He then declares, in his rich Viennese accent, that the kinds of questions he has been told he might be asked are "an abomination." He sets himself the task of offering refinements—and, not surprisingly, has excellent ideas about how to find more precise questions. Mr. Drucker also offers a glimpse of his dry humor. When someone uses the term "cash cow" in passing, he says: "You know I invented that term—for which I will roast in hell forever." When he begins the session— which is moderated by Context editor-in-chief Paul Carroll but which also involves two dozen senior executives from an array of industries—he makes a friendly suggestion. He says that if his answers wander, he should be treated as a mule. "The first thing you do with a mule," he says, "is hit it over the head with a two-by-four if you want its attention."

His suggestion is, of course, unnecessary, and Mr. Drucker proceeds to offer a broad series of observations about why innovators so often fail, about how the business world is devolving into a series of permanent, part-time relationships, and about how a business can differentiate itself in a fast-changing world. Along the way, he gets in a few digs at some prominent executives—and at himself.

CONNECTING WITH THE CUSTOMER

CONTEXT: Lots of companies talk about getting closer to the customer. You've made the point, though, that most companies don't know much about their customers. Would you elaborate?

PETER DRUCKER: Very few of us, myself included, know the customer. One reason is that all of us believe, and must believe, that the product and the service we produce is important. But 99.9% of your customers couldn't care less about your product or service. You are not that important in their universe. And that's almost impossible to accept.

The second reason is that you amass an enormous amount of information about the people who buy from you. Yet, for almost all companies, at least 70% of the people or organizations that should be your customers are not. So, if you want to understand the customer, those people who aren't your customers are the key.

Plus it ain't so easy to define a market any more. Yes, all the questionnaires, all the statistics are helpful. But you can't depend on the statistical aggregates we now call information. I have for years been preaching that nothing beats seeing executives going out twice a year for two weeks to take the place of a salesman or saleswoman on the front line. Then executives will get some glimmer of the configuration that is the value system of the customer. Without that understanding, you'll know a good deal about your product, a good deal about your service, and not a blessed thing about your customers.

CONTEXT: Once you start to know something about your customers, how do you differentiate yourself from your competitors?

DRUCKER: The customer never buys what you think you sell. And you don't know it. That's why it's so difficult to differentiate yourself. Because you have to differentiate yourself in the customer's terms, and not in your terms.

There are very few companies in the world that can turn out superior products ahead of what the competition offers. And you can't get business any more by being cheaper. Not for very long. Quality, service, yes, you can still differentiate there. But fundamentally you have to differentiate yourself by structuring yourself within the customer's business. You have to understand it.

A few organizations I know have been quite successful in this area. You may remember that maybe 15 years ago, the world's largest manufacturer of heavy equipment, Caterpillar, was down, down, down, down. From having an unprecedented percentage of the world market—65%—they went down to 22%. Today, they're back to 65% or 70%. They totally redefined the business. They didn't just restructure.

They defined the business not as making equipment but as keeping equipment running. Because that is what the customer pays for. If you were building a six-lane highway from one end of Malaysia to the other, not being able to use that huge beast, the Caterpillar, for a couple of hours would cost you more than the equipment did.

Walt Disney was very good at positioning itself in the customer's universe. Is Disney still good at it? I doubt it. The customer has changed, and Disney has not. I think that [CEO Michael] Eisner is the world's most overpaid executive. And that's a proud boast for him, because most executives are overpaid.

I think one has to work constantly on understanding: What use does the customer make of our product and service? What is his reality? Assume that there ain't no irrational customers. There are only lazy manufacturers.

BOB SHAW [senior vice president, R.R. Donnelley]: What is the role of technology as people think about their customers?

DRUCKER: If you were to ask what the two main challenges are that we face, technology is not going to help you with either.

One is to increase the dismal productivity of the new labor force—knowledge workers and service workers. In blue-collar work, the question is how the job is done. But in knowledge work, the question is what should be done. Managements haven't asked that question yet. For this challenge, technology isn't helping very much.

Look at hospital nurses, who are probably the best-trained group in the American working population. They're fabulous. Yet we don't let them do what they've been trained for, which is patient care. We use them to fill out forms.

