We interviewed Ed McCracken at SGI’s worldwide headquarters in Mountain View, California.
“CEOs have this obsession to know everything and
control everything—that is stupid.”
You were extremely successful at HP before you came to SGI. What made you leave?
I think some of it was personal. I really enjoyed running a division at HP. I was very effective at it. And then, because it was so very suc-
194
ED MCCRACKEN
cessful, we broke it up into multiple divisions and I became part of that “in-between layer” in HP, between top corporate management and the division employees.
And I found that I didn’t enjoy that job and I wasn’t extremely effective at it. I really wanted something that was a little bit more hands-on, externally focused, rather than internally focused. I really liked the customer environment. I liked the technology, and I didn’t like all the other stuff very much.
So, when SGI came about, I had a chance to come over here and meet some of the people, and saw the value-added in the company. It didn’t take long to get a whole new track going, and I was over here.
You didn’t really enjoy your growth and your progress in administration at HP?
When I joined HP I assumed that I was going to stay there for a max-imum of five years. And then it turned out that, every year, things were getting so much better for me at HP, it was hard to leave. And I was enjoying it. As long as things were moving so quickly for me at HP, it was hard to leave.
We’ve heard a lot about HP’s culture from the other CEOs that we’ve spoken with. Do you think you brought some of that culture over to SGI?
Some of it. I learned a lot at HP. It’s a great place to learn. I learned a lot about core values. Some people think our culture here is somewhat like HP, because it’s derived from it. But, it’s also quite different, in that there’s a little more intensity about being on the leading edge, that wasn’t as prevalent at HP.
Why do you think that is? Why do you think SGI is more intense, compared to HP?
Well, I think it’s maybe because of the time in which we grew up.
Hewlett and Packard, both wonderful people, came out of the World War II generation, that had a different set of values—perhaps including a little bit more institutional responsibility for the individual.
Whereas we grew up in a time where information-age individuals wanted to take responsibility for themselves; they didn’t want the company feeling responsible for them too much. And there’s so much more industry here in Silicon Valley, so that if you don’t succeed
SILICON GRAPHICS
195
here, you can go down the street and try again. It isn’t that you’re going to work for Silicon Graphics your whole life, probably. But the idea of the HP core values is, you’re probably going to work there your whole life, and, this is not part of the new world.
Plus, the computer business is changing rapidly; it’s a chaotic world, the time frames have condensed dramatically. You need to be able to make decisions quickly, and get on with them and get the job done. And that’s important. It’s also important you don’t miss a generation, or else you’re dead.
So, as the stakes have gone up, the time frames have gone down.
At HP, I always felt that I could really work hard and make anybody successful who worked for me. Here we don’t have time. If somebody’s not cutting their own way, then you just don’t have time to deal with it, as their manager. So, it’s much more of an intense culture here. Unfortunately, in the computer industry, there’s no place to hide.
Looking back, do you think you’d do anything differently regarding your transition from HP to SGI?
I really can’t with my philosophy. I almost never go back and try to rewind history. And, it [SGI] is a real success story. And if I had it to do over again, I think I’d be lucky to do the same thing. So, I don’t know.
When I came on board, it was a real disaster here. We had great people, great technology. The people here had done a project, and it was a good research project. But, the product didn’t work, and we had a hundred people. And the worst thing about it was that there were only two or three people in the company that knew the company was in crisis. They had kept it from everybody else, because they thought it would impact morale negatively.
Of course, the first thing I did was to call a company meeting, explain the whole cash situation, what our options were in terms of big layoff, or somehow saving enough money and designing our second generation in time to really solve the problems.
And then the rest is history, and that is we designed the second generation in six months without laying people off, for $500,000 a month, and we went out and raised another $10 million just in case, which we actually didn’t use before our IPO.
And, we all lived. But now a lot of people appreciated it [calling the company meeting], and we have developed this culture of sharing
196
ED MCCRACKEN
information and no secrets and disclosing everything and having everybody be part of the solution.
And you think that sharing information is one of the things that defines SGI’s culture.
Well, it’s, it’s one of them. I think the whole idea of an information-age company is to get as much connection as you can up to the nodes of the organization, so that as the nodes make decisions, they’re going to make them within an overall context. Because you want them making decisions as much as possible. Because it’s quicker and better.
How do you ensure that the information that you transmitted to the nodes is accurate? That is, when you send a message down five or six layers with 80 percent of the message accurately being transmitted, then there is an accuracy level problem.
Well, you don’t try to communicate through different levels; you communicate directly. You find every opportunity through the Web or through electronic mail, or through any means to go directly. I’m doing that with the company right now.
I mean, the old idea of the fifties organization, where, you worked as a manager with your staff, and then your staff worked for their staff, and their staff worked for their staff—it doesn’t work that way here.
