Startups & Business › Product-Market Fit

Net Promoter Score

A survey score for how likely customers are to recommend you, and its limits.

Also known as: Net Promoter Score, NPS, net promoter

Net Promoter Score (NPS) asks “how likely are you to recommend us?” on a 0–10 scale: 9–10 are promoters, 7–8 passive, 0–6 detractors; score = %promoters − %detractors. One number, trackable over time, comparable across the industry — and dangerously easy to game, misunderstand and over-weight.

100 responses:  60 promoters − 20 detractors = NPS 20
(same score from wildly different distributions — always read the split, not just the number)

Its legitimate uses: trending loyalty over time within one company, segmenting follow-up (call every detractor), and benchmarking loosely against industry ranges. It predicts referral behavior better than satisfaction scores do — intention to recommend correlates with acting on it more than “satisfaction” ever did.

The classic mistakes:

• Managing to the number. Teams optimizing survey scores (begging for 10s, timing asks post-delight) inflate NPS while the product stagnates. It is a thermometer; polishing it does not cure fever.
• Cultural blindness. Response styles differ across countries — some cultures avoid extremes, compressing scores structurally. Calibrate per market, especially across Southeast Asia’s diversity.
• Ignoring the passives. Strategy lives in moving 7s to 9s and 6s upward; obsessing over the headline while the distribution rots underneath misses the work.
• Survey spam. NPS-ing every interaction trains customers to ignore you, biasing responses toward the angry and the delighted. Cadence it (quarterly relationship surveys plus transactional triggers), and respect fatigue.
• NPS as the only loyalty metric. It measures stated intent, not behavior. Pair with retention and actual referral rates (word of mouth) — behavior overrules statements.

Use it for: trends, detractor rescue lists, and segment comparisons — never as a target that determines bonuses, and never alone. See Sean Ellis for the sharper fit question.