Think of the best team you have ever worked on, and the worst. Chances are the difference was not raw talent. On the good team, people admitted when they were stuck, meetings ended with decisions, and the work got a little better each month. On the bad one, capable people talked past each other, problems surfaced late, and the final week was a scramble.

Researchers have spent decades asking what makes a team effective. Their answers are less glamorous than “hire stars,” but much of the evidence points one way: how a team is set up, and how its members work together, tends to matter more than who is on it.

What does team effectiveness mean?

Most teams judge themselves by a single result: did we ship, hit the target, close the quarter? Harvard psychologist J. Richard Hackman and his colleagues argued that this is too narrow. In a 2005 paper with Ruth Wageman and Erin Lehman, Hackman summarized their three-part definition:¹

• The work meets the standards of the people who receive it. Quality, quantity and timeliness are judged by the team’s clients, not by the team itself.
• The team gets better at working together. An effective team is more capable at the end of a piece of work than it was at the start.
• Members learn and are well. Being on the team contributes to people’s learning and well-being rather than frustrating, alienating or deskilling them.

By this standard, a team that hits its deadline by burning people out, or finishes a project too divided to take on the next one, has not been fully effective.

Myth

An effective team is one that hits its target.

Fact

By Hackman's definition the team must also get better at working together, and its members must learn and be well.

When output falls short, the model traces it to three performance processes: how much effort members collectively put in, whether their strategy fits the task, and whether they bring enough knowledge and skill to bear.¹ Asking which of the three is missing can be more useful than asking who is to blame.

Talent or teamwork? Project Aristotle and “team IQ”

Google studied the question internally. In an effort code-named Project Aristotle, its researchers examined 180 of the company’s teams: 115 engineering project teams and 65 sales groups.³ They judged effectiveness four ways, combining ratings from executives, team leaders and team members with sales performance against quarterly quota.

According to Google’s guide, what mattered was less who was on a team than how it worked together.³ Several factors were not significantly connected with effectiveness at Google, including whether teammates sat together, consensus-based decision making, members’ extroversion, individual performance, workload, seniority, team size and tenure. Google itself cautions that these factors may still matter elsewhere, and this is an internal study, not a peer-reviewed paper.

180Google teams studied in Project AristotleSource: Google re:Work, 2015

Academic work on “collective intelligence” points in a similar direction, with more debate attached. In a 2010 Science paper, Anita Williams Woolley of Carnegie Mellon University and colleagues studied 699 people working in groups of two to five.⁴ They reported a general factor that explained how well a group performed across a range of tasks. It was not strongly correlated with members’ average or highest intelligence, but it was correlated with their average social sensitivity, how evenly people took turns in conversation and the proportion of women in the group.

That idea has been challenged. In 2017, Marcus Credé and Garett Howardson reexamined data from six published samples and concluded that support for a general collective intelligence factor was generally weak, pointing to statistical issues that may have inflated the results.⁵ In 2021, Christoph Riedl, Woolley and colleagues pooled data from 22 studies covering 5,279 people in 1,356 groups.⁶ They reported strong support for the factor and found it was predicted most strongly by the group’s collaboration process, followed by individual skill and group composition.

The fair reading: whether groups have a single “team IQ” is still debated. But the studies that find one, like Google’s internal research, point to how people work together, not only to individual talent.

Hackman’s five conditions for team effectiveness

Hackman’s model, developed from the late 1980s onward, does not treat great teamwork as something a manager can directly cause. Teams develop over time in their own ways, but certain conditions make good performance more likely without guaranteeing it. His model names five, which Wageman, Hackman and Lehman built into a survey tool tested with 2,474 members of 321 teams:¹

Two details are especially practical. First, the model holds that direction motivates best when the ends are specified clearly and firmly but the means are left to the team.¹ Dictating both underuses the team’s talent. Second, norms matter early. The authors note that the expectations a team settles into tend to stick until something fairly dramatic prompts a rethink.

1. A real team
2. Compelling direction
3. Enabling structure
4. Supportive context
5. Expert coaching

How much does design matter compared with day-to-day coaching? The evidence the paper summarizes is mixed. In Wageman’s earlier study of field service teams at Xerox, the team’s direction, structure and context together accounted for 37 percent of the variation in objectively measured performance, compared with less than 1 percent for the leaders’ coaching.¹

But in a study of intelligence analysis teams, the five conditions together accounted for 74 percent of the variation in a combined effectiveness measure, and coaching mattered much more: task-focused coaching from peers had the strongest link of any measure.¹ The authors hold that coaching cannot compensate for a badly flawed design, and note that research was still sorting out which conditions matter most for which teams.

Further reading

• 

  The Fearless Organization

  by Amy C. Edmondson

  Edmondson's book on psychological safety, the idea from her 1999 team study that this article covers.

As an Amazon Associate WiserHours earns from qualifying purchases.

