The first management conference we know of was called in 1882 by the German Post Office.
The topic—and only chief executive officers were invited—was how not to be afraid of the telephone.
Nobody showed up.
The invitees were insulted.
The idea that they should use telephones was unthinkable.
The telephone was for underlings.
I was reminded of this story in the early 1960s when I was working with IBM to make computers accessible to executives.
Back then, some of us already understood this wasn’t just another gimmick—that it was something that would profoundly, even fundamentally, change the way we organize industry and do business.
Information would become the main productivity factor.
IBM’s Tom Watson Jr. came up with a brilliant idea.
We would have a meeting for CEOs and talk about “computer literacy.”
In fact, it was on that occasion that we coined the term.
However, I immediately tried to talk Watson out of this brilliant idea.
I told him the story of the German Post Office.
“You’re at that same stage,” I said.
“Nobody’s going to show up.
It’s too weird for them.”
Twenty-five or thirty years ago that kind of meeting was indeed not possible.
Thirty years from now such meetings will not be necessary because today’s CEOs will have been succeeded by their grandchildren’s generation.
Anyone who knows this generation, who has children between the ages of ten and thirteen, won’t be surprised by what I learned while visiting my youngest daughter and her children.
My grandson, a very nice boy, is at thirteen no longer into computers.
Kid stuff, he says, except for parallel processing.
However, he keeps his hand in, you might say.
He said to me, “Grandfather, Daddy’s computer is no longer state-of-the-art.”
The joke?
My son-in-law happens to be a physics professor who runs one of the bigger nonmilitary computer installations in existence.
But my grandson was right.
When this generation grows up and moves into our jobs, we won’t have to talk about computer literacy.
Just as we no longer have to talk about how not to be afraid of the telephone.
My grandson’s five-year-old sister can dial the world over.
And does.
Of course, my grandson is not the only one who is computer literate.
In this country, his whole generation is.
This is one area where we are way ahead.
Computer literacy is in its infancy in Japan and is still unheard of in Europe.
My wife has nieces and nephews in Germany, and their children know nothing, though as it happens the parents are both scientists.
The parents work with computers, but the idea that a nine- or ten-year-old should be familiar with computers is a new one.
Even though we are way ahead on this one, we’re still not quite where we should be.
We must be computer literate in self-defense.
Ten or fifteen years hence, not only will we take computer literacy for granted, we will have become information literate as well.
And that, very few people are.
Most CEOs still believe that it’s the chief information officer’s job to identify the information the CEO requires.
This is, of course, a fallacy.
The information officer is a toolmaker; the CEO is the tool user.
Let me illustrate.
Recently, I got around to repairing the overstuffed sofa in our guest room, which I should have repaired three years ago.
At the hardware store, I asked the owner which upholstery hammer would be best for the task.
I didn’t ask him whether I should repair the sofa.
That decision was my job.
I merely asked him for the right tool.
And he gave it to me.
When I had my fax machine installed a few years ago, I had the telephone installer come in to put in a new line, and he was very helpful.
He looked around and said to me, “You may have chosen the wrong place to put it.
I think it would be awkward over there.
Why not here?
And I can easily give you a line here, too.”
But he didn’t tell me to whom to send faxes or what to say in them.
This is my job.
His job was to give me the tool.
CEOs must accept that if the computer is a tool, it is the tool user’s job to decide how to use it.
They must learn to assume “information responsibility.”
Which means asking, What information do I need to do my job?
From whom?
In what form?
When?
As well as, What information do I owe?
To whom?
In what form?
When?
Unfortunately, most people still expect the chief information officer or some other technologist to answer those questions.
This won’t do.
I teach at a small graduate school, Claremont.
About twelve years ago, we wanted to have a computer sciences building built.
When it came to raising the money, we beat out Stanford and Yale.
We got an enormous amount of corporate money because we said in our proposal, “This school will not be in business in ten years.
If we do a halfway decent job, it will have become superfluous.
In ten years there will be computer engineers, there will be people who design software.
But computer science as a separate discipline in a management school will be gone.”
We got all of this money simply because we said that in ten or fifteen years we will not have to spend a lot of time creating toolmakers.
We’ll need them, of course.
But users will know how to use the tools; toolmaking will remain important but purely technical.
The first step is to take information responsibility: What information do I need to do my job?
In what form?
The information specialists can then say, Look, you can’t get it in this form; you can get it in that form.
The answer is relatively unimportant and technical; it is the basic questions that count: When do I need it?
From whom?
What information do I owe?
We are rebuilding organizations around information.
When CEOs talk of eliminating management levels, they begin to use information as a structural element.
Many times, we quickly discover that most management levels manage nothing.
Instead, they merely amplify the faint signals emanating from the top and bottom of the corporate infrastructure.
