Net Promoter Score Calculator
NPS from 0–10 survey scores (a list or score:count pairs): promoters, passives, detractors, the score with its confidence interval, and what a target NPS needs.
NPS = % promoters − % detractors with the three shares and counts, the mean score, a confidence interval at your level, the score distribution, and the extra promoters a target NPS would take.
Example: Ten scores 10 9 9 8 7 6 10 3 9 8: 5 promoters, 3 passives, 2 detractors → NPS 30, but ±48 at 95 % with so few responses; reaching 50 would take four more promoters and no new detractors.
One number,
with its error bar.
How the score is computed, what the confidence interval means, and how the target line works.
The score
Each response on the 0–10 scale is a promoter (9–10), a passive (7–8) or a detractor (0–6). NPS = % promoters − % detractors, from −100 to +100; passives count in the base but not in the difference. Scores can be typed as a list in any order or as score:count pairs (10:40 9:20 …); anything that is not a whole number 0–10 is listed and ignored. The mean score is shown for those who prefer it.
The interval
Treating each response as +1, 0 or −1, NPS is the mean × 100 and its variance is p₊ + p₋ − (p₊ − p₋)². The interval is ± z × √(variance ÷ n) × 100 at the confidence level you choose (z = 1.96 at 95 %), by the normal approximation — fine for n above about 30 and rough below it, which the page says. A survey of ten answers has an interval about ±50 points wide: the score alone means little.
The target
For a target NPS the page solves (P + m) ÷ (n + m) − D ÷ (n + m) = target for m, the extra promoters needed with no new detractors — a way of sizing the gap, not a plan. No industry benchmark is built in: compare against your own earlier surveys or a source you trust. Nothing leaves the browser; the same four anonymous usage counts as the rest of the site apply.
SOURCES
- Reichheld, "The One Number You Need to Grow", Harvard Business Review, 2003 (NPS definition); interval from the variance of a +1 / 0 / −1 response, normal approximation
Last reviewed 22 September 2026. How results are checked: How we verify.