How ‘Paranoia’ Has Helped Intel Rise to the Top

As chief executive of Intel Corp., Andrew S. Grove has long managed according to the motto that serves as the title of his new book, “Only the Paranoid Survive.”

By constantly assuming that the competition is about to win out--even when the evidence indicates, as it has for most of the 1990s, that rivals have no chance of challenging Intel’s firm grip on the computer microprocessor market--Grove has made Intel into the world’s most profitable and fastest-growing large industrial company.

The book is focused on identifying and coping with a business phenomena that Grove calls a “strategic inflection point.” This is a make-or-break juncture for a business that’s caused by some kind of very large-scale change--a “10X change”--in the business environment. Companies that fail to understand strategic inflection points often whither and die, Grove says, while those that manage them well go on to even greater prosperity. And the basic idea of the strategic inflection point can be applied to an individual’s career as well as to a business.

The book, subtitled “How to Exploit the Crisis Points that Challenge Every Company and Career,” starts out with an account of a recent inflection point in Intel’s history: the infamous “Pentium bug” incident, in which a minor flaw in Intel’s flagship microprocessor provoked an uproar among consumers and ultimately forced the company to offer free replacements at a cost of some $475 million.

Grove spoke with Jonathan Weber, editor of The Cutting Edge, about strategic inflection points, mergers, and the impact of the Internet.

Question: In defining strategic inflection points, you say they’re not just a result of technological change but rather that they’ve existed throughout business history and in many different kinds of industries. But it does seem like they would be more frequent in technology-related industries. Is that true? Or do they come up more often for all different kinds of businesses than they did, say, 50 years ago?

Answer: They do. But it’s not just that they occur more often in technology businesses. As technology impacts a larger part of the total economy, it causes these ripples to hit these other industries at a higher frequency.

The newspaper business, for example, is not a technology business, but it is impacted by the stuff that companies like ourselves are doing. The faster the change rate in our industry, the faster these waves come at other industries, and the reach of those changes into other industries is greater.

Q: Is there always a good response to these inflection points, and as a corollary to that, is there always only one response?

A: There’s probably one best answer for a given company. It doesn’t mean that any other answer will kill you. . . . I’m in the middle of teaching a course, a brand new case on the telecommunications industry--brand new in that the Internet is now part of the case. I don’t think it’s widely appreciated what the Internet is going to do to telecommunications.

And I have a student body at Stanford that is very Internet-savvy. We had a very good discussion on this, and at the end of it I said, “OK, you’ve got homework for next time, write down the pluses and minuses of what the Internet means to a telecommunications company and then make up your mind whether five years from now it’s going to be a net positive or a net negative.” And they were almost evenly split on the net positive and net negative. They’d all gone through the same case and the same discussion. It’s hard. That’s what it’s like being in the middle of a strategic inflection point.

Q: It seems like a lot would depend on what particular companies do with this inflection point. If Pacific Bell or SBC Communications or whoever doesn’t take the actions that are required, it could be a negative for them but a positive for somebody else.

A: Like a BT-MCI.

Q: Yeah.

A: That’s one of the most appealing mergers, most logical mergers. With the stroke of a pen they created the first international telecommunications company. That’s pretty impressive.

Q: It certainly seems more logical than some of these other mergers, which I don’t really follow exactly.

A: Like ABC and Disney?

Q: Yes, among others. This whole notion of vertical integration in the entertainment production and distribution arena, I think they’re going to do this for five or 10 years and then they’re going to disaggregate it again because they don’t know how to run something like that.

A: Exactly. I believe there’s a big economic trend toward specializing in a horizontal fashion. There’s an invisible force that is rearranging industries, when they become mass-production industries, toward a horizontal arrangement. Someday, somebody’s going to get a Nobel Prize in economics for coming up with a mathematical analysis of why that is so.

Industries start vertical because you have to bundle a new service together in its completeness. Then they grow up in a vertical fashion, and when they get big enough they get horizontalized. It happened in the computer industry. It happened in the transportation industry. . . . If somebody did a study, there would be a lot more success in horizontal mergers than in vertical mergers.

Q: Let’s go back to a specific point in the book, the Pentium flaw episode. It seemed very clear at the time that you were going to have to do what you eventually did and offer to replace all the chips. Why couldn’t you see that?

A: I’m inside my skin, I’m inside this company, I’ve dealt with quality problems for 30 years and that’s not how we deal with quality problems. You analyze them and you go into the factory and you explain it to the customer. And all of that is possible while you’re dealing with people who look at data and analyze it in the same kind of fashion as you do--a bunch of other computer engineers.

You were looking at it the same way as everybody else in the world was looking at it. And you were right, you were a proxy for the consumer world and they are much bigger than us.

We should have [offered a replacement chip to anyone who wanted one] and the whole thing never would have happened. And it’s not like nobody said we should do that. . . . [Intel Chief Operating Officer Craig] Barrett was right about it, we argued incessantly over what we should do when it was starting. Barrett said, “Why don’t we give anyone a chip that wants it.” And I said no! I mean, that’s ridiculous! Where do we stop? Then somebody’s not gonna like something else about it. Do we just forever replace chips? I mean, we’ve got a specification, the stuff meets specification . . . the irony of this is that Craig is the internal operations guy and he was right on that one.

