The Dell Way Michael Dell's famous business model made his company the world's premier computer maker. Now he's branching into new fields and taking on virtually every other hardware manufacturer. Can "the Model" stand the strain?
By Kathryn Jones Additional reporting by Brian Caulfield, Lois Gilman, Ian Mount, Erick Schonfeld, and Owen Thomas

(Business 2.0) – At Dell they call it, simply, "the Model."

You can see it in action inside the Topfer Manufacturing Center, a squat, white factory a few miles south of Dell Computer's headquarters campus in Round Rock, Texas. The newest of Dell's seven plants worldwide, it is cavernous--big enough to enclose five and a half football fields--and a blur of activity: Market domination in motion. Boxes of Intel microchips and electronic components from Taiwan and Korea skitter by on double-decker conveyor belts. Workers read orders off a monitor and assemble a new Dell OptiPlex desktop computer every three to five minutes. The finished boxes, more than 25,000 on a typical day, then trundle off on other conveyors to be shipped directly to customers. The whole fandango is choreographed so tightly that the factory rarely needs more than two hours' worth of parts inventory. Two hours. Parts storage takes up roughly the space of an ordinary bedroom.

Nobody, but nobody, makes computer hardware more efficiently than Dell. Operating costs soaked up just 10 percent of Dell's $35 billion in revenue last year, compared with 21 percent of revenue at Hewlett-Packard, 25 percent at Gateway, and 46 percent at Cisco. No unnecessary costs: This is an all-but-sacred mandate of the famous "Dell direct" business model. No inventory, no middlemen to eat into profits, no agenda other than giving the customer what he or she wants. Crucially, the Dell model also insists on no more than minimal spending on research and development (1.3 percent of revenues last year, compared with 15 percent at Intel and Microsoft) and no proprietary technology of the company's own. The goal is to offer computers at irresistible prices--and to leave the cost and the risks of innovating to others.

In other words, the Dell model allowed the company to reinvent the PC industry--without inventing anything.

Now the Model faces its biggest challenge.

Dell is moving into markets where it has never been before. In some of these new fields, Dell's PC success has opened doors; in others, the company is starting from scratch. In one market, success has nothing to do with Dell's specialty of cranking out good hardware at great prices but instead hinges on providing that ephemeral thing known as services. Every inch of the way, the Model is being challenged, tweaked, tested, and pounded on. And surprisingly, its defenders say, the thing is stronger than ever.

On a December morning, Michael Dell, eponymous founder, multibillionaire, and at the ripe old age of 37 the longest-tenured CEO of any U.S. PC maker, is trying to put all this into a simple metaphor. "Some people say we're like Wal-Mart," he ventures. Dell seems to find the comparison distasteful, but it's not far off. Starting with $1,000 in seed money in 1984, the company has grown from its founder's University of Texas freshman dorm room into a force that, like the Bentonville, Ark., behemoth, dominates its industry. Dell is No. 1 in desktop PCs, No. 1 in the United States in low-end servers, and the country's No. 1 Internet retailer. The PC industry sank into the worst slump in its history 18 months ago, but Dell's revenues have nevertheless grown 14 percent to more than $35 billion, and the company is on track to earn $2 billion this year, even as its competitors run hundreds of millions of dollars in the red.

No wonder shareholders regard the Model as just short of Revelation. As of mid-December they were willing to pay 36 times earnings to own a share, a sign that they expect gangbuster growth for years to come. To Michael Dell and his right-hand man, president and chief operating officer Kevin Rollins, it is clear that those expectations can't be met with profits from PCs alone.

So, like Wal-Mart, Dell has had to expand into new businesses. During the past 18 months, the company has announced its entry into a host of new fields: network switches, PDAs, and printers, as well as the sophisticated hardware at the heart of corporate computing networks. Says Dell: "The best way to describe us now is as a broad computer systems and services company."

With the territory, however, comes a whole new set of enemies. Dell isn't just competing against the floundering PC industry. Now it's Dell against everyone who's left. Already competitors are talking trash. "Look at where they came from and where they want to go," says Tim Dougherty, director of strategy for IBM's server group. "Expertise in enterprise computing is not in Dell's skill set." Mike Winkler, HP's chief marketing officer, predicts that his company's archrival has finally overextended. "The closest analogy," he says, "is Napoleon's invasion of Russia."

