The Science of Leading

EducationBusiness

Listen

All Episodes

Google’s Team Secret: Why Interactions Beat Talent

This episode breaks down Google’s Project Aristotle and the surprising finding that team performance depends less on star talent than on how people interact. It explores psychological safety, why silence can hide problems, and the practical habits leaders can use to build stronger, more effective teams.

This show was created with Jellypod, the AI Podcast Studio. Create your own podcast with Jellypod today.

Is this your podcast and want to remove this banner? Click here.


Chapter 1

The surprising thing Google learned about team performance

Claire Monroe

Welcome to the Science of Leading. Edwin, I want to start with a finding that messes with a lot of leadership instincts: Google looked at 180 teams across engineering and sales, and the surprise was that the best team was NOT the one with the best individual talent. It was the one with the best interactions.

Edwin Carrington

Yes... and that distinction matters more than people think. Most leaders still build teams like they're drafting an all-star roster. Get the smartest engineer, the strongest seller, the high-potential operator. But Project Aristotle found that who was on the team mattered less than how those people behaved together. That is a very inconvenient truth for executives who love résumés.

Claire Monroe

Wait! 180 teams. That's a big enough sample that you can't just shrug it off as one quirky department. So the old fantasy of, like, "put all the A-players in one room and magic happens"... that really doesn't hold?

Edwin Carrington

Not reliably, no. Because brilliance can cancel itself out if people don't feel safe enough to use it. The big reveal was psychological safety, a term Amy Edmondson helped make central to management thinking. Very simply: can I speak up, admit a mistake, ask a naive question, disagree with you, and not pay a social price for it?

Claire Monroe

"Social price" is the phrase that lands for me. Because in a real workplace, people are doing that math constantly. If I challenge my boss in this meeting, do I look difficult? If I admit I missed something, do I look incompetent?

Edwin Carrington

Exactly. And when the answer feels risky, people get quiet. They withhold doubts, concerns, half-formed ideas. Which means the team loses information. And leadership, at bottom, is partly an information problem. If the truth can't travel upward, sideways, or across functions, the team will look polished right up until it fails.

Claire Monroe

That makes this almost counterintuitive in a painful way. Because the "best people" might actually be the most dangerous team if everyone's protecting status.

Edwin Carrington

Dangerous is not too strong a word. A room full of accomplished people can become a room full of performance. Everyone signaling competence, nobody exposing uncertainty. And then a team confuses confidence with coordination.

Claire Monroe

Is that why vulnerability keeps coming up in this conversation? Because I know some leaders hear that word and think, oh no, now we're doing feelings instead of performance.

Edwin Carrington

Yes, they do. But vulnerability in this context is practical. Matt Sakaguchi has a lovely example from group facilitation: he opens by sharing something personal first, which gives everyone else permission to stop performing. In organizations, the equivalent might be a leader saying, "I may be missing something here," or, "I got that wrong." Small sentence, enormous effect.

Claire Monroe

So the move isn't sentimental. It's operational. If the boss says, "I missed that," suddenly other people can say, "Actually, I think we're solving the wrong problem."

Edwin Carrington

That's right. Psychological safety is not about being nice all the time. It's about making candor possible. The team that can say, "I disagree," without humiliation will usually outperform the team that smiles, nods, and ships avoidable mistakes.

Chapter 2

How to measure a team that feels hard to measure

Claire Monroe

And this is where managers get twitchy, right? Because the next question is always: fine, but how do I measure THAT? Give me a dashboard, give me a score, give me a traffic-light system.

Edwin Carrington

Ah yes, the ancient managerial prayer: if it matters, surely it can be put in a spreadsheet. Project Aristotle did use surveys. People rated statements such as whether they felt safe expressing a different opinion, whether the team could depend on one another, whether they could navigate roadblocks. Useful, certainly. But the deeper signal was often in patterns of behavior.

Claire Monroe

Patterns like equal turn-taking?

Edwin Carrington

Yes -- equal turn-taking and social sensitivity. In strong teams, no one voice dominated every conversation. People read the room better. They noticed when someone wanted to speak, or when tension entered the meeting. It wasn't just what people said; it was whether the conversation itself had enough oxygen for multiple minds.

