The Truth about Teambuilding>
What We Really Know about Teamwork


Asking the Right Teamwork Questions

Richard Hackman and former Harvard Univ. colleague Nancy Katz are leading researchers in the psychology of small groups. Though written for other scientists, their chapter on “Group Behavior and Performance” in the an edition of the Handbook of Social Psychology points to some practical advice for nonacademics like you and me. But mostly it should raise questions about whether you are asking the right questions when you think of team best practices.

Take a section titled, “How Well Do Groups Perform?” Hackman and Katz say the study results are mixed. “In fact, it is the wrong question to ask,” they say. “It is wrong because it confounds three separate and quite distinct issues:

  • “When should groups be used, and when should they not?
  • “How does group performance compare to that of individuals?
  • “What differentiates groups that realize their full potential from those that do not?”

As suggested by #1, not all groups should be teams. “A great deal of organizational work is performed by sets of people that are called teams but that actually are coacting groups” of individual contributors, they write. Many teambuilding efforts are a waste of time in such groups, as I’ve already stated on this site. Even when the group should be a true team given its membership and tasks, other factors can interfere. “The social system context within which a task is performed also can strongly mitigate against teamwork—for example, when organizational policies, practices, or culture make it impossible to properly design and support teams…” Hackman and Katz write.

The scientific answer to Question 2 is mixed as well: “competent individuals, either working alone or (in) nominal groups, can outperform interacting groups that are poorly structured and supported,” they say. “And a great group can generate synergistic outcomes that exceed what would be produced even by extraordinarily competent individuals…” This relates to the “Baby and Bath Water Syndrome” I talk about. In many cases where genuine team development would solve many of a group’s problems, leaders and/or members resist it because of previous bad experiences with teams that were “poorly structured and supported.” Hence they prolong their suffering by throwing the baby out with the bath water.

Question 3 leads to another challenge of teambuilding. “A standard claim of cultural anthropologists is that any person is in some ways like all other people, like some other people, and like no other person,” Hackman and Katz write. “The same is true for groups…” Some research findings will apply to every team, and some only to teams similar to those in a given study. Scientists refer to this as “generalizability,” the degree to which findings from one study might apply to the general population from which study subjects were drawn. The trick for team leaders and consultants is figuring out which is which, or, how to “translate” general findings to specific teams, to use the authors’ word.

For example, take what they call “hidden profile” tasks in which team members have information about the task other members don’t know they have. Hackman and Katz say, “Groups performing hidden profile tasks (for example, making a decision about who to hire or where to locate a business) rely so heavily on shared information that they rarely come up with the best answer unless individual members are somehow prompted to share with the group the information that they uniquely hold…” Large, diverse, high-turnover, and virtual teams are likely more vulnerable to the problem. For example, “the more that a dispersed group relies on electronic technologies for communication, the more challenging it will be for members to get a good ‘take’ on who knows what,” they say. Teams also may give too much weight to a person’s input based “more on that person’s demographic attributes (e.g., gender, age, or ethnicity), position (e.g., rank, role, or office), or behavioral style (e.g., talkativeness or verbal dominance) than on the person’s actual expertise,” the authors lament.

What, then, if we are helping a largish team of 15 people, somewhat diverse, with a moderate turnover rate, whose members sit in different buildings but meet in person fortnightly? How do the studies on the hidden profile problem apply to that team? Far too many trainers go for the simple answers without thinking about these subtleties, assuming results in one study will work for every team. I worry over the advice I give about teamwork for that very reason.

In this case, Hackman and Katz (citing various studies I’ll omit) suggest these tools which I think most teams would benefit from, in part because they address other issues also:

  • “clear and challenging group goals coupled with incentives for achieving them”
  • “a division of cognitive labor that publicly identifies members’ areas of special expertise”
  • “collective accountability for the process by which the team generates its output”
  • “an organizational culture that emphasizes shared collective interests rather than individual distinctiveness and”
  • “group norms that place a higher value on collaboration and critical thinking than on achieving consensus.”

The authors aren’t saying consensus is bad, just that the other items should not be sacrificed to it. If that line makes you think of “groupthink,” hold on to your hat. Hackman and Katz state there is no clear evidence in the literature for groupthink, or for the effectiveness of brainstorming. Often this situation is due to findings that apply to specific types of teams but aren’t generalizable.

If you lead a team, Hackman and Katz have other suggestions for the hidden profile problem. Leader actions “found to be helpful are those that elicit and legitimize dissenting views; those that ask members to reflect on the group and its work, thereby prompting task representations that encourage them to draw upon one another’s knowledge; and those that invite members to explicitly plan the performance strategy the group will use in carrying out its work” (citations omitted).

Despite the debate among researchers about the details, there is solid agreement on many team best practices. Yet most teams don’t do them, instead opting for the “quick fix” of standard teambuilding. If you’re thinking of teambuilding, maybe you should ask some more questions.

Source: Hackman, R., and N. Katz (2010), “Group Behavior and Performance.” In Fiske, S., D. Gilbert, and G. Lindzey (eds.), Handbook of Social Psychology (5th ed.), Wiley: New York. (I wish to thank Dr. Bradley Staats of the Kenan-Flagler Business School, Univ. of North Carolina at Chapel Hill, for providing me with a copy.)

