(FORTUNE Magazine) – Just a week before the board of directors ousted him as CEO of Compaq Computer, Eckhard Pfeiffer was feted at a University of Houston gala to celebrate the endowment of a new chair in his name. A portrait of Pfeiffer was auctioned off. Guests received MCI 30-minute calling cards decorated with his photo. Close friend Carolyn Farb, the socialite who chaired the gala, hailed Pfeiffer as "one of the world's boldest thinkers." Cynics in Houston refer to the event as a "coronation." But Pfeiffer's reign, such as it was, would soon end.

The real ruler of Compaq, almost from its inception, has been soft-spoken, unassuming venture capitalist Ben Rosen. For years he has been answering his own phone in the small office he keeps in New York's Met Life building, high above Grand Central Terminal. Rosen takes great pride in having funded Compaq in its early years and having helped it grow into one of the world's most important companies. He now holds about $100 million of stock and has maintained an insurance policy to safeguard his baby. He has never let the CEO sit as chairman of the board. He does that himself.

The basic facts of Pfeiffer's firing are well known. The PC industry has been getting more challenging, and Compaq has been handling the challenges poorly. Since late January it has lost almost half its market capitalization (see chart). A week after Pfeiffer announced that first-quarter profits would be only half what Wall Street had been led to expect, Rosen convened the board, sans Pfeiffer. It voted him out. Now Rosen and two other directors are running the company as they seek a new CEO. The trio fits comfortably into Pfeiffer's corner office at Houston headquarters.

Pfeiffer blames Compaq's woes on aggressive pricing by IBM that destabilized an already shaky industry. We'll address that later; first, let's take a look at what happened behind the scenes during the waning of Pfeiffer's reign.

Eckhard Pfeiffer was in many ways a very effective CEO, but he suffered shortcomings of both style and substance that became more problematic as Compaq grew. Stiff and imperious, Pfeiffer has a single-mindedness that helped push Compaq into the forefront of computing. But he created a contentious atmosphere in the executive suite, as top managers battled to win his favor. Says longtime chief strategist Robert Stearns, who left last June: "It felt like we were the Olympic dream team of the PC industry, but over a period of about three years, most of the real players were either forced out or left in disgust." Among the many other top execs who left--for whatever reasons--are CFO Daryl White, North American chief Ross Cooley, server boss Gary Stimac, two general counsels, and manufacturing head Greg Petsch. Says Pfeiffer: "I delegated authority and responsibility to a point that exceeded some people's ability to fully live up to that expectation."

Pfeiffer introduced a whole series of executive perks to a company that had always had an egalitarian culture. Where parking places had never been reserved, Pfeiffer oversaw the construction of an executive parking garage. The executive floor was repeatedly remodeled, with access increasingly restricted.

These trappings of power would be mere curiosities, were they not signs of a more substantial problem. Says a former executive: "Eckhard saw himself growing in prestige by merely having a large headcount and a large revenue number." Compaq has gotten huge, that's for sure: When Pfeiffer took over in 1991 (at Rosen's request), Compaq had 21,000 employees and $3.3 billion in annual revenues; now it is the largest PC maker in the world, with 69,000 workers and annual revenues of $31.2 billion in 1998.

Pfeiffer talked constantly about being No. 1. Says Stearns, "In his quest for bigness, he lost an understanding of the customer and built what I call empty market share--large but not profitable." Jim Moore, a technology strategy consultant with GeoPartners Research in Cambridge, Mass., says Pfeiffer "raced to scale without having economies of scale." Pfeiffer denies this, insisting that he always focused on profitability. But the colossus he built does now seem unable to move nimbly enough to thrive in the fast-changing computer industry. Last year, for instance, Compaq forecast demand poorly and shipped distributors too many PCs. When resellers dumped them at fire sale prices, Compaq's operating profit disappeared for two quarters, since it protects resellers from heavy losses. Dell, of course, has no such problem: Since it builds PCs only after they're ordered, it can respond more quickly to shifts in demand.

