Your NPS, With Its Margin of Error Not Just a Number

How it works

Responses are classified on the standard NPS scale: promoters score 9-10, passives 7-8, detractors 0-6. NPS = % promoters minus % detractors, stated as a point score from -100 to +100. The 95% confidence interval uses the classical multinomial variance formula Var(NPS) = (p_promoter + p_detractor - (p_promoter - p_detractor)^2) / n. With a segment column mapped, each segment gets its own NPS and interval, and the best and worst segments are compared with a two-proportion-style z-test, only when both have at least 30 responses.

Use it whenever you have raw answers to the 0-10 'how likely are you to recommend us' question and want the score, its real margin of error, and honest segment comparisons.

Not for 1-5 star ratings or CSAT scales (NPS is defined on 0-10 only), and not for tracking NPS over time: use a time-series or control-chart analysis for trend questions.

What data do you need?

One 0-10 score per respondent, optionally with a segment. For example, a post-purchase survey:

Minimum 10 rows · Best with 100-50,000 responses and 2-8 segments

What's in the report?

Standard-library analysis: Net Promoter Score from a 0-10 likelihood-to-recommend column. Promoter / passive / detractor shares, the NPS point score with a proper 95% confidence interval, the full 0-10 score distribution, and per-segment NPS with a statistically honest best-vs-worst comparison. Works on any survey export: map the 0-10 score column, optionally a segment column.