Businessweek Archives

Digital's Turnaround: Time For Phase Two


Information Processing: COMPUTERS

DIGITAL'S TURNAROUND: TIME FOR PHASE TWO

When a man knows he is to be hanged in a fortnight, it concentrates his mind wonderfully.

--Samuel Johnson

Around Digital Equipment Corp., they call it Black Friday. Everybody seems to remember where they were when they heard the news. Pauline A. Nist, a top computer-systems executive, was in England when her secretary called to report that back at headquarters in Maynard, Mass., people were "walking around acting like the sky had fallen in." Lawrence G. Walker, now head of Digital's networking unit, says he got the news during the middle of the night in Australia "and literally fell out of bed."

Black Friday--Apr. 15, 1994--was the day Digital stunned Wall Street and even its own executives with the news that it had lost $183 million in the prior quarter. The loss far exceeded DEC's internal projections, exposing a startling weakness in its financial systems. Worse, the quarter was supposed to herald a long-awaited rebound after three years of red ink and restructurings. Instead, it laid bare the ugly truth that Digital was still in an uncontrolled downward spiral. CEO Robert B. Palmer's strategy lay in tatters. Cash on hand was down to $1 billion--when DEC was ordering layoffs and plant closings that would cost $1.2 billion. Industry watchers began to whisper the unthinkable: Digital, once the No. 2 computer company, might not survive.

But if Black Friday was DEC's darkest day, it may also have been a crucial turning point. Facing the corporate equivalent of Johnson's noose, Palmer and his team cobbled together a last-ditch turnaround plan that finally addressed some deep-seated problems. Announced last July, the plan shifted sales from DEC's sales force to hundreds of resellers--in the process, slicing thousands of jobs. It also scrapped Digital's once-innovative matrix management scheme, which had deteriorated into a politicized jumble of inefficiency and finger-pointing. Instead, it has been grouped into a series of mini-DECs, organized by product line, each responsible for its own bottom line and geared toward emulating the best in each industry segment.

The progress since then has been remarkable. Costs are down nearly 20%, and the shrunken workforce is bringing in higher revenues. There are also hot new products, including the AlphaServer 8400, a system for running networks that uses up to 12 of Digital's speedy Alpha microprocessors, and the new HiNote laptop, which has garnered rave reviews. Digital posted slim profits in the second and third quarters of fiscal 1995, the first back-to-back quarters of black ink since early 1991. The cash crunch is over and DEC's stock, which hit 18 1/4 last April, is in the 40s. "I think the future for Digital is very bright," says Palmer. "But that's not to say that we've yet completed by any means the financial turnaround or the strategic turnaround."

"NOT A STRATEGY." Indeed. Nobody is suggesting that the company's long-term success is assured. Once the king of minicomputers and a market pioneer in other areas, the company has fallen to second, third, or worse in nearly every industry segment. In effect, Palmer is trying to play catch-up on a dozen fields while quarterbacking a $14 billion giant.

Still, with Digital's finances on the mend, Palmer says he can concentrate on the strategic turnaround. Working with consultants from Bain & Co., the company has hatched a new corporate strategy intended to reignite growth. Palmer, a 54-year-old native Texan and semiconductor-industry veteran who co-founded Mostek Corp. in 1969 and moved to Digital in 1985, says the basic thrust is to leverage DEC's deep expertise in networking.

Digital, he claims, "knows more about networked computing than anybody. You can buy a PC or even a complex system from any number of vendors," he says. "But how do you get all this stuff to work effectively and reliably in your network?" The company, then, will focus its investments on hardware, software, and services that support this overarching ambition. Though he concedes that Digital "is not necessarily the best in specific" product lines, he says its strength lies in "the package--integrating, supplying, and deploying distributed computing environments."

Nobody doubts that there's a need for such expertise. But, says Gresham T. Brebach Jr., a former head of Digital's consulting unit, "There's a clear disconnect between the way they describe the strategy and the way they've arranged themselves to bring products and services to market." If networking is truly the goal, he argues, Digital should have kept an earlier structure that offered "solutions" for specific industries. "Connectivity is a historical core competency," he says. "But a core competency is not a strategy."