The 10 Ingredients of Better Teamwork

Recipes don’t always work out. Ask some of my dinner dates. But the closer you follow them, the more likely you are to get the results pictured in the cookbook. Wouldn’t it be nice if there were a recipe for cooking a high-performing, happy team?

I think there is. Like any recipe, it starts with the ingredients. In 2001, three respected researchers led by Michelle Marks of Florida International Univ. proposed such a list. In 2008, another research team put it to the test. (This points to a problem with what passes for “teambuilding”: Consultants and managers repeat as fact teamwork models that were just educated guesses when published, like “Forming, Storming,” etc.) From an initial list of 825 candidate studies, Jeffrey LePine of the Univ. of Florida and other researchers culled 157 with more than 1,500 correlations measuring factors in Marks’ model in a way they could be accurately combined. The team then ran a “meta-analysis,” meaning they analyzed the data as if the numbers had come from one big study.

The first conclusion of their article in the journal Personnel Psychology will sound a little odd: There is such a thing as “team processes.”

“Well, duh,” you say.

Imagine, though, that in 10 studies by different scientists, one-third found that stuffed animals helped team performance, one-third found those toys hurt performance, and one-third found no link. We could agree that stuffed animals exist, but we would know they have no predictable impact on team performance. Team processes are not tangible things that can be touched, however. If the researchers found no behavior patterns that allowed them to predict the performance of a team, the conclusion would be that teams operate in randomly different ways. The term “team process” would be meaningless for scientists and managers.

Fortunately, the meta-analysis confirmed there is an über-factor Marks called “team process” that impacts both measurable team performance and people’s satisfaction with the team experience. Altogether, LePine’s team concluded, a one-step change in team process translated to a one-third change in team performance and half-step change in member satisfaction. The larger the team, and the more team members had to coordinate to complete their work, the stronger the correlations were.

As Marks’ team had suggested, the data broke into three broad categories:

  • “Transition processes describe actions that teams execute between performance episodes. They have a dual focus as teams seek to reflect upon and interpret previous team accomplishments, as well as prepare for future actions.”
  • “Action processes reflect… activities that occur as the team works toward the accomplishment of its goals and objectives.”
  • “Interpersonal processes… reflect those team activities that are focused on the management of interpersonal relationships.”

These further divided into 10 subfactors that impact team results in varying ways. Using these as ingredients in your teambuilding as I describe below, you can bake a better team.

Transition Processes

Mission Analysis

“This process involves the identification and evaluation of team tasks, challenges, environmental conditions, and resources available for performing the team’s work,” the article says. Take the time as a team to create a mission statement or project goal, scope out your work, identify barriers to success, and list out needed resources.

Goal Specification

Identifying and prioritizing goals increase the odds of accomplishing the work. Based on other research, I recommend coming up with two to a maximum of five per year.

Strategy Formulation

“This involves developing courses of actions and contingency plans,” the authors say, “as well as making adjustments to plans in light of changes or expected changes in the team’s environment.” List out the steps needed to accomplish the goals you identified and to manage risks.

Action Processes

Monitoring Progress toward Goals

The article says, “This process involves members paying attention to, interpreting, and communicating information necessary for the team to gauge its progress toward its goals.” Check and report on how the team is doing at accomplishing its goals at least weekly.

Systems Monitoring

“This involves activities such as tracking team resources (e.g., money) and factors in the team environment (e.g., inventories) to ensure that the team has what it needs to accomplish its goals and objectives.” Until experience proves the team wrong, believe it when it tells you the amount of time, people, or materials it will take to accomplish what you want done.

Team Monitoring and Backup Behavior

A sign of a high-performing team, to quote the article, is “members going out of their way to assist other members in the performance of their tasks.” Encourage this by publicly praising both the helpers and the people brave enough to ask for help.


Without “synchronizing or aligning the activities of the team members with respect to their sequence and timing,” as the authors say, simply listing the tasks will not get them done. Identify the order they must be done in, and put a method in place to ensure people know when others are waiting on them.

Interpersonal Processes

Conflict Management

Pay attention to “the manner in which team members proactively and reactively deal with conflict. Effective conflict management includes showing mutual respect, willingness to compromise, and developing norms that promote cooperation and harmony,” like team rules.

Motivating and Confidence Building

These “activities that develop and maintain members’ motivation and confidence with regard to the team accomplishing its goals and objectives” are as important as the more objective activities.

Affect Management

In line with the previous statement, this “represents those activities that foster emotional balance, togetherness, and effective coping with stressful demands and frustration.” Treat emotional reactions as a fact in your decision-making, because people are not strictly rational actors in any part of life.

Again, following a recipe does not guarantee success. But it does increase the odds compared to not following it. If you want your team members to achieve better results and be happier doing it, focus your teambuilding activities on these fact-based ingredients instead of the junk food of events, assessments, and simulations.

Source: LePine, J., et al. (2008), “A Meta-Analysis of Teamwork Processes: Tests of a Multidimensional Model and Relationships with Team Effectiveness Criteria,” Personnel Psychology 61:273.

Biologist Gives Evolutionary Advice for Better Teamwork

One way to confirm whether one idea reflects reality better than another is to see if scientists in different fields, working independently, come up with results supporting that idea. If researchers with no reason to agree look at very different aspects of life and come up with the same conclusions, you are closer to a fundamental truth. Long before Columbus “discovered” the New World, educated people knew Earth was round because of evidence from what today we call the fields of astronomy, physics, logistics, and others.