I imagine that most CEOs have heard the first law of information theory: Every relay doubles the noise and cuts the message in half.
The same holds true for most management levels, which neither manage people nor make decisions.
They serve only as relays.
When we build in information as a structural element, we don’t need such levels.
This, however, creates enormous problems.
For instance, where will we look for opportunities for promotion?
Few businesses will have more than two or three layers.
Will CEOs be able to accept that more layers are a sign of poor organization?
You violate a basic rule.
Very few people get into a management job before they are twenty-six or twenty-seven.
You have to be in a job five years not only to learn it, but to prove yourself.
And yet you have to be young enough to be considered for senior management jobs before you reach fifty.
That gives you three levels of management.
If you look at the General Motors of today versus that of yesterday, you’ll see it’s slimmed down a bit.
The company used to have twenty-nine layers, which meant that nobody could really be considered for a top management job before age two hundred and eleven.
This, obviously, is part of GM’s problem.
Where will the promotion opportunities come from?
How will we reward and recognize people?
Also, how will we prepare people for jobs that are not functionally narrow?
These are big challenges.
And we don’t know the answers.
We know only that we will pay much more.
Money will become far more important because in the past thirty years, we have substituted title for money in many cases.
We have had rapid promotions in title with very little increase in salary.
That’s over.
Far more important is the change in the process.
When we learn to use information as a tool, we are learning what to use it for, what we need, in what form, when, from whom, and so on.
The moment you examine these questions, you realize that the information you need—the really important information—you cannot truly get from your information system.
Your information system gives you inside information.
But there are no results inside a business.
Many, many years ago, I coined the term profit center.
I am thoroughly ashamed of it now, because inside a business there are no profit centers, just cost centers.
Profit comes only from the outside.
When a customer returns with a repeat order and his check doesn’t bounce, you have a profit center.
Until then you have only cost centers.
When we talk about the global economy, I hope nobody believes it can be managed.
Drucker on Asia — A Dialogue Between Peter Drucker and Isao Nakauchi, The Asian Crisis and more
It can’t.
There is no information on it.
But if you are in the hospital field, you can know hospitals.
If you were to parachute into some strange place and make your way to the lights in the valley, you would be able to identify the correct building as the hospital.
Even in Inner Mongolia, I can assure you that you will know you are in a hospital.
No mistaking it.
No mistaking a school.
No mistaking a restaurant.
People who tell me they operate in a world economy are those whose shares I sell immediately.
One cannot operate where one can’t know anything.
We simply have no information at all.
You can’t know everything.
You can know only what you know.
This is why the enterprise of tomorrow is going to be very narrow in focus.
Diversification can work only if you have the information.
And you don’t have it if the competition can come in from Osaka without any warning.
We have so little information on the outside, on markets, on customers.
Nothing—as many people have learned the hard way—is changing faster than distribution channels.
And if you wait until you get the report, it’s way too late.
Technology itself provides the perfect illustration.
It’s no longer the nineteenth century, or even the twentieth, when you could assume that technologies pertaining to and affecting your industry came out of your industry.
Time has overthrown the idea behind the great research labs, of which IBM’s is probably the last.
There will be no other like it.
Most of what really had an impact on the computer and the computer industry did not come out of IBM’s lab.
IBM couldn’t use most of the brilliant things that came out of its lab in its own business.
And that holds true of Bell Labs and of the pharmaceutical labs.
Technology is no longer a series of parallel streams that in the nineteenth century underlay our academic disciplines.
Instead, it is a crisscross.
It is chaotic and therefore has to come from outside.
And about this outside we know nothing.
Here you are, a pharmaceutical manufacturer.
You are going to be made obsolete by mechanical instruments or processes—say, by a pacemaker or a bypass.
You may have the world’s best lab, but the changes in your business will not come out of your lab.
Your lab is focused on the inside, as is your information system.
In effect, we are trying to fly on one wing, the wing of inside information.
The big challenge will not be to get more or better inside information, but to add outside information.
To cite an example: Most people believe this country has a balance-of-trade deficit.
Most people are wrong but don’t know it.
The early-eighteenth-century balance-of-trade concept was developed when a bright cookie had an idea.
But this brilliant idea was limited to merchandise trade, and that is the only figure reported.
Although this country today has a merchandise trade deficit, it also happens to have an enormous service trade surplus.
The official amount is two-thirds of the merchandise trade deficit.
The actual figure is probably much bigger, because the real service trade figures are simply not there.
For instance, we have some five hundred thousand foreign students in this country; the minimum they bring in is $15,000 each.
Therefore, we have about $7 to $8 billion in foreign-exchange income from these non-American students.
It is simply not reported.
I believe we may actually have a total merchandise and service surplus, though only a very small one.
The figures aren’t there, only the concept.
