Customer Success
Net Promoter Score: Calculate NPS and Read the Results Carefully
Calculate Net Promoter Score with a worked example, understand promoters and detractors, and report response coverage without overstating loyalty.

What you will learn
- Classify NPS responses and calculate the score correctly.
- Distinguish the score from satisfaction, response coverage and actual referrals.
- Use comments and follow-through to interpret the result responsibly.
Net Promoter Score, or NPS, summarizes answers to a recommendation question on a zero-to-ten scale. Subtract the percentage of detractors from the percentage of promoters, then report the result with the response count, survey question and collection context. Use the group distribution and comments to decide what to investigate next. Recommendation intention is different from satisfaction, renewal or a recorded referral.
What does Net Promoter Score measure?
Bain’s official scoring explanation asks how likely a person is to recommend an organization to a friend or colleague. The score groups responses into promoters, passives and detractors. Preserve the exact question, scale, invitation population and survey timing when interpreting a trend.
A relationship survey asks about the broader organization or product. A question immediately after one interaction can reflect that particular experience. Do not merge the two simply because both use a recommendation scale. State the context in the report so another person can understand what respondents were evaluating.
How is NPS calculated?
What are NPS detractors?
Detractors are valid respondents who choose a rating from 0 through 6. Their percentage is subtracted from the promoter percentage. The category records a survey response; it does not explain the person’s reason or prove a future cancellation.
Standard NPS response groups
| Rating | Group | Treatment in the score |
|---|---|---|
| 9 or 10 | Promoters | Add their percentage of all valid responses |
| 7 or 8 | Passives | Remain in the denominator, but are neither added nor subtracted |
| 0 through 6 | Detractors | Subtract their percentage of all valid responses |
NPS = promoter percentage − detractor percentage. The possible result runs from −100 to +100. Report the score as points, not a percentage of people who are happy. Do not omit passives from the denominator.
Fictional example: 100 customers are invited and 40 provide valid answers. There are 22 promoters, 10 passives and eight detractors. Promoters are 22 ÷ 40 = 55%; detractors are 8 ÷ 40 = 20%. NPS is 55 − 20 = 35 points. Response coverage is separately 40 ÷ 100 = 40%.
Check the arithmetic: 22 + 10 + 8 = 40 and 55% + 25% + 20% = 100%. The 60 people who did not respond have unknown recommendation intentions. They do not become passives by default.
What does a score of 35 tell you?
In this example, the promoter share exceeds the detractor share by 35 percentage points among valid respondents. It does not tell you why someone chose a rating, how the silent majority would respond or whether a particular customer will renew. Report the three group counts, not only the final score.
The same score can hide different distributions. In another fictional set of 40 answers, 14 promoters, 26 passives and no detractors also produce 35 points. That pattern is different from 22 promoters and eight detractors. A single number cannot replace the distribution and the comments.
Use the survey-bias lesson to audit who answered and how the instrument was administered. Use retention measures for observed customer and revenue behavior. Neither recommendation intention nor a favorable comment is a substitute for those outcomes.
Compare scores only when the context is comparable
- Use the same question and rating scale, or explain the change.
- Keep the eligible population and invitation method comparable. A survey of active champions is not the same as a survey of all customers.
- Separate relationship and interaction-level surveys.
- Report valid responses and response coverage for each period.
- Check whether a changed customer mix, survey channel or collection window could explain the movement.
Avoid a universal “good NPS” threshold. An external benchmark may use a different market, population, method or period. Before citing one, confirm those definitions and the source. A modest change in a small sample deserves investigation, not a claim of a durable company-wide improvement.
Use feedback to choose an owned next step
A recommendation score becomes more useful when it leads to a specific question. Ask a neutral optional follow-up about the reason for the rating. Do not pressure respondents to improve their score or assume that a detractor wants a sales call.
Original NPS interpretation and follow-through board
| Evidence | Unresolved question | Responsible next step |
|---|---|---|
| Several comments describe unclear setup | Which part of setup lacks context? | Customer-success owner reviews the current instructions with permitted source evidence |
| Promoters mention a particular workflow | Is the benefit reproducible for other suitable customers? | Product owner validates the actual workflow before using the claim |
| Few inactive customers answered | How does the respondent group differ from the invitation group? | Research owner reports the coverage limitation and considers a suitable follow-up method |
Before the review, write a short evidence note: what was asked, who answered, which themes have supporting comments and which questions remain unresolved. Give the next owner a specific question to resolve rather than an instruction to “improve NPS.” Check that the resulting action addresses the experience behind the feedback.
Keep the review connected to a customer business review where appropriate, but do not turn a general score into an account-specific conclusion without evidence.
Copyable AI prompt
Try this with Aurora AI in Finta, Codex, or Claude. Share only material you are authorized to use, remove unnecessary personal information and review the draft.
Using [exact NPS question], [eligible invitation count], [authorized de-identified responses] and [survey dates], validate ratings 0–10, count promoters/passives/detractors, calculate group percentages and NPS, and report response coverage separately. Preserve missing and invalid responses. Summarize supported themes and follow-through questions. Do not infer nonrespondents, apply invented benchmarks, predict renewals, contact customers or modify records.Keep the score with its context
Take the survey context, group counts and open questions into a traceable customer discussion. Explore Finta’s Customer Voice Researcher for organizing permitted feedback into a reviewable brief. Keep the survey instrument, calculations and observed customer outcomes in their responsible systems, then follow through on the action a named owner accepts.
Sources and method. By Finta Editorial Team. Sources checked November 15, 2026. Worked examples and worksheets are original fictional teaching materials, not customer results.