Striking, then, was an interview I heard with an evolutionary biologist on National Public Radio. “Evolutionary Biology is a subfield in biology that is concerned with the gradual change in the traits of living organisms over generations, especially the emergence of new species,” according to Biology Online. David S. Wilson believes the concepts of evolution apply not only at the genetic level, but at the level of communities. Discussing his work to apply his knowledge of evolution to community change, he provided a list of conditions that encourage helping behaviors in groups of people working on community projects. I was stunned to hear advice that could have come straight from any evidence-based book on effective teamwork. Wilson is a professor of biology and anthropology at Binghamton Univ., part of the State University of New York.

Wilson said he has long studied why cooperation and selflessness, called “prosociality” by scientists, would have evolved. Given the concept of “survival of the fittest,” selfishness seems to make more sense. (There turn out to be good reasons to be selfless from the genetic standpoint, among them building a support network or ensuring that related genes survive through the survival of kin.) Wilson’s work in the city of Binghamton bore this out. He measured prosociality in individuals there and compared it to where they lived and their social networks. He found that “if you’re a highly prosocial teenager, you’re likely to be surrounded with social support from family, neighborhood, school, religion, (and) extracurricular activities,” Wilson said in the interview.

Prosocial behavior is a good thing to have in business teams, obviously. You may have seen here the term “organizational citizenship behaviors” or OCBs, which are prosocial actions in a business setting. It perked up my ears, then, to hear host John Ydstie ask what enhances prosociality. Here is Wilson’s list:

  1. “Number one, a strong sense of group identity, and a strong sense of what the group is about. If you don’t think of yourself as a group, and if you don’t know what the purpose of the group is, then it’s unlikely to function well as a group.”
  2. “Two, proportional costs and benefits. It cannot be the case that some people do all the work, and other people get the benefits. That’s not sustaining over the long term.”
  3. “Three, consensus decision-making. People hate being told what to do, but they’ll work hard for a decision that they agree upon.”
  4. “Four, monitoring. Most people want to cooperate but there’s always a temptation to slack a little bit… So unless you can monitor good behavior, forget about it.”
  5. “Next, graduated sanctions. If somebody does misbehave, you don’t bring the hammer down. You remind them in a nice and friendly fashion, and… escalate in those rare cases when necessary.”
  6. “Next, a fast, fair conflict resolution… in a manner that’s regarded as fair by all parties.”
  7. “Seven, autonomy—for a group to do these things, they have to have the authority to manage their own affairs.”
  8. “Finally, in a large society consisting of many groups, those groups have to be put together using those same principles.”

By this point in the hypertext, these have to sound familiar to you. Despite the eye-rolling the mention of mission statements brings sometimes, they are proven to improve performance when created by the team (not handed down to it) and actively used in decision-making. Wilson’s #2 and #4 link with my recommendations to use action items and formal project management (even for non-project teams) and to let teams self-assign tasks. Per his #3, building buy-in as part of decision-making is far more effective for creating full support on important issues than making the decision and then trying to build buy-in. His “graduated sanctions” and “fast, fair conflict resolution” match my calls for the team to create a self-enforcement procedure for team discipline and for the use of formal conflict-resolution techniques like those in my Full Scale agile™ (FuSca™) method for Agile-at-scale. Empowerment, which I harp on regularly, equals autonomy. And a “Company Culture Change” service I offered through TeamTrainers tried to establish across the enterprise team structures supporting all these concepts. FuSca devotes a major section to enterprise-wide agility. A good team can fail in a bad environment. I have seen high-performing teams get killed when their creators moved on.

Scientific crossover has popped up in many interesting ways in my consulting work. The first book-length project I did in my earlier journalism career was on the science behind romantic attraction. Much of what I found shows up in business, usually unconsciously, in hiring, work relationships, and compensation. The overlaps between the supervisor-employee relationship and the parent-child relationship are well documented. Though rarely spoken of in the workplace, no one in my classes showed surprise when I brought it up. Finally, the list of research fields represented in my bibliography ranges far beyond management and social psychology to economics, computer science, and others.

Wilson has applied his concepts to a “Design Your Own Park” project and a program to help kids at risk for dropping out, he said in the interview. A student had to have flunked three classes the prior year to enter the Regents Academy. Yet, Wilson said, “after its first year, not only did the students do much better than the comparison group, but they actually performed on a par with the average high school student on the state-mandated exam.”

What business leader wouldn’t like those kinds of results?


  • Boorstin, D. (1983), The Discoverers. Vintage Books: New York.
  • Ydstie, J. (2011), “Can Evolution Breed Better Communities?” National Public Radio Weekend Edition, 8/28/11.

Reviewing “Five Dysfunctions” Eight Years Later

I begin with an apology to all those told I had read The Five Dysfunctions of a Team by Patrick Lencioni before 2010. I thought I had. But I checked my bibliography that year and was embarrassed to realize I have been commenting based on what I had heard about the book. Fortunately, those brief comments stand after my reading it. Overall my impression is favorable, and though it has few actionable steps, I agree with someone who said it might inspire people to do real teambuilding. “It is teamwork that remains the ultimate competitive advantage,” Lencioni writes, “both because it is so powerful and so rare.”