The CEO's ultimate grab for growth was the acquisition of Digital Equipment Corp. last June, for $8.4 billion. Compaq wanted Digital's giant services organization. It needed bodies to serve its big customers, which were seeking more help in planning and running their increasingly complex corporate networks. Services also have considerably higher profit margins than PC sales. When Pfeiffer tried to buy Digital's services arm alone, he was rebuffed. So he bought the whole company, with its conservative culture and many underperforming businesses. Chief strategist Stearns opposed the deal, believing Compaq would be better off purchasing Gateway to get into the business of selling computers directly. After four years of talks, Pfeiffer and a tiny group of confidants presented the Digital deal to the rest of management as a fait accompli.

Going into the deal, Compaq had a much more gung ho and aggressive culture than Digital. (Indeed, that cultural mismatch led many to question the deal's wisdom.) Nine months later, Compaq seems to have lost its way, and its culture seems more like Digital's than the other way around. And Compaq is still saddled with many crummy Digital businesses. Though it has already laid off 15,000 after absorbing Digital's 54,900 employees, insiders predict even more cuts ahead. Despite all this, Pfeiffer says that, "Had we not had this blowup here, I think the world would have acknowledged that this is a huge integration with incredible efficiency and speed."

Pfeiffer deserves credit for having done a lot right. He built Compaq into an extremely profitable industry powerhouse. Despite his recent missteps, he presided over a spectacular rise in the company's stock price, which went up 1,072% during his tenure, vs. 243% for the S&P 500. (Of course, Dell stock rose 13,375%.)

In a long interview on April 30, Pfeiffer blamed Compaq's recent woes on "a very targeted action" by IBM to take market share by cutting prices. He really doesn't see any way he could have avoided disappointing Wall Street. And it's true the industry has its troubles. There are, in fact, signs that customer demand is anemic, and PC prices are plummeting (see chart). Says Roger Kay, a PC market researcher at International Data Corp.: "All of a sudden, the industry has woken up to the issue of cost. You can't hide a bad cost structure behind good revenue growth now."

The problem with Pfeiffer's defense is that in one way or another, each of the four other major PC makers--Dell, Gateway, Hewlett-Packard, and IBM--is coping better than Compaq. For starters, with the exception of IBM, all of them make, distribute, and market PCs more efficiently. Says Gateway CEO Ted Waitt: "There's a crisis at Compaq but not in the industry."

Dell and Gateway stand apart from the others because the bulk of their sales is made directly, typically over the phone and the Internet. That yields two advantages: tremendous cost-reducing efficiencies in distribution, and a better connection to customers, making it easier to sell additional products and services.

At Dell the sales just keep rolling in. New figures for the first quarter from IDC show that Dell sold 52% more PCs than it did a year ago, while total industry sales grew just 19%. Compaq, by contrast, lost market share, growing only 16%. Dell is aggressively beefing up its product line for big business with new high-end storage products, some of which sell for $100,000 or more each. Meanwhile, its consumer business, always peripheral to corporate sales, is now on a $3 billion annual run rate. And Dell has built a $1.4-billion-a-year services business from scratch: It now has more than 1,600 technical and service consultants working with customers. Vice chairman Kevin Rollins says he sees few limits to Dell's practice of selling directly to customers: "We believe we can sell to pretty much anyone this way." When Rollins talks about growth, he does not sound like Eckhard Pfeiffer: "Additional revenue streams are marginally interesting to us, but what's really interesting are additional profit streams."

Gateway is pioneering ways to get additional revenues from consumers and small businesses, which account for the lion's share of its sales. Though CEO Waitt is obviously worried about declining PC prices, he says the company is protected by its selling structure: "Our direct relationship with our customers makes life easier for us and will enable us to more profitably head into this new environment." Says security analyst Daniel Kunstler of J.P. Morgan: "Being very focused on consumers was once held against Gateway, but now it's apparent it brings more stable and loyal customers. What was once a curse is now a blessing."

