Employee NPS (eNPS) Calculator

Employees answer on a 0–10 scale. Promoters score 9–10, passives 7–8, detractors 0–6.

eNPS = % promoters − % detractors, from −100 to +100.

Use the 0–10 scale, not a 1–5 one. Some HR tools use five points and the resulting numbers are not comparable to anyone's benchmark.

eNPS borrows the customer Net Promoter Score and points it at your own staff. Employees answer "On a scale of 0 to 10, how likely are you to recommend [company] as a place to work?". Bain & Company, which co-created NPS, describes this as the most common wording. The score is a quick read on how many people would vouch for working here against how many would warn a friend off.

It is a pulse, not a full engagement survey. It tells you the balance of advocates and critics; it does not tell you why, which is what the follow-up questions are for.

Group every answer: promoters score 9 or 10, passives 7 or 8, detractors 0 to 6. Then eNPS = (promoters ÷ total responses × 100) − (detractors ÷ total responses × 100). The result runs from −100 to +100 and is a number, not a percentage.

Worked example (illustrative numbers): 100 employees answer. 38 score 9 or 10, 40 score 7 or 8, and 22 score 0 to 6. Promoters are 38%, detractors 22%, so eNPS = 38 − 22 = 16. The 40 passives stay in the 100 you divide by; drop them and you would get 63.3% − 36.7% = 27, eleven points of pure arithmetic error.

The calculation is identical to customer NPS. Only the question changes, so the eNPS calculator above and our NPS calculator produce the same number from the same counts.

The margin of error uses the variance of a Net Promoter Score, published by Brendan Rocks in The American Statistician in 2016: with p and d as the promoter and detractor shares, the 95% margin is 1.96 × √((p + d − (p − d)²) ÷ n) × 100.

For the 100-person example, p = 0.38 and d = 0.22: the variance is 0.60 − 0.16² = 0.5744, the standard error is 7.6 points, and the 95% interval runs from about 1 to 31. That is the honest reading of "eNPS 16": somewhere between barely positive and good.

Now a 12-person team: 5 promoters, 4 passives, 3 detractors. The calculator shows +17 with an interval of about −29 to +62. If one promoter becomes a detractor next quarter (4, 4, 4), the score drops to 0. One person's mood moved the number 17 points, which is why small-team eNPS should be read as a conversation prompt, not a KPI.

An eNPS only works if people believe nobody can trace their answer. The standard safeguard is a minimum reporting group: results for any group smaller than the threshold are hidden, not shown with a warning. Qualtrics' employee-experience product defaults to 5, and Culture Amp describes its reporting minimum as usually 5.

In practice: do not report eNPS for a team, location or tenure band with fewer than five responses, and do not add demographic questions whose groups you do not expect to clear that bar. In a 12-person team, asking for department and tenure narrows most answers to one or two people. Our guide to running a genuinely anonymous survey covers the leaks (sign-in, email fields, metadata, small-group breakdowns) in detail.

Bain's own guidance is that employee feedback should be kept confidential, with none of the individual follow-up calls you would make to a customer detractor. If you want to follow up, ask for volunteers in a separate, clearly identified question.

Bain notes that employee scores can be substantially lower than customer scores, so do not hold your eNPS against a customer NPS benchmark. We could not find a cross-company eNPS benchmark that publishes its sample, question wording and timing, which you would need to compare against it honestly. Many HR blogs repeat bands such as "10 to 30 is good", but without a traceable source.

The band label the calculator prints (below zero, needs work, good, great, world class at 0, 10, 30 and 50) is the same rule of thumb used on the customer NPS calculator. Use it as a rough reading, and use your own trend, with the interval beside each point, as the real benchmark. A score below zero always deserves attention: it means more critics than advocates.

Bain says companies most commonly ask for eNPS feedback monthly, and recommends keeping each survey to a few questions. Many smaller organizations run it quarterly instead. The trade-off is simple: more frequent pulses catch changes sooner, but each wave needs enough responses to beat the margin of error, and people stop answering surveys that never lead to visible action.

A reasonable rule: survey only as often as you can share results and one action back to the team before the next wave. Keep the wording, scale and timing identical between waves, and always publish the response count next to the score.

Skip the headline number for teams small enough that the interval spans most of the scale, or where anonymity cannot be protected; hold a structured conversation instead. eNPS also cannot diagnose: it will not tell you whether pay, a manager or workload is the problem. For that, pair it with a short set of rated statements and analyze them with our Likert scale calculator.

In Zunoform, the NPS question type gives you the 0–10 scale, an optional built-in follow-up ("What is the main reason for your score?"), and a summary tab that shows the promoter, passive and detractor split and the score. Turning on anonymous mode stops the form storing IP address, user agent, referrer, location, language and screen size for that form. Publishing a form with an NPS question needs the Pro plan; you can build and preview it on Free.

The Employee Satisfaction Survey template already includes the "recommend us as a place to work" question. Adjust it before you send it: it asks for department, which is only safe if every department clears the five-response minimum, and it offers an optional work-email field for people who want to be contacted. Remove both for a small team if you want to call the survey anonymous rather than confidential.