Running a staff survey takes an afternoon to set up and about two weeks to run. Most of what decides whether it is worth doing happens before the first link is sent and after the survey closes. This guide goes through it in order.
1. Decide what it is for
Write down, in a sentence, the decision the survey should inform. "Where should we focus our wellbeing work next year?" is a purpose. "To get a sense of how staff feel" is not, and leads to a report nobody acts on.
The purpose decides the survey:
- How staff are, and what to fix first: a wellbeing and workload survey. See how to run a staff wellbeing survey.
- Which tasks to stop, cut or simplify: a workload survey. See how to run a teacher workload survey.
- How your school compares with England: the Teacher and leader survey with national comparison, which asks the Department for Education's questions word for word.
- Why people leave: the Staff exit survey, run all year.
2. Decide who to ask
Ask everyone the decision affects. For most surveys that means all staff, not only teachers. Support staff often see problems teachers do not, and the Support staff voice survey asks about their roles in their own terms.
One exception: the national comparison for teachers and leaders is with teachers and leaders only, because that is who the Department for Education surveyed. Run it with them, and ask support staff separately.
If you want results by group, such as teaching and support staff, or by phase in an all-through school, keep the groups large. A group's results are shown only when 10 or more people in it answered.
3. Choose when
Pick two weeks away from reports, exams and other peaks, and the same point in the year each time, so next year's results can be set beside this year's. Set the closing date before you open it: the survey closes on its own at that time.
4. Tell staff before it opens
Staff answer honestly when they know three things: what it is for, who will see the results, and what will happen next. Say all three, in a briefing and in writing.
- Give staff the privacy notice. The data protection pack has a staff privacy notice template to adapt.
- Say plainly that answers are anonymous, and how. Our anonymity page sets out every rule and is written to be shared with staff.
- Say when they will hear the results, and keep to it.
5. Send it
- Personal links by email suit most staff: each person gets their own link, so no one can answer twice, and the email address is deleted as soon as the link is used.
- An open link works where not everyone has a school email address, such as some support staff. Share it in a staff room, on a poster with its QR code, or in a message.
- Give time to answer: ten minutes at the end of a briefing or meeting, so answering does not come out of staff's own time.
6. Watch the response, and remind once or twice
You can see how many have answered while the survey is open, but never who. A reminder goes only to people sent a link who have not answered yet, with a fresh link, so no one who has answered is chased. We suggest one reminder after a week, and one two days before it closes.
There is no target response rate that makes a survey valid. What matters is that the people who answered are like the people who did not. A higher response makes that more likely. A low response in a small team can also mean its results are not shown at all, because fewer than 10 answered.
7. Read the results
When the survey closes, the headline result is ready, with strengths and priorities. For each question, look at three things:
- The figure itself, and how many answered it.
- England's figure, where one is published, and whether the gap is more than chance. How the comparison works explains the test.
- Last time, if you ran the same survey before: each result is compared with the last run, and a change is called only when it is larger than chance would explain.
Read the comments last, once you know what the figures say. They often explain the figures.
8. Share the results with staff
Share them within a few weeks, as promised, including what was hard to hear. Staff will judge the next survey by how you handled this one. The full report includes briefing slides for a staff meeting, and any comment can be removed before the report is shared.
9. Act, then ask again
Choose two or three priorities, not ten. Say what will change, who is responsible and by when. Then run the same survey at the same point next year, and see whether it did.
What it costs with Benchmark.Pulse
Running a survey is free, and so is the headline result when it closes. The full report, with every chart, comments grouped into themes and briefing slides, is £149. See pricing, or read a sample report first.