NPS calculator: what is the Net Promoter Score, and how do you calculate it?

Calculate your NPS

What is NPS?

NPS is based on a single question. On the Bain & Company page, it reads: “How likely are you to recommend us to a friend or colleague?”

The answer is a value from 0 to 10. Bain divides the answers into three groups:

• Promoters (9–10): according to Bain, loyal, enthusiastic fans.
• Passives (7–8): satisfied for now. According to Bain, their repurchase and referral rates are up to 50% lower than those of promoters.
• Detractors (0–6): unhappy customers. Bain writes that they account for more than 80% of negative word of mouth.

Fred Reichheld developed the measure in 2003, when he was a partner at Bain. That same year he wrote an article in the Harvard Business Review titled “The One Number You Need to Grow”.

How to calculate NPS

The formula is short: NPS = share of promoters − share of detractors. Passives don’t enter the score directly. But they count toward the total number of answers and so affect the shares.

1. Put each answer into its group: 9–10, 7–8 or 0–6.
2. Divide the number of promoters and of detractors by the total number of answers, and multiply by 100.
3. Subtract the share of detractors from the share of promoters.

Example: of 100 people, 45 are promoters, 30 passives and 25 detractors. NPS = 45 − 25 = +20. With its default values, the calculator shows exactly this result.

Because both shares lie between 0 and 100, the score lies between −100 and +100. If everyone gives 9 or 10, it is +100; if everyone gives 6 or less, it is −100.

The score is a difference of shares. In a report I write “+20”, without a percent sign. If you write “20% NPS”, readers sometimes take it for 20% of customers.

The difference between NPS, CSAT and SUS

All three are surveys, but each measures something different. Mix them up, and you look for the answer to the wrong question.

CSAT has no single standard form. Jeff Sauro of MeasuringU writes that satisfaction can be measured with one statement or with several. The scale format also changes from study to study.

Sauro describes the SUS, by contrast, as a reliable and valid measure of perceived usability. According to NN/g, NPS correlates well with the perception of usability. You can find your own SUS score with the SUS calculator.

What is a good NPS?

I don’t give a number for this question, because none of the sources I opened gives a general threshold. Bain’s calculation page doesn’t give a number above which a score is good either.

Culture shifts the score too. NN/g cites a 2021 Qualtrics XM Institute study of 17,509 consumers in 18 countries. In Japan and South Korea, NPS was negative even for companies the participants liked. In India and Mexico, scores were 60 and above.

The most reliable comparison is therefore your own past. Ask the same question regularly to the same audience, and look at the direction of the score.

Criticism of NPS

The claim that NPS predicts growth has been tested. Keiningham, Cooil, Andreassen and Aksoy published a study in the Journal of Marketing in 2007. The data came from the Norwegian Customer Satisfaction Barometer: 21 firms, more than 15,500 interviews.

The team found no “clear superiority” of NPS over other measures in the industries Reichheld had named as examples.

Therese Fessenden of NN/g looks instead at the limits of NPS in UX. The score tells you how the experience is perceived, but not why. And NPS from a test with five people is unlikely to give you valid data.

Grouping also loses information. If someone who gave a 2 gives a 5 in the next survey, that is a real improvement. But both values count as detractors, and NPS doesn’t move. Fessenden therefore recommends reporting the mean with a margin of error.

My note

I see NPS as an alarm. If the score drops, you learn that something is broken. What is broken, you only learn by watching users.

Next to NPS I always put an open question: “Why did you give this score?” Then I read the detractors’ answers and group recurring complaints. That becomes the real to-do list.

If the sample is small, don’t report the score on its own. The question of how many users to test matters even more with NPS. With 10 answers, a single person moving from detractor to promoter shifts the score by 20 points. If you compare two versions, check the difference with the logic of the A/B test calculator.

If your score is dropping and you can’t find which screen the complaints come from, I go through every step of up to 10 flows in your application with the UX Audit Pro.