Cover Story

The New IBM

Is it new enough? Chairman John Akers has a bold plan to remake Big Blue

And it may work, if he can crack the corporate culture

By John W. Verity and Thane Peterson in Armonk and Deidre Depke and Evan I. Schwartz in New York. With bureau reports.
Business Week

December 16, 1991

At first, it looked like the kind of news that investors had long hoped IBM would make--a top-to-bottom shakeup to clear out the deadwood, ''empower'' the remaining workers, and finally begin the process of restoring Big Blue's luster. With rumors of an upheaval spreading, IBM issued a news release on Nov. 26, describing a bold plan to create autonomous businesses from the monolithic computer giant--perhaps even spinning off entire new companies with their own stock and boards of directors. At the same time, 20,000 workers would be eliminated, and not just through the traditional IBM voluntary programs that had inadvertently encouraged the best and brightest IBMers to leave. This time, IBM was going to cut back by firing the laggards.

The plan, which top executives at the company's hilltop headquarters in Armonk, N. Y., herald as ''The New IBM,'' isn't quite as radical as it first seemed. But as outlined to 500 managers flown in from divisions around the world on Dec. 4, it creates a new corporate structure that would begin by decentralizing authority and decision-making from IBM's all-powerful management committee and ultimately leave some parts of the computer giant almost freestanding, independent companies. Starting with a handful of units now, IBM will create perhaps dozens of semi-independent operations. They will be freed of big corporate overhead expenses and, over time, will assume their own profit-and-loss responsibility.


The new IBM would resemble a holding company. Headquarters executives will be there to offer guidance--if needed. But heads of the new units, in theory, will act as CEOs. They'll present annual business plans and promise IBM a certain return on funds invested. If they succeed, they'll earn a share of the profits. Some may even be allowed to float a portion of their shares in public markets.

If IBM really can disperse decision-making from Armonk, it stands a chance of unleashing enormous resources. Hidden within the troubled $67 billion goliath are the world's largest personal computer company, a software organization five times the size of Microsoft Corp., and a disk-drive maker that dwarfs industry ''leaders'' such as Seagate Corp. (chart). But many of these units have been held back by headquarters-mandated policies that stopped them from competing flat-out with nimble young upstarts. More significant, because so many important decisions have been left to headquarters, managers haven't been motivated to take risks--or responsibility for decisions that backfire.

That has already changed. IBMers--even those at the highest levels--now know that their jobs are at stake if they don't deliver the goods. CEO John F. Akers telegraphed that message with stunning clarity on Nov. 25 when he removed George H. Conrades as head of the company's debilitated U. S. sales division. Robert J. LaBant, who helped lead the company's resurgence in minicomputers, was rewarded with the top sales job. A few days later, Carl Conti, head of the company's beleaguered mainframe unit, announced that he is taking early retirement. The moves sent shock waves through IBM (page 39).

To most corporate executives, this may sound like Management 101. But at IBM, it constitutes what Akers is calling a ''fundamental redefinition'' of the business. It spells the end for the Big Blue cocoon, an organization that insulated most managers from unpleasant details like profit-and-loss statements and kept them employed even when their performance was mediocre. Industry-watchers say these days even Akers could be axed if the board finds his performance wanting.

Understandably, there are plenty of skeptics. Reshaping and decentralizing IBM, General Motors, or any other giant organization is a Sisyphean task that takes years of reprogramming, observes management consultant Tom Peters. ''Corporate culture has more to do with the mind than with the organizational chart,'' he cautions.


No IBM-watchers are more skeptical than those on Wall Street, where a series of comeback plans over the past few years has raised hopes that were inevitably dashed. Even before details of this reorganization were made public, analysts had second thoughts. The stock jumped immediately after the Nov. 26 announcement but soon fell as analysts learned that the changes would be evolutionary, not the sudden overhaul that the announcement implied. They began focusing again on the near term--on the $3 billion charge IBM will take to eliminate jobs and continuing problems in businesses ranging from mainframes to laptops. By Dec. 4, the stock had slumped to 90, its lowest point since 1983.

Wall Street's big fear is that this is another halfway measure--like the $2.5 billion writedown in 1989 that was supposed to restore the company to fighting trim. ''They've done it, but they haven't done it,'' says Paine- Webber analyst Stephen Smith of the reorganization. ''What's wrong is that there's one sales force. What needs to happen is that each of these businesses gets their own.'' To really free IBM's business units to compete, he asserts, they need their own sales teams, rather than relying on the corporation's mainframe-oriented reps.