Lencioni, head of the management consulting firm The Table Group, wrote a business fable about Kathryn, hired as CEO of a troubled software company. The leadership team is fractured, rudderless, afraid to confront problems internal or external to the team. Through a series of off-site retreats and regular meetings, she corrects the problems. One person quits and another is “quit-fired,” as I call it, but the rest pull together and turn the company around. Having written and edited fiction, I think he does a nice job of capturing the experience of true teambuilding.

He closes with a synopsis of his model and methods for using it. Writing independently at the same time, we came up with the same general approach. (I finished the original version of The SuddenTeams Program in 2001; he published the book in 2002.) For example, Kathryn opens the first off-site with an exercise he calls “Personal Histories” that mirrors the enhanced introductions I use. With one exception, she uses no typical teambuilding activities, instead focusing on changing the behaviors of people using their real-life business issues. “We’re not going to be catching each other falling out of trees,” she says. She keeps them focused on the problems to be solved. She reinforces what it takes to be a good team player, including a willingness to confront other members and put team goals ahead of personal goals.

Kathryn spends little time on theory other than Lencioni’s “Five Dysfunctions,” which build up from the first through the last:

  1. Absence of Trust
  2. Fear of Conflict
  3. Lack of Commitment
  4. Avoidance of Accountability
  5. Inattention to Results

These dysfunctions are valid, though he should have left off the “The”: there are far more than five, and none of his five are necessary to kill a team. And trust does not have to happen first to fix them. Set up systems that spotlight people’s task and behavior agreements, and trust will rise along with performance. In fact, psychology research shows trust cannot be “created”; it has to develop over time through repeated positive interactions.

There’s only one thing Lencioni got absolutely wrong according to the scientific literature. He says the Myers-Briggs Type Indicator is “scientifically valid” and has a “basis in research,” which are simply false, as I’ve already explained. He unwittingly provides one reason why personality tests have positive impacts anyway: “everyone likes to learn about—and talk about—themselves.” They also get people talking openly about what he calls “behavioral tendencies” and provide a common vocabulary. I use cheaper and less invasive ways.

A lesser warning, from my earlier discussion of surveys: His “Team Assessment” has a critical but common survey design flaw, the use of “and” in many of the measures. For a question that says “A and B,” if you agree that A “Usually” happens but B “Rarely” does, neither answer (nor a compromise “Sometimes”) is accurate.

Lencioni slams the effort to create consensus, which literally made me gasp until I realized he was misapplying the term. He really is talking about unanimous or “complete agreement” according to The American Heritage Dictionary. Consensus, the dictionary says, is only a “general agreement.” On Full Scale agile™ I show how a team can reach consensus even if a member thinks the decision is a mistake.

Lencioni was wrong to extol the virtues of “conflict,” but I was in the “Conflict is good” camp myself in 2002. The definitive study that changed my mind came out two years later. Change “conflict” to “confrontation,” and I have no problem with his positions. As Kathryn says, “Our ability to engage in passionate, unfiltered debate about what we need to do to succeed will determine our future as much as any products we develop or partnerships we sign.”

Other quotes or points that I put sticky notes by as I read included:

  • Kathryn makes someone put away his computer during meetings because “it’s easy to imagine the person… checking e-mail or working on something else.”
  • Using a sports reference she says, “If you let profit be your only guide to results, you won’t be able to know how the team is doing until the season is almost over.”
  • After off-sites, “As soon as the reality of business problems is reintroduced… people revert back to the behaviors that put them in the difficult situation in the first place.” This is why my Agile work is performed on-site in the team’s regular meetings.
  • “I’m talking about everyone adopting a set of common goals and measurements, and then actually using them to make collective decisions on a daily basis.”
  • On setting goals, Kathryn says: “If everything is important, nothing is.”
  • Kathryn failed to make clear what information from the first off-site could and should be shared with employees.
  • She eventually decides “focusing on real issues, rather than exercises, was her best bet.”
  • “As difficult as it is to build a cohesive team, it is not complicated.”
  • Like me, Lencioni downplays the value of “ropes courses and other experiential activities” unless they are “layered upon more fundamental and relevant processes.”
  • “Great teams… understand the old military axiom that a decision is better than no decision.
  • As I teach, he writes: “the most efficient means of maintaining high standards of performance on a team is peer pressure.”

A problem for managers wanting to use the book is that Kathryn has skills developed as a 7th-grade teacher. She is also an experienced meeting facilitator and coach, which most line and middle managers are not. Most managers never receive training on real teambuilding, yet they are expected by their companies to supply it. Hopefully this site will help those folks succeed.

Source: Lencioni, P., (2002). The Five Dysfunctions of a Team: A Leadership Fable. Jossey-Bass: San Francisco.

Problem-Solving Teams Created 1.5 Workers

A lot of philosophy classes are on my college transcript, taken with the incomparable Rick Miller, a Columbia Univ. Ph.D. incongruously teaching for decades at the North Carolina School of the Arts. From those and Zen writings I can provide a plethora of moral and ethical reasons why nearly every organization should create true teams with their own charters, goals, procedures, and improvement plans. But some managers will only be motivated to change by one reason: Money.

