What does it mean to say that an intelligence team has “done well”? Teams usually have a mission with a specific objective: perhaps submitting an analytic report to a client, or obtaining certain items from a denied area, or completing an investigation of an information systems intrusion. If concrete objectives such as these were not accomplished, that clearly would signal poor team performance. But does doing well equate to simply accomplishing the team’s objective and nothing more?
Of course not. There always are multiple factors that bear on a team’s overall effectiveness. What did the clients of the team’s work actually think of the product? Was it helpful to them in meeting their own objectives, or was it something to be filed and forgotten? How about other stakeholders—people who did not commission the work but were affected by what the team produced? What were the consequences for them, and what were their reactions? What happened to the team itself? Did working together build the team’s capabilities, strengthen it as a performing unit? Or did the team burn itself up in the process of getting the job done? And how about the individual team members? Did they learn some things along the way, or was being on the team merely a frustrating exercise that contributed nothing to their own development as intelligence professionals?
As is evident from these questions, any robust assessment of team effectiveness must attend simultaneously to several different types of outcomes. And, no way around it, assessing how well a team performed always involves value judgments, whether or not they are explicitly stated. This chapter first identifies the values that my colleagues and I use to assess the effectiveness of the teams that participate in our research, including intelligence community teams we have studied. Then I describe a simple checklist that anyone can use to monitor the work processes of a team in real time. The higher a team’s standing on the checklist processes, the more likely it is that the team’s eventual product, service, or decision will be first rate. The chapter ends with a discussion of what can be done to help a team develop the best possible work processes, thereby increasing the chances that it will develop into a highly effective performing unit.
My colleagues and I use three dimensions to assess the teams we study. Because these dimensions reflect our own values about what “doing well” as a team actually means, we do not count as fully effective any team that fails on any of the three. To illustrate the three dimensions, I provide below examples of intelligence analysis teams we have studied that scored especially high, and especially low, on each of them.1
1. The productive output of the team (that is, its product, service, or decision) meets or exceeds the standards of quantity, quality, and timeliness of the team’s clients—that is, of the people who receive, review, and/or use the output.
One group we studied prepared reports that, after review, usually reached the desk of a senior policy official. Word frequently came back from the official that he found the group’s analyses quite helpful. By contrast, another team generated a regular monthly report on certain international transactions—a report that, it turned out, was received and filed by the official’s assistant, and never even seen by the team’s presumptive customer.
It is the clients’ standards and assessments that count in determining team effectiveness. Not those of the team itself, except in those rare cases when the team is the client of its own work. Not those of outside researchers or evaluators, except when they are engaged to perform an assessment by those who do have legitimacy as reviewers. And not even those of the team’s manager, who only rarely is the person who actually uses a team’s output.
Good teams meet their clients’ expectations. But what if a team’s client is ill-informed, misguided about what actually is needed, or even corrupt—for example, someone who is intent on using the team’s product to justify something inappropriate? In such circumstances, the best teams expand the scope of their efforts and take initiatives that can range from reasoned argument all the way to political action to ensure that their products will be used appropriately. Actively managing one’s clients can be risky for a team but it sometimes is necessary.
It is one thing to determine whether or not a team’s product has satisfied its client (one can simply ask) but quite another to assess its objective quality. A strategy some clients use to circumvent this difficulty is to focus less on the product itself and more on the methodology the team used to generate it. Good procedures, then, serve as a surrogate for outcome quality.2 According to a study of intelligence-relevant analytic practices in the private sector, this is what many analysis-intensive businesses do. The firms surveyed reported that “both tradecraft and accuracy were important, but the common refrain was that if an analyst did it ‘our way, using our method,’ then more often than not the results would be accurate.”3
2. The social processes the team uses in carrying out the work enhance members’ capability to work together interdependently in the future.
One team’s task was to refine certain quantitative indicators of an activity of special interest to its client. Over time, members of this team developed deep knowledge of one another’s special strengths and weaknesses and became so highly skilled in coordinating their activities that members anticipated each other’s next moves and initiated follow-on activities even as their colleagues were completing previous steps. In another group, by contrast, the longer members worked together the more dissension and conflict they experienced. Eventually group work became so distressing that members could agree about only one thing—namely, that they should ask their manager to disband the group, which he subsequently did.
Effective groups operate in ways that build shared commitment, collective skills, and task-appropriate coordination strategies—not mutual antagonisms and a trail of failures from which little is learned. They become adept at detecting and correcting errors before serious damage is done and at noticing and exploiting emerging opportunities. And they periodically review how they have been operating, milking their experiences for whatever learning can be gained from them. An effective team is a more capable performing unit when it has finished a piece of work than it was when the work began.
