Anonymous employee surveys in small teams

"Anonymous" is easy to promise in a company of 2,000. In a company of 12, it's harder. If one person works the evening shift and someone writes "evenings never get told anything," everyone knows who wrote it.

Your team knows this too. In research from the software company Visier, 47% of employees said they feel pressure to hold back honest feedback in workplace surveys. The most common reason, at 37%, was not believing the survey was really anonymous (NJBIA, reporting Visier's research).

So in a small team, you have to protect anonymity on purpose. Here's what gives people away, and how to protect them.

What gives people away

• Small groups. Ask "Which team are you on?" when two people work in accounts, and you've just identified them.
• Written comments. The issue someone raises, the words they use, the way they always phrase things. In a small team, a quote can be as good as a signature.
• Knowing who answered. If you can see who hasn't answered yet, you can also work out who has. Everyone knows that, and it changes what they write.
• Timing. If results show up as they arrive, the first answer that comes in at 7am is easy to place.
• The form itself. Plenty of free forms record email addresses or names. In Google Forms and Microsoft Forms it's a setting, and it's easy to leave on by mistake.

Six rules that protect anonymity

1. Wait for enough answers. Most survey tools hide a group's results until enough people have answered. Culture Amp's default is 5 (Culture Amp). Workleap Officevibe shows multiple-choice results from 3 people, but holds back written answers until 5 (Workleap). In a team under 25, 3 is the lowest number worth using, and only for the whole team's totals, not for small groups.
2. Don't split a small team. One set of results for the whole team is safer than five tiny groups. If a group has fewer than five people, merge it with another one.
3. Share themes, not quotes. "A few people aren't sure what client information is OK to use" tells you what you need to know. The exact sentence doesn't add much, and it can point to one person. Only share a theme if more than one person raised it. A theme from one person is a quote in disguise.
4. Don't chase anyone. Send reminders to the whole team, one or two at most, then stop. The moment you ask one person "did you fill it in?", everyone hears the survey isn't anonymous.
5. Don't ask for details you don't need. Age, start date, job title. In a team of 15, any two of those can point to one person.
6. Say what you'll see before the first question goes out. Then stick to it. One broken promise ends honest answers for a long time.

Anonymous or confidential?

These two words get used as if they meant the same thing. They don't.

Most online surveys are confidential, not strictly anonymous, because the tool has to know who answered to stop reminding them. That's fine, as long as you describe it accurately. "I'll never see who said what" is a promise you can keep. "Completely anonymous" might not be.

What not to do with the answers

Don't guess who wrote something, not even as a joke. "That has to be Dave" in a team meeting tells everyone the survey isn't safe.

Don't reply to a comment as if one person wrote it. Reply to the theme, to the whole team: "Some of you said the new rules aren't clear. Here's what I'm changing." (More on that reply in Why your team stopped answering your surveys.)

WeeklyTake is built around these rules. You see team totals, never individual answers. Nothing shows until at least three people have answered. Written answers become themes, not quotes. And you can't see who answered or who skipped. That makes answers much harder to trace, but in a very small team no tool can make it impossible. Keep the questions general and the groups big.