They want to know, “How is the time and hassle going to help my bottom line?”

Fortunately, I can make that argument as well. Here are a few real-world examples:

  • At a Green Giant plant in Illinois, a line-employee team reduced average machine changeover time by more than half, realizing $793,000 in downtime and inventory savings.
  • The graphics department in Palm Beach County (FL) formed a self-directed team by splitting half of the former supervisor’s pay among the team members in exchange for taking over the supervisor’s duties. Within a year the team created a 21% increase in revenues—while saving the taxpayers the other half of the supervisor’s pay!
  • A self-directed team at Tektronix was able to produce in just three days the same number of units formerly produced in 14.
  • After four months as a self-directed team, three technical communicators at Los Alamos National Laboratory were doing as much work as the group had done before with five people—including a team leader—yet reported no sense of overload.

Another example comes from a study in which researchers tracked machine operators in a light manufacturing plant for three years. At the start of that period, the anonymous company the authors called “PARTS” introduced improvement teams similar to quality circles. QCs are thought of as a Japanese innovation, but were introduced to that country after World War II by American consultants like J. Edwards Deming to help it rebuild. In the PARTS version, teams of volunteers from different work groups were empowered to identify and correct “plant-wide issues and the specific problems that confront individual team members.” Thus says the study article by the researchers, economics professors Derek Jones at Hamilton College and Takao Kato at Colgate Univ.

One goal of the teams was “problem-solving, or devising a concrete solution to a specific problem that will result in quality improvement and productivity gains.” But interviews with the CEO and human resources leaders showed that a second goal was just as important: “providing production workers with an opportunity to participate in workplace decision-making.”

Each self-directed team consisted of an engineer and an average of seven workers, aided by a consultant. The teams met once a week for 30 to 45 minutes. Here’s the first interesting tidbit: Members’ nonproductive time only went down 15 minutes per week. The company tracked labor time tightly, including breaks, preparation and clean-up, sick time, etc. Right off the bat, the company got back a third to a half of the time it gave people to work on solving problems. The explanation may be as simple as survey results in which members reported that they worked harder after joining the teams.

Meanwhile, the machines they worked on recorded daily output, and all workers’ parts were randomly audited against technical requirements. “Though products would not be classified as ‘hi tech,’” Jones and Kato say, “customers demand high quality…” The nature of the work was such that large variations in quantity and quality were easily possible between workers using the same machines to create similar parts.

Workers on the teams increased their output by 3% on average and improved quality, as measured by rate of rejected parts, a whopping 27% compared to non-members. The researchers looked for other possible explanations besides team membership, but no other major changes took place in the plant’s technology, human resources policies, or economic conditions that might have helped or scared people into producing more.

You might be thinking people who volunteered were good performers already, so gains were easy. The researchers compared members to people not on the teams. They also looked at differences between those who volunteered for the teams and those asked by managers to join. The latter tended to be poor performers instead of top ones. The authors think managers did that in hopes participation would improve the performance of those workers, which it did.

The key means of improvement appeared to be the learning shared between the machine operators and with the one engineer on each team. “During these meetings, team members learn valuable lessons, and they become more motivated, inventive, and ‘smarter’ workers with a clearer awareness of the importance of quality control,” Jones and Kato write. Improved self-image, social connections, information- and knowledge-sharing, and understanding of the impact of productivity on the plant’s success also may have contributed, the authors say.

There were other benefits to the company of 225 people. Plant-wide productivity and quality seemed to go up, though the researchers did not look closely at that. In surveys, team members reported higher:

  • Job satisfaction.
  • Commitment to the company.
  • Trust in management.
  • Belief in their own skills.
  • Levels of communication with managers and people outside their work groups.

Not one person who got on a team quit.

Let’s do the math on this program. A total of 54 machine operators averaging $7.64 per hour eventually served on teams. Using a standard labor year of 2,078 hours, that’s total base pay of $857,300 per year. A 3% increase in productivity equals $25,719. The net cost of 0.6% downtime is $5,144, leaving a net gain of $20,575. Maybe $20,000 hardly seems worth the effort to you, but how many initiatives in your company provide a 500% net return on investment (ROI)? In this low-wage company, that was the equivalent of adding 1.5 full-time employees at a cost of 5 grand. And these figures are only for one year. The gains slowly dropped over time, such that other actions would be needed to maintain them, but at a rate that would take nearly three years to go away. The article doesn’t include enough information to calculate the cost savings from a 27% increase in quality, or from the apparent improvements among people not on teams.

This was not a progressive firm. It had no profit-sharing or stock ownership plans, and no worklife-enhancing HR policies like job rotation. The monthly “all-hands” meeting was the standard “we-talk-about-nothing-major, you-listen” waste of time. A small 401(k) contribution was technically a bonus, but so routine that workers and managers had come to expect it, wiping out the motivational value. Imagine the financial impacts if the company had turned the functional work groups into empowered teams; allowed people to cross-train and do different jobs; or introduced a goal-and-rewards system!

The effort obviously had top-level support given that the CEO hired “a full-time consultant with long experience in the introduction of teams at other firms.” Most case studies of failed change efforts point to a lack of engagement by top management as a principal cause. If you’re not the Big Boss and PARTS has convinced you to look at teams as a tool for improving your operations, I suggest using the numbers above to make the argument.