3. The group experience, on balance, contributes positively to the learning and professional development of individual team members.
One team we studied needed to draw upon state-of-the-art knowledge about certain aspects of information technology to accomplish its tasks. Members reported that working with other members was akin to attending a continuing seminar on cutting-edge developments in computer science. By contrast, members of another group spent the majority of their time monitoring systems for signs of possible trouble—essentially staring at screens that rarely showed anything amiss. Members of this group reported not just that they were bored, but also that carrying out the work actually had atrophied their professional skills.
Groups can be wonderful sites for learning—for expanding one’s knowledge, acquiring new skills, and exploring perspectives that differ from one’s own.4 Teamwork also can engender feelings of belonging, providing members a secure sense of their place in the social world. But groups gone bad stress their members, alienate them from one another, and undermine members’ confidence in their own abilities. While not denying the inevitability of rough spots in the life of any group, I nonetheless do not count as effective any team for which the impact of the group experience on members’ learning and professional development is substantially more negative than positive.
These three criteria can be used to assess the effectiveness of any intelligence team, regardless of task or setting. The relative weight of the three criteria, however, varies across times and circumstances. If, for example, a temporary task force were formed to perform a single intelligence task of extraordinary importance, then the second and third dimensions would be of little import. The opposite would be true for a group that was formed mainly to help members gain experience, learn some things, and become competent as a performing unit—as might be the case, for example, in a training course. Truly great groups continuously manage trade-offs among the three criteria as their circumstances change, sometimes focusing relentlessly on the specific piece of work to be accomplished but other times finding occasions for reflection and for activities specifically intended to strengthen individual and team capabilities.
Here is the problem. You cannot tell how well a team has done on the three criteria of effectiveness just discussed until after the work is finished—and, sometimes, not until much later than that. As will be seen in Chapter 9, it is always a good idea to take some time after a piece of work has been completed to reflect on how members might have worked together better, and about organizational supports that might have been helpful to the team. But how about in real time, when the game is afoot? What can be done then?
It turns out that ongoing monitoring of just three aspects of the group process can provide an excellent basis for predicting how well a team eventually will do. Process monitoring also can identify areas where changes in how a team currently is operating might increase its chances for success. The three key team processes are these:
1. The amount of effort members are expending in carrying out their collective work.
2. The task-appropriateness of the team’s performance strategies, the choices the team makes about how it will carry out the work.
3. The level of knowledge and skill the team is applying to the work.
Any team that expends sufficient effort in its work, deploys a performance strategy that is well aligned with task requirements, and brings ample talent to bear on its task is quite likely to achieve a high standing on the three criteria of effectiveness discussed above. A team probably will fall short, however, if it operates in ways that compromise its standing on those same three processes—that is, if members apply insufficient effort, inappropriate strategies, and/or inadequate talent to their work.
Try it out. Reflect on the red and blue teams in the PLG simulation described in Chapter 1. How did those teams differ in the level of effort expended, in the appropriateness of their performance strategies, and in their utilization of members’ talents? Now try it out on a team of your own. Think of a pair of teams on which you have served, one that performed superbly and another that was a disaster. Compare the standing of those two teams on the three performance processes. I would be quite surprised if they did not differ on at least a couple of them, if not all three.
What specifically should you focus on in assessing how well a team is managing the three key group processes? And what can you do when you identify a problem or an unexploited opportunity in how a team is managing its effort, strategy, or talent?
The first step is to focus specifically on what is working well and what is not. For each of the three performance processes, research has identified both a characteristic “process loss” that keeps a team from making good use of its full set of capabilities, and an opportunity to build positive synergy that exceeds what would result from merely adding up members’ individual contributions.5 That is, the team may operate in ways that depress its overall effort, the appropriateness of its strategy, and/or the utilization of its members’ talents. Alternatively, team processes can enhance collective effort, generate uniquely appropriate strategies, and/or actively develop members’ knowledge and skills—in effect, creating new internal resources, team capabilities that did not exist before. These process losses and synergistic gains are summarized in Figure 3-1, which is presented as a checklist you can use in making a real-time assessment of how a team is doing.
There always are some overhead costs to be paid when groups perform tasks. Merely coordinating members’ activities, for example, takes some time and energy away from productive work, resulting in a lower level of actual productivity than would be attained if members used their resources completely efficiently. The most pernicious effortrelated process loss, however, is social loafing—the tendency we all have to slack off a bit when working in groups, to exert less effort on team tasks than we do when performing work that is ours alone. Social loafing occurs because individuals usually can “hide” to some extent in a team. Moreover, each team member may feel less personally responsible for collective outcomes because there are multiple hands on the wheel.