Michael Dell has heard all that before. Competitors once dismissed him as an underage computer geek who'd never amount to anything against the likes of Compaq and IBM, and clucked that Dell's direct model could never move off the desktop into the complex world of corporate computing. Both predictions were wrong. The point is, Dell says, that the Model, like the company, is broader than people give it credit for. "We have a pretty simple system," he says. "The most important thing is to satisfy our customers. The second most important is to be profitable. If we don't do the first one well, the second one won't happen."

Moving In From the Edge: Servers Opportunity: $50 billion Competitors: HP, IBM, Sun

Until the mid-1990s, Dell existed almost exclusively at the edges of business computing, its desktop PCs and laptops the last stop between the information core of an enterprise and its employees. Now Dell wants a much larger share of the industrial-strength hardware at the heart of corporate systems, where the stakes--and the profit margins--are higher. These are the machines on which entire businesses rise and fall.

In the mid-1990s, Dell took its first steps in from the desktop, tackling low-end servers, $5,000 to $25,000 machines just powerful enough to serve up webpages or run small-office e-mail systems. Like PCs, they ran Microsoft Windows on Intel chips, and Dell could build them cheaper than anyone. "When it comes to assembling servers, we enjoy all the same advantages from our procurement, logistics, and manufacturing capabilities as we do making PCs," says Randy Groves, vice president and general manager of Dell's enterprise systems group. Dell's enterprise revenues, almost nonexistent in 1994, accounted for 13 percent of the company's total intake in 1998. Three years later Dell passed Compaq as the top provider of Intel-based servers, with 31 percent of the market.

There is more at work here than the tendency of recession-squeezed IT managers to choose a cheaper alternative. Many of the servers that Dell replaced ran on proprietary technology, like IBM's Power4 microprocessors or various proprietary flavors of the Unix operating system. If you wanted upgrades or new applications, you had little choice but to get them from the manufacturer. Dell's model eschews these arrangements, which the founder calls "proprietary prisons." Instead, Dell machines stick with the de facto industry standards, like Intel microprocessors and Windows or Linux operating systems, which don't lock users into any particular hardware.

That was a key selling feature to Jeff Davis, a senior systems programmer at the oil company Amerada Hess, where he runs a cluster of several hundred Dell workstations to model undersea geological features in search of oil deposits. Although Dell is his vendor of choice, he likes the fact that the company's machines are built from standardized parts, so he can mix and match boxes from a variety of vendors to hold down his costs. "We can pretty much pick a vendor based entirely on the merits," he says, "and on whether they've got the right price." He spends $300,000 a year upgrading and maintaining the cluster of workstations; they replace an IBM supercomputer that cost $1.5 million a year to lease and operate.

In the Dell model, sticking with industry standards is not simply a matter of building customer goodwill. It's more like obeying the Second Law of Thermodynamics: You don't really have a choice. "In the long run, all technology tends toward low-cost standards," Dell explains. The Model is predicated on it. After all, in a world dominated by standard platforms, the hardware running the platform becomes a commodity, and the most important reason to choose one vendor over another is price. And when it comes to price, Dell can compete with anyone.

Land Without Standards: Storage Systems Opportunity: $22 billion Competitors: EMC, Hitachi, HP

Having made inroads into servers, the next logical step for Dell was into storage systems, the advanced computers that house an enterprise's most crucial data. But in contrast to the PC and server industries, there are no standard storage technologies, and Dell floundered as it tried to apply the Model. The company first tried unsuccessfully to develop a system in-house but quickly realized that it didn't have the expertise. It then tried reselling other companies' systems. In a rare acquisition, Dell purchased network storage specialist ConvergeNet Technologies for $332 million in 1999, only to find that ConvergeNet's elegant but complex technology made a poor fit with Dell's commodity-producer business model. Within a few years, Dell had written off the entire investment.

