The NPS formula
Net Promoter Score comes from a single survey question: how likely are you to recommend us to a friend or colleague?, answered on a scale of 0 to 10. Responses split into three groups:
- Promoters scored 9 or 10: enthusiastic, and the ones who actually recommend you.
- Passives scored 7 or 8: satisfied, but not enough to say anything to anyone.
- Detractors scored 0 to 6: unhappy enough to put others off.
The score is the difference between two of them:
NPS = % promoters − % detractors
Passives are deliberately left out of the calculation, though they still count toward your total responses. That is the part people most often get wrong by hand. It is also why a business where almost everyone is mildly satisfied can score close to zero.
NPS is not a percentage
The result is a whole number between −100 and +100, not a percentage. It is the difference between two percentages, which is a different thing. Writing it as “30%” implies it can be compared with CSAT or a satisfaction rate, and it cannot. A score of −100 means every single respondent was a detractor; +100 means every one was a promoter. Both are effectively unheard of in real data sets.
What is a good NPS score?
The widely used general bands are:
- Below 0: more detractors than promoters. Needs attention.
- 0 to 30: good. Most businesses sit here.
- 30 to 70: great. Customers are recommending you unprompted.
- Above 70: exceptional, and rare.
Treat those as orientation rather than a target. Industry averages get quoted confidently but are collected under wildly different survey methods, sample sizes and question wording, so a cross-industry comparison is usually meaningless. Your score against your own previous quarter, on the same survey, tells you far more.
What the number does not tell you
NPS compresses everything into one figure, which is its appeal and its weakness. A few things worth holding on to:
- Small samples swing hard. With twenty responses, one detractor moves the score by five points. Below roughly thirty responses, treat it as a signal rather than a measurement.
- Who answers is not random. People with strong feelings respond more, so scores tend to be more polarised than your actual customer base.
- The comments are the useful part. The score tells you whether something changed; only the written answers tell you what to do about it.
- Timing changes everything. Asking straight after a resolved complaint gives a very different answer to asking mid-problem.
None of this makes NPS useless. It makes it a tracking metric rather than a diagnosis: good for noticing a change, poor for explaining one.
NPS vs CSAT
CSAT asks how satisfied someone was with one specific interaction, and reports the share who answered positively as a percentage from 0 to 100. NPS asks about overall willingness to recommend and runs −100 to +100. They measure different things on different scales, so they cannot be compared, averaged, or converted. Most teams run CSAT on individual touchpoints and NPS on the relationship as a whole. The calculator above does both. Switch with the toggle.
Frequently asked questions
How do you calculate NPS?
Subtract the percentage of detractors (people scoring 0–6) from the percentage of promoters (9–10). Passives, who score 7–8, count toward your total response number but are left out of the score itself. So 50% promoters and 20% detractors gives an NPS of 30.
Is NPS a percentage?
No, and this trips people up constantly. NPS is a whole number between −100 and +100, not a percentage. It is the difference between two percentages. Writing it as “30%” is wrong and makes it look comparable to CSAT, which it is not.
What is a good NPS score?
As a general guide: below 0 needs attention, 0–30 is good, 30–70 is great, and above 70 is exceptional. Context matters more than the band, though. Comparing your score against your own previous quarter is far more useful than comparing it to an industry average collected under a different survey method.
Why do passives not count toward the score?
By design. NPS measures the gap between advocacy and dissatisfaction, and passives are neither. They are satisfied but not enthusiastic. They still matter commercially: converting passives to promoters is usually the fastest way to move the score, because they are already close.
How many responses do I need for the score to mean anything?
There is no hard threshold, but a score from a handful of responses moves wildly with each new one and should not be reported as a trend. As a rule of thumb, treat a score built on fewer than about 30 responses as a rough signal rather than a measurement.
What is the difference between NPS and CSAT?
CSAT asks how satisfied someone was with a specific interaction and is reported as a percentage from 0 to 100. NPS asks how likely they are to recommend you overall and runs from −100 to +100. They answer different questions and cannot be compared or averaged together.
Is my data sent anywhere when I use this calculator?
No. The whole calculation runs in your browser. Nothing you type or paste is transmitted, stored or logged, which is why the tool works with no signup and no account.