U.S. News 5/22/00
A chat with Andrew Grove and Craig Barrett
The CEO and chairman of Intel talk tech with U.S.News & World Report
usnews.com
By William J. Holstein
ANDREW GROVE
Though he thought he'd be easing back by now, Andrew Grove, the co-founder of Intel and author of Only the Paranoid Survive, is still chairman of the company at age 63. Here is the transcript of an interview with U.S.News & World Report, conducted on April 18.
Will financial volatility damage the health of the technology sector?
There's a reentrant circle that reinforces itself. Technology has created a lot of investor interest because it plays such a direct role in the economy. The valuations that investors therefore place on these stocks has become a role in the growth of technology. Volatility as such has very little to do with it. What's much more important is the valuation of these shares, and the valuation climate has been exceptionally good in recent years.
Wouldn't there be a slowdown in how technology is changing the economy if the value of these stocks dropped?
I don't think so. If there were a serious and lasting correction, it would result in a slowdown in the formation of new companies. But the implementation of new technologies is typically done by mature companies, which finance it out of their own internal cash flow. Typically at Intel, we aren't selling stock to buy computers with the proceeds.
Aren't there speculative excesses in the markets that need to be purged?
There are always excesses. When money is available, it is used sometimes for good, sometimes for excesses like M.B.A.'s who get $30 million on the basis of a business plan. At the end of the day, I can't make a judgment on what's excessive and what's not. The market has its own logic, its own determinant. Things are worth what other people will pay for them. That's what the market does.
Doesn't a shakeout imply that start- ups will be buying fewer Intel products?
The bulk of the money spent by E-commerce companies has been for brand advertising. There are going to be fewer ad pages in U.S.News & World Report much sooner than there are going to be fewer computers bought. I've been mesmerized by the fact that old-media industries, including television and broadcast companies, have been beneficiaries of the new media that was supposed to put them out of business. The investment money has been going into advertising spending, which can easily be dialed back.
You're a believer in the new-economy argument, right?
No. What makes you think I'm a believer in the new economy?
Well, your company has helped create it.
I'm a believer in technology, and I'm a believer in technological transformation of a lot of commerce. But the phrase "new economy" is so ill defined and imprecise. I can't be a believer in it unless you define it.
In the sense that our economy behaves in different ways than it used to?
It's more productive.
So it's real? It's sustainable?
Oh, absolutely. Productivity gains are sustainable. I think we'll see business cycles in the future just as we've seen business cycles in the past. But productivity growth allows us to grow at a higher rate without increasing inflation. That part, I believe. The role of technology in the economy has grown over a period of time. That has been recognized slower than it has actually happened. In the past two years, the recognition has caught up.
Turning more to business, when was Intel's relationship with Microsoft born, the so-called Wintel alliance? Was it the result of an explicit deal or did it happen by accident?
It's very simple. IBM was in a big hurry to get into the PC business in 1980 and, in a departure from previous practices, purchased microprocessors and other components on the commercial market, as well as a commercially available operating system to accelerate their entry. We [Intel and Microsoft] were lucky or good enough to be chosen independently of each other.
So there was never a time when you at Intel said to the folks in Seattle, "Let's do business together"?
Microsoft, at the time, didn't have a product ready. They were a tool company and language compiler. They were working on products but not operating systems.
Now it seems that you are trying to diversify from your reliance on the PC, which means you are moving away from that old alliance into a brave new world. Is that right?
Well, for 15 years, the PC was the environment in which our products were used. Today, there is network computing and Internet computing. Personal computers are enjoying faster growth rates than they have in many years as a result of their being the dominant access device to this world of networked computing. But they are no longer as dominant as they were when they were a stand-alone tool. What is dominant today is a much broader configuration of factors that involves servers, personal computers, wireless devices, but most important, networking implements that connect all these devices into a coherent network. We accept the market environment as our guiding factor. It's a lot more complex. A lot bigger. It doesn't take the place of the personal computer; it supplements it. We have to accept that guidance for our strategic activities. We have to change along with the environment that has changed on us.