FIGURE 3-1 Group Process Checklist
The process gain for effort develops when members become highly committed to their team, proud of it, and willing to work especially hard to make it one of the best. In a phrase, they have developed team spirit. When this happens, members may exhibit high, task-focused effort even if objective performance conditions are less than ideal. For example, members may develop such a strong can-do attitude that each new adversity is framed as yet another challenge to be surmounted.
A team’s strategy is the set of choices that members make about how to carry out the work. For example, an operations team might decide to divide itself into three subgroups, each of which would initially work on one subtask, with the launch of the full operation deferred until all the subtasks were completed. Or a team performing a task that requires a creative solution might choose to free associate about possibilities at its first meeting, reflect for a week about the ideas that came up, and then reconvene to draft the product. Or an analytic team might study the requirements of its task and then choose a particular structured technique that members think would best facilitate their work on that analytic problem. All of these are choices about task performance strategy.
If a piece of work is familiar to a team, or even if it merely seems familiar, a previously developed performance strategy is likely to kick in and guide team behavior. Reliance on established habitual routines is highly efficient because members do not have to actively deliberate anew about how to proceed with each piece of work. But such routines also invite significant process losses, especially when members are so focused on executing them that they fail to notice that the task or situation has changed.
This process loss apparently played a role in the crash of Air Florida Flight 90 into the 14th Street Bridge shortly after takeoff from Washington National Airport on a snowy January afternoon in 1982. The National Transportation Safety Board investigation concluded:
… the probable cause of the accident was the flight crew’s failure to use engine anti-ice during ground operation and take-off, and their decision to take off with snow/ice on the airfoil surfaces of the aircraft, and the captain’s failure to reject the takeoff during the early stage when his attention was called to anomalous engine instrument readings.6
The cockpit voice recorder for the period just after the engines were started shows how habitual routines may have contributed to the tragedy. The captain had called for the after-start checklist, a standard procedure intended to make sure an aircraft is set up properly for taxi. As is typical, the first officer read each checklist item and the captain responded after checking the appropriate indicator in the cockpit.
CAPTAIN: Generators
FIRST OFFICER: Pitot heat
CAPTAIN: On
FIRST OFFICER: Anti-ice
CAPTAIN: Off
FIRST OFFICER: Air-conditioning pressurization
CAPTAIN: Packs on flight
FIRST OFFICER: APU
CAPTAIN: Running
FIRST OFFICER: Start levers
CAPTAIN: Idle
FIRST OFFICER: Door warning lights
CAPTAIN: Out
The checklist is a routine, run every time engines are started. The standard response to the “Anti-ice?” query that I italicized is indeed “Off,” especially in the summer and for crews that typically operate in warm or dry climates—as was the case for this crew. The oft-repeated litany may have become so ingrained for these crew members that they did not even consider the possibility that their situation required a nonroutine response to a routine query.
The crew had a second chance to save the flight several minutes later, as the takeoff roll began. The first officer, who was the pilot actually flying the aircraft, noted that something seemed wrong (“God, look at that thing [the airspeed indicator], that don’t seem right, does it?”) but the captain did not respond. When he repeated his concern, the captain provided reassurance (“Yes it is, there’s eighty [knots]”). Even though the first officer was not convinced (“Naw, I don’t think that’s right”) he continued down the runway. The takeoff routine was still not broken, despite the availability of data indicating that it was not proceeding normally. Less than a minute later the aircraft crashed into the bridge. Although the consequences are rarely as dramatic as they were for the Air Florida flight, mindless reliance on habitual routines compromises the performance of many different types of task-performing teams.
By contrast, teams sometimes develop ways of interacting that result in truly original or insightful ways of proceeding with the work, creating synergistic process gains. For example, a group might find a way to exploit some resources that everyone else has overlooked, invent a way to get around a seemingly insurmountable performance obstacle, or come up with a novel way to generate ideas for solving a difficult problem. Developing innovative performance strategies involves two different activities. First, the team scans both its external environment and its internal resources to identify problems and opportunities. And then it actively considers a variety of ways it might circumvent the problems and exploit the opportunities, eventually choosing the one that seems best to team members. When group norms support such activities (see Chapter 7), a team can generate a genuinely innovative way of proceeding with the work—a performance strategy that did not exist before the group invented it.
