WHO'S AHEAD IN THE COMPUTER WARS Look whose earnings are down: IBM, Apple, Digital, and more. Not all big companies will survive the rough 1990s. But the smart and the tough stand to prosper.
By Brenton R. Schlender REPORTER ASSOCIATE Mark Alpert

(FORTUNE Magazine) – DOES THIS INDUSTRY sound as if it's in trouble? Last year computer hardware and software sales in the U.S. grew by about 10% -- triple the rate of the overall economy -- and accounted for 2% of the $5.24 trillion GNP. Worldwide, computer makers, racing along at a 13% clip, rang up more than a quarter of a trillion dollars in revenue. The industry is still the fastest-growing commercial endeavor in history -- by far. But it has seen better days. Sales growth in 1989 skidded in nearly every segment of the industry, from stately mainframes to spunky PCs. IBM, the bellwether that competes across the board and around the world and that brings in more than a quarter of the entire industry's revenues, saw its growth slow from 10% in 1988 to 5%. In December, Big Blue announced a $2.4 billion restructuring that caused a 35% drop in profits for the year and ultimately will trim 10,000 from its U.S. payroll. Unisys and Data General also took big write-downs to cover the costs of laying off people and closing plants in 1989, and posted losses for the year. Recently, Digital Equipment Corp. announced a 44% quarterly drop in earnings on flat revenues, and Apple Computer released figures showing an 11% quarterly decline in profits. Last year's sales of publicly held U.S. computer companies as a group grew only 5%. Wall Street is so put off that computer stocks, once the ultimate growth issues, trade at price/earnings ratios similar to those of stodgy utility companies and shrinking Wall Street investment firms. Has the computer industry, like the postwar baby boom it was born with, reached middle age? Not exactly, though its adolescent growth spurt is probably over. Have computer makers given the world more data-processing power than it really needs? Definitely not. Will the next decade see a huge shakeout among computer makers, leaving only a few bloodied survivors? No, although competition will be fierce and a few companies will fail. Are the industry's go-go days gone-gone? No, even the gloomiest analysts say that existing markets are far from saturated and that fresh technologies will create booming new ones. Well, what's everyone so worried about? The problem is this: The industry is caught in a catch-22 of its own making. Competition and technology have combined to help manufacturers drive down the cost of computing power and, with it, the price of the machines. As a result, companies have to sell many more cheaper computers just to stay even with the previous year's sales and earnings, let alone grow. Without revenue and profit improvements, computer makers cannot develop the technologies necessary to keep up in this accelerating race. Says Jimmy Treybig, founder and CEO of Tandem Computers, one of the healthier companies: ''The economics of our business are frightening. We can deliver a lot more power for a lot less money each year, but the industry isn't growing as fast as it once did. The bottom line is, if your company doesn't have big growth you have a horrible problem.'' While prices are coming down and performance is charging up, innovative new computers are blurring the distinctions between the traditional computer markets. A PC doesn't have to be just a word processor or a number cruncher anymore. With a little extra computing power, it can compete with its big brother, the technical workstation, for the hearts and minds of scientists and engineers. A bunch of PCs tied together by a network and linked to a data- filled electronic library called a file server can take over whole corporate departments and small businesses. This setup -- $3,000 for each PC plus $25,000 for the file server and the network -- does the job previously handled by larger minicomputers that cost from $100,000 to $500,000. In fact, it does more: You can play games or keep your personal calendar on a PC; you can't with a mini. As PCs are moving up in the world, the technical workstations are traveling down, dropping below the $10,000 price barrier and appealing to customers with broader interests and lower IQs than nerds. Since these machines were designed for a bunch of scientists to use together, they are networking naturals. That makes them worthy competitors of the PCs in the office. Not to be outdone, the latest minicomputers, harnessing several of the same microprocessors that drive PCs and workstations, have the horsepower of million-dollar mainframes for less than a fifth the cost. These machines are called ''multiprocessor microcomputers,'' and you might as well get used to the mouthful name: These will be hot machines in the 1990s. CONFUSED? Now you know how customers feel. But they are getting smarter and more demanding and are also recoiling from the plethora of new technology that has hit the market in recent years, particularly machines that haven't delivered the benefits their manufacturers claimed. Apple Computer CEO John Sculley tersely concludes, ''This industry may have hyped itself into a recession.''

