What this covers
Colin Fisher and Russ Roberts examine what determines whether teams succeed or fail, arguing that structure—how tasks are designed, how group size is set, how purpose is communicated—matters far more than individual leadership charisma or trust-building exercises. The conversation moves through organizational research and historical cases, showing that the free rider problem dissolves when people identify with a shared purpose and do work that feels whole and significant, and that even the best coaching cannot salvage a poorly structured team. They distinguish between relational trust (the kind you have with family) and task-based trust (confidence someone will do their job to your standard), arguing the latter drives organizational performance while the former gets most of the attention in corporate culture.
The discussion ranges across Hackman's Work Design Theory, which shows how task significance and autonomy motivate collaboration; Edmondson's hospital studies, which found that the safest-reporting teams actually made no fewer errors but simply had psychological safety to report them; and concrete cases from the Manhattan Project to investment banking that illustrate how connecting workers to meaningful purpose outpaces monetary incentives. Fisher and Roberts explore why group size creeps upward (asking people to leave is uncomfortable), why the same person can be a star performer on one team but fail when hired alone (as Groysberg discovered), and why social loafing grows as accountability diffuses. They also push back on importing leadership lessons from sports, where task design is already solved, and argue that much celebrated individual achievement—from Steve Jobs to scientific discoveries—rests on great teams whose contributions get erased by cultural narratives of genius.
Fisher argues that team performance is governed primarily by structure—well-designed tasks, clear and meaningful purpose, right group size—rather than by individual leaders or trust-building exercises, and that the best trust and motivation emerge from doing real work together.
- Free-riding is solved by shared purpose and well-designed tasks, not just incentives
- Task-based trust matters more than socio-emotional trust and is built by doing the work
- Structure beats coaching: poorly structured teams can't be coached into success
Structure—task design, team composition, and goal clarity—matters far more than coaching or individual leadership personality.
- Sports coaches enjoy huge advantages middle managers lack: well-designed tasks, clear outcomes, and pre-structured roles—which is why people play sports for fun on weekends but never voluntarily do their assigned group projects—so learning leadership lessons from sports overlooks the task-design work coaches never have to do.
“what all football coaches, all sports coaches, have at their disposal that you don't have if you're a middle manager, is you have really well-designed tasks.”
- Per Hackman's Work Design Theory, tasks are one of the most powerful forces determining collaboration: work motivates when it gives people whole pieces they can see through to the end user (task identity), is important to someone (task significance), grants autonomy over the process, and provides feedback from the work itself—and groups handed impoverished tasks lacking these qualities predictably fail to work hard.
“this research comes largely from my mentor, Richard Hackman, who came up with Work Design Theory... If you give people whole pieces of work where they can see the outcome of their labor and its effect on whoever happens to be using it... It's got to have task significance. And that we also have to have autonomy over our work processes.”
- Managers are implicitly taught they should be saviors who cajole people into trying harder through personality and charisma, but no amount of charisma overcomes bad work: when people don't know what they're working on, why it matters, or how to coordinate, and tasks are dreary and poorly designed, you will always see reduced effort.
“No amount of charisma is going to overcome bad work. You're always going to see people not trying as hard as they could.”
- When the effects of team structure (composition, size, goal and task quality, social norms) are pitted against coaching (ongoing process intervention and motivation), structure wins: Wageman's Xerox study found well-structured teams benefited from good coaching and resisted bad coaching, while poorly structured teams could not be helped by good coaching and were severely harmed by bad coaching—so fixing structure must come before coaching.
“for teams that were well structured, they benefited from good coaching... But, the poorly structured teams didn't benefit from good coaching. You could coach them up all you wanted: they still didn't perform very well.”
Leadership is a set of functions groups need, not an intrinsic trait of celebrated individuals; great outcomes emerge from teams, not solo heroes.
- Rather than treating leadership as an intrinsic trait of celebrated figures, it is better understood as the functions a group needs achieved—sometimes articulating a grand vision, sometimes a mid-level official rallying a team to maintain services—so the unheralded leaders who simply 'do what needs doing' deserve recognition alongside the Steve Jobses of the world.
“rather than thinking of it as something that's intrinsic, that's a trait of those people, I like to think of it more in terms of the functions that we need to achieve for any group.”