Everybody here is trying to learn—analyzing the black box, trying to understand how it works, and its advantages and disadvantages.
And in the end—everybody shares the same kind of data.
In terms of learning, you were quoted as saying that the military industry was driving a lot of your technology, but now it is Hollywood that’s doing it—
Yeah—Hollywood certainly is an important part of our future, and I think we’ve attracted Hollywood for two reasons; one is we enjoy it, and all of our employees enjoy it, and the second is that Hollywood is full of people who are not satisfied with today’s technology.
One of the criticisms we’ve dealt with since the beginning of the company is all these people out there who say, “Having that great technology is so expensive, but who needs it? You can do whatever you want with a PC.”
SILICON GRAPHICS
197
And, people have been saying that now for twelve years. People in Hollywood know better. They know you don’t do these kinds of things with PCs. Our market tends to be characterized by people who know that you can’t do these things with PCs. Technology drivers.
We’ve tried to define your driving characteristics and came up with: unassuming, analytical, and humble—(All laughed.) —in a pretty big way. I went to a speech that you gave recently, and there were about seven or eight times where you just said, “I don’t know.” And I thought that was such a departure from most of the—
—from Scott? [McNealy of Sun Microsystems]
(They laughed again.) Well, without naming names. I decided it was pretty unusual—
—You know, CEOs? CEOs have this obsession to know everything and control everything. I mean, that’s stupid. You can’t know everything. You can’t control everything. In the inf
ormation-age organization, you want people in the nodes to know everything associated with what they’re doing.
We have a learning organization. So, if you feel like you know all the answers, you can’t learn, and you probably don’t belong at SGI.
Understanding those new paradigms before other people do is our corporate strategy. We have to be able to learn. You have to be the best technologist working with the best customers in a learning mode.
And the result of that learning is understanding new paradigms, and then implementing that existing technology, putting that right into a silicon, if appropriate, and getting that into the marketplace five years before the commodity players. That’s what gives you the fifty percent gross margin and lets you spend ten to twelve percent on R&D and twenty percent on selling, and have a model of innovation.
So what are some of these new paradigms?
It can go all the way from specific things like morphing and texture mapping and graphics, to the importance of the Web, to parallel processing with software that’s programmable by mere mortals. It’s ample range, but, our company has consistently introduced the kinds of technology, the kinds of systems that are at least five years from the commodity players.
198
ED MCCRACKEN
When you have to constantly introduce these new technologies and at the same time keep growing as a company, how do you find the right people? You recently hired a lot of people, and in the process stated that you didn’t want a “bozo invasion.” How did you manage to do that?
When you hire lots of people you do make some mistakes. But, generally, I think, we have done an excellent job. The issue we ended up having was that it took a lot of time. So, we didn’t really plan as much for having our managers so heavily involved in recruiting that they had less time to do everything else that they had been doing. So we found that there’s a tremendous delta [change] between growing at 30 to 35 percent, and growing 40 to 50 percent. That delta is seemingly pretty innocent, but it really isn’t because we have all the natural feedback in our company that grows 30 percent. But if you try to go 40, 50 percent, you’ve got to really work at it, and you can only really work at a handful of things, as a manager. We’ve had a few issues like that that we’ve kind of worked our way out of now. We’ve gone through the pits of hiring all these people and trying to integrate them, and taking all the time to do that.
We’ve slowed down now. We’re tuning up the machine, and you can feel the machine getting more aggressive and more tuned now than it was six months ago.
You’ve had phenomenal growth over the last few years. How much of SGI’s success do you think is just luck? Some of the CEOs that we’ve spoken with feel that it’s all luck. Do you believe that?
Well, I certainly agree with, whatever you call it—karma or whatever else—there’s a lot of that in everything. I grew up on a farm in Iowa, and I really appreciate the farming mentality because you work really hard—and then you let the weather happen. And some years are good and some years are bad. You don’t have much responsibility for it. It’s what you do, rather than the result, which is important.
But, I think our company also gets a lot of really great people.
I’ve never seen the kind of talent we have in any other company. So, you combine that with the fact that we’ve had ideas that are truly revolutionary, that just really changed the industry.
With every person we hired, it seemed like we had new options.
We hired the multiprocessing team almost by accident. During a period of time when the computer industry was collapsing, including Silicon Valley, every year some company would fail or be close to fail-
SILICON GRAPHICS
199
ing, and we would be able to reach in and grab the top five percent of the people and bring them into Silicon Graphics, with a set of technologies associated with the new people.
That’s probably the luckiest thing that happened to us, actually.
It hasn’t been true in the last two years; we’ve had to get back to college and recruit people, because the industry is so strong right now.
We actually were better off as a company when the industry was less strong—now everybody does well. Everybody does well, and so everybody’s growing and everybody’s being successful.