The other challenge is to figure out what's going on outside the company. Inside the company, there are only costs. The greatest mistake I made in a long life, and I made lots of them, was to invent the phrase "profit center." There ain't no profit centers inside a business; there are only cost centers. The only profit center is a customer whose check hasn't bounced.

But we know nothing about the outside. The biggest change that technology could bring would not be a faster computer. It would be concepts for getting information about the outside. These concepts are very, very slow in coming.

THE FUTURE OF PLANNING

CONTEXT: Many of the executives I have talked to recently seem to take a definitive rather than an opportunistic position on whether it's smarter to be a first mover or a fast follower. What's your viewpoint?

DRUCKER: When you look at the fate of the ones who have a truly new concept, they are very, very rarely the most successful competitors.

I'll go back to the early days of the computer. I was the advisor to the old Tom Watson [of IBM] on the conversion from the calculator to the computer in the early 1940s. I'm that old. Nobody—nobody—anticipated that businesses would have the slightest interest in the computer. In the Depression, I used to know a great many businessmen who spent sleepless nights wondering how they could get the money to pay their people. I don't remember anyone who, when he got the money, worried more than a second how to spend it. [So there seemed to be no reason to think anyone would buy a computer.]

Univac pioneered the computer for scientific and military uses. But then business showed interest. IBM saw that it owned the business market. It swallowed its pride, copied the Univac, and took leadership away from Univac. That's very typical.

The pioneer is convinced that it knows what market a new thing is designed for. But this rarely is the market that subsequently picks up the product.

The second reason is that the first version of anything that's truly new is rarely adequate. A lot of things have been engineered into it for which there is no use, and a lot of the things that are needed nobody foresaw. That's true of many things. I always stress to my clients that they have to be ready to change. You have to be prepared to make your innovation obsolete. But many companies aren't prepared to do that.

Examples: IBM has never innovated anything. But when Apple came out with the PC, IBM, doing something unusual for a big old company, immediately saw a huge market. They did something about it. Eighteen months later, IBM had the PC market, and not Apple.

You've heard of Tagamet. What you don't perhaps know is that it was not the first, but the second, ulcer drug based on the same scientific principles. The first one, brought on the market by a major German company, had side effects. It took care of the ulcer, but for a week you couldn't drink any liquor and lost your sense of taste. Tagamet just manipulated the formulation and took the market for itself. Later still, a truly creative imitator took over. And so it goes.

CONTEXT: Haven't we learned anything about how to innovate better?

DRUCKER: Let me see if you want to know the typical stages in a true innovation. In the first stage, the innovator is unwilling to accept that the market he designed the product for is not the right market. Are you interested in the classic case? Battles in the 19th century were the bloodiest ever to that time, beginning with the Crimea in 1855 and the American Civil War. But early anesthetics like chloroform couldn't be used on the battlefield, so the race was on for new anesthetics. The first one was cocaine, which was supposed to be nonaddictive; it's actually highly addictive. Then for 30 years, governments spent huge sums to develop a nonaddictive anesthetic. In 1906, a German chemist synthesized novocaine. All of us have had it at the dentist. It has never been used in major surgery. But for 20 years, until he died, that German chemist, who got unimaginably rich on the patent, traveled the world preaching to all the dental students, forbidding them to use his noble discovery for such a mundane purpose as filling teeth. That's the extreme. But I've seen lots of other cases that weren't that far away from it.

The second stage, the takeoff stage, is the one where we have learned something. After 40 years of preaching, we have convinced everybody that cash flow is more important than profit. (Everybody except Wall Street, and Wall Street will never learn that lesson. There are no slower learners than securities analysts. I can testify to that; I was one. So they and the companies they promote still run into credit crunches, which kill about four of every 10 new businesses and cripple most of the rest.)

The third stage of any successful innovation is when a business outgrows its management, its structure. We are in a health care crisis precisely because health care has been so successful in this century. We are in an educational crisis because education has been so successful. Between 1950 and 1980, the number of people in colleges increased 10-fold. The colleges are still running the same way, and it just no longer works.

Fourth and finally, there is succession. The hardest thing to do is succeed the founder.