Moreover, DEC's claim to network superiority is open to question. A decade ago, it was the acknowledged leader, but the rest of the world has not stood still. "There is no way on the planet that Digital is the best at networking," says Philip C. Anderson, a professor at Dartmouth's Tuck Business School who follows the computer industry. "They can do a good computer network, but the idea that they're better than the next 10 or 15 companies is ridiculous."

Some observers feel that Digital's networking plans, even if successful, will do little more than add a few margin points to what is really a hardware-oriented strategy. The new Digital, they argue, looks like the Digital of the 1970s--when it was known as a "hot box" company and before it tried to take on IBM. Then, as now, it sold high-performance hardware mostly through third parties, with services and software playing a supporting role.

There's nothing wrong with a hardware strategy--it made DEC the huge success it became in the 1980s. But to keep up with today's hot-box companies--Compaq, Silicon Graphics, and Sun Microsystems, to name a few--means gearing the company to a low-margin, commodity-like environment. Even today's slimmed-down Digital may not be ready for that. DEC's revenue of $235,000 per employee looks downright inefficient next to IBM's $317,000 or Sun's $403,000. "If all they want to be is a volume hardware player, they still have too many people," warns Lynn Berg, a research director at Gartner Group.

COMPUTING ON THE GO. Palmer bristles at the idea that Digital is reverting to simply a hot hardware company. "That would leave very little opportunity for good margins or to grow the business," he says. Digital executives claim they already have many of the pieces to support their networking vision and promise to reveal more in coming months. They plan to push a raft of Digital software products that allow computers running different operating systems to communicate and share programs. And the company recently introduced a service to maintain and continually upgrade a customer's PCs for a fixed monthly fee, taking away the headache of trying to outguess the pace and direction of PC technology at the endpoints of the net.

Digital also is focusing on several potentially hot new markets involving complex networks, notably video servers, the Internet, and mobile computing. The company claims it has won about half of the worldwide contracts for video servers, complex systems that will deliver TV programming on demand. And later this fall, Digital plans to unveil a software program aimed at helping make a mobile PC act like it's connected to a corporate network, even when it's not.

While Digital waits for its still-vague networking approach to pay off, it will have to rely on its mainstay: hardware. Rivals such as Compaq Computer Corp. and HP have cast their fate with the industry-dominant Intel architecture, but Digital has adopted a twin-track approach. The first track is the building of PCs and larger servers using Intel Corp. chips. DEC's PC unit has leapt from an asterisk on industry market share charts three years ago to No. 11 last year and is predicting a sizzling 60% growth rate for fiscal 1996. But Digital executives say they've dialed back their ambition to break into the top five this year, and some concede the PC unit is unlikely ever to show big profits, although it does provide badly needed entry-level products to bring DEC new customers.

ALPHA AND OMEGA. Digital's second track--its real edge in hardware--remains its super-fast Alpha microprocessor. The chip was the first with 64-bit architecture, which means it grabs data from a computer's memory much faster than industry-standard 32-bit chips. Bringing it out was a gamble for Digital, because the new architecture required software makers to adapt programs at great expense. Alpha also required huge investment in R&D and chipmaking that would not be recouped if the microprocessor didn't catch on in a big way.

After an agonizing lag, Alpha's speed is beginning to attract key software. "The price-performance is impressive," says Ron Robbins, president of RTP Technology Corp., a Fair Lawn (N.J.) reseller. "We're taking databases off mainframes and putting them on Alpha, and crunching data in 2 minutes vs. 22." In April, Oracle Systems released an Alpha version of its popular Oracle 7 database, a combination that Oracle claims produces by far the industry's best performance. John B. Jones Jr., an analyst at Salomon Brothers Inc., expects Digital's Alpha systems sales will soar 84% in fiscal 1995, ending June 30, to $1.7 billion, and may rise another 55% by next June. While that's impressive, many of the sales simply replace older Digital hardware.

Alpha still faces a long-term problem: The chip hasn't won a single influential convert among other computer makers. That could ultimately prove fatal when it comes time to fund the mind-boggling cost of succeeding generations of chips. Digital executives concede that they will have to find a solution, perhaps by finding a partner to share the cost of the next chip plant. But they say a new $450 million plant in Hudson, Mass., will carry Alpha into the next century.

