What Business Leaders Get Wrong About Psychological Safety

Psychological safety has become one of the most cited concepts in workplace culture over the last decade. After Google’s Project Aristotle research identified it as the most important factor in team performance, the concept moved from academic research into every major corporate training curriculum and culture deck in the country.

And yet most of the leaders I’ve seen try to implement it get it wrong. Not because they lack good intentions. Because they misunderstand what it actually is and how it actually gets built.

What psychological safety actually means

Amy Edmondson at Harvard Business School has done the foundational research on this. Her 1999 paper defined psychological safety as “a shared belief held by members of a team that the team is safe for interpersonal risk-taking.” The key phrase is “interpersonal risk-taking.” It’s not about comfort. It’s about whether people believe they can speak up, push back, ask a dumb question, or admit a mistake without being punished for it.

That’s a specific thing. It’s not the same as having a friendly team culture. It’s not the same as employees being happy. You can have a team where people genuinely like each other and still have very low psychological safety, because the moment something goes wrong or someone disagrees with the boss, the interpersonal risk of speaking up feels too high.

Mistake one: confusing comfort with safety

The first mistake leaders make is optimizing for comfort rather than safety. They hold team lunches, celebrate birthdays, and run engagement surveys. Those things aren’t bad, but they don’t create psychological safety. What creates psychological safety is what happens when someone raises a problem.

If a team member tells you that a project is behind schedule and your response, even a subtle one, signals that you’re disappointed in them rather than grateful for the early warning, you’ve just trained everyone on that team to not tell you when projects are behind schedule. That’s the opposite of what you need.

I’ve seen this play out across businesses of every size. When I started building from a 10-person crew at Penebaker Enterprises to a large multi-state roofing operation, the teams that outperformed weren’t always the ones with the best morale scores. They were the ones where problems got surfaced fast, where people felt safe enough to say “I think we’re making a mistake here” before it was too late to do anything about it. The comfort-focused teams often had higher engagement scores and slower error detection. The safety-focused teams caught problems early and fixed them before they compounded.

Mistake two: thinking it’s a top-down rollout

The second mistake is treating psychological safety like a policy you can announce and implement. I’ve watched companies roll out “speak-up culture” initiatives with all-hands presentations, intranet pages, and posters on the wall. And then a few months later, someone gets publicly embarrassed for raising a concern in a meeting, and the whole initiative is dead. People don’t forget those moments. They file them away.

Psychological safety is built through behavior, not announcements. Specifically, it’s built through the repeated pattern of how leaders respond when something goes wrong or when someone disagrees. Every one of those moments is a data point. Teams are constantly calibrating whether it’s safe to be honest, based on what they’ve observed.

The pattern that builds it: leader acknowledges problem without blame, asks what can be learned, focuses energy on the path forward. The pattern that destroys it: leader visibly frustrated, asks whose fault it is, punishes the messenger even slightly. You only have to see the second pattern a few times to stop bringing problems forward.

Mistake three: treating it as separate from accountability

This is the most common confusion I see among leaders who are trying to get this right. They think psychological safety means being soft on performance. That if you create an environment where people feel safe to fail, you’re lowering standards. Edmondson addressed this directly. Psychological safety is not the same as low accountability. In fact, the research suggests the combination of high safety and high accountability is what produces the highest performance.

Think about it operationally. If someone on your team knows they’re going to miss a deadline, do you want them to tell you this week or next month when it’s too late? The answer is obvious. But they’ll only tell you this week if they believe the response to “I’m going to miss this deadline” is “okay, what do you need” rather than “why are you missing this deadline.”

Accountability and safety are not opposites. High-performing teams hold people accountable for outcomes while maintaining safety for honesty about the process. The distinction is between punishing failure and addressing it. You can expect excellent performance and also create an environment where people tell you when something is off track. Those things reinforce each other when you get the model right.

What actually builds it

From watching this work and fail across different organizations, here’s what I’ve seen make the difference:

The leader goes first. If you want your team to admit mistakes, admit yours. If you want them to share concerns, share yours. The permission structure flows from the top. When I’ve publicly acknowledged a call I made that didn’t work, something shifts in the room. People start sharing things they weren’t sharing before. This isn’t a one-time move. It’s a pattern you have to maintain.

Reward the early warning, not just the outcome. One of the most powerful things a leader can do is visibly thank someone for surfacing a problem early, even when the news was bad. “Thank you for telling me this before it got worse” does more to build psychological safety than almost anything else. It directly signals that honesty is valued.

Ask genuine questions rather than performing them. Leaders often ask “what do you think?” without really wanting an answer that challenges their view. People can tell when a question is an invitation and when it’s a formality. If you ask and then dismiss the answer, you’ve done the opposite of what you intended. Ask when you’re actually ready to hear a response that pushes back, and respond in a way that shows you heard it.

Separate the post-mortem from the performance review. After-action reviews work when they’re genuinely about learning, not judgment. If team members suspect that candor in a post-mortem will be used against them in their next review, they won’t be candid. Keep the two conversations functionally separate.

The operational case, not just the cultural one

I want to make the operational case for this clearly, because some leaders hear “psychological safety” and think it’s primarily a values conversation. It’s not. It’s an information problem.

Leaders make decisions based on information. The quality of those decisions depends on the quality of the information they receive. If your team doesn’t feel safe telling you bad news, you will consistently get incomplete information, and your decisions will be correspondingly worse. Teams with high psychological safety give leaders better information, faster, because they’re not filtering what’s safe to say.

That’s the core of why Google’s research found it was the top predictor of team performance. It’s not that safe teams feel better about their work. It’s that they catch problems earlier, challenge bad ideas before they get expensive, and generate better solutions because more people are actually contributing to the problem-solving.

If you want to build teams that outperform, start here. Not with the culture poster. With the next time someone brings you bad news. The way you respond in that moment is your real policy.

For more on how this connects to building resilient teams under pressure, that’s where I’d point you next.

---

Want the whole story, not just this piece?
Read my story

Last updated: June 28, 2026