Building Psychological Safety in Teams: A Practical Guide

Psychological safety is the shared belief within a team that it's safe to take an interpersonal risk — speaking up with a half-formed idea, admitting a mistake, asking a question that might sound obvious, or disagreeing with someone more senior — without being embarrassed, dismissed or quietly penalised for it. The term was defined by Harvard Business School researcher Amy Edmondson in the 1990s, and it has since become one of the most consistently replicated predictors of team performance, notably identified by Google's own internal research as the single biggest factor separating its highest-performing teams from the rest.

It's worth being precise about what it isn't. Psychological safety isn't comfort, and it isn't the absence of conflict or challenge. The most psychologically safe teams often have the most direct conversations — people say what they actually think, including when it's uncomfortable. What's different is that the discomfort comes from honest disagreement rather than from fear of how it will be received.

Why it predicts performance so reliably

Most of the value a team produces depends on information that isn't yet visible to the leader: the early warning sign of a project going wrong, the idea nobody's mentioned because it might sound naive, the mistake that could be corrected cheaply now or expensively later. In a team without psychological safety, that information stays hidden until it can't be ignored — which is usually well after it would have been useful to know. In a safe team, it surfaces early, when there's still time to do something about it.

There's also a learning effect. Teams that can openly discuss what went wrong actually improve; teams that can't tend to repeat the same mistakes because nobody names them clearly enough to fix.

What erodes it

Psychological safety is easier to damage than to build, and the damage often comes from small, ordinary moments rather than dramatic ones. A leader who visibly deflates when challenged. A mistake met with blame rather than curiosity about what happened. An idea that gets talked over before it's fully out. None of these need to be intentional — they just need to happen enough times for people to quietly conclude that speaking up isn't worth the risk. Once that conclusion sets in, it's remarkably sticky, because the safest response to uncertainty is always silence.

What actually builds it

Respond well the first few times it's tested. How a leader reacts the first time someone admits an error or pushes back on a decision sets the tone far more than anything said in the abstract about "an open culture". React with curiosity rather than defensiveness, and people notice.

Ask genuine questions and mean it. "What am I missing?" or "What's the strongest argument against this?" only works if the leader visibly takes the answer seriously, including when it's inconvenient. Asked as a formality, it teaches people the opposite lesson.

Separate the mistake from the person. Treating an error as information to learn from, rather than evidence of someone's competence, makes it far more likely that the next mistake gets reported early rather than hidden.

Model uncertainty yourself. A leader who can say "I'm not sure" or "I got that wrong" gives everyone else permission to do the same. Teams tend to mirror the risk tolerance their leader displays.

Make space in meetings deliberately. Quieter voices and more junior people often need an explicit invitation — a direct question, a pause held a beat longer than feels natural — rather than an open floor, which tends to go to whoever is most comfortable speaking first.

The balance with high standards

The teams with the strongest psychological safety are rarely the most relaxed about quality — they're often the most demanding, because people trust that challenge is about the work and not about them personally. Edmondson's own framing pairs psychological safety with accountability: safety without high standards drifts into comfort, and high standards without safety produces anxiety and silence. The combination is what produces genuine performance.

Where this shows up in leadership development

Most of what builds or erodes psychological safety happens in a leader's everyday reactions — in the moment, often under some pressure — which makes it a genuinely hard thing to change through intention alone. It's exactly the kind of pattern that's easier to shift with structured reflection and honest outside feedback on how those reactions actually land. If building this in your own team is something you're working on, a short conversation is a useful place to start.