The offsite ran two days, and by the second afternoon I had filled half a notebook with sentences that were said in the room and then revised at the coffee break. The team had been built with real care. The finance chief had taken a division public. The head of product could read a market a year before it moved, and the operations lead was someone the CEO kept calling the best he had ever hired. Each of them, on their own, was the kind of person you would fight to keep. Asked to write down the company's single most important goal for the next two quarters, they handed back a stack of cards on which almost no two answers matched.

Nobody in that room was underperforming, and that was the part that took me a while to see. The CEO had done what most leaders do when they put a senior team together. He found the strongest individual he could for each seat and assumed the strength would add up. It feels like the safe move. The research on how teams actually perform keeps finding that it does not hold.

McKinsey spent more than two years building a diagnostic for team effectiveness, running it across 110 teams and more than 900 people in 42 countries, and one of the first things their data broke was the assumption the CEO was working from. A team stacked with the best individual talent in every seat does not reliably turn into a better team, and sometimes it turns into a worse one.1 The behaviors that separate a strong team from a struggling one account for most of the distance between them, somewhere between 69 and 76 percent, and very little of that has to do with raw individual horsepower.1

The team that exempts itself

Here is what struck me as strange, having spent years coaching the teams that sit below these tables. The executive floor is usually where the language of teamwork gets manufactured. The delivery teams two and three levels down are the ones made to run retrospectives, write working agreements, name who owns which decision, and sit through facilitation when their collaboration jams. The leaders who require all of that rarely run themselves the same way. The top team meets, decides, scatters, and calls the standing weekly its teamwork. It is the one group in the building judged on outcomes and never examined as a team.

The executive team is the only team in the company expected to win without ever being examined the way it makes everyone beneath it submit to examination. It authors the requirement and writes itself an exemption.

The price of that exemption shows up in the same data. Teams that scored above average on trust were 3.3 times more efficient and 5.1 times more likely to deliver their results than teams that scored below it.1 Trust at that altitude has nothing to do with warmth. It is whether a head of sales can tell a head of product, while the CEO is sitting right there, that the current roadmap will cost them the quarter, and have it land as information instead of as a shot across the bow. Plenty of senior teams cannot manage it. They perform agreement while the leader watches and carry the actual disagreement into the hallway, which is most of what I was writing down in that notebook.

Some of the trouble runs underneath trust. When the researchers checked whether members even agreed on what kind of team they were on, meaning how much their work genuinely depended on one another, only 46 percent of teams agreed.1 The rest were, in the study's language, playing different sports at the same table. One executive treats the weekly meeting as the place to defend a number. Another treats it as the place to make an enterprise call. They walk out each certain a decision was reached, having reached two different ones.

Why the strongest people make it harder

The talent that earns someone a seat at that table is the same talent that makes the table hard to run. People reach the executive floor by owning a function and defending it well, rewarded for a decade or longer for being right inside their lane and for protecting what their lane needs. Then they arrive at the one table where the assignment inverts, where the work is to surrender ground for an answer that serves the whole company, and the instincts that carried them there start working against them. McKinsey found that team leaders rate their own team's health higher than the members do, across almost every measure.1 The person most sure the top team is functioning well is usually the person running it.

What changes a senior team rarely arrives in the form of a ropes course or a personality inventory. The thing that moves it looks a lot like what the teams downstairs were handed years ago. Give the group a real problem it cannot carve into separate parts, one shared outcome that no single member can claim on their own. Keep them inside it long enough to feel where the friction actually lives. Then name out loud who decides what, because friction loses room to operate once the decision rights stop being ambiguous. And the leader has to move first, inviting hard feedback onto their own calls before expecting anyone else to take it.

The leadership team tends to be the last team anyone treats as a team, and the first one whose dysfunction the rest of the organization inherits. If you sit on one, the uncomfortable question is the useful one. When did your top team last examine how it works together with the same honesty you demand from every team you have asked to get better? I would be curious what you find when you put the question on the table.