Those four stages represent the danger points for new businesses, new services, new ideas. One can prepare for them, one can avoid them, but very few successful innovators or entrepreneurs have the mindset to prepare and avoid, so they run into the same problems.

MITCH HILL [chief financial officer, Imagineering division of Walt Disney]: Our 2,500-person workforce is about one-third consultants hired for a specific project responsibility. We also use many casual and part-time employees, making full-time, card-carrying Imagineers rarer than they were five to 10 years ago. Over time, I see nothing that will reverse this trend, which I think is common throughout corporate America. What's the endgame of this trend?

DRUCKER: For maybe 100-plus years, the trend was inexorably toward a society of large corporations. An employee society. In the 1970s, that trend turned. What we have increasingly now are alliances, networks, partnerships, consultant relationships, temp relations.

Members of an executive management program I run decided the future is as a society of networks. I think they had good reasons. We will probably maintain and even strengthen the core company. Otherwise, especially as we go global, communication becomes too difficult. But within these limitations, we are clearly moving toward a flexible structure.

There is a fellow in my executive management program, about 55, a former vice president of a very big company, a distinguished metallurgist who is now on his own and has six permanent clients, each of them keeping him on a retainer, including his former employer. I said to him, "George, why did you leave the XYZ Company?" He said: "In that company, I had major metallurgical challenges three times a year. The rest of the time, I wrote memos. Now, with six clients, each having three or four metallurgical problems a year, I am fully busy doing what I love to do, which is advanced metallurgy. I am not a consultant. When there is a metallurgical problem, I am in charge and am a member of their management for the six weeks it may take to lick that problem."

I am not saying that this man is typical. But he is also no longer atypical.

So we are moving into a society where, instead of command-and-control, you have partnership. But don't kid yourself. In many cases, partnerships are permanent part-time relationships. I have worked with the same clients more or less for 30 years. Some of these clients I see every three years. Next week, I will spend the whole week with an old client who first became my client before World War II. They do not see me as an outsider, and I don't see them as temporary. They are permanent clients, but not full-time clients.

JAY TUTHILL [chief executive, Tuthill Industries]: In the manufacturing world, is there an optimal facility size?

DRUCKER: The answer to this question is, Don't be dogmatic. If you are going into China, you have to be centralized. This is true of anybody who has tried to go into China. Even GE. GE is very proud of its 16 business units, each of which has a very big international business. In China, however, there aren't the 16 units; there is only GE. That is simply because the Chinese will not accept anything else. The Chinese will not work with an underling. And unless you are the boss, you are an underling. Even the head of a decentralized business is an underling. Plus the fact that if you go into China decentralized, you have to bribe so many more people.

On the other hand, don't go into India as a joint venture, except on a very narrow basis. Because your Indian partners are very narrow.

If you have to give service to a Toyota all over the world, you can't be small. No way. In other services, you want a business that operates in a narrow niche, fills that niche, and is highly focused. You don't want it to be big.

Let's take the mutual funds business, which has been one of the fastest-growing businesses these last 15 years. There are giants like Fidelity, and there are niche players. Both types do well. The in-between players have not done well—of course, this has been a very good time for everybody. Either you are so big, like Fidelity, that your advertising can blanket the country, and so big that no broker/dealer can do without you, or you are a niche player. In between, there is nothing that differentiates you.

On the other hand, venture capital firms, which have been highly successful, are all extremely specialized. The biggest, I think, has 80 senior people. Because in venture capital the premium is on being very close to your customer, being part of your customer's management, really.

So size itself isn't the issue. Look at the nature of your business.

THE FUTURE OF THE CORPORATION

BOB EVANSON [senior vice president, McGraw-Hill]: What are your thoughts on the best organization for global growth when the corporation has many small units that operate both domestically and internationally? Does matrix management really work?

DRUCKER: The answer is: No, matrix management does not work, but do you have a choice? I hate matrix management. There is an old proverb of Roman law that says, The slave who has three masters is a free man. Matrix management means that nobody is accountable for anything. I happen to believe in accountability.