- Behind every scientific discovery or great company associated with a single celebrated individual lies a great group; our culture's 'cult of individualism' makes single heroes an easier handle for stories, but digging into the reality reveals collective accomplishments, and almost nobody does great things on their own.
“every time there's a scientific discovery or a great company that's really associated with a single individual, when you start digging underneath that, you find a great group that was around them. That, really, nobody does this stuff on their own.”
- Real teams are the primary vehicle by which human problems get solved—the big scientific breakthroughs, great musical albums, admired sports teams, and world-changing discoveries are almost entirely accomplished by groups and teams rather than lone individuals.
“real teams are primarily the vehicle by which human problems get solved. That, the big breakthroughs, the scientific breakthroughs, the great musical albums, all the sports teams we look up to, the discoveries that are going to change the world--those are almost entirely done by groups and teams”
- Groysberg's study of investment analysts found that when firms hired away star performers, the star's performance was only portable if they also brought their admin team; hiring the star alone meant a long ramp-up and often the high performance never reappeared—showing team and support, not just individual talent, drive results.
“They found the portable star's performance was really only portable when you hired their team. When you just took the star, it took them a long time to get off the ground. And, in a lot of cases, you never see that performance become portable.”
People work harder when they identify with teammates and shared purpose; diffused accountability and weak group identity fuel free-riding.
- Human beings dislike feeling controlled, so when we believe someone else is master of our fate we try less hard; a micromanaging boss who dictates every step also shifts blame for failure onto themselves, removing the worker's ownership of the outcome.
“human beings don't like feeling controlled. So, when we believe that someone else is the master of our own fate and our own destiny, we tend to try less hard.”
- Groups can be both the cause of and the solution to free-rider problems: when we don't care about the task and don't identify with anonymous others, we try less hard, but when we strongly identify with the other people in the group and the shared purpose, we will actually try harder than we would working alone.
“if we really strongly identify with the other people in the group, the purpose that we're working towards, then we actually might try more than we would if we were working alone. So, groups can be both the cause of and the solution to these kinds of effort and free rider problems.”
- It is baked into human psychology that when accountability for an outcome gets diffused across larger groups of people, individuals try less hard—demonstrated by Latané's cheering experiments where people produced less output when they merely believed others were performing the same task with them.
“when accountability for any outcome gets diffused across larger groups of people, that we don't try as hard. And that there's some great psychological experiments of this... the classic study by Bibb Latané and colleagues is about cheering... when you believe there's other people, you try less hard. You have less output.”
- When teammates endure pain for each other in high-stakes settings, caring and compassion matter, but it cannot be purely dyadic—trying hard only for a specific friend and not for others; effective performance requires that compassion infuse the whole group through a shared identity, which first requires even knowing who is part of your group.
“usually, we don't want that to be purely dyadic. In other words, I don't want it to be, 'Oh, Russ is my best friend, so I'm going to try really, really hard when it's Russ, but I'm not going to try really, really hard when it's Dan.'”
- Bands that are one-hit wonders often succeed once and then turn on each other, usually because after success the pressure for individual members to claim more credit for themselves overrides the collective goal—illustrating how post-success ego dynamics destroy teams.
“I talk about some bands that were one-hit wonders. They succeed, and then they completely turn on each other, usually because the pressure for individual members to get more of the credit for themselves overrides the collective goal after they've succeeded once.”
Abstract goals require vivid shared vision; task-based trust, not relational trust, drives organizational performance; small bounded teams outperform large groups.
- A 'group' can be any broad social category, but a 'team' is a small, bounded, interdependent group that has come together to accomplish a shared goal; rather than a sharp demarcation, there is a continuum of 'groupiness' (scientifically 'group entitativity') along which crowds can gradually become teams when they develop shared purpose.
“a team is a small, bounded, interdependent group that's come together to accomplish a goal of some sort, that there's some purpose to their being together.”
- There is an inherent tension in team goals: the most important goals (a more just world, eliminating poverty) tend to be abstract and hard to measure, while easily measurable goals (complete this video game level in five minutes) tend to be less significant—so leaders must navigate between importance and measurability.
