NPS (Net Promoter Score)

Net Promoter Score (NPS)

NPS sorts respondents on a 0-10 scale: promoters (9-10), passives (7-8) and detractors (0-6). You compute it as NPS = %promoters − %detractors, which can land anywhere from -100 to +100, and the passives never enter the math. Say you have 60% promoters, 25% passives and 15% detractors — that works out to NPS = 45. As for the bands, anything above 50 counts as excellent and above 70 is elite (Apple and Tesla territory), 30-50 is solid, and below 0 is rough. Fred Reichheld put it forward in HBR (2003). Industry took to it quickly, though academics keep poking at it, mostly over how arbitrary the scale is and the way passives get weighted in a non-linear fashion.

Applications

It shows up across customer experience, voice of customer programs, quarterly NPS tracking, sector benchmarks from SAS/Bain reports, and as a customer success KPI. Transactional NPS pins down specific moments like post-onboarding or post-support, whereas relational NPS reads how people feel about the brand overall.

FAQ

Why ignore passives? The original framework only counts the extremes as something you can act on. Passives sit on the fence and aren't going to recommend you or warn people off.

Is NPS comparable across sectors? Only carefully. B2B SaaS tends to average 30-40, while banks and ISPs sit well below that. Hold yourself to your own sector's benchmark, not to Apple.

How big a sample do I need? For relational NPS, shoot for 200+ responses. Go smaller and your confidence intervals widen out, and the quarter-over-quarter numbers start jumping around on noise.