In October 2001, Dell accepted that the closest thing to a standard was the systems built by industry leader EMC, and agreed to co-market its midrange ($30,000 to $500,000) Clariion storage systems until 2006. The deal brought EMC a partner in a sector where it had been hard-pressed by Compaq; Dell got into a key enterprise market without the cost of (any further) R&D. The partners say the deal is working: They've added 1,500 customers so far.

As Dell ponders the lessons of jumping into a sector without clear standards, the storage industry seems gradually to be moving exactly as the Model would predict. Standards are winding their deliberate way through industry committees, but in the meantime, the technology is taking matters into its own hands and loosening the hold of proprietary systems. New technologies that link small devices to storage networks are beginning to replace the proprietary big iron of companies like EMC. If cheap networked boxes become the storage standard, Dell is the logical winner. Indeed, in a December survey conducted by investment research group ChangeWave, 17 percent of Dell's current corporate clients said they expected to buy Dell storage systems in the next six months.

Looking to Supersize: Networking Opportunity: $13 billion Competitors: Cisco, Enterasys, Nortel, 3Com

In the networking world, Dell's natural targets are the routers and switches that shunt information through corporate networks. Here, Dell is moving more cautiously. Its sole offering in the category is its PowerConnect switch family, so-called Layer 2 devices, the simplest, most commodity-like of their kind. At $20 per port, their positioning is all about price. (According to the Yankee Group, Cisco's entry in Layer 2 costs $100 per port, while a 3Com switch costs $38.) So far, most sales have been add-ons to other deals, a tactic that Dave Smith, 3Com's vice president for sales and service, derides as the "Want fries with that?" ploy. "If you buy a bunch of servers," Smith says, "they'll ask, 'Want some networking gear to go with that?'" Still, Smith is taking Dell seriously: He has instituted a "Match Dell" program, in which 3Com grants its resellers discounts of as much as 25 percent in return for their offering 3Com goods at prices that match Dell's.

Many analysts say Dell will have trouble moving beyond basic networking hardware. While standard protocols like Ethernet determine how data flows through networks, the software and chips that distinguish the smartest switches from the rest are jealously proprietary. To challenge Cisco in higher-end equipment, Dell would have to build its own, and that would be folly, says Jim Slaby, an analyst at the Giga Information Group. "You'd have to duplicate 15 years of software development and thousands of man-years of R&D. It would be like creating a word processor to compete with Microsoft Word."

But big R&D investments are not part of the Model anyway. And as is happening in storage, the industry's leaders could well be on the verge of losing their proprietary grip on networking hardware. Intel and Broadcom are building instructions into networking chips that make the equivalent of years of R&D available to any interested hardware maker. Dell is waiting.

The Human Touch: Services Opportunity: $350 billion Competitors: Accenture, HP, IBM, others

Perhaps Dell's biggest enterprise challenge lies in applying the Model to services. The company has a big services group--some 8,000 employees strong--and Dell sees it playing an expanding role in future growth. But how? Everyone from IBM to the newly independent consulting arms of disgraced accounting firms wants in on the $350 billion IT consulting game. And at first blush, the Model offers no real competitive edge. "Dell has amazing production efficiencies," says Gary Chapman, head of the 21st Century Project at the University of Texas at Austin. "But managing a workforce of consultants is a whole different thing. It's much easier to control costs when you're only dealing with widgets."

But when asked whether Dell can do in services what it did in PCs, Kevin Rollins says, "Absolutely." The Model isn't just about making cheap boxes; it's also about freeing customers from overpriced relationships. "Our direct model has basic principles: Don't let anyone come between us and the customer. Keep clear communication, and no extra costs." Those tenets apply even when Dell is selling services, Rollins says, and they enable the company to give service customers a better experience than the competition. That is an opportunity Dell is programmed to seize.

There's a difference, though, between principles and experience--particularly in services, where knowledge is the only asset that counts. There's no reason to suppose that mere principles can make Dell a success when it's literally years behind its competitors.

No reason but one: Betting against Dell has been a fool's errand for nearly 20 years. And if the company succeeds in dominating enterprise services as it has dominated PCs--well, Rollins will have been right. It must have been part of the Model all along.