Psychological Safety: The Invisible Glue That Holds Great Teams Together

This is Part 3 of a series. Start at the beginning with Engineering Team Success: A Relatable Story of Process Change.

▼Table of Contents

• What Is Psychological Safety?
• The Proof Is in the Data
• What It Is NOT
• What Destroys It
• What Jane Did Next
• How to Build It as a Leader
• If You're Leading Leaders, Add These
• The Takeaway

Jane had the right process — part one took care of that. She had the right people — part two covered that too. Marcus was gone, the team had exhaled, and on paper, things were looking up.

But Jane noticed something.

The room was still quiet.

Not the good kind of quiet — the focused, flow-state quiet that engineers dream about. Not chaos-quiet or checked-out-quiet either. Something subtler, and harder to name. You can feel it. Engineers were hesitating before they spoke in planning sessions. Junior team members were still prefacing their ideas with "this might be a dumb question, but..." Disagreements were swallowed instead of surfaced. Risks went unvoiced until it was too late.

The process was healthy. The people were right. But the team wasn't flying yet.

What Jane was missing was the invisible glue — the foundation beneath every other leadership initiative — and without it, all the right people and all the right process in the world will still underperform. We've all heard the phrase thrown around, it's called psychological safety. And it changes everything.

What Is Psychological Safety?

The term was coined by Harvard Business School professor Amy Edmondson, who has spent decades studying teams and why some succeed spectacularly while others fall flat despite every advantage. Her definition is simple and worth pondering over:

It is not about being comfortable. It is not about being friends with your teammates. It is not about a workplace free of challenge, pressure, or hard feedback.

Psychological safety is about risk and fear response. Specifically, the belief that raising a concern, admitting confusion, sharing a half-baked idea, or calling out a problem nobody wants to talk about will not result in embarrassment, rejection, or retaliation. Engineers already quietly fighting imposter syndrome don't need another reason to stay silent — and in a low-safety environment, they have plenty.

In a safe environment, people speak up and ideas flow. When not, they go quiet. And quiet teams miss things. Quiet teams have preventable failures. Quiet teams lose their best ideas before they ever reach the surface. And often, it's the bad ideas that inspire the good ones.

The Proof Is in the Data

If you still need convincing, let me point you to good ol' Google.

In 2012, Google launched Project Aristotle — one of the most comprehensive studies ever conducted on what separates high-performing teams from the rest. They studied 180 of their own teams over several years, measuring dozens of variables: skills mix, personality types, seniority, tenure, team size, structure — everything you'd expect to matter.

The result? Almost none of the expected factors explained the difference.

What did matter — by a significant margin — was how team members felt around each other. Whether people believed it was safe to take risks without fear of judgment or retaliation.

Let that sink in. One of the most analytically rigorous companies on the planet — a company that could measure nearly anything (and probably does) — concluded that the most critical driver of team performance is fundamentally human.

Google employed an army of data scientists, statisticians, and researchers who could measure nearly anything. Their answer was essentially: be nice to each other and don't make people feel dumb. Sometimes billion-dollar research programs confirm what kindergarten teachers have known since forever.

This isn't soft. This is strategy.

What It Is NOT

Before going further, let's clear up the misconceptions — because psychological safety is widely misunderstood and, honestly, sometimes used as an excuse to avoid hard conversations or dodge accountability.

Psychological safety is not:

• A guarantee that all ideas are good ideas. (They're not.)
• Permission to skip difficult feedback. (You still have to give it — kindly and directly.)
• An absence of standards or accountability. (Accountability, done right, is a sign of respect.)
• The same as team harmony or being "nice". (The most psychologically safe teams often disagree loudly... and often, it is a lot of fun!)

The distinction that matters: psychological safety is about the interpersonal environment, not the performance expectations. You can hold a team to incredibly high standards — you should — and still create a space where people feel safe to say "I don't understand" or "I think we're making a mistake."

In fact, the highest-performing teams tend to have both — high standards and high psychological safety. That combination is where the magic lives.

What Destroys It

Here's the uncomfortable truth: most leaders are eroding psychological safety without realizing it.

Not through malice. Through habits and ignorance.

• Dismissing ideas quickly. When someone shares a half-formed thought and it gets shot down in seconds — without curiosity, without exploration — the lesson everyone in the room absorbs is: don't share half-formed thoughts. And that's exactly where the best ideas tend to live.
• Reacting badly to bad news. If the messenger gets punished, there will be no more messengers. And the problems that don't get surfaced are the ones that quietly grow into crises.
• Sarcasm or condescension. Even once. Even from a leader who thinks they're being lighthearted (I am guilty of this). The team remembers.
• Rewarding the loudest voice, not the best idea. When dominance wins over merit, quieter team members stop competing. The quieter ones are often the ones with the most considered perspective.
• Making someone feel foolish for not knowing something. The words "obviously" or "we've been over this" do enormous damage — especially to newer team members who are still learning what's safe to ask.