Q: What process didn’t you have in place or what kind of mechanism should you have had that would have let you see that?

A: It was an attitude issue. I was being self-righteous, “I’m right, I’m right, go away,” that kind of thing. It was largely an attitude issue and it was largely me.

Q: At the end of the book, you talk about the Internet and evaluating whether that’s an inflection point for Intel. You go through this test that you set up at the beginning to determine whether a circumstance qualifies, and then you say according to the test it isn’t an inflection point, but that actually it really is. That seems to undermine the notion that there’s a systematic way to analyze these situations.

A: I think one should go through this test, but it’s not a formula. According to the formula, the Internet shouldn’t matter to us. But it’s just incredible to me that half the world is going to change and for us it’s to be business as usual. I don’t believe my own analysis.

I really do think it’s going to be a big deal. I think it’s going to change personal computing. There are a bunch of models on how it’s going to change it and I don’t buy any of those models, but it’s going to change it. And the only way we’ll know how it will change it is not sitting on our ass and speculating but by building some of these [new kinds of network computers] devices and introducing them or having someone introduce them and working with people who don’t necessarily see the world the same way as we do.

(Some companies, led by Oracle Corp. and Sun Microsystems, are developing a new class of stripped-down personal computers known as “network PCs” that are less expensive and easier to use than traditional PCs. They take advantage of software programs and content residing on the Internet.)

Q: But you talk in the book about how you can’t hedge your bets, you need to make your decision and direct all of your resources at a single, clear objective. Isn’t what you’re doing by designing “network computers” and other Internet devices hedging your bets?

A: First let chaos rein and then rein in chaos. We are in the early phases of letting chaos rein. You can’t make these definitive decisions too early, or they are likely to be wrong. You have to get some miles on your tires with experimentation before you can decide.

Some of the stuff you can write off. When the whole thing was a home device like [Oracle Corp. Chief Executive] Larry Ellison described a year and a half ago, that was fairly easy to dismiss. . . . But all the reasons we dismissed it as a consumer device are different if it’s a corporate computing device.

We ourselves believe in managing computers down the wire [on a network]. Now how much difference is there between managing PCs down the wire and network computing? Why, it gets a little semantic.

Now if it’s network computing and you started from scratch, would you build a network computer the same way as you build a PC? Well, probably not. How will you build it different? And you start arguing. But now we’re arguing the right question. We are past the question of whether it’s a big phenomenon. We’re past the question of whether the computer’s going to be different. All we are arguing is to what extent and in what way it will be different.

Q: So a certain kind of experimentation is not an implementation of a decision but part of a decision-making process.

A: That’s correct.

Q: In terms of Intel’s strategy in dealing with the Internet, one thing you said in passing in the book that really surprised me was you see media and advertising as an opportunity for Intel.

A: No, no, no, no. Not as an opportunity for Intel. It’s an applications driver [for the PC].

Q: You’re looking to fuel chip sales.

A: Right.

Q: Which is different from, for example, Microsoft. They really think they’re going to be a media company. They’re not just doing media to sell software that makes media work.

A: Microsoft probably sees themselves as a software company that sees the boundary between software of one kind and digital content software of another kind blurring.

Q: What do you think about that?

A: I don’t know. It somewhat depends on the response of the incumbent media company companies. It goes against the rules of the horizontal industry that you cannot be in two horizontal bars in a substantial fashion. I mean, it’s like we couldn’t be a branded PC company and supply every PC company on a daily basis.

Q: But that’s what they want to do. They want to have outside suppliers of content and then they want to compete with them too.

A: I see tension in that firm, and if the incumbent media companies choose to really be significant digital-era firms, the tension’s going to be unbearable.

Q: Do you see the incumbent media companies doing that?

A: They haven’t asked my opinion. But no, I don’t. I see early-stage experimentation but no commitment. I can’t look at anybody and say these people are going to be kings of the digital media or die. And I think it’s time. If you look at the “computer-watching” population, it’s the equivalent of a major network, 15 to 20 million people. So do you see investments commensurate with that? You see that at Microsoft, but elsewhere? I’m not sure.

Q: Obviously, the Internet is the big thing out there, but is there anything else you see as a “10X” force, or a force comparable in magnitude to the Internet?

A: I do, but it’s a subject all to itself. It’s called globalization. This has turned into a cliche and everybody forgets what the hell it means. And what it really means is there’s one marketplace for labor, one marketplace for goods, one marketplace for everything. An increasingly large portion of work that can be done, that needs to be done in a modern society can be done practically anywhere in the world. That comes with the Internet.

That has been happening in material goods for the last several decades, and we have seen a major impact in terms of stagnant wages and standard of living and all this. And it’s now moving to the knowledge area and the impact is going to be horrendous. There are going to be winners and losers, a very major societal impact. I won’t be surprised if this gets to be a big subject in the next election in the same way immigration has. It’s a big deal.