“there's an inherent tension in things that are really important--that we want to have a more just and peaceful world... Those are, as you say, inherently less specific, more abstract, more difficult to figure out how we're going to measure them than much more concrete goals”
- When a goal is abstract and hard to measure, leaders should paint a vivid, shared picture of a future where the goal is achieved—as JFK did by charging NASA with putting a man on the moon within 10 years—so the organization can rally around it and then deconstruct it into smaller, appropriate sub-goals.
“I talk about JFK charging NASA with putting a man on the moon within 10 years. Now, there's lots of little goals that go underneath that, but that's a vivid picture that everyone at the organization could rally around.”
- When managers inherit a 20-person team they cannot do all the work with everyone, but egos and politics make them avoid excluding people; the practical solution is to make small three-to-seven-member task forces or committees the default mode for doing discrete work, treating full 20-person meetings as the exception rather than the rule.
“saying, 'Well, here's a discrete task that we're going to do, and I'm going to have a task force or committee do that, and that task force or committee will have three to seven members.' Just making that the default mode by which we actually do the work”
- The word 'trust' conflates two distinct things: relational trust (socio-emotional confidence to share deep secrets, typical of family and friends) and instrumental or task-based trust (confidence that someone will do a task to your standard); these two are far less correlated than people assume, and task-based trust is what matters most in organizations.
“there's what psychologists would call relational trust, which means I trust you with my deepest, darkest secrets... But, there's another kind of trust, which we would call instrumental trust or task-based trust. That means when I ask you to do a task, you are going to do it to the standard that would satisfy me.”
- The ideal team size is three to seven members; meetings with around 20 people become largely one-way communication where you cannot hear everyone's perspectives or information, so they are a recipe for accomplishing nothing.
“spoiler: three to seven is the ideal size.”
Psychological safety—freedom from fear of blame or ostracism—is essential for teams to report honestly, admit mistakes, and coordinate effectively.
- Leaders have enormous influence over how psychologically safe a group is, especially in how easily safety is destroyed: it takes only a couple of instances of blaming people or yelling to make them much more hesitant to speak up or admit mistakes, even in life-or-death situations.
“It just takes blaming people a couple of times, yelling once or twice, and then people become much more hesitant to speak up. They become hesitant to admit when they've made mistakes, even when it's life or death.”
- Because our primitive brains are so afraid of ostracism, we are very hesitant to tell teammates when they aren't delivering what we need; overcoming this fear to create an environment where people can ask questions, make mistakes, and give difficult feedback without fear—Edmondson's psychological safety—is essential for getting work done.
“Because our primitive brains are so afraid of ostracism, we're very hesitant to do that. We don't like to feel like, 'Oh, this other person is not going to like me, and so I'm not going to give them feedback.'”
- In Edmondson's study of hospital nursing teams, the best-structured teams (clearest goals, best tasks, autonomy) reported MORE adverse drug events than poorly structured teams—the opposite of Hackman's theory—but investigation revealed they were not making more mistakes, only reporting more, because their psychologically safe climate made staff willing to report dangerous events the poorly structured teams concealed.
“The teams with the best structure... were reporting more adverse drug events than the teams that were what we would call poorly structured... it wasn't that the well-structured teams were making more mistakes. They were just reporting more mistakes.”
Organizations rarely manage group dynamics well and fail at foundational basics like clarifying team membership and designing coherent tasks.
- Despite groups being central to human work and life, humans are systematically poor at managing group dynamics: we ignore the influences groups have on us, conform in ways incompatible with our own goals and well-being, and make a mess of things when we don't pay attention to these unseen forces.
“for how important they are in human behavior, we're shockingly bad at them: that we ignore the influences they have on us, that we're pushed to conform in ways that are incompatible with our own goals and our own well-being”
- Team-building exercises are based on the flawed premise that trust must precede working together; in reality this reverses the correct logic—rather than building trust first through off-site trust falls and secret-sharing, teams should do the work first and develop the needed trust through doing it.
“this comes back to the idea that, before we can work together, we've got to have trust first... But, this turns out not to be very true.”
- Even knowing who is part of your group is shockingly uncommon in organizations: one study found only 7% of top management teams agreed on who the members were, meaning 93% had at least some disagreement on how many people are even on the team.
“There's one study that said only 7% of top management teams know who the other members of the top management team are. They all give different numbers of people who are even on the team.”