Facing an industry slowdown -- probably not a recession -- for the next two or three years, computer makers conclude that they must change quickly or they will fade away. For some, that simply means adding product lines and attacking new markets. But for others it portends an anguishing identity crisis as they try to find a secure place in the industry's new order. In addition to corporate soul-searching, the computer industry in the 1990s will see the following trends: a push for hardware and software that will link existing machines from different companies together; the development of industry standards so that new machines from different companies will talk to each other from the very beginning; and a continued emphasis on low-cost manufacturing. Two brand-new markets will emerge: image processing, which electronically stores all kinds of paperwork from handwritten letters to charts to photographs and drawings, and on-line transaction processing, which will spread the technology of airline reservations systems and automated teller machines to just about any business. Much of the action will be overseas, though as the recently announced merger of the computer operations of Siemens and Nixdorf shows, the industry in Europe already seems to be consolidating. While the U.S. computer market continues to grow at a 10% annual rate, international sales will do a bit better -- 12% or so. Computer companies that can take advantage of these trends will win; those that ignore them won't be around for the turn of the century. It was all so much simpler in the early and middle Eighties, when most companies comfortably kept to their own market segments and sold proprietary equipment that effectively locked out other vendors. ''The PC changed all that,'' says Ray Thomas, information systems manager of Hadson Corp., a defense contractor and energy company in Oklahoma City. ''When lots of business managers became exposed to computing, they discovered a whole range of new problems they could use computers to solve. The key is connecting everything together so that the PCs tap into the company's bigger computers the same way minicomputers can hook up with mainframes.'' MICHAEL SIMMONS, an executive vice president at BankAmerica, makes the point: ''By the end of this year, the computer power on desktops will be 40 times the power in the old glass computer rooms. We have the computer power, so now the issue is networking, not more computers, and that creates a problem for the makers of iron.'' Steve Jobs, co-founder of Apple Computer and now chairman of Next Inc., makes the trend sound like advertising copy: ''While the Eighties were the age of personal computing, the Nineties will be the age of interpersonal computing.'' Companies are only now recognizing the value of hooking PCs together. Says Cheryl Currid, director of applied information technology for Coca-Cola Foods in Houston: ''The beauty of networks is in how we've seen people's work become more meaningful. For example, our secretaries can handle the daily schedules of more executives.'' And her division can get by with fewer secretaries. What about all the computers sitting around corporate America, effectively antisocial and mute? Hewlett-Packard has put together a comprehensive and innovative strategy to get those computers, from PCs to mainframes, to talk with each other. Over three days in January, the company introduced an unprecedented 26 new computers and workstations that should help get conversation started. The key component is a piece of H-P software called New Wave Office that allows groups of people on PC networks to share different programs and data -- often supplied by one of Hewlett-Packard's popular minicomputers -- and makes electronic mail as easy to use as a telephone. H-P is also at the fore of the ''open systems'' movement that is trying to establish standard computer, software, and networking designs for the entire industry. Customers have been clamoring for years for the ability to mix and match equipment from a variety of vendors. Hewlett-Packard chose to supply it even though, as one senior manager says, ''open systems don't necessarily make much business sense because they ultimately drive down prices and margins even faster.'' The company felt it had no choice. Says CEO John Young: ''The balance of power has shifted from the vendor to the customer, and you have to respect that.'' You also have to respect the results of Young's new strategy: In each of the past two years, H-P's computer sales have grown 25%. DATA GENERAL and Unisys have joined the stampede to open systems. But customers claim they haven't seen many benefits -- yet. Says Mark Teflian, vice president of information systems for Covia, the company that runs United Airlines' reservations system: ''We got to the end of the Eighties and looked around and saw that everything the computer makers are promising is still in pieces. All the technology is there for very strong growth for the computer industry, but there's no cohesive way to put those pieces together. We've had to do it ourselves.'' Manufacturers with the solutions to the jigsaw puzzle like Hewlett-Packard, should prosper in the next decade. Since computer makers will find it hard to keep growing the way they used to, they will try to cut costs to preserve their profit margins. One way: reducing the money spent on manufacturing. IBM and DEC have historically produced almost all their