Why most team surveys fail, and five fixes

Team surveys rarely fail at the point of collecting answers. The questions go out, most people fill them in, a report is produced. They fail before that, in how the survey is set up, and after it, in what nobody does with the result. We have watched this happen in large organisations and small ones, and the causes are remarkably consistent.

Here are the five we see most often, and what to do about each one.

Five ways a team survey fails

1. There is no anonymity floor

People answer honestly only when they believe the answer cannot be traced back to them. A survey in a team of four with a breakdown by role is a survey with names on it, whatever the cover letter says. Once someone has been identified from a "confidential" survey, that team is done with surveys for years.

The stakes are higher than the survey itself. Amy Edmondson's 1999 study of work teams found that psychological safety, the shared belief that a team is safe for interpersonal risk-taking, goes together with learning behaviour, which in turn relates to team performance. A survey that quietly punishes candour damages the thing it is trying to measure.

2. Averages hide the split

A team score of 3.0 on "we resolve conflict constructively" can mean everyone chose 3, or that half the team chose 5 and the other half chose 1. The first team is lukewarm. The second is in a cold war. The average cannot tell them apart, and most survey reports show only the average.

The most damaging version of this is a high score with low agreement: the leader and a few others rate something as great while the rest of the team quietly disagrees. Nothing in a green dashboard tile warns you.

3. Nobody owns the result

Organisation-wide surveys are commissioned by HR, reported to the executive team and cascaded to managers as a slide deck. By the time a team lead sees their numbers, the result belongs to everyone and therefore to no one. The lead did not ask for it, did not choose the questions, and has no obligation to act.

4. There is no follow-up conversation

The most common fate of a team survey is a thank-you email. The second most common is a 40-slide readout that nobody remembers. Both tell the team that answering was pointless, which is why response rates drop with every round.

5. Too long, too rarely

The annual 80-question survey has two problems. Nobody wants to fill it in, so answers get sloppy halfway through. And a year is far too long between measurements: by the time the result arrives, half the context has changed and you cannot tell whether anything you did made a difference.

The five in one place

None of these fixes needs a budget. They need a decision about how the survey is run. If you are choosing questions, our guide to team survey questions covers what to ask and what to leave out, and the psychological safety survey guide goes deeper on the first point.