Our biggest challenge will continue to be obtaining this outside information so that we can make good decisions.
This relates to the domestic market, the way customers are changing and the way distribution systems are changing.
It also relates to technology and competition, because both can put you out of business.
When the pacemaker hit, the market for the most profitable cardiac medicine disappeared in five years.
It was only after the market had disappeared that people stopped to ask what had happened.
We need outside information, and we will have to learn.
But it is complicated because most businesses have two information systems.
One is organized around the data stream; the other, far older one, around the accounting system.
The accounting system, though, is a five-hundred year-old information system that is in terrible shape.
The changes we will see in information technologies over the next twenty years are nothing compared to the changes we will see in accounting.
We have already begun to observe changes in manufacturing cost accounting, whose roots go back to the 1920s and which is totally obsolete.
But that is only for manufacturing, not service.
Manufacturing today accounts for 23 percent of the GNP and perhaps 16 percent of employment.
Thus, for the vast majority of businesses we have no accounting that’s worth anything.
The problem with service-business accounting is simple.
Whether it’s a department store or a university or a hospital, we know how much money comes in and we know how much money goes out.
We even know where it goes.
But we cannot relate expenditures to results.
Nobody knows how.
At present, these two systems are separate.
They will not be separate for our grandchildren’s generation.
Today’s CEOs still depend on the accounting model.
I don’t know of a single business that bases its decisions on the data processing stream.
Everyone bases their decisions on the accounting model, even though most of us have learned how easy this model is to manipulate.
We know where we can trust it and where we can’t.
We have all fallen through thin ice often enough not to walk on every part of it.
We have learned to depend on cash flow because any second-year accounting student can manipulate a P&L.
By the next generation, when the data processing stream is more familiar, we will be able to merge the two, or at least make them compatible, which they aren’t today.
We teach them separately in the schools.
We have an accounting major and a computer science major, and the two don’t talk to each other.
Both departments are, as a rule, headed by people who know little about information.
The person who heads up your accounting system knows government requirements.
The head of data processing came up in hardware.
Neither knows information.
We will have to bring the two together, but we don’t yet know how.
My own guess is that ten years from now a medium-size company, let alone a large one, will have two different people filling two positions that one person handles today.
It will have a chief financial officer, who will not manage anybody.
This person will manage the corporation’s money, the biggest part of which will be managing foreign exchange—tough now and soon to become much worse.
And the company will have a chief information officer, who will manage its information systems.
The company will need both.
They look at the world and the business quite differently.
Neither of these people, though, is focused on the wealth-producing capacity of the business or on tomorrow’s decisions.
They are both focused on what happened.
Not on what might happen or could be made to happen.
We have an enormous job ahead of us.
We need to make ourselves and our businesses information literate.
That job will begin with the individual.
We must become tool users.
We need to look on information as a tool for a specific job, which few people do.
(Most who approach information in this way are not in business; those who have gone the furthest are in the military.)
Our second big job is to use our data processing capacity to understand what is happening on the outside.
The available data are usually in poor form and of dubious reliability.
The only companies that have any information of this kind are the big Japanese trading houses.
They have information about the outside (what they have about Brazil is amazing), but it took them forty years and a great deal of money to acquire it.
For most CEOs, the most important information is not about customers but about noncustomers.
FYI only: Convincing the Swing Vote: How to Lure ‘Non-customers’
This is the group in which change will occur.
Let’s take a look at that endangered species, the American department store.
Nobody knew more about their customers than did these stores.
Until the 1980s they held on to their customers.
But they had no information about noncustomers.
They had 28 percent of the retail market, the largest single share.
However, this meant that 72 percent didn’t shop at the department stores.
And the department stores had no information on these people.
And they couldn’t have cared less.
Thus, they were unaware that new customers—especially the affluent—do not shop in department stores.
Nobody knows why.
They just don’t.
By the end of the 1980s, though, these noncustomers had become the dominant influence group.
They began to determine how all of us shop.
But nobody in the department store world knew this because they had been looking at their own customers.
After a time, they knew more and more about less and less.
We must begin to organize information from the outside, where the true profit centers exist.
We will have to build a system that gives this information to those who make the decisions.
And we will have to bring together the accounting and data processing systems, which is something few people are interested in doing.
We’re at the beginning.
If you are not computer literate, do not expect anyone in your organization to respect you.
The young people in your businesses take it for granted; they expect at least literacy from the boss.
My five-year-old granddaughter would have no respect for me if I told her, “You know, I’m afraid of the telephone.”
She wouldn’t even believe me.
Times change, and we must change with them.
We are moving from minimal computer literacy—knowing little more than the ABCs and multiplication tables of computing—to the point where we can really do something with computers.
That will be an exciting and challenging prospect for the years ahead.
We are just entering that stream.
It’s going to be a fast one.
1998