Claire Monroe

So let me try to say it back. A manager wants one clean number. But the better question is, who speaks in our meetings, who stays silent, who gets interrupted, and whether disagreement actually changes the discussion. Is that basically it?

Edwin Carrington

Almost. I'd add one more thing: whether people can raise a problem early. That's a crucial tell. If roadblocks only appear after deadlines slip, the team may have compliance, not safety.

Claire Monroe

That one stings. Because every leader has had that moment where they say, "Why didn't anyone tell me sooner?" And the uncomfortable answer is... they probably did the math and decided it wasn't safe or useful to tell you.

Edwin Carrington

Precisely. Or they believed you only wanted good news. Leaders teach teams what is discussable. If every hard conversation is punished -- even subtly, with a sigh or a defensive explanation -- people learn to bring you polished surfaces.

Claire Monroe

And polished surfaces are kind of the enemy here. They look efficient. They feel calm. Meanwhile nobody's saying, "This timeline makes no sense," or, "Sales promised something product can't deliver."

Edwin Carrington

Mm-hm. So yes, use the survey. Ask the questions. But don't worship the score. Let it start a conversation. "Do you feel safe disagreeing?" is useful. More useful still is, "Tell me about the last time you disagreed here. What happened?" That's where the truth lives.

Chapter 3

What actually improves team effectiveness

Claire Monroe

So if a leader's listening and thinking, okay, I buy it, what do I actually work on? Google landed on five factors: psychological safety, dependability, structure and clarity, meaning, and impact. I love that list because it's not vague culture wallpaper. It's pretty concrete.

Edwin Carrington

It is. And the order matters. Psychological safety creates voice. Dependability creates trust in execution. Structure and clarity answer the question, "Who does what by when?" Meaning answers, "Why does this matter to me?" Impact answers, "Does our work change anything beyond this room?" When those five are present, teams stop wasting energy on confusion and self-protection.

Claire Monroe

Let's make that practical. If I'm leading a team Monday morning, one move is modeling vulnerability -- saying, "Here's what I'm unsure about." Another is clarifying roles, because fuzzy ownership creates a lot of fake harmony. Everyone nods, nobody owns it.

Edwin Carrington

Yes. And I'd add normalizing dissent. Not occasional dissent -- normalizing it. You can literally ask, "What are we missing?" or, "Who sees this differently?" Then wait long enough for an answer. Leaders often ask for candor and then rescue the silence too quickly.

Claire Monroe

That pause is brutal. Five seconds feels like an hour.

Edwin Carrington

It does. But that five seconds may be the distance between performative agreement and real thinking. Another practical move is inviting quieter voices in without ambushing them. "Claire, we haven't heard from you yet -- what concern do you see?" Done warmly, that changes the room.

Claire Monroe

And then meaning and impact -- those seem softer, but I don't think they are. If someone knows exactly how their work affects a customer, or a patient, or even just the next team downstream, they care differently.

Edwin Carrington

They do. People give more when the work is legible. Not just assigned, legible. And that's why team effectiveness is not a personality trait. It's not, "We're just a high-performing bunch," or, "We're awkward but brilliant." It's designed. Daily. In how meetings are run, how mistakes are handled, how roles are defined, how dissent is treated. Which means every leader is shaping team performance long before the quarterly numbers arrive.

Claire Monroe

Yeah... and that's the part I keep coming back to. If effectiveness is designed, then every awkward meeting is giving you data. Not about whether your people are good enough — about what your team has been taught is safe to say. Edwin, if there’s one practical action step a team can take this week, what should it be?

Edwin Carrington

Start small and make it real. In your next team meeting, ask one question: “What are we not saying out loud?” Then go around the room and let each person answer once without interruption or debate. Just collect what you hear. After that, choose one item the team can act on before the week is over. That simple rhythm tells people their voice matters, and it turns a vague conversation about culture into a visible team habit. And if you want a practical way to measure and improve team performance, visit OAD.ai and try it for free. It’s a simple next step if you want to turn team conversations into real insight. Thanks for listening to The Science of Leading.

Claire Monroe

Bye for now — and we’ll see you next time.