In the end, Kenneth H. Olsen simply had too soft a heart. Despite his infamously gruff exterior, the 66-year-old founder of Digital Equipment Corp. still wanted to run the $14 billion minicomputer company as a giant family business, with him as the benevolent father. Even as the company's payroll rose to 126,000 employees, he still made sure that everybody got a turkey at Thanksgiving. After 35 years, he wasn't about to change.
But the computer industry had changed--drastically and irrevocably. With powerful desktop computers stealing minicomputer sales, the DEC of holiday turkeys, lifetime employment, and profit margins fat enough to pay for it all was a thing of the past. The times called for a new DEC, a far leaner and meaner operation. With the old DEC hemorrhaging, on July 16, Olsen finally was pressured to resign.
In a conference call, Olsen told directors that he will step down on Oct. 1. At the same time, he nominated a new CEO to carry out the drastic cost-cutting he couldn't stomach--and to rebuild DEC around a promising new microprocessor technology called Alpha.
TECHNICAL PROWESS. Olsen's handpicked successor, Robert B. Palmer, is a 180-degree turn from the cantankerous, unpredictable founder. Palmer, 51, is regarded as a well-organized and articulate manager. And in contrast to the intensely puritanical Olsen, Palmer is a debonair divorce who drives a Porsche Targa.
But, like Olsen, the new boss is also an entrepreneur who made it on his technical prowess. Palmer worked his way through engineering school and, while still in his twenties, co-founded memory-chip maker Mostek Corp.. Palmer's background in the cutthroat semiconductor business has conditioned him to continuously push down costs. A onetime chip designer, he has also been a driving force in developing DEC's Alpha chip, the key to turning the company around.
Palmer's appointment was approved by the board on July 22. He had already won over directors during his frequent appearances at board meetings. "The presentations he's made are most impressive," says Thomas L. Phillips, a director and retired chairman of Raytheon Co. While Olsen dragged his feet on cost-cutting, Palmer showed the board what could be done. After taking over manufacturing in late 1990, Palmer quickly shuttered 10 factories. His staff cuts account for more than half of the 19,000 positions DEC has eliminated in the last two years. Just last month, Palmer convinced directors to allocate $425 million for a chip plant that he argued is crucial for the Alpha effort and DEC's rebound.
These contacts with the board helped propel Palmer past longtime DEC executives John F. Smith, senior vice-president for operations, and Software Vice-President David L. Stone, who had been seen as likely successors to Olsen. Palmer "sells himself well and sells his programs well," says MasPar Computer President Jeffrey C. Kalb, who brought Palmer to DEC in 1985 to run semiconductor operations. Neither Palmer nor Olsen would speak with BUSINESS WEEK.
RUMORS FLY. With Palmer's appointment, investors and customers breathed a collective sigh of relief. Despite losses, Olsen had held out all this spring against the cost-cutting moves that top executives and his board demanded. Analysts predicted more losses, and by late June, the stock had drifted to a nine-year low of 33 1/4, a dismal skid from an all-time high of 199 (chart). As the company closed the books on fiscal 1992, ended June 30, rumors began to spread on Wall Street that the board had reached its limit and was ready to force the legendary founder out.
Insiders say that Olsen and the board had clashed over everything from sales commissions to the company's direction. Rumors of his departure boosted DEC's stock to 38. When the news finallycame, the stock jumped another four points, and some analysts issued new purchase recommendations. "A lot of people believed Ken was the problem, not the solution," says Marc G. Schulman, an analyst at UBS SecuritiesInc.
The euphoria quickly passed as the enormity of Palmer's task became apparent. The question remains whether it's too late to do anything more than avert disaster. By delaying DEC's day of reckoning, Olsen had raised the costs to staggering heights. DEC is not only years behind rivals in cutting bloated costs and driving into new markets but it is also no longer even profitable. Minicomputer sales have been stagnant for years, and VAX revenues are now plunging as customers wait to buy replacement machines based on the new Alpha chip, due out late this year. The result: DEC is expected to show a $300 million operating loss for the quarter ended June 30, its third quarterly loss in a row. Schulman forecasts operating losses in the next two quarters totaling as much as $208 million.