Gateway doesn't sell just PCs. It sells a broad range of software, including packages for doctors, lawyers, and other small-business people. It sells Internet access; it gets additional revenues from customers who pay in installments; it gets ad revenues from its Website. Says Waitt: "Everybody's looking for some form of ongoing revenue stream from their installed base. As prices come down, the box itself is a smaller percentage of the overall solution you offer customers."

Gateway uses its close connection to sell customers on hot, new, expensive features. Stories of buyers phoning in and finding themselves persuaded to buy a $3,000 PC for their mother are not uncommon. To further the connection to customers, Waitt has launched 155 Gateway Country Stores, mostly in the U.S. The stores carry no inventory, but customers get to test machines before they buy, all with the helpful assistance of salespeople only too happy to make in-person pitches.

Hewlett-Packard has many of the same problems as Compaq: It's a behemoth that needs to be nimbler; it would like to sell directly to customers but is afraid of alienating resellers; and its CEO, Lew Platt, is an ambitious man who for years was enamored of growth, aiming to make HP one of the "FORTUNE 10." (It is now No. 14 on the FORTUNE 500.) Perhaps not coincidentally, HP's CEO is headed out the door, albeit in far more decorous fashion--Platt is overseeing the breakup of HP into a computer and imaging company and a measurement company, and will leave after the board has found a new CEO for the computer half.

Platt modulated his talk of growth after last year's disastrous second quarter, when HP lost money on PCs for the first time since 1994. Platt heeded the warning. With Duane Zitzner, who now heads the PC division, he immediately refocused the group on profitable growth. Now HP is working assiduously to reduce both product and parts inventories and improve its manufacturing operations. It is also selling PCs directly, even as it vows to retain its long-standing connection with resellers. Says Webb McKinney, head of the group that sells PCs to businesses: "We think we can pass Dell eventually, and we're going to do it by growing share and profit simultaneously." HP already has the largest PC market share in U.S. retail stores, having seized that spot from Compaq in January.

Unlike Compaq, HP has an incredibly strong profit center: Its printer business is by far the market leader. A large percentage of Hewlett-Packard home PCs are sold with a printer. And printers have a wonderful ancillary profit stream of their own: ink. Merrill Lynch analyst Steven Milunovich estimates that printers account for 55% of HP's profits--with 40% coming from ink and cartridges, and just 15% coming from the machines.

On the surface, IBM's PC business is considerably worse off than Compaq's. The division lost $990 million last year and $89 million in the first quarter of this year--although that was down from $458 million a year earlier. But there's a silver lining here. After selling PCs to a company, IBM can often also sell services to the customer, thus locking in profits for years. A well-managed Compaq could eventually do the same thing. Jim Pertzborn, the newly appointed general manager of IBM's PC group, recently told CNET's News.Com, "It is fundamental for us to be in the PC business to provide solutions to customers." (IBM declined to comment for this story.) Pricing well-reviewed products aggressively, IBM gained PC market share in the first quarter, with sales up 30%, according to IDC.

IBM's price cutting was a factor Pfeiffer pointed to when he tried to rationalize Compaq's disappointing quarter. But he can't get off the hook just by blaming industry conditions. The fact is, each of his big competitors, even money-losing IBM, has a better strategy for dealing with today's malady of declining prices and erratic demand.

The big lesson is simple: PC makers need to find other ways to make money, even if by some miracle PC prices eventually rise from the dead. (For more on the PC's future, see our interview with Andy Grove in the box that follows.) Says Jim Liang, who co-heads technology investment banking at Morgan Stanley: "I don't think we'll see pure-play PC companies that are growth companies anymore." Pfeiffer understood that: It's why he wanted Digital's services arm. But in his rush for growth, he bloated Compaq by buying the rest of Digital--and in this fast-moving market, bloat is really, really bad news.

REPORTER ASSOCIATE Patty de Llosa