Skip to Content

Direct from Dell

At 19, he revolutionized the selling of PCs. AT 36 he’s ready to take on HP, Sun, EMC and Cisco.

Even in this era of business prodigies, his youthful success is virtually unrivaled. Michael Dell recently turned 36 years old, and already he’s considered a sage on matters of business and technology. Dell Computer, which he founded at age 19, was the best-performing American stock of the 1990s. Long the front-runner in U.S. personal computing sales, the company took the global lead from Compaq in the first quarter of 2001-and with $32 billion in revenues last year, it’s bigger than Microsoft.

So how do you build the world’s largest PC company? Dell points to such old-fashioned strategies as giving people what they want and for a lower price; for that reason, he admires the work of Henry Ford and Wal-Mart’s Sam Walton. But he also revolutionized the selling of personal computers, using a direct-business model whose fundamental tenets include taking custom orders directly from customers, thereby reducing inventory and streamlining distribution. That model rocked the world of competitors like IBM, Compaq Computer and Apple Computer and changed forever the economics of the business.

Despite such accomplishments, Dell’s chairman and CEO remains focused on the next big thing. Recently, with PC sales growth slowing, he moved into storage and Internet servers-taking on the likes of Hewlett-Packard, Sun Microsystems, EMC and Cisco Systems. Technology Review editor at large Robert Buderi met Dell at his Austin, TX, headquarters and asked about the company’s brilliant legacy of growth, the “pragmatic” R&D intended to keep the flame burning, the future of computing and his hopes for the now slumping economy that this year has seen his firm endure its first layoffs-between 4,700 and 5,700 jobs, more than 10 percent of its work force.

TR: People get damned and praised for the wrong reasons sometimes. What do you feel you’ve done as an innovator-and what’s been misunderstood?
Dell: People look at Dell and they see the customer-facing aspects of the direct-business model, the one-to-one relation-ships. What is not really understood is that behind these relationships lies the entire value chain: invention, development, design, manufacturing, logistics, service, delivery, sales. The value created for our customers is a function of integrating all those things.

If you go back to the dot-com craze of a year or two ago, a lot of people were saying, “Oh, we know how we’re going to succeed-we’ll create a Web site just like, and we’ll be the Dell of the whatever industry.” The problem with that is that Dell’s business is not a Web site. While we do half our business on the Web, it’s just the front end to this business system. Dell has filed well over a thousand patents, we have four or five thousand engineers, and we spend $700 million in R&D. And we’re always asking ourselves, “How can we optimize the set of conditions we’re faced with-whether it’s supply chain, customer relationships, time, speed or cost?”

Remember that the computer industry when Dell entered it had gross margins of 40 percent plus. On top of that, you had dealers with margins of 20 to 30 percent. So the end user was paying a pretty incredible premium over the cost of goods for the product. And then Dell came along, and I fully acknowledge if it hadn’t been Dell it would have been somebody else, and the industry gross margins have gone down quite dramatically. Which means that parts are far more affordable, which means that the industry has grown a whole lot faster than it ever could have, which means that people have greater and faster access to technology. And I believe fundamentally the market grew faster because we forced a condition to exist-and businesses that didn’t add value were forced to move on.

TR: What’s the hype in computing, and what is real? Is the PC dead?
Dell: The industry tends to get fascinated with new concepts and ideas-and one of the biggest problems we see is technology in search of a problem. Take a very contemporary example. Bill Gates has been talking about tablet computers [portable devices with, ideally, the simplicity of a Palm Pilot but the computing power of a laptop-and between the two in size]. Now, maybe tablet computers will be successful this year or next year, maybe they won’t. But the idea of a tablet computer is not new. The pen computer industry was here 10 years ago, and there were trade shows and companies specifi-cally created for that. Didn’t work, wasn’t ready.

It’s a great example of technology push instead of customer pull. Another good example is the Internet appliance, which captured a lot of attention about a year ago. And here again, you had this fascination with something that’s new-sounds like a good idea, but maybe isn’t. So when you go back and look at the bone pile of things that might have been-the “almost famous” category of technologies-it’s a pretty big mess.

And that’s why we embrace this philosophy of time to volume. At Dell, we invest only in technology our customers want, rather than trying to guess what they might want. And we only manufacture what our customers ask us to make, when they ask us. Dell has five days of inventory, while at HP, they have 63 days of inventory. They also have something called a distribution channel. This is when you go to CompUSA and you look on the shelf and see all the computers stacked up. Those are monuments to the failed forecasts of what you didn’t want to buy. And in the distribution channel, there’s another 25 to 30 days of inventory, so you add that to the 63, and you get 85, 90 days of inventory compared to our five.

Now, what that means, first of all, is a time-to-market advantage. I can get you the freshest, latest, greatest Pentium 4 and all associated operating system, et cetera, 85 days faster than HP. It’s also a massive economic advantage, because the value of components and manufacturing materials declines about one percent per week. So the net of this is they’re all screwed up.

The other problem with a long feedback loop is the distortion in the signal. If you tell me what you want to buy and I hear it 90 days later, by the time I get the signal you want something else. This is another area where our business model has a big advantage.

TR: Still, PC sales have been slowing-increasing the need to find new growth areas. What was it that attracted you to servers and storage?
Dell: About five or six years ago, we saw that the application world was becoming increasingly server-centric-we couldn’t have anticipated Internet-centric, but that’s where it was ultimately going-and that more and more of the profit was shifting from the client [the PC] to the server. We also saw that our competitors, most notably at the time Compaq, were dramatically overcharging customers for those products.

TR: The same thing you’d seen in PCs?
Dell: Same phenomenon. Five years ago we had two percent market share in the U.S. server market; Compaq had 44 percent. Last quarter [the fiscal quarter that ended Jan. 31, 2001] we had 22 percent, and they had 26 percent. We’re growing four and a half times faster than the market, while they’re shrinking. Soon we’ll pass Compaq in the United States [this spring, Gartner Dataquest reported that Dell had indeed surpassed its rival in the U.S. server market], and at this point our U.S. server business is more than four times larger than HP’s. So we’ve just gone in totally opposite directions from our competitors.

What we see in storage, with our friends at EMC, is the same thing all over again. There’s more of a software component to storage-in the management of large volumes and the virtualization of different logical units of storage across a wide variety of heterogeneous server platforms. But there’s nothing about those products or technologies that scares us. The industry keeps moving towards broad standards, and that creates a real dilemma for companies who try to create and lead an industry by maintaining proprietary platforms and high margins.

The unmistakable trend here is that industry or open standards are beneficial to customers. If you look at any proprie-tary computer company in history, whether it’s Silicon Graphics, Data General, Digital, Prime or fill in the blank, the fate is almost always the same. Their

higher margins make them a prime target for nonproprietary competitors. Because in an industry or open-standards environment, companies can focus on what they’re good at and not spend resources on what they’re not. So innovation is high, and costs are low-and this is the challenge a lot of these companies have.
Our business model is in many ways the polar opposite of the proprietary computing model. If our margins are going up, we get worried, and we wonder whether we’re doing everything we can to deliver more value to our customers. Because if we’re not, we know somebody else will.

TR: What do you think is the real role of the Internet in our economy today?
Dell: At the root of any economic system is the cost of transactions. You have something you want to sell, I want to buy it, and what that transaction ultimately costs is tied to the cost of communicating information. The Internet is the latest evolution of communication technology-tremendously powerful because it enhances the flow of information. So basically it’s like a big vacuum that sucks friction out of the economy.

And there are a number of businesses that may have existed in the past on the idea that the customer has somewhat limited information, so they’re going to buy their product. In other words, you live in a small town. The customer has no other ways to get it, that’s the price. You don’t have a choice. You can’t really fly to another city and get it, that’s not practical. You don’t have other sources of information. Now all that’s changing. So the Internet can help those companies that thrive on information, like a Dell. Businesses that don’t thrive on information, that thrive on lack of information, they’re in danger.

TR: Dell’s R&D expenditures, about $700 million, are small compared to Microsoft’s $4 billion. They’re also only about 2.5 percent of revenues-minuscule for high-tech firms. Do you worry you’re not investing enough?
Dell: It’s not like the nuclear-arms race. We have to pick the amount that’s appropriate for what we want to do. Could we spend $1.75 billion? Absolutely. Could we spend $2.75 billion? Even easier. We could spend $5 billion on R&D. But the harder question is, Where is your point of diminishing returns? In a typical, classical development world, you’d say, well, we’re just going to tie this thing to the revenue, which I contend is really nuts. And so what we don’t do here is say, “Revenues are going to be $40 billion, and five percent of $40 billion is $2 billion, so let’s figure out how to spend $2 billion on R&D.”

Instead we say, “Okay, we want to have these eight-way servers. We’re going to have these storage programs. We’re going to have a full lineup of notebooks and desktops and workstations. We’re going to have these software management tools. What will it cost us to do all this?” And if that number is $842 million, then that’s the budget. And if we add three things and it goes up to $860 million, or we take out four things and it goes to $820 million, that’s the budget.

As for comparing Dell to Microsoft, it doesn’t give you the full picture. Microsoft is our partner, and we leverage the R&D work they’re doing in operating systems.

TR: So you’re actually getting the benefit of their R&D. You’re piggybacking?
Dell: Exactly. If you look at the original orientation of the industry, each computer company tried to do everything itself-its own silicon, processors, operating system, power supplies. Digital was in the business of making power supplies up until three or four years ago. Now, you say, why do you want to design power supplies? Well, maybe it’s fun, or maybe it’s what we did last year-or maybe we just think we’re better than anybody else. Maybe we got this disease called Not Invented Here.

