Build Psychological Safety With Small Risks

If you want to know how to build psychological safety on a team, start smaller than a kickoff speech. A leader can announce that every idea is welcome, and most teams will nod and then stay quiet, because people judge safety by what happens to the first person who tries something and misses. Psychological safety gets built from small interpersonal risks that a team takes together and that turn out fine. This post is for L&D leads, HR partners, and team leaders who have already said the words and are still waiting for someone to speak up.

What does psychological safety actually mean?

Psychological safety is a belief, shared across a team, that the team is "safe for interpersonal risk taking." Harvard Business School professor Amy Edmondson defined it that way in a 1999 study of 51 work teams in a manufacturing company, published in Administrative Science Quarterly. Her paper found that teams with more psychological safety showed more learning behavior, and that learning behavior accounted for the link between safety and team performance.

It was a field study at one company, so it shows the pattern. It does not tell a manager what to do on Monday morning, which is the question most readers arrive with.

Why doesn't announcing safety at a kickoff work?

A statement is weak evidence, and people update their beliefs from evidence. Everyone on a team is running a private estimate: if I say the half-formed thing, what happens to me? A kickoff slide answers with words. The first meeting afterward answers with an event. If the first person to admit a mistake gets a raised eyebrow, the slide is forgotten by lunch.

That is our position, and you can argue with it. The practical consequence is that safety needs a sequence of small events, each one cheap, each one ending with the team building on the risk instead of punishing it. One brave moment does not do it. Repetition does.

Is there research on building it on purpose?

A little. One study we found tests a routine instead of a speech. Researchers at Bern University of Applied Sciences, working with Zurich University of Applied Sciences, gave 50 teams of 4 to 22 people at Swiss companies one email every Monday for six months. Each email held a roughly 15-minute exercise: practice a new behavior, reflect on one, or work something out together. Twelve comparison teams only filled in questionnaires. According to the Bern press release, surveys every two months showed significant gains in psychological safety, performance, and innovative strength for the trained teams, and no significant change in the comparison group.

Read that with care. A press release does not give effect sizes, the survey instruments, or whether teams were assigned to groups at random. It is a promising result, not a settled one. What we take from it is the shape: short, weekly, done with the team, and spread over months.

What does a small, rehearsed risk look like?

In a typical freestyle+ improv-based psychological safety training session, the risks start tiny and grow. Each one is a rep of Bravery, the B in the freestyle+ BOLD Mindset (Bravery, Optimism, Listening, Discovery).

The room opens with agreements: surprise yourself, make everyone look and feel great, listen with curiosity, take care of yourself, and serve the moment. The agreements set the rules for the exercises, and the exercises are where people find out whether the rules hold.

One-word story

In pairs, people build a sentence one word at a time, alternating. Nobody can prepare. In a typical session, someone says a wrong or odd word within the first minute, and nothing bad follows. For a person who dreads being visibly wrong at work, that is a low-cost first piece of evidence.

"Yes, but" versus "yes, and"

One person offers a simple work idea, such as using AI to improve onboarding. The partner answers first with "yes, but," then the pair restarts and the partner answers with "yes, and." Offering an idea is the risk. The debrief asks how the two conversations felt different, and people usually notice that the second conversation made them more willing to keep going.

One small thing back at work

The session ends with each person choosing one small experiment to try with their team that week. The aim is to carry the habit out of the room while it is still small.

On teams adopting AI, the same risk shows up as trying a half-formed prompt where colleagues can watch. freestyle+ AI adoption training is built around that problem.

How can you try this without hiring anyone?

Four moves cost almost nothing, and a team can start them this week.

• Go first, and go small. The most senior person in the meeting shares a draft they are unsure about, or a call they would make differently, and asks the group to add to it before anyone critiques it.

• Build before you judge. For one meeting, start every response to an early idea with "yes, and." Critique is welcome afterward. It just has to come second.

• Keep it weekly and short. Fifteen minutes on the same day each week is easier to sustain than a half-day retreat once a year, and it matches the cadence in the Bern study.

• Watch the second risk. The first person who speaks up shows the team what is possible. The second person is the real test: if they speak up and nothing bad happens, the belief starts to move.

Who is this not for?

Teams where speaking up has actually been punished. Exercises do not outweigh a manager who retaliates, and that problem needs coaching or an HR process before it needs a workshop.

It is also not for anyone expecting a one-time fix. A workshop gives a team a first set of reps and a shared vocabulary. Safety holds only if the habit continues afterward.

One more limit, stated plainly: none of the research above tests improv exercises specifically. The case for improv is mechanical. It is a fast way to run many cheap risks with a partner whose job is to build. That is our reasoning, not a finding.

What is a good next step?

If a version of this would help your team, start with a free 15-minute call. We will ask what has gone quiet, and tell you plainly whether a workshop fits. You can also read what teams say afterward on the workshop reviews page.