What is a good NPS score? Benchmarks, their limits and how to check yours

Any Net Promoter Score above 0 means you have more promoters than detractors. Beyond that, "good" depends on your industry, whether you sell to businesses or consumers, the countries you survey and how you ran the survey. As one reference point, Survicate's 2025 benchmark of its customers' surveys found a median NPS of 42 overall, 38 for B2B companies and 49 for B2C companies. Survicate calls a score of 30 or more "doing well" and above 70 "among the best."

Published benchmarks disagree with each other, though, and a score from a small sample can move 20 points by chance. The most useful comparison is usually your own score over time, measured the same way each time, and only when the change is larger than your margin of error.

If you need the basics of how the score is calculated, start with Net Promoter Score.

What published benchmarks say

Two survey-software vendors publish detailed benchmarks from their own customers' data. Both are vendor figures, and both describe their methods.

Survicate also reports the spread within its data: the bottom 10% of companies scored -4 in 2025 and the top 10% scored 78. That spread is a reminder that an industry median describes the middle of a wide range.

Bain, which developed NPS, offers certified competitive benchmarks, including through its NPS Prism benchmarking service.

Why the benchmarks disagree

Look at software sold to businesses. Survicate's median for B2B software is 29. Retently's average for the category it calls B2B Software and SaaS is 41. That is a 12-point gap for similar labels. Several differences can explain it:

• Different customers. Each vendor reports on companies that chose its product, not a random sample of companies.
• Median versus average. Survicate reports medians, which it says are less affected by extreme scores. Retently reports averages.
• Different categories. What counts as "software" or "consulting" differs between the two reports.
• Different surveys. Timing, channel and who gets asked vary from company to company within each dataset.

Country matters too. NN/g summarizes a 2021 Qualtrics XM Institute study of 17,509 consumers across 18 countries. When people rated companies they liked, respondents in India and Mexico produced scores of 60 or more, while respondents in Japan produced -47. A company surveying mostly Japanese customers could be doing well with a score that would look poor elsewhere.

The type of score matters as well. Bain distinguishes relationship, experience and competitive benchmark scores and says relationship and experience scores should not be compared directly. A score collected right after a resolved support ticket is not comparable with an annual survey of all customers.

Sanity-check your own score before comparing it

1. Check the calculation. Use the 0 to 10 scale, count 9 and 10 as promoters and 0 to 6 as detractors, and keep passives in the total. Leaving passives out of the total inflates the score.

2. Work out the margin of error. Take a hypothetical B2B software company with 40 responses: 18 promoters, 14 passives and 8 detractors. Its NPS is 18 ÷ 40 = 45% minus 8 ÷ 40 = 20%, or 25. Using the adjusted-Wald method that MeasuringU recommends for NPS confidence intervals:

• Adjusted sample: 40 + 3 = 43
• Adjusted promoters: 18.75 ÷ 43 = 0.436
• Adjusted detractors: 8.75 ÷ 43 = 0.203
• Adjusted NPS: 0.436 − 0.203 = 0.233
• Variance: 0.436 + 0.203 − 0.233² = 0.585
• Standard error: √(0.585 ÷ 43) = 0.117
• 95% margin of error: 1.96 × 0.117 = 0.229, or about ±23 points

The true score could plausibly be anywhere from about 0 to 46. That range covers Survicate's B2B software median of 29, its overall B2B median of 38 and Retently's B2B software average of 41. With 40 responses, this company cannot say whether it is above or below any of those benchmarks.

3. Check who answered. Bain notes that when you survey only your own customers, people who dislike doing business with you may not bother to respond. Compare response rates across customer segments. A score built mostly from your happiest accounts will look better than your customer base as a whole.

4. Check how and when you asked. A survey sent after successful onboarding calls will score differently from one sent at renewal. Record the trigger, the channel and the audience alongside the number.

5. Check for gaming. NN/g gives examples of inflated scores: staff mentioning the survey only after a good interaction, or customers offered a coupon in exchange for a rating. If anyone's bonus depends on the score, look closely.

6. Keep the method fixed. AAPOR's advice for surveys in general applies here: "If you want to measure change, don't change the measure." Same question, same scale, same audience, same timing.

Decide what to do with your score

Treat an external benchmark as a rough reference and your own trend as the main one. Survicate itself recommends comparing against your own previous results after the first survey.

Then read the answers to "What is the main reason for your score?" and group them. The groups tell you more than the number: a score of 25 driven by slow support calls points to different work from a 25 driven by missing features. Call a few detractors, passives and promoters to hear the full story behind their answers. Customer survey questions has a short follow-up call guide, and Net Promoter Score covers the metric's other limits.

Your next step

Before sharing your score, write three numbers next to it: the number of responses, the margin of error and the benchmark you are comparing with, including who published it and how it was measured.

If you want to understand how customers of competing products see the category, a benchmark cannot tell you that, but conversations can. Instant Expert can find people who match a description you write, such as "operations leads who switched help desk software in the last year." You review who it finds, it sends your invitations, and you pay only for calls that get booked. The directory page for operations professionals in customer support software shows the kind of people you can reach.