Somebody on your leadership team is going to say it: “We should do something fun for the team this quarter.”

They’re right. And you already know how it usually goes. The escape room. The cooking class. The facilitator with the laminated workbook and the beach ball. Your people show up, they’re good sports, they laugh in the right places, and everybody agrees on the drive home that it was better than expected.

Two weeks later, nothing is different. Same pitches. Same client calls. Same senior associate answering a question the client didn’t ask.

I’ve bought that afternoon. More than once. I’ve stood at the back of a room I paid for, watching smart people have a nice time, and known by the second coffee that not one client conversation was going to change because of it. The team liked me for it. That was the whole return.

This post is about why that happens, and what a team event looks like when the fun is the byproduct of building a skill instead of a substitute for one.

Your skepticism is correct, and it’s backed by data

If you suspect most team building is theater, you’re in good company. When Google spent two years studying what made its own teams effective, the researchers tested whether people who socialized outside of work made better teams. They didn’t.

What mattered was whether people felt safe enough to ask a basic question, admit a mistake, or push back on an idea without being embarrassed for it.

Socializing didn’t move that number. Practicing candor did.

Amy Edmondson, who has spent her career on this, put it plainly in Harvard Business Review last year: psychological safety is not about being nice or comfortable. It’s about candor, and it works alongside high standards, not instead of them. Which is the problem with the beach ball. It’s comfortable. It’s nice. It asks nothing of anyone. And so it builds nothing.

Here’s what both pieces of research point at, whether they mean to or not: the thing that makes a team better is people taking a small interpersonal risk in front of each other and surviving it. That’s a skill. Skills get built by reps, not by dinner.

What “fun” looks like when it’s doing work

The best team event I’ve been part of had no icebreaker. It had a volunteer.

A senior professional stands up with a real client situation, one that’s live, one they’re not sure they handled well. I play it back twice. First the version most of us default to under pressure: the fast answer, the credentials, the “what we’d recommend is.” Then the trusted version: the question that should have come first, the silence after it, the frame the client can grab and run with.

I’m always the one who gets it wrong. Your people never have to. They watch the two versions land on a situation they recognize, and the room does the thing rooms do when they see themselves: it laughs, and then it gets very quiet.

That’s the entertainment. Not a game bolted on to make the training bearable. The skill itself, demonstrated on something that matters to the people in the chairs, is the thing they’ll talk about at dinner.

Then they practice. In pairs, on their own client situations, with a script for the ask and a script for the frame. It’s uncomfortable. I tell them the discomfort means we’ve found the right difficulty. Nobody has ever quoted the beach ball at dinner. People quote “that’s about us. I mean, me.” for months.

The test for any team event

Before you book anything, ask three questions.

Will anyone do something Monday that they didn’t do Friday? If the answer is “they’ll feel closer,” that’s a nice afternoon, not a team event. Closeness that doesn’t change a client conversation is a memory, not a return.

Is anyone taking a real risk in the room? A volunteer with a live situation is a risk. A trivia round is not. The research says the risk is where the value is. Design for it.

Can you see the skill in the room? If you can’t name the behavior your people are practicing, there isn’t one. “Ask before you answer” is a behavior. “Collaboration” is a poster.

Lots of team building fails all three and gets rebooked anyway, because the alternative feels like no fun at all. It isn’t. The alternative is fun that happens to be useful.

What this looks like as an event

A Trust Skills Session runs two hours. Four to twenty-four people. In your room or off-site. One real situation from a volunteer, my two versions, then everyone practices on their own client. Every person leaves with one behavior to work on and a way to count it. I’ve written up what happens minute by minute if you want to picture your team in it, and there’s a look inside one session with an investment banking team if you want to see the moment the room turned.

If you’re planning an offsite, here’s how the session fits a full day without becoming the boring part of it.

Book the escape room too, if you want. Just don’t confuse it with development.

Firms don’t grow because their people got smarter. They grow because their people got trusted.