One of the most common and pernicious process losses encountered by task-performing teams is the inappropriate “weighting” of members’ inputs. The credence given to a member’s ideas sometimes depends far too much on that person’s demographic attributes (such as gender, age, or ethnicity), position in the broader community (such as rank, role, or home organization), or behavioral style (such as talkativeness). When a team gives more weight to such factors than to what the person actually knows about the work, it wastes one of its most precious commodities—the talents and experiences of its members.
Assessing which members have the special expertise needed for a given part of the collective work is not easy because what members know, or know how to do, is not nearly as apparent to teammates as are each member’s surface attributes. In the absence of clear data about expertise, the human tendency is to turn to surrogates that are visible. But, as was evident in the PLG simulation described in Chapter 1, those surrogates invite use of social stereotypes and almost inevitably result in inefficient weighting of certain members’ contributions. Although good weighting processes help teams make better use of the expertise put in the group when it was composed, such processes are much easier to advocate than to execute when fast-paced work is being executed in a rapidly changing environment. In such circumstances, teams are especially likely to rely on surface attributes in weighting members’ contributions because constant change obscures information about members’ actual expertise.7
The process gain for knowledge and skill comes when team members develop a pattern of interaction that fosters learning from one another, thereby increasing the total pool of knowledge available for work on the team task. The practice of cross-training, often encouraged in self-managing work teams in industry, is an example of such behavior, as are more informal activities that involve the sharing of knowledge, expertise, and experience among members. Cross-functional and cross-organizational teams are especially good sites for generating synergistic process gains of this kind, since such teams always have a diversity of member knowledge, skills, and experiences. Yet, even in relatively homogeneous teams, members can learn from one another and, indeed, sometimes can generate entirely new understandings that expand a team’s overall capability.
Return briefly to the examples of the well- and poorly performing teams you considered earlier. How did those teams stand on the specific process losses and potential synergies shown in Figure 3-1? Do the specifics in that checklist give you any greater insight into what went wrong, and what went right, in the groups you were thinking about? And do they suggest any interventions that you might have made, as a team leader or member, that could have helped the poorly performing team do better?
You may have noticed that the key processes have everything to do with generating high-quality task processes and nothing to do with fostering interpersonal harmony. To see why improving team effectiveness depends more on task than on interpersonal processes, consider a team that is having performance problems. That team also is quite likely to exhibit interpersonal difficulties such as communications breakdowns, conflict among members, leadership struggles, and so on. The most natural thing in the world is to infer that the observed interpersonal troubles are causing the performance problems and, therefore, that a good way to improve team performance would be to fix them. As reasonable as this inference may seem, it is neither logical nor correct. Although serious interpersonal conflicts can of course undermine team performance on occasion, it does not follow that the best response would be to help members improve their interpersonal relationships.
Research on interpersonally oriented interventions such as process consultation, team building, and other group development activities shows that participants do find them engaging and that, when competently led, they positively affect members’ attitudes about their teams. But they do not reliably improve team performance. Moreover, experimental studies comparing teams that receive task-focused intervention with others that receive interpersonally focused interventions generally have found the former to significantly outperform the latter.8
It may even be that the causal arrow points in the opposite direction—that is, that performance drives interpersonal processes (or, at least, perceptions of those processes) rather than vice versa. In an experimental study, social psychologist Barry Staw gave teams false feedback about their performance and then asked members to provide “objective” descriptions of their group process. Members who were told their groups had done well reported more harmonious and better communications, among other differences, than did members who were told their groups had performed poorly.9
Despite these research findings, many senior managers continue to focus their attention and actions on how smoothly and harmoniously their organizations operate. Indeed, they sometimes explicitly recognize and reward those leaders whose units show improvements on organization-wide satisfaction surveys. If it is true that high workforce morale is more a consequence than a cause of excellent performance, they may have it exactly backward.
Lay observers, as well as more than a few leadership scholars, tend to view leaders as the dominant influence on team performance. Consider, for example, how we explain an athletic team that has winning season after winning season. “That John Wooden at UCLA,” we exclaim. “What a basketball coach he was!” Or reflect on a team that has had a few losing seasons: It is the coach who is fired. Overattribution to leaders of responsibility for team outcomes occurs for both favorable and unfavorable outcomes. It is so powerful that team members themselves succumb to it. And it is so pervasive, at least in Western cultures, that my colleague Ruth Wageman and I have given it a name: the leader attribution error.10
Our tendency to view leaders as the main influence on how well teams do is understandable because their actions are much more visible than structural or contextual factors that also may be strongly shaping team outcomes. And under some conditions, leaders’ actions really do spell the difference between success and failure—the leader is the cause and team performance is the effect. But here is the rub: research has shown that leader behavior makes the most constructive difference for teams that are reasonably well structured and supported in the first place. If a team is poorly composed, has an ambiguous or unimportant purpose, and operates in an organization that discourages rather than supports teamwork, there is no way that a leader’s hands-on interventions with that team can turn things around.