own components. On the other hand, many makers of microprocessor-based computers -- PCs, workstations, and the new breed of multiprocessor minicomputers -- use off-the-shelf parts and rely on others to make and assemble most of their circuitboards and components. Hence their manufacturing overhead is much, much lower. Says Edson D. de Castro, Data General's chairman: ''Apple produces almost four times Data General's revenue with about the same number of employees.'' No surprise, then, that so many companies, including IBM and DEC, are consolidating manufacturing operations, eliminating jobs, and even looking to a few outside suppliers. Some computer executives believe the reliance on other manufacturing sources, while an advantage now, could eventually backfire. ''Too much of the infrastructure has moved offshore,'' says Jobs. ''I worry that someday there won't even be a manufacturing industry in the U.S. What will that do to our historical dominance of this industry?'' His Next Inc. is fighting that trend. Although it buys its chips, monitors, keyboards, and disk drives from other suppliers, many of them foreign, the company creates its own circuitboards and assembles its sleek workstations from beginning to end at a fully automated plant in Fremont, California. Given all these trends, some clear, others conflicting, how will the big names in computing fare in the Nineties? Here's a survey, segment by segment, beginning with the company that crosses over all the segments: -- IBM. From PCs to workstations to minicomputers to mainframes, Big Blue is in more markets than any of its competitors. That is an advantage, since it gives the company the big picture of the industry and enables IBM to balance disappointment in one area with success in another. But the broad product line means many more competitors, quite a few of whom are focused on a single market. IBM will find it hard to hold its share of the business in the 1990s. According to Infocorp, a market research firm in Cupertino, California, IBM's share of industry revenues has slipped from 37% in 1985 to less than 30%, a decline that almost certainly will continue. While price erosion in the industry has affected almost all the contestants, it has probably hurt IBM the most. The company's latest restructuring attempt, its second in three years, is aimed at improving profit margins in the face of continuing price reductions. Says Robert Djurdjevic, president of Annex Research, a Phoenix firm: ''IBM stands out in the industry for its awesome size -- especially around its waist. This will probably not be the last we hear about IBM layoffs.'' According to Pierre Hessler, IBM's director of marketing and services, Big Blue's top priority is to change from being ''a product-oriented marketing organization to a customer-solutions-oriented organization.'' That means more emphasis on selling computer combinations tailored to specific industries, like banking or process manufacturing, instead of pitching separate machines piecemeal. Working with the customer should also mean more revenues from software and services, which will grow faster than hardware in the 1990s. Says Hessler: ''If we can do all that, faster revenue growth and bigger profit margins will follow.'' Another source of optimism: IBM's businesses overseas, which are increasing at 15% annual rates or more. Hessler says that international sales, now about 55% of IBM's revenues, should total 60% in the next few years. -- PERSONAL COMPUTERS. This segment was the star performer of the Eighties, coming from nowhere to account for 39%, the biggest share, of the industry's dollar sales. During the first half of the decade annual growth topped 70%, and in the latter half it averaged nearly 24%. Analysts say the PC market will continue to expand in the early Nineties at about a 13% rate, then taper off to 9% by the year 2000.

The slowing doesn't alarm Rod Canion, CEO of Compaq Computer. Says he: ''The problem is that the past few years have been so strong that 10% to 15% growth rates are disappointing. We should instead look at those as more normal growth rates.'' But Compaq, whose revenues swelled 51% for the first three quarters of 1989, will not be content to keep pace with the rest of the industry. It recently introduced a $16,000 to $26,000 supercharged PC called the SystemPro, which packs the processing wallop of a minicomputer at a fraction of the price. It can act as a technical workstation, a PC network file server, even a departmental minicomputer. ''We think that once customers get a taste of this, they won't like the taste of old-fashioned minicomputer systems,'' says Canion. Analysts see a bright future for SystemPro and for Compaq in the 1990s. Apple Computer, the technological trend setter for PCs in the Eighties, intends to stick to what it knows best in the Nineties. Says Sculley: ''We have no reason to focus our attention beyond individual business users. If anything, we'll be narrowing our focus.'' That means Apple won't make bigger systems and will rely less on the home and education markets. In fact, as demand fades for Apple's lowest-priced computers, the 12-year-old Apple II and the four-year-old Macintosh Plus, the company is not making any visible effort to replace them with new under-$1,000 computers. Where will Apple get its shine? Says Sculley: ''We still have a very small share of the overall business PC market, so there's lots of room for us