Enps survey: how it works and how to set it up

An eNPS survey asks your staff one question: on a scale of nought to ten, how likely are you to recommend this company as a place to work? The score it produces is a quick reading of morale, and when revenue has stalled, morale is often the first thing to slip and the last thing anyone measures. This article answers one question: how do you run an employee net promoter score survey on the levelpartner.ai platform so the answers are honest, the score is repeatable and something changes as a result?

What the score measures and how it is worked out

Everyone answers the same nought to ten question. People who answer nine or ten are promoters. Seven or eight are passives. Nought to six are detractors. The score is the share of promoters minus the share of detractors, so it runs from minus one hundred to plus one hundred. Passives count towards the total number of responses but not towards either side.

The number on its own says little. What matters is its direction from one round to the next, and the follow-up question that tells you why people answered as they did.

Honesty depends on anonymity

Staff answer honestly only if they believe nobody can trace the answer to them. That belief is fragile. A survey sent from a manager's email, a question about department in a team of four, or a results meeting where someone says "I can guess who wrote this" will ruin it for every later round.

On the platform, surveys normally collect contact details so answers attach to a person. For an eNPS survey you want the opposite. Leave out name, email and phone, share one generic link rather than personal links, and test that a submission arrives without creating or matching a contact. If your set-up cannot do that, do not promise anonymity. Promise confidentiality instead, limit who can see the results, and say so plainly.

Before you start

You need access to Sites and permission to share a link with all staff. Agree who sees the raw responses, ideally one person outside the line management chain. Decide whether to break results down by team, and only do so for teams of ten or more. Pick a fixed schedule, such as once a quarter, and put the next three dates in the calendar now.

How to set it up

1. Go to Sites, then Surveys, and create a survey named for the round.
2. In the Survey Builder, leave out name, email and phone fields.
3. Add the nought to ten question with your fixed wording.
4. Add one open question asking for the main reason behind the score.
5. Add a team question only if every team has ten people or more.
6. Write a closing message saying when results will be shared.
7. Save, preview and submit a test response.
8. Confirm the test appears under Sites, Surveys, then Submissions.
9. Share the survey link with all staff through your usual internal channel.
10. Send one reminder halfway through, then close the survey on time.
11. Work out the score and share it within two weeks.

What to do with the answers

The score tells you whether things are getting better or worse. The comments tell you why. Read every comment, group them into three or four themes, and pick one theme to act on before the next round. Tell staff which theme you chose and what you are doing about it.

Do not try to fix everything. A company that promises five changes after a survey and delivers one has taught its staff that the survey is theatre. A company that promises one change and delivers it will see more honest answers next time.

Time it takes and what it replaces

Setting up the first round takes about an hour. Each later round takes ten minutes to copy and relaunch, plus two or three hours to read comments and prepare the summary. It replaces the long annual engagement survey that most staff stop finishing halfway, and the exit interview as your only source of honest feedback.

When not to run one

Skip eNPS if you have fewer than about fifteen staff, because anonymity is almost impossible and a quiet conversation works better. Do not run it in the middle of a restructure, when every answer reflects uncertainty rather than the workplace. And do not start if nobody has the time to act on the results.

For other questionnaires to clients or prospects, building an online survey that drives follow-up is the place to start. To keep the survey readable on a phone, see the guide to form layout and labels. If the eNPS survey sits alongside other internal forms, keeping a growing set of forms under control helps you name and own them.

Troubleshooting

Few people respond. Staff do not believe it is anonymous or do not believe it changes anything. Say exactly who sees the answers and share what changed after the last round.

Responses create contacts. A contact field is still on the survey. Remove it, test again and delete the test contact.

Comments are short and vague. The open question is too general. Ask for "the main reason for your score" rather than "any other comments".

Scores swing wildly between rounds. Response numbers are too low to be stable. Push for more responses before reading much into the movement.

Frequently asked questions

What is a good eNPS score?

Any positive score means more promoters than detractors. Compare against your own last round rather than other companies, since sectors and survey methods differ.

How often should we run an eNPS survey?

Once a quarter suits most firms. More often and staff tire of it; less often and you miss changes until they become problems.

Should the eNPS survey be anonymous?

Yes, wherever you can make it truly anonymous. If you cannot, call it confidential and explain who sees the answers.

Can we add more questions?

Keep it to the score, the reason and at most one team question. Longer staff surveys have their place, but they are a different tool.