So rather than prescribe the “right” leadership styles for facilitating competent teamwork, this book identifies the structural and contextual conditions that, when in place, increase the chances that a team will get off on a good track—and, importantly, that members will actually be able to use the competent coaching and teaching that the best team leaders provide.
Thinking about enabling conditions prompts leadership strategies that differ greatly from those that derive from conventional cause-effect models. To illustrate, let me draw on an analogy I have used elsewhere—namely, the two different strategies a pilot can use in landing an aircraft. One strategy is to manage the system hands-on in real time. The pilot actively flies the airplane down, continuously adjusting heading, sink rate, and airspeed with the objective of arriving at the runway threshold just above stall speed, ready to flare the aircraft and touch down smoothly. The alternative strategy is to get the aircraft stabilized on approach while still far from the field, making small corrections as needed to heading, power, or aircraft configuration to keep the plane “in the groove.” It is well known among pilots that the safer strategy is the second one; indeed, the prudent action for a pilot who winds up in the first situation is to go around and try the approach again.
To be stabilized on approach is to have the basic conditions in place such that the natural course of events leads to the desired outcome—in this case, a good landing. The same way of thinking applies in many other domains of human endeavor. Consider, for example, constantly tinkering with a nation’s interest rates, money supply, and tax policies, versus establishing fundamentally sound economic conditions and letting the economy run itself. Or micro-managing the upbringing of a child, versus creating a good family context that promotes healthy but mostly autonomous development. In all of these instances, the better strategy is to devote the first and greater portion of one’s energies to establishing conditions that lead naturally to the desired outcomes and the lesser portion to online process management (this proposition is explored in detail in Chapter 10).
These considerations bear directly on how teams in intelligence organizations are designed and led. There is nothing that anyone— manager, leader, or member—can do to make an intelligence team be great, to ensure its effectiveness. But what can be done is to create and sustain six specific conditions that smooth a team’s path toward its objectives. And then, with those conditions in place, leaders can draw on their own special strengths and styles to help their teams use well the full array of organizational resources and supports that are available to them.
The chapters in Part II of this book explore the six conditions that, our research has found, increase the chances that a team will achieve a high standing on the criteria of effectiveness discussed earlier in this chapter. The six enabling conditions are: (1) creating a real team (rather than a team in name only), (2) specifying a compelling direction or purpose for the team, (3) putting the right number of the right people on the team, (4) specifying clear norms of conduct for team behavior, (5) providing a supportive organizational context, and (6) making competent team-focused coaching available to the team.
How much of a difference do these conditions actually make in how well teams perform? A number of empirical studies have addressed this question. In a study of field service teams at Xerox, for example, organizational psychologist Ruth Wageman found that the way teams were designed and organizationally supported accounted for significantly more variation both in the level of team self-management and in team performance effectiveness than did team leaders’ hands-on coaching.11 In our previously mentioned study of intelligence community analytic teams, Michael O’Connor and I found that the enabling conditions controlled 74 percent of the variation in an independent, multi-attribute measure of overall team effectiveness, an extremely strong effect. Finally, a cross-national study of senior leadership teams documented that the enabling conditions strongly predicted team effectiveness ratings made by panels of outside assessors.12 Clearly, the enabling conditions make a considerable difference for a variety of team types—a finding that is entirely consistent with the differential performance of the red teams (for which the conditions were solidly present) and the blue teams (for which they were not) in the PLG simulations described in Chapter 1.13
The process losses described earlier in this chapter always put an intelligence team at risk of wasting at least some portion of its resources, including members’ knowledge, experience, and readiness to work together. Moreover, a team that encounters serious process losses also is more likely to be tripped up by one of the well-established vulnerabilities of group work discussed by intelligence community veteran Richards Heuer. In his book Small Group Processes for Intelligence Analysis, Heuer both describes what can lead a group astray (for example, counterproductive group members, meeting hijacking, premature consensus, groupthink, polarization of views, social loafing, and problems in computer-mediated communication) and provides some sound advice about how teams can avoid these dysfunctions.14
The presence of the enabling conditions demonstrably lessens the chances that a team will fall victim to process losses that can compromise its effectiveness. Just as importantly, those conditions open up the possibility of synergistic process gains, enabling a team to turn in a performance that outstrips anything that could be obtained by merely stitching together the separate contributions of individual team members. The chapters that follow explore in detail how this happens and lay out some concrete steps that team managers, leaders, and members can take to increase the likelihood that it will.