What the Net Promoter Score measures
The Net Promoter Score comes from a single survey question that Bain & Company words as "How likely are you to recommend us to a friend or colleague?", answered on a scale from 0 to 10. Answers are sorted into three groups:
| Group | Answers | Counts toward the score |
|---|---|---|
| Promoters | 9 and 10 | Added |
| Passives | 7 and 8 | Only in the total |
| Detractors | 0 to 6 | Subtracted |
The result runs from −100, if everyone is a detractor, to +100, if everyone is a promoter. Take the example in the calculator: of 100 answers, 50 are promoters, 30 passives and 20 detractors. That is 50% minus 20%, an NPS of +30.
Margin of error
A survey only asks some of your customers, so the score is an estimate. To see how far it could be off, score each answer +1 for a promoter, 0 for a passive and −1 for a detractor. The NPS divided by 100 is then simply the average of those scores. With p the share of promoters and d the share of detractors, that average has a variance of (p + d) − (p − d)², using the formula for the variance of a discrete random variable in OpenStax Introductory Statistics, section 4.2. The standard error of an average is the standard deviation divided by √n (section 7.1), and the error bound is z times the standard error (section 8.1), with z = 1.96 for 95% confidence.
For the example, the margin at 95% confidence is ±15.3 points, so the true score is likely between +14.7 and +45.3. With only 20 answers, 10 promoters, 6 passives and 4 detractors, the score is also +30, but the margin is ±34.2 points. As Albert Madansky points out, promoters and detractors come from the same answers, so their percentages are not independent; the formula above already accounts for that. It is an approximation that works best with larger samples; when every answer falls in one group, it gives a margin of 0, which only means the sample shows no spread.
How many answers you need
Solving the margin of error formula for n gives the number of answers for a target margin: n = z² × (p + d − (p − d)²) ÷ (MOE ÷ 100)². The calculator uses your current shares. For the example, about 938 answers would bring the margin down to ±5 points at 95% confidence. To plan a survey from scratch, the sample size calculator works out the answers you need for a percentage.
What moves the score
Each answer is worth 100 ÷ n points. Turning a detractor into a passive, or a passive into a promoter, raises the score by that much; turning a detractor into a promoter raises it by twice that. With 100 answers, each one is worth 1 points. That is also why scores from small surveys jump around: with 20 answers, a single person changes the score by 5 points.
A score on its own says little. Compare it with your earlier surveys run the same way, and look at the comments people leave alongside their number. To track whether recommendations turn into sales, pair the score with the conversion rate calculator.
How to use the calculator
- 1Choose Count per score and type how many people gave each answer from 0 to 10, or choose Paste scores and paste the column of answers, or choose Group totals if you only have the three totals.
- 2Pick a confidence level for the margin of error. 95% is the usual choice.
- 3Read the score, the shares of promoters, passives and detractors, the likely range and the chart of answers. Use Copy result or Copy link to share it.
Net Promoter, NPS and Net Promoter Score are trademarks of Bain & Company, Inc., NICE Systems, Inc. and Fred Reichheld. This calculator is not affiliated with them; it uses the published definition of the score.
Sources
- Bain & Company, Net Promoter System: Measuring your Net Promoter Score (question, groups, formula, trademarks)
- OpenStax, Introductory Statistics 2e, 4.2 Mean or expected value and standard deviation
- OpenStax, Introductory Statistics 2e, 7.1 The central limit theorem for sample means
- OpenStax, Introductory Statistics 2e, 8.1 A single population mean using the normal distribution (error bound, z = 1.96 for 95%)
- Albert Madansky, Statistical inference about Net Promoter Scores, Quirk’s (2019)
Frequently asked questions
How do you calculate NPS?
Ask how likely people are to recommend you on a scale from 0 to 10. Count answers of 9 and 10 as promoters, 7 and 8 as passives and 0 to 6 as detractors. The score is the percentage of promoters minus the percentage of detractors. Passives count in the total but not in either group.
What is the range of the score?
From −100, when every answer is a detractor, to +100, when every answer is a promoter. It is shown as a whole number, not a percentage, even though it is the difference of two percentages.
What is a good NPS?
This calculator does not rate scores as good or bad, because we found no published standard that fits every survey. The most useful comparison is with your own earlier surveys, asked the same way and to the same kind of customers. Before reading much into a change, check whether it is larger than the margin of error.
How many responses do I need?
It depends on how precise you want the score to be. With 50% promoters and 20% detractors, 100 answers give a margin of error of about ±15.3 points at 95% confidence, and about 938 answers would bring it to ±5 points. The calculator shows the number for your own answers.
How is the margin of error calculated?
Each answer is scored +1 for a promoter, 0 for a passive and −1 for a detractor. The NPS divided by 100 is the average of these scores, so the usual standard error of a mean applies: the margin is z × √((p + d − (p − d)²) ÷ n) × 100 points, where p and d are the shares of promoters and detractors and z is 1.96 for 95% confidence.
Why do passives matter if they do not count?
They count in the total. More passives at the same number of promoters and detractors pull both percentages down, so the score moves toward zero. Turning a passive into a promoter raises the score by 100 ÷ n points, and turning a detractor into a promoter by twice that.