Never has the level of performance expected of the small group of managers at the top of large organizations been higher. Today’s challenges demand not only great personal effort, but also first-class collective leadership. Never, therefore, has the reality at most companies, including yours, been more troubling. Very few so-called "teams at the top" really work. Even fewer are real teams.

Elsewhere in these organizations, as you well know, plenty of teams do work. Self-managing—that is, boss-less, not leaderless—teams are enabling companies of all sizes to increase productivity, quality, and bottom-line results. Cross-functional teams are helping them to remove internal structural barriers, break down functional "silos," and ratchet up innovation, speed of response, and service performance. Front-line sales teams are prompting them to combine product/service skills with account development sensitivity in ways that improve both customer loyalty and profit margins.

Why this huge difference in experience? It cannot be simple caution on your part. You have, certainly, seen enough large-scale team efforts fail to be immune to the faddish belief that teams provide a sure-fire panacea for all corporate ills. But you have seen many local—and quite valuable—successes as well. The explanation must lie elsewhere.

All or nothing

In our research for The Wisdom of Teams, as in our work with senior executives both then and since, we have often been struck by the many assumptions these leaders make about the relevance of team performance to their own immediate circumstances.¹ Although rarely explicit and almost never discussed, their beliefs about the purpose, membership, end products, role mix, leadership, and—most important—sustainability of teams at the top show great consistency. But it is a consistency of error. In fact, it is precisely such assumptions—plausible and cogent, but wrong—that disincline these groups of managers to seek team performance levels for themselves in the first place. Or, even if they do seek them, that so often bring their efforts to grief.

Misleading assumptions

Careers spent in traditional "command and control" environments can easily lead managers to assume, for example, that:

• the animating purpose or mission of any team at the top is—and should be—virtually indistinguishable from the mission of the company as a whole;
• all of a CEO’s "direct reports" must be members of the team;
• since what the team does is review, discuss, decide, inform, and delegate, the true end products of its work are guidelines and standards for others;
• what each member does as part of the team should be consistent with his or her organizational position and role; and
• formal team leadership, to be credible, must be exercised by the most senior member—that is, by the CEO.

However commonsensical at first glance, these points of view inevitably lead managers to operate, de facto, as a working group and not as a real team. They may say of themselves that they are—or are genuinely trying to be—a team. They may honestly believe that that is what they are doing. They may even be confident that they are doing it successfully. But they are not. Indeed, they cannot be—not if their actions and attitudes are shaped by any one or more of these assumptions.

A real team is more than just the sum of its parts

When a group of people has no separate and distinct purpose as a group, when its membership is defined not by relevant skill but by official position, when it generates no joint work product, and when its leadership is determined by external hierarchy instead of internal need—such a group is not a real team. This is not mere semantics. A team, by our definition, is a small number of people with complementary skills who are committed to a common purpose, performance goals, and approach for which they hold themselves mutually accountable. As a result, they deliver extra performance benefits. A real team is more than just the sum of its parts.

A working group is the sum of its parts. It has no shared work product because there is no genuinely shared work

By contrast, a working group makes its contribution through the aggregate, independent, and discrete contributions of its individual members as individuals. It is the sum of its parts. It has no shared work product because there is no genuinely shared work. Nor is there any shared purpose (save the overall health of the company) or goal or approach. And, of course, when these are absent, neither can there be any mutual accountability. The point here is not that real teams are good and working groups, bad. It is, rather, that they add different kinds—and levels—of value.

In fact, the top managers at many of the world’s most successful companies—General Electric, JP Morgan, Hewlett-Packard, and Emerson Electric, to cite just a few—operate, explicitly and intentionally, in working group mode. It is hard, demanding work. They share values, insights, information, and best practices; make and syndicate collective decisions; and reinforce a strong performance ethic among themselves and throughout their companies. What they add, as a group, helps each member do his or her own job better. What they support, as a group, are the efforts of each member to reach individual performance goals and deliver on individual accountabilities.