Maybe I’m missing something, but how could anyone turn down a 500% ROI?

Source: Jones, D., and T. Kato (2011), “The Impact of Teams on Output, Quality, and Downtime: An Empirical Analysis using Individual Panel Data,” Industrial and Labor Relations Review 64(2):215.

The Science of Agile Teamwork

The benefits of empowered teamwork are one subset of the benefits of Agile project management. Having moved from general project management and teamwork consulting into Agile transformations, I now have seen this firsthand many times. A talk by an internationally known “Agilist” confirmed that belief earlier in my career.

If you know about Agile, you might want to skip this and the next paragraph as I explain it to newcomers. Agile is an approach to project management, a philosophy that is more adaptive to change and expressed through a statement called the “Agile Manifesto” and 12 supporting principles. Since Agile came out of the software world and the “Scrum” method is the most popular there, I’ll use an example incorporating those to explain. Traditional “waterfall” methods start a software project with an attempt to capture everything the end user needs to do and estimate the time and money needed. Once that information is approved, next comes design of the software. Then the software is created (“coded”). Next comes the testing phase, during which mistakes (“bugs”) are fixed until the client accepts the software. Finally comes rollout of the product. Each step flows downward into the next, hence the “waterfall” allusion.

In Scrum, an initial set of requirements is collected and prioritized by a “product owner” working with the client. These are broken into small chunks, feature by feature, and captured in a “product backlog.” The product owner, a facilitator called a “Scrum Master,” and the other team members take a short period to do some initial design work. Then they decide how many of the features they can finish within a pre-selected time, usually two to four weeks. Once the team decides what to do, that set cannot be changed short of an emergency. The team then does a mini-waterfall of sorts, finishing the selected set of features to the degree they could be released to the customer (whether or not they are right then). After a demo and lessons-learned review, they repeat the whole process with the next set of features, starting the next business day. They’re done with the whole thing when the customer says they’re done. (For a detailed explanation, see “The Agile Difference” on my Full Scale agile site.)

At a meeting of the local chapter of the Association of IT Professionals, Robert Galen spoke on “Mature Agile Teams—Sixteen Essential Patterns.” Galen was director of research and development at iContact, an e-mail marketing company, and has his own Agile consulting practice. Along with making me feel better about two points over which I parted ways with one Agile company—for those keeping score, I was right on one and half-right on the other—he provided a number of points about teamwork that work in any environment.

One of his 16 “patterns” was “Truly Collaborative Work.” Examples on his slide included, “Developers willingly engage in Testing.” This is not common in waterfall projects, and I have witnessed how it improves quality and cooperation. Another point was, “Members help each other out.” Science has shown that “organizational citizenship behaviors” improve team performance. “Listening to each other; mutual respect” appeared as well. In poorly performing teams, people listen at each other, listening but not really hearing (hence the Active Listening class I offered through TeamTrainers).

“Behaving Like a Team” was another of Galen’s patterns. I especially liked his point about “Providing each other congruent feedback.” Scrum promotes a practice I suggest for all teams in my teamwork book, daily “stand-up” meetings. These are conducted literally standing up, for a maximum of 15 minutes. Each member reports on only three things: what I did in the prior work day, what I plan on doing the next day, and any blocks I’ve run into. The last item becomes a top-priority action item for the Scrum Master. Galen said members of effective teams also participate in “Passionate debate.” He added “conflict,” but I later suggested the word “confrontation.” As discussed earlier in this hypertext, the scientific evidence shows that conflict of any type harms teams, but members must be willing to confront each other to make better decisions. Galen also said members will spend personal time together and succeed or fail “as a team.”

The research literature supports the use of self-managed or self-directed teams in most circumstances. Under his pattern, “Quality on all fronts,” Galen said Agile teams are “Self-inspecting; self-policing; self-learning.”

He did an excellent job of defining the role of the supervisor of a self-managed team. My oft-repeated summary is, “Tell the team what direction you need it to go, give it its boundaries, and get out of the way.” Then you fall in behind, making sure the team has the resources it needs and nudging it to stay on course. Another speaker that night had a great analogy. Josh Anderson, Agile Coach at Teradata, likened this to raising the bumpers when you take kids bowling, so their balls stay out of the gutters. Galen’s related pattern was “Saying NO as a Leader.” He emphasized that managers can’t just walk away from the team, and added in bullet points:

  • “Sometimes direction is required.”
  • “Courage to tell it like it is.”
  • “Behind the scenes, 1:1 Coaching…”

Finally, he emphasized, members’ first loyalty must be to the team. This made me uncomfortable, because plenty of teams have failed by focusing too much on themselves. But Scrum’s emphasis on including at each step the customer’s representative (the product owner), and often the customer, is a perfect way to align team cohesion with business goals.

What We Know about Virtual Teams

We are more alike than different, but we notice the differences more. This truth about individuals applies to teams as well, for top leadership and production teams, medical and construction teams, and virtual teams as compared to standard face-to-face or “collocated” teams. The tasks, jargon, education, and communication methods differ, but the group dynamics are the same. For example, a working paper from the Massachusetts Institute of Technology (source #5 below) reports that an analysis of 97 studies over 18 years found better virtual team (VT) design was linked to higher output quality. This hypertext makes clear that is true about teams in general. Also related to higher VT quality were collaborative behaviors, effective coordination and communication, trust, team cohesion, low conflict, team efficacy, and a shared understanding of the task, all of which holds true for collocated teams. But of course, there are some significant differences.

