ARMONK, N.Y. -- Thomas Watson, the man who created International Business Machines Corp., constantly reminded his employees to think. John Akers, IBM's current chairman, is a little more demanding. He wants the company to be perfect.

"To have a business of $70 billion and 375,000 people committed to near perfect execution by mid-decade is fairly unusual," says Akers. "I don't think any other company has a stated goal to be as close to perfect as possible."

Under the rubric of "market-driven quality," Akers is asking his people to shoot for a 10-fold reduction in product and service defects by 1992 -- and then for another 10-fold decrease by 1994. This isn't just rhetoric. IBM just won a Malcolm Baldrige Quality Award for its AS 400 minicomputer line, and the company is investing money, time and training to reinforce the message.

That's fine. Quality should rank right up there with apple pie and motherhood as a basic American value. But this "corpocratic" quest for perfection obscures the more interesting challenges that face the world's largest and most profitable computer company: What does "market driven" mean in an industry where customers fear they're not getting an adequate return on their information technology investment? What does "perfection" mean when you're struggling to link a dozen disparate technologies into a coherent system that somebody wants to buy?

IBM's real challenge -- as Akers well knows -- isn't becoming perfect; it's to swiftly and cost-effectively take advantage of all the opportunities it has before it.

In fact, in an interview in his office, Akers freely and bluntly acknowledges this. He does what three years ago would have been unthinkable: He compares IBM unfavorably to one of its competitors, Sun Microsystems Inc., the hot Silicon Valley workstation company.

"Let's use Sun as an example," says Akers. "It brings performance to the market at a very fast clip and has been able to do that better than anyone else. That's why they're doing so well: People like performance improvements at what's seen as a reasonable price. So if IBM wants to be successful vis-a`-vis Sun, we have to do what they're doing at least as well ... to even have a chance to be even with them."

Of course, IBM isn't interested in being even with Sun. "We ought to be able to do better than Sun," says Akers. "I think there are opportunities for us that are unique technically and relationships-wise."

For one, he asserts, IBM should be better at networking the workstations to what is probably an existing installed base of IBM machines. The problem, though, is that networking, linking and integrating disparate hardware, software and telecommunications technologies is far harder than most people expected. Five or six years ago, most organizations had systems dedicated to solving specific problems. Today, distributed computing means that different computers are trying to talk with each other about a wide variety of unanticipated problems.

"We're trying to do things that have never been done before," Akers acknowledges. "We're doing invention on the fly with customers. It's visible, it's awkward, and it doesn't always work the first time. It used to be that everything was clickety-click by the time it was installed in the customer's office. Those days are gone."

So much for perfection. The reality is that because IBM has more customers and more technologies than anyone else, it has a greater challenge -- and opportunity -- in linking them all. The issue isn't creating "defect-free" systems for customers. It's making sure that IBM's systems engineers and market representatives learn from each other so that they don't have to reinvent the wheel at every new customer site.

These have been very difficult times for IBM. Its historic 15 percent annual growth rate has decayed. "We are at a point where we have learned how to have a highly successful business at an 8 percent to 10 percent growth rate," Akers says.

IBM's chairman describes a company whose priority is now process -- getting things done and getting them done right. Executing the fundamentals. Getting the customer as intimately involved in the development cycle as in the sales cycle.

IBM has decided that it is now going to lead the industry by exerting influence instead of control. Indeed, the company is going to lead its customers by following them. By no stretch of the imagination is this an exciting strategy that will lift the industry to new heights. But then, nobody has ever claimed that IBM was exciting -- only that it cares very much about being very successful by doing well for its customers.

Michael Schrage is a columnist for the Los Angeles Times.