How NPS works
Net Promoter Score comes from one question: how likely are you to recommend us to a friend or colleague, answered on a scale running from zero to ten. Answers are then sorted into three groups. Nines and tens are counted as promoters, sevens and eights as passives, and everything below that as detractors. The score is the share of promoters minus the share of detractors. Passives are dropped from the sum entirely, which is why the result can be negative and why it moves in bigger jumps than the underlying answers do.
The question is almost always followed by an open box asking why. That free-text answer is the part worth having. The score tells you the direction of travel; the comments tell you what to change.
Why NPS matters
Its value is consistency rather than accuracy. Because the question and the grouping never change, you can watch the same measurement move across months, across markets and across product lines without arguing about methodology. That makes it a reasonable early warning system: sentiment usually turns before renewals and repeat orders do.
It also gives non-marketing colleagues something they can hold. A finance director or a clinic owner will engage with one number and a page of customer sentences far more readily than with a dashboard of channel metrics, and that engagement is often what gets a fix funded.
Common mistakes with NPS
The biggest is who answers. If the survey goes only to people you enjoyed working with, or only to customers who are still active, the score measures your sampling rather than your service. Timing does the same damage: asking immediately after a good delivery and never after a complaint produces a number that cannot fall.
The second is treating the scale as universal. How people use a nought-to-ten scale differs by language, culture and market, so an international business comparing Nepal against Australia on one league table is often comparing survey habits, not satisfaction. Compare each market against its own history instead.
The third is worth stating plainly: using the score to decide who gets invited to leave a public review is review gating, and it breaks Google’s review policies. Ask everyone or ask no one.
How to act on it
Read the comments first and the number second. Group the free-text answers into causes, count how often each cause appears, and take the largest one to whoever owns that part of the business. A score with no cause attached to it cannot be acted on by anybody.
Close the loop with detractors individually — a reply and a fix recovers more revenue than a reporting slide does. Then pair the trend with a measure of what people actually did, such as retention rate, and keep any genuine review-collection work inside a proper reputation management process rather than bolting it onto the survey.