And, Smith says, cutting 20,000 jobs is not enough. That move, he estimates, will restore just one point of gross profit margin, which has dropped from 56% to 51.5% since 1986. Even with the $1 billion that IBM says it will lop off expenses next year, he estimates that it will only earn $7 per share, down 33% from 1990.

The new IBM may not be gestating as quickly as investors might like, but IBM is reaping praise for at last moving in the right direction. ''The restructuring is a heroic effort to rebuild a leadership role for a company that is one of the industry's greatest success stories,'' says Michael Mirow, vice-president for planning at Germany's Siemens. ''It will be a very difficult task.'' Adds John Sculley, chairman of Apple Computer Inc.: ''If IBM does this well, it could really surprise everyone and come out a juggernaut again.''

Clearly, IBM has little choice but to try bold measures. In the past six years, it has seemed all but powerless to respond to the accelerating pace of change in the computer industry. The company that started off the 1980s by declaring itself America's best defense against a surging Japanese computer and electronics industry limped into the 1990s. It had fallen behind in PCs, dropping from a 27% market share to 16.5%. It was a late starter in the hot new markets for workstations and laptops. And it failed in a massive software-development effort that was intended to keep its mainframes relevant in an era of superfast microchips. All in all, IBM's share of the world computer market dropped from 36% to 23% in the 1980s, according to the Yardstick Service of market researcher Gartner Group.

As rivals chipped away at its markets, IBM's profits sank. Starting in 1987, the company began eliminating jobs through attrition and costly early-retirement programs. In all, 53,000 jobs were cut. Still, IBM's earnings have continued to decline (chart). Worse, with economic slowdowns crippling demand in all its major markets, revenue is expected to decline 2%, to $67 billion, the first revenue drop since the 1940s.


That dismal picture, say company insiders, finally spurred Akers to get to the heart of the matter. In an era of rapid technological change, where the life cycle of a notebook computer is as short as three months, for instance, IBM's old management structure is an albatross. Endless meetings are required to review new product plans, making it nearly impossible for IBMers to bring innovative machines to market quickly.

Yet getting to market first with new technology, above all else, is the name of the game. ''You can be the most brilliant designer in the world. But if someone else uses a newer generation of technology, their box will be smaller, it'll run faster, and it will cost less,'' IBM President Jack D. Kuehler often says. And by being first that other guy earns enough to move quickly to the next technology and beat you again, he notes.

Despite a 1988 reorganization that was supposed to speed up decisions by giving more authority to six group executives, even now few moves are made without headquarters approval. Former executives say Armonk policy continues to be dominated by the mainframe computer division, which still delivers half of IBM's overall profits. As a result, products or marketing initiatives that are even vaguely competitive with the mainframe have often been derailed, former IBMers say. ''The mainframe guys have been a ball and chain around IBM's neck,'' says James N. Porter, a disk-drive industry consultant.

Take one recent example. Last year, IBM's PC unit was ready to launch a new model to get it back into the fast-growing laptop market after an embarrassingly poor showing a few years earlier. To be competitive, division chief James Cannavino priced the machine at $4,995. Armonk, however, insisted on a $5,995 tag, to hit corporate profit-margin targets. That made it easy for dozens of laptop makers to undercut IBM's price. Now the price has been cut to $3,645. But the laptop has never recovered from its slow start.

Theoretically, that won't happen in the new organization. Cannavino's $14 billion PC group, one of several that will be given relatively wide latitude, will be permitted to set its own pricing in all regions where PCs and workstations are sold. And for the first time, Cannavino will make decisions about how PCs and workstations are manufactured and developed. The new structure gives his unit control over long-term development so that PC researchers won't have to compete with mainframe developers for resources.

The price for this freedom will be accountability. Starting with the 1992 annual report, insiders say, IBM will break out the profitability of various groups. Managers will be measured against profit goals. They'll be free to reach those targets and have a big incentive to succeed. In the new IBM, employees who fall short won't be around indefinitely.