Now, as already noted, the faulty assumptions that get made about the opportunities for a real team at the top often lead managers, willy-nilly, into a working group mode of operation (no matter what they call it) on specific issues or projects. Rarely do they discuss adopting real team discipline instead. This matters because it sacrifices, out of hand, the potential levels of performance that only real teams can achieve. More troubling, it usually forecloses the possibility of experimenting with team discipline later on.

Managers often assume that, at the top of their organizations, they face a bi-modal choice: either they work as a team or they do not

This is because the final misleading assumption managers often make is that, at the very top of their organizations, they face a bi-modal choice: either they work as a team or they do not. Period. It’s a black-and-white decision: all or nothing. Our experience, however, suggests that this is not the case. On the contrary, thinking in "all or nothing" terms limits arbitrarily—and unnecessarily—the performance options available over time to senior management groups.

Keeping these options open is important. The best way we have found to do this is to pay careful attention to five guiding principles:

1. Define teams in terms of performance opportunities

Performance challenges—not the desire for teamwork—produce the kind of commitment essential to teams

The least likely way for a group to become a team is to make becoming a team its primary objective. This is why so many leadership groups that label themselves teams never achieve the desired levels of performance. Clear and compelling performance challenges requiring the collective effort and multiple skills of a small group are what produce the kind of commitment essential to teams—not the desire for teamwork, cooperation, or togetherness.

Seeing the trees

Consider, for example, the experience of the top management group at ForestCo, a leading forest products company. During the past several years, ForestCo, like others in its industry, has used team performance at the plant level to boost dramatically the productivity and quality of its mills. These results convinced the group at the top that the company would be a lot better off if they, too, were more of a team. In addition, since many of them—including the CEO—were nearing retirement, all believed that establishing a real team at the top was critical to an effective leadership transition. After all, the CEO, COO, and CFO already thought—and spoke—of themselves as a team. Why not build on success?

Long years of comfortable hierarchical interaction prompted all of the dozen executives in the group to assume that, to become a team, each of them had to be on it. But no matter how hard they tried, they could not identify any team performance challenge or collective work product that truly required the contributions of all twelve. Indeed, the more they focused on becoming a single team, the more frustrated they became. As the COO lamented, "This team thing is just not working the same way at the top as it does in the mills."

Some challenges are best accomplished through the formal roles and responsibilities of individual executives

This growing frustration, however, bore important fruit. First, the twelve came to realize that their purposes as a group—the concrete tasks they needed to accomplish—were not the same as the company’s overall mission. Second, they came to see that raising the level of organizational performance above the plant level counted far more than building a single team at the top. And third, they discovered that some challenges—improving the performance of particular business units, say, or spending capital more efficiently—were best accomplished through the formal roles and responsibilities of individual executives. Other challenges, however, required a different approach.

Reshaping the corporate mission or integrating different marketing approaches: goals of this sort required the joint, real-time contributions of several people. No single individual had the requisite knowledge, skills, and judgment. For these performance challenges—and these only—real value would accrue from a team approach similar to one that was succeeding in the mills.

More important, however, the group finally saw both that more than one kind of team situation existed, and that none of them required the contributions of all 12 members. In other words, different skill sets and, thus, different combinations of people, were needed to address each separate performance challenge.

Freed at last from the self-defeating effort to create a single, all-purpose team, the top managers of ForestCo finally understood that their core, shared purpose as a group was to provide the collective leadership necessary to enhance the company’s overall operating performance. Moreover, they saw that because much of what they had to do to achieve this purpose was best done through individual assignment and responsibility, only limited parts of it genuinely required teams. On balance, then, they would primarily be a working group—but a much more effective working group than before. What they had lost was their team illusion. What they had gained was performance.

