertise. Formed 12 task forces at a strategic planning session in late May in order to study growth opportunities.
   TECHNOLOGY Set up a new technology-assessment system in which teams from around the company will come together to make judgments about technology.
   COST-CUTTING Ordered 35,000 additional layoffs and established nine task forces to trim $1.75 billion from overhead in such areas as information systems, inventory management, purchasing, and transportation.
   NEW HIRES Brought in a new chief financial officer, Jerome York, from Chrysler and a new senior vice-president for human resources and administration, Gerald Czarnecki, from Bank of America-Hawaii.
   MANAGEMENT STRUCTURE Set up a new 11-member executive committee to promote greater corporate cooperation and formed a 34-member worldwide management council that will discuss operating results and company practices and problems.
   PAY Set up a new compensation program that rewards top managers for companywide performance rather than an individual unit's.

Like any outsider, Lou Gerstner had to learn a new language when he joined IBM. But he struck back on Sept. 13, banishing the term LOB -- line of business -- to make IBM more "customer-friendly." Product units will now be called divisions. A look at IBM's lexicon shows that Big Blue's CEO has a lot more work to do.
   BUCK IT UP TO CHQ To send an issue to the corporate headquarters for resolution.
   CRISP UP To improve a foil (overhead slide) presentation.
   DOWN-LEVEL VERSION A document that's no longer relevant, as in "he's working from a down-level version."
PROFS Professional Office Systems, IBM's internal electronic mail system. Also used as a verb, as in "PROFS me on that."
   FLATTEN To resolve an issue, as in "we have to flatten this before tomorrow's meeting."
   PUSHBACK A nonconcurrence, as in "I took the issue past Al, but I got a lot of pushback."
   ALL OVER THE FLOOR A situation that has gone out of control, as in "this product announcement is all over the floor."
   GOAT-ROPING A gathering of the key players needed to resolve an issue, as in "we'll have to have a goat-roping on this PC pricing strategy."
   RESWIZZLE To improve something, as in "Frank's boss asked him to reswizzle his foils." Synonym: To tweak.
   THE APPLE ORCHARD (OR, THE ORCHARD) IBM's Armonk (N.Y.) corporate headquarters, which was once an apple orchard.
   HYPO A high-potential employee, generally destined for management.
   SPEEDS AND FEEDS Technical specifications of products, as in "Give me the speeds and feeds of that new workstation."
   LEVEL-SET To give a brief summary at the outset of a meeting, as in: "Before we begin, let me level-set.''
   BIG IRON Mainframe computers.
   TAKE IT OFF-LINE To discuss an issue, usually on a one-on-one basis, after a meeting.
   BOIL THE OCEAN To exercise vigorously every possible option to resolve an issue, as in "he's really boiling the ocean on that problem."
   NONCONCUR To disagree.
   TALK TO THE FOILS To give a presentation using an overhead projector.
   ONE PERFORMER An employee with IBM's top performance rating, as in "Mary is a real one performer." Opposite: Four check, an employee with IBM's lowest performance rating who could be fired if he doesn't improve in a brief time period, known as the "measured mile."


Copyright 1993 McGraw-Hill, Inc.