Net Promoter Score (NPS) calculator

How to calculate NPS

• NPS formula

  NPS = % promoters − % detractors, or (P − D) / N × 100, where N is the number of all responses. Passives (7–8) are not in the subtraction but count in N. The result runs from −100 (all detractors) to +100 (all promoters).

• Three groups

  Promoters (9–10) are the most likely to recommend you. Passives (7–8) are satisfied but not enthusiastic, and moving them to 9–10 raises the score. Detractors (0–6) are unhappy customers, and their reasons show what to fix first.

• Promoters needed for a target NPS

  From NPS = (P − D) / N × 100, the promoters you need are P = target / 100 × N + D. With 200 responses, 90 promoters, 60 passives and 50 detractors, NPS is 20. For NPS 30 among the same people you need 0.30 × 200 + 50 = 110 promoters: 20 passives would have to move to 9–10.

• Margin of error of NPS

  Count each promoter as +1, passive as 0 and detractor as −1: NPS is their mean × 100. Its standard error is √((p + d − (p − d)²) / n), with p and d as shares, and the 95% margin is 1.96 × SE × 100. 100 responses with 60% promoters and 15% detractors give ±14.5 points.

• NPS on a 1–5 scale

  NPS exists only on the 0–10 scale. There is no standard way to convert 1–5 answers into promoters and detractors, so any mapping gives a number that cannot be compared with real NPS. For 1–5 data, report the share of 4s and 5s (top-2 box) with the CSAT or top-box calculator.

• Relational and transactional NPS

  Relational NPS asks about the company overall, usually quarterly or twice a year. Transactional NPS is sent after a purchase or support contact and runs continuously. Keep the question, timing and audience the same between waves so the trend means something.

NPS calculation examples

1. Good NPS

   • Total responses: 100. Promoters (9–10): 60. Passives (7–8): 25. Detractors (0–6): 15.

   NPS = 60% − 15% = +45: a good result by this calculator's bands, close to the excellent zone.

2. Negative NPS

   • Total responses: 200. Promoters: 40. Passives: 60. Detractors: 100.

   NPS = 20% − 50% = −30. Half of the customers are detractors: read their reasons first.

3. Neutral audience

   • 150 responses: promoters 30, passives 90, detractors 30.

   NPS = 20% − 20% = 0. Many customers in the middle and few at either end.

4. Small sample

   • 30 responses: promoters 12, passives 10, detractors 8.

   NPS = 40% − 26.7% = +13.3. The 95% margin of error is about ±29 points, so the true score could be anywhere from negative to above 40.

5. Before and after a change

   • Before: 50 promoters, 30 passives, 20 detractors, so NPS +30. After: 55/28/17, so NPS +38. Enter three numbers per wave.

   The score rose by 8 points, but with 100 responses per wave each NPS has a margin of about ±15 points. Collect more responses before calling it an improvement.

6. How many new promoters reach NPS 30

   • Now: 200 responses, 90 promoters, 60 passives, 50 detractors (NPS 20). How many additional responses at 9–10 lift it to 30?

   Solve (90 + k − 50) / (200 + k) = 0.30: k = 28.6, so 29 new promoters. Check: (119 − 50) / 229 = +30.1.

What to avoid

1. Interpreting NPS without context

   Compare with your own trend and with competitors measured the same way, not with abstract "good" numbers from the web.

2. Reading noise as a trend

   With 100 responses the margin of error of NPS can be ±15 points. To compare segments or waves, check the margin of error first.

3. Calling a 1–5 scale NPS

   Mapping 5 to promoters and 1–3 to detractors produces a number that looks like NPS but cannot be compared with anyone's NPS. Ask the 0–10 question, or report the top-2 box.

4. Ignoring passives in qualitative analysis

   Scores 7–8 mean "satisfied but not enthusiastic". Their reasons show what would move them to 9–10.

5. Measuring once without a plan

   You need regular waves (quarterly or twice a year) under the same conditions, or the trend cannot be read.