NPS calculator: Net Promoter Score

What is Net Promoter Score (NPS)?

Net Promoter Score measures how likely customers are to recommend you, from one question: “How likely is it that you would recommend [product] to a friend or colleague?”, answered on a 0–10 scale. People answering 9 or 10 are promoters, 7 or 8 passives and 0 to 6 detractors. NPS is the percentage of promoters minus the percentage of detractors.

It was introduced by Fred Reichheld in Harvard Business Review in 2003. Its appeal is simplicity: one question, one number, comparable over time. Its weakness is the same simplicity — a single score hides why people answered the way they did, so it’s most useful alongside a follow-up “why?” and a look at which customers are detractors.

NPS formula

Worked example

1. A SaaS surveys its customers and gets 200 responses: 120 answer 9 or 10, 50 answer 7 or 8, and 30 answer 0 to 6.
2. Promoters = 120 ÷ 200 = 60%. Detractors = 30 ÷ 200 = 15%. NPS = 60 − 15 = +45.
3. Margin of error at 95% confidence: 1.96 × √((0.60 + 0.15 − 0.45²) ÷ 200) × 100 ≈ ±10.3 points. The true score is likely between about +35 and +55.
4. To reach +50 with new responses only: (120 + x − 30) ÷ (200 + x) = 0.5 gives x = 20 more promoters. Turning detractors into promoters is faster: 5 conversions give (125 − 25) ÷ 200 = +50.

The margin of error matters more than most teams expect. If next quarter’s survey reads +40 from a similar number of responses, that’s within the noise — not evidence that anything got worse.

What is a good NPS?

Any score above 0 means more promoters than detractors, but “good” depends heavily on the industry, the country, how and when you ask, and who you ask — so published benchmark averages are hard to compare with your own number. More reliable ways to judge it:

• Compare with your own past surveys run the same way — same question, same timing, same kind of customer — and only treat differences larger than the margin of error as real.
• Split by segment: plan, tenure, the channel a customer came from. A healthy overall score can hide a segment full of detractors.
• Read the detractors’ comments. The reasons behind 0–6 answers are usually more actionable than the score.

Common NPS mistakes

• Leaving passives out of the total. NPS divides by all responses; dropping 7s and 8s inflates both percentages and changes the score.
• Reporting NPS as a percentage. It’s a score from −100 to +100, not a percent — “an NPS of 45”, not “45%”.
• Treating small changes as real. With a few hundred responses, a move of several points is normal noise. Check the margin of error before celebrating or worrying.
• Surveying only happy moments. Asking right after a successful action or a support win biases answers upward. Survey on a schedule, not after positive events.
• Changing the question or scale. A 1–5 scale or a different wording isn’t NPS and can’t be compared with past scores.
• Ignoring who doesn’t respond. Unhappy customers often don’t answer at all, so a low response rate can make NPS look better than reality.

How to connect NPS to behavior with VisitTrack

VisitTrack doesn’t run surveys — use your survey or email tool for that. What it can add is context: where promoters and detractors came from, and what they did in your product.

1. When someone answers, record their group as a custom event — window.visitrack("nps_promoter"), nps_passive or nps_detractor — or send it from your server (server events) if the survey runs by email.
2. Make each event a goal to count responses by group on the Goals tab.
3. Identify signed-in users with their score as a trait, for example window.visitrack.identify(user.id, { nps: 9 }), so each person’s first-touch source and activity are in the People view next to their answer.
4. Compare where promoters and detractors first came from — channels whose customers become detractors are worth spending less on, whatever their CAC.