2. Assign team members on the basis of skills, not status

Not everyone has to be a member; the question is, who has the skills needed to meet each specific performance challenge

Once a top management group focuses not on becoming a team but on attaining team-like levels of extra performance, they can easily overcome the counterproductive assumption that everyone has to be a member. The relevant question becomes, instead, who—both among and beyond the group—has the skills needed to meet each specific performance challenge. Establishing team membership in this way, rather than by position or reporting relationship, boosts total performance capacity, deals effectively with realistic time constraints, and provides important new opportunities for personal growth and development.

Banking on skills

The experience of the "New Ops Team" at ElectroFin—which provides a wide array of financial services to customers in more than three dozen countries around the world—exemplifies these benefits. During the past 20 years, the systems and operations approach followed by each branch had diverged significantly from that followed by the others. In fact, so driven by disparate local issues and near-term performance needs had these choices been that improving the company’s performance clearly required the real-time contributions of people with a variety of skills.

ElectroFin’s top group instinctively recognized that they needed a team to tackle the challenge of standardization. But they also recognized that they themselves were not going to tackle it as a team. Some members lacked the relevant skills; others, the necessary time. Equally important, the group did not include some crucial players. As finally assembled, the New Ops Team included the line managers responsible for Asia and South America, the CFO, the head of ElectroFin’s analyst group, an outside advisor to the CEO, and the head of systems. Of these, only three were part of the top management group, and the team leader was not the most senior person.

The formal positions of team members are much less critical than the skills they bring to the challenge

The formal positions of the team members were much less critical than the skills they brought to the challenge. Not all those skills, however, were in place at the beginning: some had to be developed as things went along. From the start, for example, the team enjoyed a solid mix of line management, finance, systems, analytical, and interpersonal skills, as well as multiple geographic perspectives. But their understanding of company operations and systems was limited. The team had to build this—and it did. Within two months, the team had figured out how best to enhance "old" systems while focusing scarce resources on high-potential new systems. Within six, they had shaped a broad vision for standardization that key people throughout the company embraced.

But these were not the only benefits. As the top executives not assigned to the New Ops Team learned to trust it to value their opinions, they also learned that they did not have to spend valuable personal time in meetings simply to "protect their interests." This provided a powerful lesson: if not everyone has to be on a particular team, the top group has the potential—and the available time—to form several such teams. In other words, their full team performance capacity was much greater than they had originally thought possible.

3. Choose leaders on the basis of attitudes and beliefs, not seniority or position

Nearly every team at the top starts with the premise that its leader has to be the CEO or the equivalent SOPA—an old Navy term meaning "Senior Officer Present Afloat." Any other assumption seems crazy. Who else but the SOPA should be the leader?

Who, indeed? In practice, the most salient qualification for team leadership is attitude, not formal position or seniority. This is because effective team leaders must believe deeply in the purpose of the team and in the potential of each person on it. They must also believe that they, personally, do not have to make all the important decisions, nor be the only ones in control. There will be times, of course, when leaders must actively take control to keep a team on course, as well as times when they can rely on others to move the herd roughly west.

This may all sound like what managers and executives usually believe. It is not. As a rule, the higher up the organizational ladder you go, the harder it is to find people who honestly hold these beliefs. Many good managers are not good team leaders because they lack the flexibility to step back so that other members can grow, as well as to step in when necessary—often to do real work "below" their normal organizational level in order to complement the efforts of other team members.

The insidious thing about command and control management is that it works

Seniority is irrelevant here. In fact, the most senior or functionally expert person on a team will often find it especially difficult to exercise the patience necessary for this kind of leadership. Why? Because, as our colleague Marvin Bower has aptly noted, "The insidious thing about command and control management is that it works!" After years of practice, hierarchically-

oriented managers find it perfectly natural to make all key decisions, delegate all real work responsibilities, and monitor all individual performances when cast as a team leader. And so do the people to whom they delegate and whose performances they monitor.