Why psychological safety matters for teams

One idea runs through much of this research: psychological safety. Organizational researcher Amy Edmondson introduced the team version of the concept in 1999, defining it as a shared belief that the team is safe for interpersonal risk taking.² It is easy to mistake for niceness. Edmondson wrote that the term does not mean permissiveness or constant good feeling, but confidence that the team will not embarrass, reject or punish someone for speaking up.

The study

Moderate evidence

Edmondson (1999), Administrative Science Quarterly

Her study of 51 work teams in a manufacturing company found that psychological safety was associated with learning behavior, which she described as asking questions, seeking feedback, experimenting, reflecting on results and discussing errors or unexpected outcomes.² That learning behavior, in turn, helped explain the link between psychological safety and team performance.

Her data also hinted at where safety comes from. Support from the wider organization, such as resources and information, was significantly related to it, and coaching from the team leader came close to statistical significance.²

Google’s guide ranks psychological safety first of five dynamics of effective teams, followed by dependability, structure and clarity, meaning, and impact.³ It reports that people on teams with what it calls a stronger team culture were less likely to leave, brought in more revenue and were rated as effective twice as often by executives. Those are internal figures that show an association, not proof of cause and effect.

Two habits of effective teams

Sharing what only you know

Teams exist partly to pool knowledge no single person has. Jessica Mesmer-Magnus and Leslie DeChurch pulled together 72 studies covering 4,795 groups and 17,279 people.⁷ Information sharing predicted team performance under every condition they examined, and it was also linked to cohesion, satisfaction with decisions and knowledge integration.

The catch is when sharing broke down. Teams shared less when information was divided among members rather than held in common, when members depended on one another’s information and when members were more different from one another.⁷ So sharing fell in the very situations where pooling knowledge matters most. Structured discussion and cooperation were among the factors linked to more sharing.

Debriefing: looking back together

Debriefs, also called after-action reviews, have been used by the U.S. Army for decades. Scott Tannenbaum and Christopher Cerasoli combined 46 samples with 2,136 participants and found that, on average, debriefs improved effectiveness by about 25 percent compared with control groups.⁸ The effect was similar for teams and individuals, in simulated and real settings, and in medical and nonmedical work. Debriefs worked better when who took part, what they focused on and what was measured were aligned. The authors also saw signs, though less conclusive, that facilitation and structure help.

~25%average improvement in effectiveness from debriefsSource: Tannenbaum & Cerasoli, 2013

How to improve team effectiveness

1. Check that you are a real team

Write down who is on the team and the outcome you share. If people disagree about who is in, members mostly work on separate tasks, or membership changes every month, sort that out first. Google’s guide draws a similar line: work groups mainly share information along a reporting line, while teams plan work, solve problems and review progress together.³

2. Write down the ends, not the means

In a sentence or two, state what the team must deliver, for whom and by when, and how those clients will judge it. Then leave the team room to decide how to get there. Planning the work itself is a project management job.

3. Keep it small and set norms early

Add people only when the work needs their skills. Team size did not stand out in Google’s analysis, but its guide cites other research favoring teams of fewer than 10 members.³ In the first weeks, agree on a few explicit norms: how decisions get made, how to flag a problem and how quickly messages need answers. How those norms play out in meetings is where most of the day-to-day communication happens.

4. Make speaking up normal

In Edmondson’s data, leader coaching was less firmly tied to safety than organizational support, but it is something a leader directly controls.² Open reviews by asking what could go wrong, respond to bad news with questions rather than blame, and say so when you got something wrong.

5. Structure how information gets shared

Before a decision, ask each person to state what they know that others might not, or collect short written inputs before the meeting. The meta-analysis linked structured discussion to more sharing.⁷ Going around the room also evens out who talks, one of the features Woolley’s 2010 study linked to collective intelligence.⁴

6. Debrief after every milestone

Hold a short session built on the reflection, discussion and goal setting Tannenbaum and Cerasoli describe.⁸ Keep the discussion focused on the result you want to improve, consider asking someone to facilitate, and end with one or two changes, each with an owner.

1. 1What did we set out to do?reflection
2. 2What happened, and why?discussion
3. 3What will we do differently?goal setting

Then repeat from “What did we set out to do?”

7. Fix the context, not only the people

Check whether rewards and reviews quietly discourage collaboration, and whether the team has the information and training it needs. Much of that sits with the manager, which is why leadership habits matter as much as team design.

WiserHours

A team check-up

• Name the team and the outcome we share
• State the ends, not the means
• Agree explicit norms early
• Open reviews by asking what could go wrong
• Ask each person what only they know
• Debrief after every milestone
• Check that rewards do not discourage collaboration

Find this checklist online:
https://wiserhours.com/team-productivity/what-makes-a-team-effective/#checklist-team-effectiveness-conditions

The bottom line

Decades of research point away from star players and toward the conditions around them. Effective teams tend to have a real, stable membership, a clear and meaningful direction, a sensible structure and an organization that supports them. Inside those conditions, they build a climate where people can speak up, share what only they know and look back honestly at their work. None of it guarantees success, and some findings are still debated. But these are the levers a team and its leader can actually pull.