Are you changing fast enough or too fast?
We can probably judge that in five years' time. We are doing a whole lot of things. We bought 12 companies in 1999, some of them substantial. Can we productively make them part of a whole? If so, we didn't change too fast. If some of the integration effort doesn't work, then obviously we changed too fast. We have to constantly balance the need to adjust to this market picture rapidly, but not so fast that the execution of what we do falls substantially behind our strategy. We are a nuts-and-bolts company. We don't lead by press releases. We lead by products. It would be very easy for us to run ahead of ourselves. I don't know whether we have or not. We won't know for some years.
Do you foresee these other areas replacing the PC or enhancing it?
I can't see anything replacing the PC except the possibility of Internet home terminals, which so far haven't amounted to very much. The networked environment means there are a large number of server computers doing the back-office jobs that these personal computers generate. The communications equipment into which our chips go connect all these computers to each other. The network-computing phenomenon supplements the growth of PCs. They are enhancing each other.
Microsoft seems to be trying to extend its software from the PC to new fields. But Intel seems to be willing to embrace more open standards. Is that right?
That dates back to 1990 or so, when we tried to introduce the PCI Bus. To be perfectly frank, we tried to keep it proprietary. But the market [i.e., customers] told us they wouldn't cooperate with us. We accepted that message and made the PCI Bus an industry-collaborative standard.
Does the Internet mean that the era of being able to impose a proprietary technology over broad numbers of users is over?
Connectivity standards always have to be open. From railroad gauges to air traffic systems to the international telecom standards to TCP/IP, all communications standards are only valuable if everyone can join in that community to communicate on the basis of the same rules. As the market environment for information technology moves from a stand-alone world to a very connected world, the rules of the connected world will dominate more.
At the same time, the stand-alone components that plug into this network will not open up. I'm not going to name names, but any number of companies?particularly of the consumer-electronics heritage? are building proprietary boxes, propriety standards, and proprietary operating systems. The requirement for those products, whether it's a cell phone or a pager or a game station, is that they have to connect to a network in an open-standard fashion. We all participate in a network, and the rules of the network are different than the rules of stand-alone devices.
Once you accept the power of the network, it has huge implications for how you run your business.
That's correct.
How will the Intel culture have to change to make this ransition?
The biggest change is that we will have to become more flexible as we acquire companies that each have their own cultural needs and requirements. As we buy these companies, they tend to have two, three, or four locations and are geographically dispersed. Intel's method of operation was to establish large campuses. We had a large presence in New Mexico, we had a large presence in Oregon, we had a large presence in Phoenix. When we acquire companies, they might have some very good engineers living in a town in Ireland, or they might have a design center someplace else. When we acquire the technology, we acquire these sites. The notion that we can pick them up and move them and consolidate them into one of our locations is out of the question. So we have to learn how to deal with engineering resources scattered around a lot smaller and a lot more numerous locations than in the past.
And each of these companies has some characteristic or means of operating. We have to find a balance. How well do we integrate them, and how closely do we integrate them? Quite clearly, technologically speaking, we have to have close integration marketingwise when we are presenting products to customers. But as for internal operations, we have to offer more leeway to people to continue to operate in their own style. That's new for us. It's been a point of pride for us that you could basically land at any Intel location anywhere in the world and get the same product feel and the same conduct and culture as any other location. We have to figure our way through a fairly subtle set of distinctions where we are very excruciatingly clear about strategic issues and much more tolerant about process issues like culture.
<i.>ow many more companies will you buy, and how fast?
We intend to continue at the same general pace, if we find the right candidates and those candidates want to become part of Intel.
government is not concerned? Nobody is saying Intel is too powerful?
We're not acquiring companies in microprocessors. We're acquiring companies in areas where there is incredible competition, like networking-related silicon processing and network-equipment devices, where we are starting from a very low level of participation in a fragmented industry. The fact that we have a strong presence in microprocessors doesn't carry over. I wish it did, but it doesn't.
long before you can really show financial results?