Please note that we are not saying that seniority or functional expertise disqualifies people as potential team leaders. On the contrary, anyone with the right attitude can be effective. But we are saying that the assumption that the most senior or functionally expert person must lead the team is seriously flawed.

4. Enforce regular team discipline on all would-be teams

Once top management groups, like those at ForestCo and ElectroFin, really break free of the single team trap, they feel an almost palpable sense of relief. Their self-confidence grows as they move collectively to determine which of the performance challenges they face can best be met through individual assignments, and which through teams. The bad news is that most such groups do not come naturally or easily to practise the discipline without which team performance is impossible—at any organizational level.

A major performance challenge facing General Electric Motors, for example, was the development of new products and processes that would sustain both growth and competitive advantage across five separate business units. The high-level team (called "MPE") established to strengthen the ways in which the work of each of these business units incorporated improvements in materials, manufacturing processes, and design engineering faced a significant performance challenge of its own. Because the MPE team’s results were mostly reflected in the "the numbers of others"—that is, in the operating results of the individual business units—how could the team hold themselves mutually accountable? Effective team discipline required them to do so. But must that rule apply even in a situation like this?

Without a doubt, yes. In theory, of course, this question might pose a dilemma for the casual observer; in practice, it was not a real problem for the team. They knew they had to measure their performance against what each of the business units achieved by sharing in MPE’s efforts, insights, and influence. The available numbers were not pure, nor were they consistently separable in terms of whose idea produced which result. Hence, to disciples of strict individual accountability, it might appear that no one was responsible. In the event, however, both the business units and MPE do hold themselves fully accountable. Neatness does not matter. A sense of ownership for results does.

5. Recognize the value, as well as the limits, of the "working group" approach

Most clusters of top managers operate most of the time as a working group rather than as a team

Most clusters of top managers operate most of the time as a working group rather than as a team. But this does not mean that they cannot—or should not—explore a combination of both approaches suited to the particular mix of skill requirements, time demands, and performance challenges they face.

Hewlett-Packard, for example, is often thought to be a "team-based" organization. This is only partially true. Certainly, many of HP’s smaller businesses are run by teams at the top. Others, however, are run by effective working groups. Still others focus on enabling teams down the line. Moreover, at the top—at least prior to Lew Platt’s recent designation as CEO—it was clear that the corporate leadership functioned primarily as a working group.

Yet from this mix of approaches, HP has produced real team-like performance at multiple organizational levels and across different businesses. Because of their strong and balanced performance ethic, leadership groups at HP gravitate naturally toward the kind of working approach for which each particular challenge calls.

The general pattern, here and elsewhere, is for a leadership group to begin as a working group and then to identify the areas in which some of its members may need to move into a real team mode—some of which will be ongoing, others one-time events. Thus, it is critical to consciously address—and re-address—the choice between team and working group. Finding the optimum mix of leadership configurations is not a one-time event.

New CEOs always face this kind of challenge. Many, however, fall into the trap of believing they must decide quickly on a single approach that will continue over time. Whatever its appeal in terms of consistency, this belief usually leads to a working group simply labeling themselves a team and implicitly sacrificing the performance upside, over time, of a more flexible approach. Some leadership groups do this instinctively (particularly in companies that have a strong performance ethic already in place), but most are constrained by the comfort and seeming efficiency of the single working group approach.

The notion we would like to leave with you is simply put: the options for boosting team-like performance at the top level of your corporation are more various and flexible than you may appreciate. There is no value in artificially limiting the range of choice. Keeping that range open, however, means being willing to change comfortable behavior patterns, test and try out new approaches, reach out for real team achievements, and enforce real team discipline. Merely applying team labels and rhetoric can be counterproductive. The goal, after all, is not to paper over lackluster performance with fine language, but to improve—by being more sensitive to the range of possible approaches—performance at the top.

We would also like to leave with you a word of caution as you explore various team possibilities at the top: for any real team option, it is imperative to apply the "team basics" discipline that leads to team levels of extra performance.

Regards,