Net Promoter Score Calculator: Formula and Examples

A net promoter score calculator answers a specific question: among customers who responded to your survey, how much does the share of promoters exceed the share of detractors? Enter the response counts below to calculate NPS, see the group percentages, and check your arithmetic. Passives stay in the total. That small detail can change the result considerably.

How the net promoter score calculator works

NPS starts with a recommendation question, usually asking how likely someone is to recommend your business to a friend or colleague. Respondents choose a rating from 0 to 10. Bain's explanation of NPS scoring groups ratings of 9 or 10 as promoters, 7 or 8 as passives, and 0 through 6 as detractors.

The formula uses the percentage of all valid responses in each group:

NPS = percentage of promoters - percentage of detractors

Using counts, the same formula is 100 × (promoters - detractors) / total responses. Total responses means promoters plus passives plus detractors. Do not subtract the raw counts and stop there, unless you happened to receive exactly 100 responses.

The result ranges from -100 to +100. All detractors gives -100; all promoters gives +100. Report it as a score, such as "NPS 60," rather than "60% satisfied." NPS is not the percentage of happy customers, and an individual customer's rating is not an NPS result.

SurveyMonkey's NPS documentation confirms the scale and subtraction, but its rounding sequence differs: it rounds the group percentages to whole numbers before subtracting them. This calculator subtracts using full precision and then rounds the final score. A small difference from that platform can therefore be a rounding issue rather than a data error.

Net promoter score calculator examples you can check

Suppose your survey receives 70 promoter responses, 20 passive responses, and 10 detractor responses. There are 100 valid responses. Promoters make up 70% of that total and detractors make up 10%, so NPS is 60.

Now imagine the same number of promoters and detractors, but 120 passives instead of 20. There are 200 responses. Promoters account for 35% and detractors for 5%, so NPS is 30. The gap in counts is unchanged, but it describes a smaller share of the surveyed group. This is why passives belong in the denominator.

A score of zero needs context. It can mean every respondent was passive, as in the fourth row, or that promoter and detractor counts were equal. Those are different feedback patterns, despite the identical score. Check the distribution before deciding what to do next.

For the last row, the calculation is 100 × (2 - 1) / 3 = 33.333... . The tool displays 33 as the headline score and 33.33 as the calculated value. Neither display makes three responses a reliable picture of an entire customer base.

Prepare your survey counts before calculating

Start with a defined audience and reporting window. For example, you might invite customers who completed onboarding during September 2026 and report the valid responses collected within a consistent follow-up window. Record that rule before sending invitations, so a later team member can repeat the process.

Count only valid answers to the recommendation question. A blank answer is not a zero rating and does not belong in the detractor group. A rating of zero, however, is a valid detractor response. If your export includes duplicate submissions, apply a documented deduplication rule before counting groups.

Decide whether you are measuring the overall customer relationship or a particular interaction. A survey after a support conversation answers a different question from a survey sent to your active customer base. Label the scope in the report. Do not quietly combine those surveys to produce a larger response count.

Keep invitation counts separate from response counts. If 500 eligible customers receive invitations and 100 provide valid answers, the calculator's denominator is 100, not 500. Separately reporting the 20% valid-response rate helps readers understand how much of the invited audience is represented.

Read the score without overclaiming

A positive NPS means the promoter share exceeds the detractor share among respondents. A negative NPS means the reverse. It does not tell you which operational issue caused a low rating, whether customers will actually refer someone, or how much revenue will change.

Avoid declaring one universal "good" score. Audience selection, survey timing, industry, and collection method can all affect how a comparison should be read. If you use an external benchmark, check how it was collected and whether its audience resembles yours. Your own consistently measured trend is often a more useful starting point than an unrelated headline number.

Always show the sample size. With ten responses, moving one person from detractor to promoter changes NPS by 20 points. With 100 responses, the same move changes it by 2 points. These are arithmetic examples, not estimates of statistical significance. This calculator does not provide confidence intervals or significance testing.

Also ask who did not reply. Qualtrics' measurement guide notes that customers at the extremes may be more likely to share their opinions. More responses can help, but they do not automatically remove response bias. Keep the invitation approach consistent, and avoid asking only your most enthusiastic customers.

Use NPS alongside behavior. Review retention, repeat purchases, referrals, and support themes rather than assuming a score increase proves business growth. Our guide to customer retention strategies can help turn those observations into a practical follow-up plan.

Combine survey results using the underlying counts

If two comparable survey groups need a combined score, add their promoter, passive, and detractor counts first. Then calculate NPS from the combined totals. A simple average of the two scores gives each group equal influence, even when their sample sizes differ.

For example, Group A has 80 promoters, 10 passives, and 10 detractors, giving NPS 70 from 100 responses. Group B has 2 promoters, 2 passives, and 6 detractors, giving NPS -40 from 10 responses. The simple average is 15. The combined counts are 82 promoters, 12 passives, and 16 detractors, so the combined NPS is 100 × 66 / 110 = 60.

That calculation is appropriate only if combining the groups answers a sensible question. Keep results separate when question wording, survey timing, or audience definitions differ. A pooled result can also hide a weak segment, so retain the individual group scores next to the total.

Build a small NPS reporting template

You do not need a complicated report. Give the reader enough context to understand what the score covers and which action follows from it. Copy these fields into a worksheet or dashboard:

• Reporting period and survey type.
• Eligible audience, invitation method, and number invited.
• Valid responses, response rate, and deduplication rule.
• Promoter, passive, and detractor counts and percentages.
• NPS result and rounding method.
• Recurring feedback themes, action owner, and review date.

Add a neutral follow-up question such as "What is the main reason for your rating?" Keep comments separate from calculation inputs, and do not paste names, email addresses, or other personal information into this tool. For collecting richer explanations, adapt the questions in our customer interview template.

Make actions specific. If several comments describe confusing onboarding instructions, assign a review of those instructions and define how you will check the change. Do not make improving the score itself the only task. That encourages teams to chase a rating instead of addressing the experience.

Common questions about NPS calculations

Do passives count?

Yes. Passives are included in total valid responses but contribute nothing to the promoter-minus-detractor numerator. Excluding them can overstate the positive or negative result.

Can NPS be negative?

Yes. If detractors outnumber promoters, the result is negative. For 15 promoters, 5 passives, and 30 detractors, NPS is -30. That describes the response balance; it is not a negative revenue figure.

What happens with no responses?

NPS is undefined when all three counts are zero. The calculator shows an error rather than returning zero. Zero is a legitimate result for equal promoter and detractor counts, so using it for an empty survey would be misleading.

Why does my platform show a slightly different score?

Check the export filters, excluded responses, reporting window, and rounding sequence. Compare the underlying group counts before changing your formula. If you report platform results and calculator results together, label the rounding method so small differences are not mistaken for changes in customer opinion.

Should I average customer ratings instead?

An average 0-10 rating is a separate metric. NPS groups the ratings and subtracts the detractor percentage from the promoter percentage. Keep both labels clear if your report includes both calculations.