"Team building" has become a buzzword in American business. The results are not overly impressive.

Ford Motor Co. began more than 10 years ago to build teams to design its new models. It now reports "serious problems," and the gap in development time between Ford and its Japanese competitors has harshly narrowed. General Motors' Saturn Division was going to replace the traditional assembly line with team work in its "factory of the future." But the plant has been steadily moving back toward the Detroit-style assembly line. Procter & Gamble launched a team-building campaign with great fanfare several years ago. Now P&G is moving back to individual accountability for developing and marketing new products.

One reason- perhaps the major one- for these near-failures is the all-but-universal belief among executives that there is just one kind of team. There actually are three- each different in its structure, in the behavior it demands from its members, in its strengths, in its vulnerabilities, its limitations, its requirements, but above all, in what it can do and should be used for.

The first kind of team is the baseball team. The surgical team that performs an open-heart operation and Henry Ford's assembly line are both "baseball teams." So is the team Detroit traditionally sets up to design a new car.

Fixed Positions

The players on the team: they do not play as a team. They hate fixed positions they never leave. The second baseman never runs to assist the pitcher; the anesthesiologist never comes to the aid of the surgical nurse. "Up at bat, you are totally alone," is an old baseball saying. In the traditional Detroit design team, marketing people rarely saw designers and were never consulted by them. Designers did their work and passed it on to the development engineers, who in turn did their work and passed it on to manufacturing, which in turn did its work and passed it on to marketing.

The second kind of team is the football team. The symphony orchestra and the hospital unit that rallies round a patient who goes into shock at 3 a.m. are "football teams," as are Japanese auto makers' design teams. The players on the football team or in the symphony orchestra, like those on the baseball team, have fixed positions. The oboe never comes to the aid of the violas, however badly they might flounder. But on these teams players play as a team. The Japanese auto makers' design teams, which Detroit and P&G rushed to imitate, are football-type teams. To use an engineering term, the designers, engineers, manufacturing people and marketing people work "in parallel." The traditional Detroit team worked "in series."

Third, there is the tennis doubles team- the kind Saturn management hoped would replace the traditional assembly line. It is also the sort of team that plays in a jazz combo, the team of senior executives who form the "president's office" in big companies, or the team that is most likely to produce a genuine innovation like the personal computer 15 years ago.

On the doubles team, players have a primary rather than a fixed position. They are supposed to "cover" their teammates, adjusting to their teammates' strengths and weaknesses and to the changing demands of the "game."

Business executives and the management literature have little good to say these days about the baseball-style team, whether in the office or on the factory floor. There is even a failure to recognize such teams as teams at all. But this kind of team has enormous strengths. Each member can be evaluated separately, can have clear and specific goals, can be held accountable, can be measured- as witness the statistics a true aficionado reels off about every major-leaguer in baseball history. Each member can be trained and developed to the fullest extent of the individual's strengths. And because the members do not have to adjust to anybody else on the team, every position can be staffed with a "star," no matter how temperamental, jealous or limelight-hogging each of them might be.

But the baseball team is inflexible. It works well when the game has been played many times and when the sequence of its actions is thoroughly understood by everyone. That is what made this kind of team right for Detroit in the past.

As recently as 20 years ago, to be fast and flexible in automotive design was the last thing Detroit needed or wanted. Traditional mass production required long runs with minimum changes. And since the resale value of the "good used car"- one less than three years old- was a key factor for the new-car buyer, it was a serious mistake to bring out a new design (which would depreciate the old car) more than every five years. Sales and market share took a dip on several occasions when Chrysler prematurely introduced a new, brilliant design.

The Japanese did not invent "flexible mass production": IBM was probably the first to use it, around 1960. But when the Japanese auto industry adopted it, it made possible the introduction of a new car model in parallel with a successful old one. And then the baseball team did indeed become the wrong team for Detroit, and for mass-production industry as a whole. The design process then had to be restructured as a football team.

The football team does have the flexibility Detroit now needs. But if has far more stringent requirements than the baseball team. It needs a "score"- whether it's the play the coach signals to the huddle on the field or the Mozart symphony everyone in the orchestra puts on his music stand. The specifications with which the Japanese begin their design of a new car model- or a new consumer-electronics product- are far more stringent and detailed than anything Detroit is used to in respect to style, technology, performance, weight, price and so on. And they are far more closely adhered to.

In the traditional "baseball" design team every position- engineering, manufacturing, marketing- does its job its own way. In the football team or the symphony orchestra, there is no such permissiveness. The word of the coach or the conductor is law. Players are beholden to this one boss alone for their order, their rewards, their appraisals, their promotions.

The individual engineer on the Japanese design team is a member of his company's engineering department. But he is on the design team because the team's leader has asked for him- not because the chief engineer sent him there. He can consult engineering and get advice. But his orders come from the design-team chief, who also appraises his performance. If there are stars on these teams, they are featured only if the score calls for a solo. Otherwise they subordinate themselves to the team.

Even more stringent are the requirements of the doubles team- the kind that GM's Saturn Division hoped to develop in its "flexible-manufacturing" plant, and that any such plant does indeed need. This team must be quite small, with five to seven members at most. The members have to be trained together and must work together for quite some time before they fully function as a team. There must be one clear goal for the entire team and yet considerable flexibility with respect to the individual member's work and performance. And in this kind of team only the team "performs"; individual members "contribute."

All three of these kinds of teams are true teams. But they are so different- in the behavior they require, in what they do best, and in what they cannot do at all- that they cannot be hybrids. One kind of team can play only one way. And it is very difficult to change from one kind of team to another.

Gradual change cannot work. There has to be a total break with the past, however traumatic it may be. This means that people cannot report to both their old boss and to the new coach, conductor or team leader. And their rewards, their compensation, their appraisals and their promotions must be totally dependent on their performance in their new roles on their new teams. But this is so unpopular that the temptation to compromise is always great.

Teams Are Tools

At Ford, for instance, the financial people have been left under the control of the financial staff and report to it rather than to the new design teams. GM's Saturn division has tried to maintain the authority of the traditional bosses- the first-line supervisors and the shop stewards- rather than hand decision-making power over to the work teams. This, however, is like playing baseball and a tennis doubles match with the same people, on the same field, and at the same time. It can only result in frustration and non-performance. And a similar confusion seems to have prevailed at P&G.

Teams, in other words, are tools. As such, each team design has its own uses, its own characteristics, its own requirements, its own limitations. Team work is neither "good" nor "desirable"- it is a fact. Wherever people work together or play together they do so as a team. Which team to use for what purpose is a crucial, difficult, and risky decision that is even harder to unmake. Managements have yet to learn how to make it.