It gets a little worrisome when it’s survival of the fittest, and it becomes increasingly difficult to have opportunities to actually not only hire great people, but also gain more market-share points. It’s really hard to gain market-share points right now. We’re doing it, but it’s harder because everybody else is growing 15 to 20 percent a year, and we’re growing 30 percent a year. It was a lot nicer when they were growing zero and we were 30 percent.
What about your different management style? There is a lot about you in the press about being meditative, introverted, and this is in contrast to most of the CEOs that we’ve met with. Why do you think you’ve been successful with this leadership style?
To me, the important thing about being a CEO—about being a manager at any level—is to learn who it is that you are and bring that into the job. It isn’t that there’s one model of a successful manager; there are a lot of models. I hope the book really deals with that. And the important thing is to understand who it is that you are and amplify that, and bring that into the job, and be comfortable with it.
So, I bring who it is I am, and Scott [McNealy] brings who it is he is, and Andy [Grove] brings who it is he is, and they probably all work, and that’s fine. My wife taught this course at the business school on leadership styles with Michael Ray [a professor at the Stanford business school], and they talked about the enneagram—
the nine different leadership styles, etc. And I think it’s really pretty good stuff. What it really teaches you is that you have to develop your own leadership style.
It’s different from when I went to business school, when they had a class teaching that there was one leadership style. We listened to the equivalent of Jack Welch. Everybody would try to be like Jack—well, that’s stupid. We’re not all like Jack Welch.
200
ED MCCRACKEN
Do you think that certain leadership styles are suited for entrepreneurs, and does Ed McCracken have that type of style?
You know, I doubt it. I think I’m really good when I take an idea—
working with a true entrepreneur, like the Jim Clarks of the world, the valley is full of them—that’s beginning to gel with a small team, and then I can really work with that and turn it into something that gets pretty big, and leverage it along the way. I’ve done that over and over; I’ve done that at HP, I’ve done that here. I really enjoy doing that. But, am I someone who can go off and team up with some people in a garage somewhere and start a company? Probably not.
Probably not me.
Why not?
Well, I’d probably find it boring.
Boring?!
I just wouldn’t get much stimulation out of it. I find it boring.
We’ve talked about the skills and assets that you bring to SGI. How do you think you fail SGI? What skills do you lack and how do you compensate for that?
Well, I mean, you’ve mentioned some of them. I’m quiet, I’m introverted, and I like to play with ideas. And I’ve figured out in my mind the right direction to go, I lose a lot of interest in it. So I try to surround myself with people that are much more interested in execution, implementation, getting the job done, schedules, all those kinds of things, because, in the end, I’m not as well cut out for that.
And again, I think that managers need to know themselves and not try to do everything themselves. They need to surround themselves with people that will complement them, rather than duplicate them.
We’re going to ask you some tough questions now—
—I thought you had been.
SILICON GRAPHICS
201
I remember hearing you on NPR one day talking about shareholder lawsuits. You basically said that a
lot of shareholder lawsuits are really unjustified, that they cost so much for the company, in terms of time and effort, that companies are forced to settle. But, how do you explain the plaintiff ’s premise that you guys really are—
—Crooks.
Right. And you’re not letting shareholders know the whole story, and that they have every right to know because you’re working for them.
I mean, I can’t imagine that every CEO in Silicon Valley is a crook. I think almost every company in Silicon Valley has been sued multiple times—companies like Hewlett-Packard and Intel and ourselves. And when somebody is sued, and when there is a big settlement, the people who get the money aren’t the individual shareholders. They’re really Bill Lerach [a prominent plaintiff ’s attorney] and his friends, who have made so much money out of this, it’s unbelievable. So they’ve had quite a business going for themselves.
There are rumors that their business has dropped down considerably after the recent legislation limiting shareholder lawsuits.
Well, it has dropped down somewhat. I’d just like to take a very visible stand on this. Four years ago, I really got upset with the whole thing. We settled our first case for a few hundred thousand dollars.
But that’s the last one we settled. From then on, we decided to take everything to the mat. I testified in front of the Senate on it; I’ve been very active on this issue. The House and Senate passed it, and Clinton vetoed it—which I was disappointed in—but they overrode his veto.
Why do you think Clinton vetoed it?
Well, you’d have to ask him. But the plaintiff ’s bar is really active in Democratic politics. I’ve never met a Democrat in Washington, D.C.
who hasn’t received major funding from the plaintiff ’s bar.
So, I think it was practical politics. He didn’t fight the veto override very much; in fact, he made, near as I could tell, zero phone
202
ED MCCRACKEN
calls on that. So it was easy to override the veto. There may have been a deal involved, for all I know.
In the Company of Giants Page 23