So, Palmer's first order of business will be to preside over a jarring layoff. Instead of paring workers gradually and giving them months to search for new jobs, as Olsen had done, Palmer is expected to move decisively. Some 15,000 employees, including 8,000 in the U.S., will be fired and given a week to clear out. The numbers are so big that the company plans to stagger the layoffs over the next few weeks to avoid overloading unemployment offices in New England. Severance pay and other downsizing costs are expected to push the fourth-quarter net loss as high as $2 billion.
With DEC running an operating loss, these charges are tearing through the company's cash. By June, 1993, estimates PaineWebber Inc. computer analyst Stephen K. Smith, the company will have burned up $1 billion of its current $1.5 billion in cash. As a result, DEC will have to borrow just to fund operations over the next few quarters.
PAINFULLY LATE. To turn DEC around over the long term, Palmer has to make Alpha a hit. The technology, a so-called RISC (reduced instruction-set computing) microprocessor, will replace the VAX minicomputers, based on technology first designed in the mid-1970s. In recent years, DEC's share of the mini market has fallen to 14%, while IBM's hot-selling AS/400 series has given Big Blue a 27% share. DEC is also counting on Alphato help grab a share of the fast-growing workstation market, where it trails Sun Microsystems Inc. and Hewlett-Pack-ard Co.
On paper, at least, Alpha is the fastest RISC microprocessor yet. "Alpha is clearly a leapfrog technology," says Wes Melling, program director of midrange computing research at consultants Gartner Group Inc. And he says DEC's huge library of VAX software should make Alpha-based machines appealing in the broad office-computing market, in contrast to other RISC machines, which have been used primarily by engineers. Another potential boost for Alpha will be Microsoft Corp.'s Windows NT, a new operating system that the software giant is now adapting for the Alpha chip. That should allow Alpha machines to run all sorts of popular PC software. That's in addition to the VAX software and OSF/1, a version of Unix.
Still, DEC's entry is painfully late into a market where IBM, Hewlett-Packard, and Sun have been shipping products for several years. And these companies have also signed up other computer makers to use their RISC chips. On IBM's team, for instance, are Apple Computer, Motorola, and Groupe Bull of France. DEC, by contrast, has only one sizable Alpha partner so far, Italy's Olivetti. Japan's Kubota Corp. and supercomputer maker Cray Research Inc. plan to use the Alpha chip, but not in great numbers.
To leapfrog his competition, Palmer will have to shore up DEC's badly demoralized organization. Olsen's increasingly idiosyncratic management style over the past few years has driven dozens of exasperated executives and managers from the company. The most recent, defection was Pier Carlo Falotti, President of DEC Europe, who has been named President of ASK Computer Systems Inc., a California software maker. Falloti "was very unhappy" says Bruno D'Avanzo, a former DEC Europe executive and now a managing director at Olivetti. "Olsen's erratic habits really got to him."
MARKETING FAILURE. A series of reorganization and strategy shifts have also left customers confused. In the 1980s, DEC surged in the market with a clearcut message: Its VAXSes, ranging from desktop to mainframe-caliber machines, all ran the same software and could therefore be easily networked anywhere in the world. That message was lost as DEC cast about the new growth strategies in recent years. Peter Daboul, vice president for information systems at Massachusetts Mutual Life Insurance Co., says now he is in "a wait and see mode" regarding DEC. Although it uses four large VAXes at the home office and smaller ones in its branches, Mass Mutual has no plans to buy more DEC gear. "The key is how their new captain carries out this course correction," Daboul says.
Indeed, Palmer's success is anything but certain. His background in microchips and operations give him a good foundation. But DEC also badly needs marketing skills--an area where Palmer has yet to make his mark. "He doesn't have that much background in computers or for that matter, in marketing or sales," says MasPar's Kalb. He believes that DEC somehow lost the ability to communicate the value of its VAX software. That in turn, has made it increasingly difficult to sell its relatively high-priced machines in a sea of cheap hardware. "It was a failure of marketing," Kalb explains.
If Palmer's rapid rise, as insiders say, stemmed from his ability to sell himself to top management, DEC's resurgence may well depend upon his ability to sell DEC to customers who have grown skeptical. In Olsen's old Dec, the hard sell was frowned upon. The puritanical founder though such tactics were unseemly. For the new DEC, they may mean survival.