In a small way, IBM has already seen autonomy work. For example, the Rochester (Minn.)-based division that builds hard disk drives for IBM PCs and other machines has been set loose to pursue new markets. It now sells to other computer makers, including Apple. The division, which is being combined with mainframe disk and tape units, will get more autonomy now.

A big plus from the new structure may be greater leeway in doing business with other companies. IBM has entered into hundreds of technology, marketing, and consulting partnerships in the past decade, but its track record has been spotty, often because headquarters bureaucracy slowed things down. There are ''too many people, too many layers of management,'' complains Gerald Cohen, CEO of IBM software partner Information Builders Inc. The decentralization ''makes me feel better and better about the partnership with IBM,'' says Apple's Sculley.

But there are serious risks for IBM. Freeing a bunch of Baby Blues could diffuse IBM's focus. One of IBM's great strengths has been its ability to cobble together comprehensive information systems. Each of the IBM pieces might not be the best of breed, but customers have been willing to sacrifice some performance and pay a higher price to get IBM to handle the complexity. That's one reason why IBM's sales force is expected to remain intact. ''We'd be very concerned if we had to spend all our time coordinating among IBM business units to get something done,'' says Steve Bouch, director of technical services at British Telecommunications.

Still, IBM has little choice but to unfetter its businesses. Markets such as software, services, PCs, and workstations have long-term growth potential, while IBM's mainframe business is looking increasingly vulnerable. To move its largest ES/9000, it is cutting prices by as much as 50%. Even with discounts, however, customers aren't very interested because they can get the same computing power for one-third to one-half less by buying a used IBM processor. ''There's no compelling reason to buy the product,'' says analyst Stephen Cohen of Soundview Financial Group.

The biggest threat to mainframes, however, is from new, superfast microcomputers. The fastest RISC (reduced instruction-set computing) chips now match the speed of million-dollar mainframes. Machines built from gangs of microprocessors can run circles around them. Thus, the question for IBM is: ''Do we eat our own children, or let others eat them for us?'' says Marc Schulman, an analyst at UBS Securities Inc. If IBM is serious about letting its product groups compete freely, he says, the market will soon see cheap mainframe-caliber machines built from the RISC chips used in IBM's workstations.

The most liberating effect of IBM's reorganization could be in software. With broad, strategic goals in mind--mainly boosting mainframe demand--IBM has pushed convoluted ideas such as Systems Application Architecture (SAA), a scheme that was supposed to give its different computer lines a sheen of compatibility. The effort has yet to reach that goal. ''I haven't even heard the term SAA from IBM in a year,'' says John Imlay, chairman of Dun & Bradstreet Software Services Inc.

Now, IBM units will presumably be free to pursue ''open,'' industry-standard software more vigorously. ''The world's voting with its dollars, and it has said it wants standards independent of IBM,'' says John B. Jones Jr., who tracks IBM at Montgomery Securities.


Furthermore, once they assume profit-and-loss responsibility, IBM's software executives are less likely to spend millions pursuing corporate strategies that don't mesh with the market. The prime example is OS/2, an ambitious project to replace the MS-DOS operating system of the original PC. IBM wanted OS/2 to help tie PCs to mainframes. But it was too complex for most customers. Microsoft quit the OS/2 effort and is now cleaning up with Windows, a simpler MS-DOS upgrade. Still, IBM continues to spend heavily on OS/2. ''It's incredible how they've been able to fund this,'' says Microsoft Chairman William H. Gates III. ''They never would have spent so much on OS/2 if software were a separate company.''

Can Big Blue really change its ways? Even IBM well-wishers question how far the company will go. Indeed, it was Akers who undid earlier decentralization efforts by abandoning his predecessor's ''independent business units.'' The biggest, Entry Systems Div., had ridden the IBM PC business from a 12-man operation in Boca Raton, Fla., to a $5.5 billion, market-dominating business in 1985--mostly by breaking IBM rules. As soon as he became CEO, Akers drew the division deep into the corporate fold.

And no matter how sincere Akers' intentions are now, he and other people molded by lifetimes at IBM may not be equipped to create a truly new Big Blue. ''It will take someone from the outside to change the culture,'' concludes Gilles Tugendhat, a former IBM manager and now president of Paris-based ECS, the leading independent lessor of IBM equipment in Europe. That's the kind of shakeup Akers still hopes to avoid.


Copyright 1991 McGraw-Hill, Inc.