Net Promoter Score (NPS)

Net Promoter Score is a customer loyalty metric derived from a single question: how likely is the respondent to recommend the company to a friend or colleague, on a scale from zero to ten. Respondents scoring nine or ten are classified as promoters, seven or eight as passives, and zero to six as detractors. The score is the percentage of promoters minus the percentage of detractors, producing a number between negative 100 and positive 100. It was introduced by Fred Reichheld in 2003 and adopted very widely, particularly at executive level.

Its popularity rests on simplicity and comparability. One question is cheap to ask and easy to sustain over time, the resulting number is easy to report and track, and published benchmarks exist for most industries. For organizations that previously measured customer sentiment inconsistently or not at all, adopting a standard instrument represents a genuine improvement, and the discipline of asking regularly matters more than the specific formula.

The methodological criticisms are substantial and well established. Collapsing an eleven-point scale into three buckets discards information, and treating a seven as equivalent to a zero is difficult to justify. The same score can arise from entirely different distributions, so movement in the headline number can conceal offsetting changes. Comparative research has found that simpler measures such as mean satisfaction often predict business outcomes at least as well. Cross-cultural comparison is particularly unreliable, since response tendencies on numeric scales vary systematically between countries, making international NPS comparisons largely meaningless without adjustment.

The more damaging problems in practice are behavioral rather than statistical. When NPS becomes a target attached to compensation, it is gamed: staff coach customers on how to answer, surveys are sent selectively to satisfied customers, timing is chosen to follow positive interactions, and the score rises without any change in the underlying experience. Response rates are also typically low and skewed toward strong opinions, so the sample rarely represents the customer base.

Its value increases considerably when it is treated as a prompt rather than as a score. The follow-up question asking why a respondent gave their rating produces the actionable material, and the verbatim responses are usually worth more than the number they accompany. Segmenting by product, plan, tenure, region, and acquisition channel identifies where sentiment differs, which is where action can be taken. Closing the loop with detractors serves both recovery and diagnosis.

Where the score is genuinely useful is in tracking a single population over time under stable conditions, and even then the interpretation should focus on movement in the underlying distribution rather than in the headline figure. A stable score concealing a growing detractor group offset by a growing promoter group indicates a polarizing change, which is a different situation from a stable score with stable distribution and calls for a different response. Reporting the full distribution alongside the score, and segmenting it by the dimensions the business can act on, converts a number that invites arguments about methodology into a diagnostic that points somewhere specific.

The pragmatic position for most organizations is to keep NPS if it is already embedded and executives rely on it, but not to treat it as the primary measure of anything. It works best as one input alongside behavioral measures such as retention, repeat purchase, and expansion, which are harder to game and more directly connected to revenue. In a growth management programme, the verbatim comments feed the qualitative pipeline alongside other user research sources, while the score itself is best understood as a rough trend indicator rather than as evidence for any specific decision.