to get new business without moving into new markets.'' He firmly believes that competitors, especially IBM, will need as many as three years to catch up with the Macintosh's snazzy graphics and ease of use. Industry analysts aren't so sure. They are waiting to see how the market reacts to the flood of software applications expected this year for IBM's OS/2 operating system. Apple recently announced a cost-cutting program that will include layoffs in the Apple USA division and the taking back of top executive's company cars. Its strength for the 1990s is the clarity of Sculley's vision; its weakness is that the vision may be wrong. -- WORKSTATIONS. The vigorous market for these $10,000 to $50,000 machines is hard to contain. As the technology has become cheaper and the software easier to use, workstations have begun to creep into the PC market below and the minicomputer market above. Because they were originally designed to work as parts of networks, they are the centerpiece of open systems. In short, workstations are the hot boxes for the 1990s. Sun Microsystems, which didn't even exist a decade ago, is probably the best-positioned workstation maker for the 1990s. It has a tightly focused product line and is slowly phasing out older systems that aren't fully compatible with its flagship SparcStation line of workstations and file servers. Says CEO Scott McNealy: ''We're getting all our wood behind one arrow.'' Though the workstation will be the wonder horse of the decade, McNealy prefers to think of his company in broader terms: ''We're now a general- purpose computing company, not just a maker of technical workstations.'' While McNealy is happy to take business away from PCs, he claims that Sun has a different approach to the desktop market: ''We're not interested in replacing typewriters. What we do best is automate groups of users, and we've bet the farm on it.'' The results so far: Sun's revenues, which increased 68% to $1.77 billion in fiscal 1989, slowed in the first quarter of fiscal 1990 to a still enviable 39% rate. Hewlett-Packard, which acquired Apollo Computer last year, is another force to be reckoned with in the workstation business and shares many of Sun's , notions about automating work groups. And don't count out DEC, the originator of workstations, or IBM, which keeps saying that it will announce its entries in this market soon. -- MINICOMPUTERS. The reports of their death have been highly exaggerated. True, they have been squeezed from below by the PC and workstation networks and from above by the mighty mainframes. The segment's share of overall industry sales has shrunk to 27% from 34% in 1984. Consequently, this segment has more walking wounded, particularly Data General and Wang, than any other part of the computer industry. But don't write off these two companies yet. Data General, which is coming off four years of losses, got a head start in the open systems technology, and analysts like the fact that its new computer line is competitively priced and backed by loads of software. Wang, the word-processing pioneer, had a loss of $62 million in its first quarter of fiscal 1990, on top of a $424.3 million deficit for the entire fiscal 1989. New CEO Richard Miller has laid off 2,000 employees (out of 25,000) and embarked on a strategy called Operation Customer. The idea is to dispatch dozens of Wang engineers who will spend time with clients to find out what they want. Says Miller: ''The computer industry is a mature business now, and the only way to grow is to get close to customers and steal business from your competitors.'' Good strategy, but perhaps a little late. An idea that may be better is for Wang to throw itself into image processing -- making computers that store all kinds of documents -- a market it helped develop. A less optimistic but perhaps more likely scenario: Wang may be acquired by one of the industry's strongmen. Sequent Computer Systems of Beaverton, Oregon, one of the fastest-growing minicomputer makers in the late 1980s, has a shot at hitting the big time in the 1990s. Sequent makes open system computers that harness several of the same microprocessor chips that power PCs. They are so powerful that they can process instant on-line transactions for hotel reservation systems and retailers at about $300,000, a fraction of the cost of the bigger systems used by airlines and banks. Says Casey Powell, Sequent's CEO: ''We are competing directly with human labor, rather than simply trying to make people more efficient. Given the shortages looming in the labor population, we think we're in the right place at the right time.'' He may be right. In just seven years, Sequent's revenues have hit $100 million. ''Our goal,'' says the irrepressible Powell, ''is to reach $1 billion by 1993.'' Tandem, whose sales grew 24% to $1.63 billion in fiscal 1989, also makes on- line transaction processing systems. The company almost single-handedly created automated teller systems and has specialized in machines that execute highly complex transactions. Now it is moving into Sequent's market with a new lower-priced line of products, and also into mainframe territory. Tandem recently won a competition with IBM to supply the California Department of Motor Vehicles with a system to track driver and vehicle records. Says CEO Treybig: ''Tandem changes its vision continually. It's important to move on to the next market when you're at the peak of your success.''