(A special note for the technically brilliant architect who answers every junior question with a long pause and a slow nod, as if computing whether the question deserves a response: don't be Marcus.)

One of the most telling signals that psychological safety is low? Retrospectives that go suspiciously smooth. When people say "everything's fine" and you know it isn't. When the real conversations happen in hallways (or virtual hallways) and one-on-ones instead of the room where decisions are being made.

Sound familiar?

What Jane Did Next

Jane recognized the pattern. Months of Marcus had quietly trained the team to keep their heads down. Ideas that weren't immediately airtight stayed internal. Risks that needed surfacing were buried. The team was functional — but not yet free.

She started with herself.

In the next planning session, Jane shared a mistake she had made — a decision she had gotten wrong the previous quarter, what it cost the team, and what she had learned from it. Not with fanfare. Not with false humility. Straightforwardly and without self-protection.

She introduced a standing question at the start of every retrospective: "What is something you've been hesitant to bring up?" Not "what went wrong" — that question invites polished post-mortems. She wanted the raw, uncomfortable, half-formed stuff. The things people almost said but didn't.

She started reacting to bad news differently. When an engineer surfaced a problem — a risk spotted too late, a bug in production, a missed deadline — her first response became: "Thank you for telling me early. Let's figure this out." She meant it. Over time, people believed her.

Slowly, the room got noisier. In the best possible way.

How to Build It as a Leader

Psychological safety is not built in a team offsite or a culture deck. It is built — and rebuilt — in small moments, every single day.

• Model vulnerability first. Share your own uncertainty, mistakes, and open questions. You set the ceiling for what is acceptable to bring into the room. If you present yourself as always knowing the answer, the team will do the same — and everyone loses.
• Get curious before you get critical. When someone shares an idea you don't immediately agree with, lead with questions before conclusions. "Help me understand your thinking" does more for a team culture than a hundred values-statement slides.
• Separate learning from blame. When something goes wrong, focus on what happened and why, not who is responsible. Blame finds the person. Learning finds the system. One of them builds a better team. Learn from itand move forward.

• Actively invite dissent. Don't just make space for disagreement — solicit it. "What are we missing?" "Who sees this differently?" "What's the strongest argument against this plan?" These questions signal that challenge is not only allowed, it is valued.
• Protect the people who speak up. Nothing builds psychological safety faster than watching someone raise a concern and having leadership take it seriously. And nothing destroys it faster than watching someone be vulnerable and get ignored — or worse, penalized for it.
• Leaders speak last. If you share your view first, people often assume your idea is best or that theirs is wrong. Hierarchy affects how people respond. State the problem, ask for input, listen, and hold your own opinion until others have spoken. You’ll hear more, and often better ideas.

If You're Leading Leaders, Add These

The tactics above apply to any manager. But if you're a VP or senior leader managing other managers, there are additional levers worth pulling — because the posture you model with your EMs is the posture they'll model with their ICs. It travels downward.

• Model vulnerability with your managers, not just your teams. If your EMs never see you uncertain, confused, or wrong, they'll perform the same invulnerability for their ICs. The posture travels downward.
• Reward your managers for surfacing problems, not just solving them. If you only celebrate fixes, you'll train your leaders to hide fires until they have the extinguisher ready. By then, the fire is usually bigger.
• Make it safe to escalate bad news fast. Your highest leverage act as a senior leader is making it genuinely safe for your managers to bring you bad news early. Say it. Mean it. Prove it — repeatedly, because it only takes one bad reaction to undo months of trust.
• Run skip-levels that aren't interrogations. The goal isn't to catch your managers in something — it's to stay connected to the reality your managers are navigating. Be the kind of leader people are glad to have as a skip-level.

The Takeaway

Jane now had the full picture. Process gave her team a foundation. The right people gave her the horsepower. But psychological safety? That was the fuel.

Without it, even the most skilled team underperforms — not because they lack the answers, but because no one feels safe enough to share them. Quietly brilliant teams are still quiet. And quiet teams are leaving enormous value untapped.

The good news is this one is entirely within your control as a leader. You don't need budget approval to make someone feel heard. You don't need an org restructure to respond to bad news with curiosity instead of defensiveness. You don't need a new tool to ask a room, "What are we missing?"

You just need to mean it. Every time.

Build the environment where your team can do their best thinking. Protect the people willing to be honest. Reward the behavior you want to see more of. And keep building — because psychological safety isn't a checkbox. It is a daily practice.

Jane got there. Her team started to soar. And so will yours. The work is quiet, daily, and cumulative — but there is no leadership investment that compounds faster.

---

Part 4 of this series will tackle the question Jane faced next: how do you keep a high-performing team performing when the pressure is at its highest — leading through chaos.