Running a survey
How often should you measure psychological safety?
Often enough to see what changed, and never so often that nothing could have. On the interval, the signs you have it wrong, and what measuring again shows.
Less often than most survey tools would like, and more deliberately. A psychological safety survey is an intervention: it surfaces things, gives people words for them, and sets an expectation that something will happen. The right interval is the one that lets something happen.
Measure, act, then measure again
The first survey gives a team a shared picture and a short list of things to try. The second shows what changed after the team tried them. Run the second before anything has had time to change and you have measured the weather, not the climate; leave it so long that nobody remembers the first and you have two snapshots instead of a story.
For most teams that puts the gap at three to six months. Three if the team is working through a specific, visible change and wants to know whether it landed. Six if the first survey started a broader conversation that needs time to turn into habits. Teams change, and six months on, a survey shows movement rather than a snapshot.
Signs you are measuring too often
Nothing was tried between surveys. If the team didn’t act on the last report, the next one will say the same thing, and people will notice that answering changed nothing. That teaches futility, which is the opposite of what you wanted.
Response rates are falling. Survey fatigue is real, and the people who stop answering first are rarely the ones who feel safest.
People talk about the score. When a team starts managing the number rather than the conditions, the measurement has become a target. Slow down and talk about the spread instead.
Signs you are measuring too rarely
A new lead, a merger, a reorganisation, a serious incident or a run of departures each change a team’s conditions. An annual survey will catch none of them in time to be useful. If the team has changed, measure; don’t wait for the calendar.
What the second report shows
Surveys of the same team join up automatically. The second report carries a section on what changed since the first, theme by theme, written as a prompt for a conversation about what happened in between rather than a scoreboard. One caution comes with it: scores on the who’s-in-the-room statements can fall as awareness improves, because people start noticing a gradient they had stopped seeing. A fall there after a good conversation is often progress.
Keep the statements you want to compare the same between surveys, and change the ones that no longer fit. The report compares like with like and leaves the rest.
Across an organisation
Resist synchronising every team to the same week so that the results can be lined up. Lining them up is benchmarking by another name, and it takes the timing decision away from the people who know when their team is ready. Let teams measure when they can act on the result, and pool the patterns, not the scores.
What it costs to measure again
With Measure, nothing extra. The plans are priced by team, not by survey, so a team can measure as often as it should without the bill deciding the interval. The plans.
Questions
How often should a team run a psychological safety survey?
Usually every three to six months: long enough for the team to act on the last report and for something to change, short enough that the story holds together. Measure again sooner if the team’s conditions change, with a new lead, a reorganisation or a serious incident.
Is a quarterly psychological safety survey too often?
Not if the team acts between surveys and response rates hold. It is too often if nothing was tried since the last one, if fewer people answer each time, or if people have started talking about the score rather than the conditions.
Should every team in an organisation survey at the same time?
No. Synchronising teams so the results can be lined up is benchmarking by another name, and it takes the decision away from the people who know when their team is ready to act. Let teams choose their moment and pool the patterns, not the scores.
What if the score goes down the second time?
Read the spread before the number, and look at the who’s-in-the-room statements. Scores there often fall after a good conversation, because people start noticing a gradient they had stopped seeing. A fall can be awareness improving, and the second report is a prompt for a conversation about what happened in between.