FURTHER PROOF that the minicomputer market is mostly alive and well: The big guys are having fun. IBM has had tremendous success the past two years with its new line of AS/400 minis. And DEC continues to be the largest, best- positioned player. Besides continuing to enhance its popular proprietary VAX line of minicomputers, DEC has joined the open systems bandwagon. Says senior vice president John F. Smith: ''You have to be open because that's what customers want.'' DEC is also climbing up a rung by offering the VAX 9000, a mainframe whose price starts at $1.24 million. As Smith says: ''The mainframe market may be growing by only about 3% a year, but it's so big that any sales we get can be tremendously significant.'' Where will the sales come from? DEC has a loyal customer base, whose existing VAX machines will be perfectly compatible with the 9000. -- MAINFRAMES. This segment shares the same bad rap that minicomputers often get: Companies face high manufacturing overhead, a slow-growing market, and competition from much cheaper, smaller machines. Still, customers contend that the mainframe is central to corporate computing. Says Teflian of Covia: ''What people are calling the death of the mainframe and the birth of the PC is really the emergence of totally heterogeneous systems. Our job is to use the benefits of both worlds, not replace one with the other.'' Millions of PCs on millions of desks have led to information anarchy, and mainframes are just about the only machines that can restore order. Says IBM's Hessler: ''Now one sees the pendulum swinging back after having gone from centralization to enormous decentralization, and that return is good for the mainframe industry.'' John C. Lewis, CEO of Amdahl Corp., a $1.48-billion-in-sales maker of IBM- compatible mainframes, says that demand for mainframe computing power is holding up. ''The important measure is the growth of MIPS -- millions of instructions per second -- sold. Five years ago the growth rate was 45% to 50%, and now it's 30% to 35%. But we think it will hold steady there.'' Since the ratio of price to performance is improving at about the same rate, how does Amdahl plan to grow? By taking business away from someone else. Says Lewis, who furloughed 5% of Amdahl's 8,700 employees in November: ''In a market this big, it doesn't take much of an expansion of your share to generate a lot of new revenues.'' Unfortunately for Amdahl, that's the strategy everyone else in this aging industry seems to be taking too. -- SUPERCOMPUTERS. Last year was the year of the shakeout in this market, with such companies as Control Data Corp. and Evans & Sutherland Computer dropping out, and others merging. For the survivors, business will be nasty, because the supercomputer market is so narrow. The turmoil also gave Cray Research, the granddaddy of the industry, the opportunity to let its founder, the legendary Seymour Cray, start up a new company called Cray Computer. Says Marcelo Gumucio, president and chief operating officer of Cray Research: ''Nineteen-eighty-nine was a validation that the supercomputer industry is a difficult market to get into and out of.'' Even with fewer competitors, Gumucio doesn't think his company will grow much. Cray's revenues for 1989 are expected to be about what they were in 1988, $756 million, though the company will probably report a 23% decline in profits to $120 million. Gumucio expects sales next year to remain flat. Research's growth strategy is tied closely to the spinoff. Says Gumucio: ''Seymour wanted to build his new machines one at a time. That's fine for a startup, but we want to sell hundreds of supercomputers a year.'' To do that, Cray Research is broadening its product line from its traditional $15 million machines to somewhat less expensive models. It also is weighing the possibility of building mini-supercomputers -- smaller, less complicated, easier-to-maintain Cray machines priced under a million dollars. That's the market that Convex Computer of Richardson, Texas, has tapped with great success. Convex sells its machines to oil companies, aerospace designers, and scientists who cannot afford supercomputers. For 1989, its seventh year, Convex should ring up profits of $17 million on revenues of more than $150 million. GIVEN the industry's maturation, its confusing, intersecting product lines, and the lack of any emerging superhot concept -- like personal computing in the 1980s -- no wonder Wall Street has been unkind to computer stocks. ''They're about as out of favor as shares of an industry could be right now,'' says Richard Shaffer, publisher of Technologic Computer Letter. Nevertheless, he is surprised that computer stocks have been hammered so hard. ''This is still a growth industry with plenty of life left in it. It's just that the growth won't come as regularly as the marks on a ruler.'' Or as fast as in the past. Or to all the companies in the industry. At the dawn of the next century, who will emerge the victors? They will not necessarily be the companies with the best technology. Instead, the winners will be the ones with the smartest marketing and distribution, the leanest manufacturing, the clearest sense of purpose, and the best understanding of what customers, both in the U.S. and abroad, want. Computer companies that prosper in the next decade will do so at the expense of their rivals. Two words will describe them: tough and smart.