Don't pin me down on a date, but we started in the network area in 1990 or 1991. It took us 10 years to reach $1 billion in size, including some acquisitions and internal growth. It doesn't take a lot to generate announcements and press releases. But it does take a lot to organically build businesses. We are moving into new areas at a pretty aggressive pace. The question is whether we're moving too fast or too slow. I can't look you in the eye and tell you we are moving at just the right speed, because how do I know?
you using Intel Capital, your venture-capital arm, to take stakes in comð panies that you will eventually absorb?
No. We haven't had too many examples where we actually bought a company that we invested in. There's a very tired clich‚ in business language, and I use it. We look at the investments as being part of the "ecosystem" of each of the new areas where we are trying to expand. That's the strategic benefit. We make each of these areas more valuable to us.
like this term "ecosystem"?
I take no ownership of that word. It's actually not a bad word. But I cringe every time I sit through a presentation and every second sentence has the word "ecosystem" in it.
How hard is it to absorb some of the new companies you're buying?
Very recently, I went to San Diego to visit two or three companies that we acquired. I was warned before I landed that I shouldn't be shocked if I stumbled over dogs in one of the facilities, which I thought was pretty funny. I didn't actually see any dogs. They were put away during my visit. However, the general impression of the visits is that our new employees were very happy to be part of Intel and liked the technical and marketing relationships. But they hated our operational rules, such as the requirement that each building should have 24-hour security and that each lab should have a keycard system. I was completely taken aback by the things that I consider second nature on our part, how much of an irritant they were [to newly acquired companies]. So we will have to loosen up the uniformity of our practices. We will have to go a lot easier on them than we would on our own facilities.
How much longer are you going to be actively involved in all this?
I enjoy it. I don't have any particular age at which I will retire. I did that once when I was 50 years old. I was made CEO when I was 50, and I said I would work until 55. Gordon Moore, who was then the chairman, said that's like the California speed limit, [which many motorists ignore]. He turned out to be right.
So you're still fascinated by the battle?
In a way, more than before. Because the change in the PC environment and the Intel environment is intellectually very, very stimulating. The PC was on some kind of technical saturation curve. Yes, we were making new microprocessors faster, but they were things we'd been doing for 15 years. Here, you have to wrap your mind around what the Internet environment really means. Does it mean a third-generation dot com or a second-generation dot com, or does it mean a brick-and-mortar company moving into Internet space, or all of the above? For example, we now need to segment our sales force because different approaches and different sales techniques and different merchandizing techniques are necessary to reach these new classes of customers. Figuring these things out is spectacularly complicated and very challenging. It's kind of an intellectual new lease on life.
--------------------------------------------------------------------------------
CRAIG BARRETT
Intel Chief Executive Officer Craig Barrett has pushed to change the culture of Intel, in part by organizing a series of strategy sessions for the top 400 executives.
Why did you decide to launch such a sweeping change of Intel?
I'm kind of a pragmatic engineering guy. It looked like we were pushing a $30 billion company. We had one very strong product line, namely microprocessors. That business we anticipate will continue to be strong and vibrant, but it wasn't going to grow at the same rate over the next 10 years as it had been. We had a decade of 30 percent compounded growth. So we needed to look for other opportunities. In the same time frame, the Internet came along. The issue was: How do you change the culture of a very highly focused, almost single-product company into a variety of disciplines and products?
How do you manage a company that seems to be going in so many different directions?
It's truly amazing, isn't it? One of the reasons for the one-week training sessions was to pose that same question to our top executives. I'd kick off the meeting and go back at the end of the week and ask them, "What do we need to do differently?" We then compiled the suggestions from the organization. You could compartmentalize the answers. One was: If you want to do something differently, you have to organize differently. You have to disperse the organizational power. When we were a single-product company, Andy Grove was the chief product engineer. We had to push the decision making on these new areas to lower levels. We had to raise the new levels up on the organizational charts to be on par with the microprocessor business. We had to take some highly respected senior managers and focus them on doing some different things. We had to let people know we were serious. We had to demonstrate that we would take risks and offer internal venture capital for ideas that bubbled up. We also convinced the board of directors that we would have to supplement this internal development with outside acquisitions. We did $6 billion in acquisitions last year. We're on track to do more than that this year."
