Employee Net Promoter Score

For a long time the only way to know how a company felt from the inside was one survey at the end of the year. A thirty-question engagement survey went out once, and the tallied, interpreted results came back about two months later. Three people could quit in those two months and the survey would say nothing about it.

So instinct filled the gap. A founder would ask team leads how things were going and get back "seems fine." There was no way to tell the team that said it because things really were fine from the team that said it because saying otherwise felt risky. Then someone good would resign, and that resignation was the first hard signal that anything had been wrong.

The employee Net Promoter Score (eNPS) narrows that gap by cutting the survey down to one question. It takes the Net Promoter Score (NPS), already in use on customers, and turns it on employees instead. It gives up on telling you what the problem is. In exchange, it's cheap enough to run every quarter or every month, which means it can tell you when things started going wrong.

Ask people to answer "would you recommend this company to someone you know as a place to work?" on a scale from 0 to 10, then subtract the share of detractors from the share of promoters.

The calculation has three steps.

1. Sort the answers into three groups by score. 9 and 10 are promoters, 7 and 8 are passives, and 0 through 6 are detractors.
2. Work out what share of all respondents each group is. If 40 people answer and 18 give a 9 or 10, 12 give a 7 or 8, and 10 give a 6 or lower, that's 45 percent, 30 percent and 25 percent.
3. Subtract detractors from promoters. 45 minus 25 puts this company's eNPS at 20.

Passives drop out of the subtraction. Someone who gives a 7 or an 8 would neither talk a friend into joining nor warn them off, so they don't push the score either way. They don't disappear, though — they stay in the denominator and pull both shares down.

The result runs from -100 to +100. Everyone answering 10 gives +100; everyone answering 6 or below gives -100. It isn't a percentage, so a result of 20 is read as twenty, not twenty percent.

The cutoff catches people out: a 7 is a passive, not a detractor. Seven out of ten sounds decent, but the question asks whether things are good enough to recommend, and people who feel that way answer 9 or higher. Anything at 8 or below already says something is catching.

No absolute threshold separates good from bad. The typical range differs by industry. Conventional wisdom says anything above 0 is acceptable and +30 is healthy, but the moment you set your number beside another company's you're comparing two different scales.

The value comes from lining up one company's scores over time. Four straight quarters at 18, 20, 19 and 21 say the organization is steady. A run of 22, 20 and then 9 says something happened last quarter that's worth finding. The bend in the line carries more information than the number itself.

Check the respondent count before reading the score. In a 45-person company where 30 people answer, one person moving from promoter to detractor shifts the result by 6.7 points. At that size a 5-point drop from last quarter usually means nothing, while a 20-point drop means more than three people changed their minds and there is something real to look into. The smaller the organization, the coarser the scale, and reading it well means accounting for that.

Watch the response rate alongside it. If only 15 of 45 people answer and the score goes up, the likelier explanation isn't that the company improved but that the unhappy people stopped responding. When fewer than half reply, the response rate is the more urgent signal.

"Our eNPS dropped from 22 last quarter to 9. I think we need to run a full engagement survey."

We know something got worse; now we need the longer survey to tell us what.

"Don't break eNPS out by team. A team of six flips the number every time one person changes their answer."

The sample is too small, so the result is mostly noise.

"The score held steady but our response rate fell from 70 percent to 40 percent."

More people stopped answering, so take the score less seriously than it looks.

At Pagedeck, Grace sends this one question to the whole company every quarter. When the number fell from 21 to 9, she didn't guess at the cause in the room — she scheduled an engagement survey for the following week. eNPS tells you when things turned; it doesn't tell you where. The survey came back pointing at decision-making that had become invisible since headcount doubled, and from there the company had something specific to work on.

How the number gets shared matters too. Grace presents the score at the all-hands together with the change from last quarter and what the company intends to do about it. Post the score alone and next quarter's response rate falls, because people close a survey as soon as they decide answering it changes nothing.

The most common mistake is treating eNPS as a diagnostic. A low number feels like a finding, so it's an easy read. But a single question says nothing about why the number is low. A company where pay is the complaint and a company where managers are the complaint can both land on 5. Used as a diagnosis, a low score leads straight to a remedy picked by guesswork — more team dinners, say, for a problem that has nothing to do with team dinners. The number is closer to a thermometer. It tells you there's a fever; naming the illness takes a different test.

Turning it into a target creates the same trouble. Attach "get eNPS to 30 this year" to a manager's review and the manager has an easier path than changing the organization: warm up the room before the survey goes out, or nudge the people likely to score well into responding. The score climbs and people keep leaving. If you want eNPS to rise, the thing to raise is the underlying condition — psychological safety, or how close the team is to burnout — and the score's job is to confirm whether that moved.