eNPS: measuring the employee Net Promoter Score

The question

“How likely is it that you would recommend [organisation] as a place to work to a friend?”

Use the same 0 to 10 scale as customer NPS, with the ends labelled Not at all likely and Extremely likely. Keep the wording fixed so each round can be compared with the last.

Calculating eNPS

With 60 answers the margin of error is wide. Using the formula in the NPS guide (p = 0.30, d = 0.25, n = 60), it comes to about ±19 points, so a small change from one round to the next tells you very little on its own.

Anonymity is not optional

People answer honestly when they believe the answer cannot be traced back to them. Say plainly how the survey is run, and make sure the tool really does keep identity out of the answers rather than merely hiding it.

Small groups are the other risk. If a team of three sees its own eNPS, anyone can work out who the detractor was. Choose a minimum group size, such as five, and never report a group smaller than it, in any breakdown.

What to ask alongside it

• “What is the main reason for your score?”
• “Do you expect to still be working here in 12 months?”
• “What is one thing leadership should start, stop or continue?”
• A few broad, optional groups, such as area of the organisation and length of service, kept wide enough to protect people.

How often to ask

eNPS works as a pulse. A quarterly round gives a trend without wearing people out; ask more often only if you will act between rounds. Asking more often than you respond teaches people that answering changes nothing.

Reading the result

• Compare with your own earlier rounds, not with other organisations' published scores.
• Report the number who answered next to the score.
• Break the score down only into groups above your minimum size.
• Read the reasons before you discuss the number.
• Tell people what you heard and what will change, and when.