But you're trying to do all this with precisely the same people who have a commitment to the old ways and the old culture.
If you looked at some other companies where they had a strong culture, that strong culture may have retarded [change]. If we had any advantage, it's that we were open with top-level management, solicited their opinions, and then in a very overt way made changes. We moved senior people around. We sent a message to 70,000 employees: "This is where we're headed." People could not say we're just talking, we weren't walking the walk. We had credibility. Once you have credibility, the cultural similarities helped. We spoke the same language. We could move very rapidly. It may be the advantage we had is that we were able to pinpoint three or four things we were going to do and do them rapidly.
As you move forward, how is your relationship with Microsoft going to shift?
Microsoft and Intel have always had a generally aligned set of strategies. But we've always had friction points and differences of opinion. Microsoft wants their software to run on all hardware, including Intel hardware. Intel wants all software, including Microsoft software, to run on our hardware. Strategically, we're kind of aligned trying to move the computer industry forward. But when you get below the great big strategic vector, we have differences, and we've had them forever. Most recently, Linux, the open system software, is a challenge to Microsoft in some areas, and we've been very public that we want Linux to run best on Intel architecture, and we've made some investments in Linux companies. As the Interð net has come along and the server backbone that runs the Internet [has grown], neither Intel nor Microsoft was particularly relevant. Windows has not been the operating system of choice. As we have moved into that space, we've worked with other people. We still get along very well with Microsoft. The desktop is still very much a Windows desktop, and we're still very much aligned in the desktop. But in hand-helds and servers, we continue to diverge more. Microsoft's CE promoted hardware other than Intel's. Their Game Box, to compete with PlayStation 2, is based on an Intel microprocessor. Generally we're aligned, but we have differences of opinion.
Is it possible that we're seeing just the beginning of that divergence?
Sure, there's the potential for that. In the network space or cellular-communications space, there are different operating systems in those spaces that are not Microsoft.
Coming back to the task of remaking your own culture, are you going to be flexible enough to absorb these new companies you're acquiring?
We have become more flexible in our location planning. When you're out acquiring people spread around the universe, and you're used to big campuses, and suddenly you have 40 people in Charlottesville and 20 in upstate New York and 40 in Sausalito, Calif., you have to be more flexible. I think our employees were looking at this to see if we were serious about changing.
When we started, we gave [the acquired companies] a smothering embrace. We went in and said, "Good, you're Intel employees. We need to upgrade your human-resources system, your cafeteria systems, everything." We quickly recognized we were making acquisitions for the people and the intellectual property that resided in their heads. What we had to do was not smother these people with bureaucracy but to find out what technologies they had and what plans they had.
At Santa Clara headquarters, we've been pretty rigorous in saying that if you're in the area, you work here. But we're starting to look at satellite facilities. If you live in San Francisco, why should you spend an hour commuting each way? Maybe we should have a satellite-hotel kind of facility in San Francisco when you don't need to be in Santa Clara. We have to understand the more freelance style of these acquisitions. In a start-up company, proð tocol doesn't mean a lot. You do what you need to do to accomplish a result. Big companies impose the structure and bureaucracy on top of the pragmatism. It's healthy for us to look at these companies?not only can they learn from us but we can learn from them.
How are you and Andy Grove dividing your jobs these days?
It's been a very graceful extension of how we interacted before. I was in the chief operating officer role and in the president's job under Andy for several years. Our offices are only about 20 yards away. We can walk down the hallway and chat with each other. On a weekly basis, we have a list of "to do" items among us. We partition the tasks very cleanly between us so that there's not much overlap. There's never an issue of one person superseding or undermining the other. As he suggested when he became chairman, Andy has spent time focusing on the big trends. He spends a lot of time talking to start-up companies and bringing it back and articulating it in our management meetings. It's been very symbiotic. |