You no doubt think of a VT as one in which people are scattered about the planet in three time zones, never seeing each other in person, and struggling to use the same language. That is more of a worst-case scenario, however. Over time researchers have come to focus more on the means of communication in their definition of a virtual team: virtuality in this case is measured by how much of your communication is electronic. In some lab experiments the people were in the same room. But they were unable to see each other and only communicated by computer. After one CEO declared no one could e-mail on Fridays, two virtual correspondents found out they worked across the hall from each other!

Much of the research therefore has focused on how “computer-mediated communication” (CMC) impacts a virtual team. A 2004 review (#8 below) of the 93 studies on VTs published to date found:

  • Technology that engages more senses improves VT performance: For example, “the addition of video resources results in significant improvements to the quality of a team’s decisions.”
  • Higher technical expertise is linked to VT success, presumably because technical people are better at using CMC programs.
  • The lack of nonverbal cues in most CMCs is part of why:
    • VTs take longer to make decisions.
    • VT members have a harder time learning each others’ areas of expertise or predicting what another member will do.

The MIT paper found no overall relationship (good or bad) between the use of CMC and the quality of a team’s output. Higher reliance on CMC clearly hurt output quantity in their review, however. The 2004 review team agreed, saying virtual work “increases the amount of time required to accomplish tasks.”

That finding about “richer” media that involves more senses has been a consistent one over the years. Web cam is better than phone; phone is better than e-mail. Another study of 54 successful teams in 26 countries discussed in Harvard Business Review (HBR; #7) found that team managers felt e-mail was a lousy way to collaborate. Instant messaging (IM) worked for some teams but bothered others; the 2004 review found IM increased frustration levels. Since IM essentially demands that you multitask, and multitasking harms productivity, that makes sense. The big winner, HBR said, was virtual workspaces that went well beyond the typical shared folder of documents. These feature:

  • Threaded discussions that were facilitated and summarized.
  • Prominent display and tracking of the team’s mission, goals, and tasks.
  • Team member profiles including not only contact information but pictures, “accomplishments, areas of expertise, and interests…”
  • Information on stakeholders.
  • Meeting agendas, background information, and minutes.

The teams discussed in HBR used conference calls primarily to work out issues, not for status reporting. This fits my long-held position that team meetings should be held for decision-making, not as information exchanges. (In case any Agile readers are wondering, I consider daily “scrums” to be decision-making meetings, because issues are often raised and resolved.)

I can’t resist throwing in this anecdote from HBR that helps explain why I warn you about taking lessons from success stories. The four researchers asked for executives’ permission to study failed teams for comparison to successful ones, to see what the differences were. “They balked,” the scientists write. “No one wanted to talk about failures.” Beware the case study, including those in HBR. I’m using this HBR case study because its conclusions fit the evidence from the other sources.

In ten years of looking at the issue of virtual teams, I have never seen proof that VTs are less likely to succeed than a collocated team doing the same task. But I would label the scientists as pessimistic. When you look at the downsides, it is easy to see why:

  • In VTs it is “harder for members to establish a unified sense of purpose…” and to do project management. (#7)
  • The overall amount of communication is usually lower. (#8)
  • Virtual communication raises unique issues, “such as how to react to no participation and the multiple meanings attributed to silence.” (#8)
  • “Likewise, human-computer interface mistakes and technology failures… hamper communications,” but these errors often are mistakenly blamed on personal failings such as a poor work ethic. (#8)
  • VT members are more likely to use inappropriate behavior such as swearing and name-calling. (#8)
  • Greater reliance on electronic communications reduced team cohesion and trust and raised conflict, though again the richer the media, the better. (#9)
  • Heavier reliance on CMC was linked to lower coordination and reduced communication quality. (#9)

Add this to CMC’s impact on output quantity, and you see my point.

Another measure of team performance is whether team members want to stay on the team, referred to as “team viability.” One literature review found that overall, VT members are less satisfied with the team experience than are collocated team members. In the MIT review, which used a very different way of judging the data, the authors were surprised to find no impact on worker satisfaction. This is clearly is an area where the famous scientific cliché applies: “More study is needed.”

Given the communication issues, you won’t be surprised to learn virtual teams suffer more conflicts and find them harder to resolve. Two Stanford researchers (#4) looked over the relevant studies and determined:

  • Task, process, and personal conflicts are higher among VT members than collocated members, for various reasons such as:
    • Higher diversity and differences in perspective and social norms.
    • The difficulty of establishing a good working rhythm when separated by time and distance.
    • Reduced personal knowledge of each other.
    • Having to coordinate via technology, which reduces nonverbal context and makes it less likely important information gets to everyone who needs it.
  • Task conflicts are more likely to become personal conflicts in VTs because trust is harder to build when you don’t interact often with someone.
  • Task conflict hurts virtual teams more than collocated ones because resolving conflict electronically is difficult.

