CONTEXT: The business environment has felt different since April, when dot-com stock-market valuations plunged to what seem to be somewhat more reasonable levels. Should businesses now approach e-commerce and, more generally, innovation any differently?

JIM CLARK: I don’t think that chief executives at long-established businesses can get complacent. I think they should still be plenty worried. If I were a CEO at one of those businesses, I’d worry all the time. Innovation is still going to occur, and they need to keep up.

Because of decreased valuations, now might be a great time for big, slow-moving companies to think about Internet acquisitions. In other words, do what John Chambers does as CEO of Cisco Systems. He acquires more companies than you can imagine. That’s what I’d do; I’d go out and acquire a company. For instance, now might be the perfect moment for Wal-Mart to acquire Amazon.com. It might be a lot cheaper than trying to build a big Internet presence itself.

Chambers doesn’t blink at spending $5 billion for a company. Do you think Wal-Mart would do that? Now that he’s going to join its board, maybe Wal-Mart might consider doing just that.

CONTEXT: When you’re trying to innovate, how many ideas do you go through before you come up with something that you think you could turn into a company? And how do you sift through your ideas?

CLARK: Usually, it’s less the idea than it is the people. There is no real way to tell whether an idea is good until you have people thinking about it, working on it, and proving it. Really good people get enthusiastic about what seems like a good idea, and that makes them even more likely to be successful. If you have a team that has done good stuff before, they can take even a mediocre idea and make a good company.

As to the actual ideas, I really don’t have that many. I’ve thought about lots of things, often too late. Online auctions, for example. I had that idea but didn’t run with it because I didn’t realize how big it could get. When I heard about eBay, I thought, it can’t get that big.

What really happens is that if I find myself fixated on one thing that bugs me personally, sometimes that’s the germ of an idea for a company. Typically, I find myself exploring why something is such a mess. I’ll just continue asking, "Why doesn’t someone fix this?" until I wind myself up to the point where I say, "OK, damn it, I’m going to fix it."

I’m usually thinking about technology at the same time as everything else, because technology is at the root of what I do. I also have a good instinct for market timing, for whether an idea is too early.

CONTEXT: In your book you wrote about how it’s easy to get complacent and how you have to always be looking ahead. Other than by being frustrated by things that bug you, how do you do that?

CLARK: If one could quantify that, then everyone would be able to do it. I think it’s a combination of experience, expertise (meaning knowing technology well enough), and market awareness.

Looking back at my experience at Silicon Graphics, it was fairly obvious that on the technology side you had to know two things. First was Moore’s Law. Simplified, it posited that semiconductors were going to get faster and cheaper. The second thing was that people get wedded to software, just as they get accustomed to having the accelerator under the right foot.

Around 1985, it hit me that the personal computer was just taking off, and desktop applications were compelling people to buy more and more PCs. Unfortunately, people couldn’t get those common applications on our machine. I looked at us as being like Apple. I told our management team we could be an alternative computer company, but we couldn’t do it unless we got the price of our computers really low so that the masses could buy them. You want as many people as possible to get accustomed to using your equipment because it gives you longevity. It gives you insurance.

While we were making super-high margins selling a few machines to the high-end researchers and academics, it just struck me as a dead end. I said, "We have to move down-market." That’s the only insight that I seemed to have that other people ignored.

Our first attempt at a reasonably inexpensive machine was way, way too late. We came out with something we called a Personal Iris. It was intended to be a $5,000 machine. By the time the product was done it had a $20,000 price tag. When Silicon Graphics finally did get to about an $8,000 price, it was ’95 or ’96. Too late. Today, I have more net worth than the whole company does.

CONTEXT: How do you explain this organizational inertia? Here you were, the founder, inventor, chairman, and a big shareholder, and you couldn’t get the company to change.

CLARK: It was because I wasn’t the CEO. I think powerful CEOs get companies to do what’s important. The CEO was Ed McCracken, and he didn’t have the same religion of urgency I did.

I don’t think I was the right person to be the CEO. But it would have been really nice if he had consulted with me and treated me as a partner. He didn’t know how to do that. We had numerous heart-to-heart talks, where he’d say, "I’m not going to get your permission." "Ed," I’d say, "it’s not getting permission. It’s just treating me as a partner. I do have technological insight into these things." His preference was to go off and deal with the management team and not have me involved in those meetings.

CONTEXT: What was the lesson?

CLARK: That it is extremely important not to marginalize the visionaries. In a technology company, you need the leader, the visionary, the passionate person. And that person should be on equal footing with the guy who is the manager.

You see it invariably. At Apple, John Sculley couldn’t deal with sharing the spotlight with Steve Jobs. When Jobs went to the board and tried to push Sculley out, the board said, "No." Huge mistake. Sculley may be a great manager, but he didn’t have one ounce of instinct about what was going on in the industry. The man with the instinct and the passion was Steve Jobs. Just look at what he’s done with the company since he got back in charge.

I think my situation paralleled Steve Jobs’s. I had the instinct and the passion. It wasn’t passion for the limelight. It was passion for success for the company.

If a company—a start-up especially—has a passionate leader, the company had better try to keep him. That means getting a CEO who can share the leadership, the glory, and whatever else and listen to the founder.

Nevertheless, everyone must eventually step aside for those coming along. That means every company also needs to have a succession plan, a process of creating successive leaders. There have been longstanding, healthy companies like General Electric and Hewlett-Packard that excel at this.

I was willing to be part of the process of being rolled over. I would have stepped aside as chairman if McCracken, in becoming chairman, had been willing to have someone else be CEO. But the board wouldn’t listen to me. They considered me to be a problem. Maybe I was. But I was right.

CONTEXT: Most entrepreneurs and most executives have, if they’re lucky, one huge success in their careers. How have you continually reinvented yourself and been successful as a serial entrepreneur?

CLARK: I have a nightmare about being in a rut. I like frontiers. I like new things. That may be the engineering need, the creativity, the researcher. I couldn’t stand to write about the same thing in two different papers. I would try to write a nice paper about an idea and then move on to a new thing. I’ve always liked to change fields. I’ve found that rewards came "out of left field" when I changed from physics to computer science and from being an academic to a businessman.

I like to build things, make them work, and get them done. I put an enormous amount of will and energy into making sure that I achieve a result. But I don’t like repetition.

The point is that if you have a CEO who likes things the way they are—and just likes having his orderly, weekly staff meetings, and takes his vacation once a year—you have a problem waiting to happen. I think a CEO needs to be restless.

CONTEXT: What other traits must a leader have?

CLARK: A leader must always strive to get the very, very best people—people who threaten you because they’re smarter than you.

You also have to make sure you don’t have high turnover. I have a rule: Never lose the first good person.

In the Silicon Graphics culture, you were almost ostracized if you ever left and tried to do something else. It didn’t even occur to anyone to leave. People would have wondered, "You left such a great company with such a great group of smart people doing great fun things. How could you do that? You must be a little nuts." As an example, when I left and had success with Netscape, the floodgates opened and they never shut. So, never lose that first good person.

CONTEXT: You say that many of your businesses have been set up to scratch itches that bugged you for some time, but that didn’t happen with Netscape. After you left Silicon Graphics in 1994, you were driven by the desire to start another business. You quickly settled on a colleague, Marc Andreessen, but the two of you had a couple of false starts and took some time to come up with the Netscape idea. Starting a business based on the business’s or the founders’ needs, rather than customers’ needs, is a classically bad idea, yet you succeeded where others failed. Why?

CLARK: Let’s go back to 1991, when I was still at Silicon Graphics. I commissioned a study of the cable industry. It helped me become aware of fiber-optic networks and a bunch of things that are pretty much in vogue these days. Partly because Silicon Graphics had been in the digital-animation business, my technology background and instincts came to the fore. I said this is going to be a natural. There’s going to be television merging with computers, and networks expanding.

I wrote a paper about my ideas in 1992 and was invited to talk at an industry conference. Afterward, Time Warner got in touch and we cooked up the Time Warner project in Orlando, Fla. We built this big, interactive, television system that actually worked. It required a $15,000 computer as the "TV" receiver, but it worked.

That was one bit of experience I’d picked up before Netscape.

The other was that I had become extremely aware of the value of the PC. It started out as a toylike device, but brilliant people like Bill Gates went for market share and, by making the market so big, made the PC powerful. I had lived with the growing strength of the PC for 10 years at Silicon Graphics.

When I became aware of the Web, it became crystal clear to me that the Internet was to the future of communications what the PC was to computing back in its early days.

The Internet was really primitive. But I said, "If this thing breaks out and becomes a real commercial network, it’s the embryonic form of the interactive network I had been envisioning all along." I thought, "My God, here is my opportunity to grab something really big, make it happen quickly, grab market share, and just run like hell in a big, big market."

CONTEXT: So, while you may have been driven by a personal need, you also had a powerful insight about technology. Tell me how you think about your current businesses.

CLARK: Well, myCFO was a service I created out of personal frustration. I built this for myself. Taxes have to be done. I also needed help dealing with paperwork, all my bills. And I wanted access to all my financial information online, through one consolidated view.

Although myCFO intends to be an online business, we’re setting up offices because the nature of our business is people, relationships.

Shutterfly is another example of something that used to bug me: film developing. Until recently, you had to go to a physical store to turn in your film. This was terribly antiquated. Shutterfly lets people upload their digital photos to our site. With wireless technology, someday soon you’ll be able to zap your film to the Shutterfly site and then continue to take pictures. It will be practically instantaneous. You won’t have to burn gasoline to go back and pick up your prints. You can get them mailed to you where you are. You can get them sent to your family. You can get enlargements. I think that’s a great idea.

Shutterfly is still not profitable, and it has the most competition of the companies I’m involved with at the moment, but it is averaging a staggering number of prints per day and the growth rate has been 50% to 80% a month. That suggests to me that it is on to something.

CONTEXT: What about Healtheon? It seems to have hit some barriers.

CLARK: The main one is bureaucracy. In the health-care industry, there is an established way of doing things and no feeling of urgency to change. Doctors aren’t a competitive lot. They already give patients what they think is the best service they can. They’d love to decrease the cost of running their practices, but it’s not clear to them how Healtheon can help. Submitting a claim electronically doesn’t necessarily save them money. What it does is save the insurance companies money.

So, there’s lots of administrative bureaucracy that’s in place. At Healtheon, I’ve learned that a poor way to start a business is to think you’re going to attack an existing industry that has waste and eliminate it quickly. There may be hundreds of billions of dollars of waste in health care, but all that is part of someone’s top line.

Healtheon has been a long, five-year slog. But I’ll do well financially, and I think the company is poised to break out and do really well. When all the integration is finished following the series of mergers that Healtheon has done, I think the company’s strength will become evident to everyone.

I’d like Healtheon to be viewed unconditionally as successful. Silicon Graphics was a marked success, but then it got battered. Netscape was acquired. Healtheon was my third. I want it to get up and be a very successful company. I don’t want to be viewed as a person who starts mediocre companies and then abandons them.

CONTEXT: Is innovating harder these days? You and others have shown how much potential the Internet has, so now everyone is looking in the same space that you are.

CLARK: It makes it more difficult. [Smiles.]

CONTEXT: So, where is the next "new, new thing"?

CLARK: DNA Sciences is the company that excites me right now. Its objective is to personalize medicine by finding the correlations between diseases, genes, and treatments.

Its Web site allows people to donate DNA to the Gene Trust. By getting donations from people with known disease conditions, it hopes to find the correlations with genetics. The company has on its board of directors Dr. James D. Watson, who won the Nobel Prize for his co-discovery of the structure of DNA, and it has built a genotyping facility that will allow the typing of the genes related to specific disease states.

If DNA Sciences discovers how to test for and treat diseases more efficiently and personally, it can build a very valuable business.