Meanwhile, they count on Microsoft Corp.'s Windows NT operating system to fuel Alpha sales. NT, which is catching on quickly as an operating system for network servers, runs on a variety of chip designs. In theory, that will help eliminate Alpha's biggest handicap: lack of applications software. Then, Alpha's superior performance will really stand out, says Ed Caldwell, chief of Digital's semiconductor division. "We will be able to so outperform Intel and everybody else that we'll be able to take over the high-end desktop, period."

Will it work? Some outsiders point out that the best technology rarely wins on its own. Such non-technical issues as marketing ability, market momentum, and availability of applications software usually determine long-term winners. "There's absolutely no history of things migrating toward the fastest chip available," says Gartner Group's Lynn Berg. Computer power is advancing so rapidly, she says, that most customers will stick with a chip that's slower--as long as it runs popular software. Her conclusion: "There's a lot of wishful thinking going on" at Digital.

Not as much as there once was. Before Black Friday, Palmer and his executives were so confident that Alpha would overwhelm the competition that they never reexamined their basic assumptions about the business. They spent heavily on building a sales force to sell end-to-end solutions in specific industries, such as health care and financial services, and clung to nonstrategic businesses, such as making disk drives.

Black Friday made it clear that Digital had neither time nor money for nonessentials. Suddenly faced with questions about his leadership from shareholders and Digital's board, Palmer swung into action. An engineer by training--he's the kind of guy who keeps a log of his daily running in hopes of logging 55,000 kilometers by age 55--Palmer ordered drastic cuts. Gone are the consulting unit, the disk drive operation, and database software.

Former executives say the CEO latched onto a rescue plan hatched last spring by Bain consultants and Enrico Pesatori, an ex-Olivetti executive and former CEO of Zenith Data Systems, who had relaunched Digital into PCs with surprising success. His secret: the PC unit was a stand-alone entity, aloof from Digital's matrix setup. "The matrix structure was a major part of our problem," says Charles F. Christ, who heads the components unit. In the matrix, managers from various disciplines met in committees that were supposed to come up with better ideas through cross-pollination. But responsibilities were blurred, and when Digital started its slide "we ended up with a lot of endless discussion about what ought to be done," he says. The system masked costs and forced managers to waste time wrangling over internal procedures.

Only after the computer systems unit was separated from the matrix, says Pesatori, did managers discover they were losing 30 cents on every dollar of sales. Sales and administration costs were 2 1/2 times higher than one rival's. Pesatori, who now heads a division overseeing both Alpha-based servers and PCs, decided the company could no longer afford to sell to 10,000 accounts. He shrank the sales force to focus on just 1,000 top clients, boosting the percentage of sales through indirect channels from 38% last April to 50% today. Pesatori says the unit's costs are competitive with industry averages and will break even by the end of June.

It's a long way from Black Friday. But to a large degree, the accomplishments of the past year have merely brought Digital to the point where it's ready to compete again. If Palmer can't make DEC stand out with his networking strategy, the company risks following the path of another former industry No. 2, Unisys Corp., which now mostly serves its old base of customers and is watching its revenue slowly shrink. "We are not going to turn ourselves into a new Unisys," vows Pesatori. "Digital is a company that's too proud, where people believe they will make a difference in information technology." Once again, Digital's mission is to deliver on that promise.

THE PALMER METHOD

"I think the future for Digital is very bright... The company knows more about networked computing than anybody"

DONE

-- Refocused DEC core Alpha hardware and networking, jettisoning database software, consulting, and disk storage.

-- Shifted most sales to indirect distribution channels, slashing costs by signing on computer resellers as key selling partners.

-- Relaunched PC business

IN PROGRESS

-- Playing a hardware game but avoiding commodity margins by adding value in

networking.

-- Positioning company to take advantage of key new trends, such as mobile computing and video on demandBy Mark Maremont in Maynard, Mass.


Burger King's Young Buns
LIMITED-TIME OFFER SUBSCRIBE NOW

(enter your email)
(enter up to 5 email addresses, separated by commas)

Max 250 characters

 
blog comments powered by Disqus