But what choice do you have? Here you are, let's say, in Germany. Call the company Unilever. You are in 37 businesses. You are Germany's leading food merchant, from fish to margarine to soap to what have you. Each of these has to be run as a separate business. And yet each has also to be run as a German business.

The U.S. is about the only place where decisions don't go up very high. But the moment you go to Mexico, the moment you go to China, the moment you go to Germany, let alone the moment you go to France (God help you), at that moment you have to have a very prestigious national presence. And yet, the business is local. That means there is constant friction, constant complaint.

JUDI NORTH [who at the time was president, consumer services, at Bell South Communications, and who currently is chief executive at VSI Enterprises]: Morale in my company is poor as we enter a competitive world and are downsizing. Is that problem a given as the role of the corporation changes, or can leadership deal with that?

DRUCKER: The question is, Can we do anything about the morale of our company after we have spent 10 years destroying it?

Look, I've lived a very long life, and I've seen a lot of stupidity. But very little of it beats the stupidity with which we have been downsizing. Very bluntly.

Don't be surprised that morale is very low. The contempt for top management is dreadful. And the present generation of management is not going to regain the trust of their people. It is our greatest disadvantage in this country today.

One reason is that it has become so crystal clear to middle-management people that the financial people who control the company are contemptuous of honest work, truly contemptuous. And far too many people in our top management have shown themselves to be sadists and take pleasure in inflicting pain. You can't expect people to trust you if you get a $12 million bonus for laying off 12,000 people. So please, Judi, please don't believe in your ability to restore morale. Just try to build a little trust in what management is saying. It will take years.

THE FUTURE OF THE INDIVIDUAL

BOB KRAMER [chief information officer of Profit Recovery Group]: As technology more often mediates human interaction, I fear that we are causing an erosion of the social atmosphere. What are your thoughts on the social impacts of technology?"

DRUCKER: Communities are being built very fast, only they are not being built within the company. In 1950, I thought that the modern corporation could replace the traditional communities that we have largely dissolved. The Japanese, as you perhaps know, followed my advice and built lifetime employment on it. I think they now regret trying this.

But the Internet builds new communities. There is a physicist I know, very distinguished, the head of research at a major American university. His community is 50, maybe 100 physicists— in Chicago, in Japan, in Italy, in Germany, at MIT. It is a community. If any of these people needs a job, he finds them a job.

For the older people, who came in at a time when you worked the rest of your life at GE or wherever, networking this way is very hard. For the young ones, it's very easy. They like it at GE, perhaps. They are well-paid there. The work is nice. But they all have their resume in the bottom drawer. They all have the names of former GE people on their Rolodex. And they call up these people twice a month to say hello, so that if and when they need their help, they have it. This is why we have so little unemployment, by the way.

We are more and more coming to the point of view that the most valuable employees are beholden to their craft, to their knowledge area, maybe to their immediate colleagues. Their employer is a term almost of no consequence.

But the need for community is as great as ever, if not greater. So the Internet is building new communities.

CONTEXT: To end on a possibly positive note, I'll refer back to what you said about dismal morale and to what you just said about the growth of noncorporate communities. Might there not be a huge advantage for companies that offer their people a more meaningful and lengthy relationship?

DRUCKER: Can you afford to do that? The Japanese are trying to, and I'm not sure they can succeed. Change is so rapid—not just the technology, but the economy and markets. And the people with knowledge are so conscious of their salability, their employability. Can you really create more than a contractual relationship? Can you restore that belief in the lifelong mutual commitment?

People prefer stability. But most know that preferring stability is like preferring good health: You can't guarantee it. All you can do in your company—and you'd better do it—is find out the strengths of the people who work for you and place them where their strengths can produce results, so that there is satisfaction. At present, we are focused on the weaknesses of people. When I talk to my clients about Joe, they say Joe cannot do this and Joe cannot do that. Then I get annoyed and say, if Joe cannot do anything, why is he still your employee? If you want to build a relationship with Joe, find out what Joe's good at and make sure he gets to do that kind of work. That way you build the desire for achievement, the respect for management. And that's what people want to have. We just make it awfully hard for them.

CONTEXT: Well, if not completely upbeat, that was thoughtful. Peter, thanks very much.