This is the problem a lot of companies got into. Our approach is more pragmatic: if we can buy something that’s very similar to something we can create ourselves, we believe it might not be valuable for us to create it. On the other hand, if we’re thinking about creating something that nobody else has, that’s worth doing. And I can point to hundreds of unique inventions or ideas that we have driven. We had the first color notebook that was powered by batteries. We had the first 486 machine to ship. We had the first system to ship with the EISA [Extended Industry-Standard Architecture] bus. Right now our notebook team is continuing to drive very, very hard on size, weight, wireless integration-we were the first to integrate wireless into notebooks, with integrated antennas.

TR: Do you have a central research lab? Also, about two years ago, you started Dell Ventures, which makes strategic investments in startups and other firms. Where does that fit into this broader picture of how you assemble research?

Dell: We’ve always had some small group that has done research. It’s gotten to be a formal organization in the last couple of years-a Dell Labs group. But it’s not a huge group. Their role is to think beyond the product road maps, which end between 18 months and two years out. They also have a technology council, where we review key promising areas and determine whether we should continue research in those, or whether we should skip them and learn something else. And they do quite a bit of experimentation with things. So they’re off playing around with handhelds and wireless and voice, you name it.

Dell Ventures can also play a significant role in pioneering and testing new concepts, new technologies, new ideas. For us, Dell is not an ingredient company; we’re a systems company. We provide solutions to our customers. So, take batteries as an example. We don’t have any factories that make batteries. We don’t have any plans to make them. But we know where all the batteries are made. We consume 20-plus percent of the worldwide production of batteries that go into products like ours and have some very small number of people who understand the chemistries that go into batteries. And if we felt that the industry wasn’t investing enough, battery development would become an issue for us.

But, instead, we see a lot of experimentation. And when we see, for example, one company with a lot of production capabilities and another with a lot of technology, we put them together. We just did this with a CRT [cathode-ray tube] innovation. We found a company that had a semiconductor replacement that offered smaller size, longer life, reduced cost, better focus and lower voltage levels than the cathode we’d been using. We invested in the company, and we hooked them up with our CRT providers to get this thing designed in. We don’t want to buy the company. Why buy the cow when all you want is the milk? We’ll take a piece of that company, make sure that invention gets into our products, and probably our products first, so we get some differential advantage-and go on and look for the next one.

TR: What has you most excited about the future of computing?
Dell: A lot of the things we’ve done at Dell, in terms of supply chain and collaboration and information flow-integrating these kinds of systems-has helped to democratize the market by bringing down prices. Owning a personal computer is now widely accessible.

But the industry is really very much in its infancy. We still have a long way to go in terms of improving our products and making them easier to use. In the future it’s going to be easier to connect disparate systems together and get to the next level of productivity-and that’s very, very exciting. There will always be big, underpenetrated markets just looking for improved productivity, regardless of economic conditions. So I have lots of optimism for productivity in the future.

I’m also excited about wireless Internet access. We live in a mobile world. People are constantly on the move, but their fundamental needs don’t go away when they leave their homes or offices. They want to stay productive, and they want to keep in touch. Wireless technology lets them do that.

TR: What’s the biggest lesson you’ve learned about innovation?
Dell: To encourage it and allow it to happen. If you try to do it all from the center of the company, it doesn’t work. You have to break things down into focused units that have clear objectives, so everyone in the company is driving to innovate within their particular customer area. We have to carefully decide how we’re going to allocate our resources. And we have to allow for a measure of experimentation. When you stop experimenting, you’re dead-because then you have no ideas, you have no breakthrough thinking.

A couple of years ago, one of our teams came in and said, we got this idea-just bear with us a second-we want to take a 15-inch LCD [liquid crystal display] screen and put it in a notebook computer. I said, “What are you talking about? That’s not a notebook, that’s a portable desktop.” They said, “No, no, this will be a high-end notebook, and it’s going to be 10 percent of the market.” Well, it was risky, both in technological terms and market acceptance. Would the product be too heavy? Did our customers really want a large screen on a notebook? But we took the risk, and the idea turned out to be a huge success. We have plenty of examples where we’ve innovated and failed, but we have more where we’ve succeeded.

TR: What’s your personal role in all this?
Dell: When I see things that I think are important, I’ll push on-with wireless, say, and also by nurturing things when they’re in the skunk works stage and don’t necessarily have the organization’s full support.

I’m the agitator for progress and change.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

How scientists traced a mysterious covid case back to six toilets

When wastewater surveillance turns into a hunt for a single infected individual, the ethics get tricky.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.