Runs on PeoplePlaybooks / Measures · 3 min read
Surveys that change something
A survey is a promise. Asking people what's wrong and then doing nothing is worse than not asking.
The problem
“We run engagement surveys every year and nothing changes”
In the model
In 30 seconds
- Every survey is a promise to act. Break it twice and people stop answering honestly.
- Short and often beats long and annual. Five questions a month tell you more than fifty once a year.
- The number starts the conversation. The team finishes it.
The situation
The annual engagement survey lands. Response rates are falling. The results arrive three months later as a dashboard, and the action plan is "improve communication". Next year, the same scores.
What's really going on
Surveys fail for people reasons, not statistical ones. People stop telling the truth when nothing visible happens, when they fear being identified, or when results come back too late to matter. Meanwhile managers read the scores as a judgement on them, and get defensive instead of curious.
What to do
- Decide what you'll do with it before you ask. If you can't act on a question, don't ask it.
- Go short and frequent. A monthly pulse of three to five questions. Keep two or three the same every time so you can see trends.
- Share results within a week, with the team, unedited. Good and bad.
- Ask the team what the numbers mean. "Trust dropped this month. What's behind that?" The score tells you where to look. People tell you why.
- Pick one thing, and close the loop. "You said X. We did Y. Here's what changed." Say it out loud.
- Protect anonymity properly. Never show results for groups of fewer than five people. Never try to work out who said what.
- Combine with hard data. Pair survey scores with things like sickness, overtime, attrition and delivery. When they disagree, investigate.
The questions worth asking
| Question | What it tells you |
|---|---|
| "I know what's expected of me at work." | Clarity |
| "I can raise problems here without worrying about how it'll be received." | Trust |
| "The pace I work at is one I could keep up for a year." | Energy |
| "I've learnt something new in the last month." | Capability |
| "Would you recommend this team to a friend as a place to work?" | Overall: the simplest single measure |
The six statements on the team scan work well as a monthly pulse too.
What bad looks like
- A 60-question survey once a year, with results six months later.
- "Improve communication" as the only action.
- A manager asking around to find out who scored them low.
- Scores used to rank managers against each other, so everyone games them.
How you'll know it's working
- Response rate holds steady or rises over six months.
- Visible action: the team can name one thing that changed because of the last survey.
- Speed: results shared within a week, every time.
- Care check: people write honest comments, including uncomfortable ones.
In the frameworks
In other industries
- Healthcare the annual staff survey says "not enough time with patients". The ward that acts on it changes one rota rule within a month.
- Accountancy a busy-season pulse catches burnout in week three, not in the leavers' exit interviews.
- Manufacturing a monthly shift pulse that includes "I feel safe to stop the line" tells you more than near-miss counts.
- Small business five questions on a shared form, read together over coffee.
What to say
When you share results with your team:
"Here's what you told us, all of it. I'm not going to explain any of it away. Help me understand the two that worry me most, then let's pick one thing to change before the next pulse."
Try this week
Take your last survey's lowest score to your team and ask, "What's really behind this?"
Think about it
What did your team tell you in the last survey, and could they say what changed because of it?