Another reason for more conflicts is the problem of factions. Any team, virtual or not, that breaks into subgroups based on one or more traits is in potential trouble. When the faction members sit together in different locations the issues compound, in part because the faction members come to identify with the faction more than with the team. By coincidence, the two virtual teams I led as a project manager in Seattle had people there, in California, and in one or two cities in India. Besides the geographic diversity, I had to deal with the worst possible time difference, since India is 12 time zones from Seattle. The cultural issues were worse, especially between Seattle and… California. Just kidding. Mostly.

The subteams in India were consistently ethnic Indians, Hindu, young, and mostly male. At least they had shared perspectives. A subteam in Seattle might have one faction of older white Americans along with one or more factions reflecting a mix of demographics.

Those of you who lead virtual teams may be depressed. I do apologize. Allow me to make it up to you by providing the suggestions the scholars agree on for helping virtual teams perform better. The unanimous decision is that occasional face-to-face meetings are well worth the cost. My educated guess is that any team that will last at least six months should meet in person at least once at the start of the team’s life and every six months after that. Notice I said “at least.” Harold Geenen, widely credited for transforming ITT Corporation, gathered his global leadership teams monthly (#2).

Team structure improves performance better than daily coaching, and this may be more important to VTs than collocated teams. Researchers who reviewed VT studies done through 2004 said setting goals is very important for commitment, decision quality and other outcomes, and “formalizing work processes and strategies has been found to be critical for VT performance” (#8).

For example, a pharmaceutical team with 15 members split among London and Germany developed major problems in communication, task coordination, customer requirements, and task assignments. The team responded by starting face-to-face meetings three times a year that included workshops to clarify roles and procedures in addition to social activities. Members also met in person with a U.K. service team that was holding up their work. The problems eased. The researchers who wrote about the team suggested, “Shared goals, clear roles and regular performance feedback are particularly important… and need to be spelt out explicitly.” They also said, “Standards and processes… need to be agreed at the outset…” (#1).

So, hold an initial retreat with a virtual team to put formal structure in place (or to right a VT veering off course). In a couple of days you can get pretty far in drafting agreements on the team purpose, SMART goals, team roles and rules, meeting and decision-making procedures, and some work procedures (see “Self-Directed Agile” in FuSca for the steps). If your boss’s answer to a retreat is “no way,” set up a series of video or Web conferences to do so, in smaller chunks over the first few weeks.

Make sure team members learn about each others’ backgrounds and skills. Knowing who knows what helps any team, but is harder to learn when you don’t see everyone every day. A study of cross-cultural teams suggested this could help reduce stereotyping as well (#3). At the kickoff retreat, hold a social event with a “get-to-know-you” exercise.

As a manager, you will have to put in more time coordinating than you do for a collocated team. “Team leaders rarely let a day go by when members did not communicate with one another,” the HBR article on successful VTs said (#7). “One team leader reported being on the phone with his team for ten to 15 hours a week.” A better alternative for project teams is to become Agile, which along with other benefits will eliminate this need and enact many of the recommendations in this topic.

Given that you literally can’t look over people’s shoulders, empowering the team members to make most decisions on their own seems a no-brainer. But that means managers must “spend considerable time coaching individual team members off-line and work diligently to ensure that all team members feel fully informed…” according to a study on VT empowerment (#6). Toward that end, provide plenty of training and follow-up coaching on the communication technology your team uses. Don’t assume everyone knows how to use the advanced features.

You may have to personally replace some of the energy generated by people sitting in a room together. A study of 41 teams in nine countries found that an “inspirational leadership” style improved performance. This kind of manager “makes everyone in the team enthusiastic about the team’s assignments,” has a sense of mission, encourages people to share ideas, and “excites us with his/her visions of what we may accomplish if we work together as a team,” the study said (#5).

Then again, that sounds like the kind of manager I’d like to work for anywhere—across the hall or across the globe.


  1. Axtell, C., et al. (2004), “From a Distance,” People Management 10(6):38.
  2. Geenan, H. (1984). Managing. Avon Books: New York.
  3. Grosse, C. (02), “Managing Communication within Virtual Intercultural Teams,” Business Communication Quarterly 65(4):22.
  4. Hinds, P., and D. Bailey (03), “Out of Sight, Out of Sync: Understanding Conflict in Distributed Teams,” Organization Science 14(6):615.
  5. Joshi, A., M. Lazarova, and H. Liao (09), “Getting Everyone on Board: The Role of Inspirational Leadership in Geographically Dispersed Teams,” Organization Science 20(1):240.
  6. Kirkman, B., et al. (2004), “The Impact of Team Empowerment on Virtual Team Performance: The Moderating Role of Face-to-Face Interaction,” Academy of Management Journal 47(2):175.
  7. Majchrzak, A., et al. (2004), “Can Absence Make a Team Grow Stronger?” Harvard Business Review (May):131.
  8. Martins, L., L. Gilson, and M.T. Maynard (2004), “Virtual Teams: What Do We Know and Where Do We Go from Here?” Journal of Management 30(6):805.
  9. Mortensen, M., O. Caya, and A. Pinsonneault (2009), “Virtual Teams Demystified: An Integrative Framework for Understanding Virtual Teams and a Synthesis of Research,” MIT Sloan School of Management Working Paper 4738-09 (quotes © Mark Mortensen, Olivier Caya, Alain Pinsonneault).
  10. Other virtual team sources listed in the FuSca bibliography.

Section Menu | ← Back | → Next

Tell the world: