NPS
Also: NPS, Net Promoter Score, Net Promoter, Score Net de Recommandation, taux de recommandation net
Net Promoter Score: a loyalty metric asking how likely customers are to recommend you (0 to 10). NPS = % Promoters minus % Detractors.
What it is
Net Promoter Score (NPS) is a customer loyalty and satisfaction metric based on a single question: *"How likely are you to recommend [company/product] to a colleague or friend?"* Respondents answer on a 0 to 10 scale.
Responses are grouped into three segments:
- Promoters (9 to 10): loyal enthusiasts likely to refer others and repurchase.
- Passives (7 to 8): satisfied but unenthusiastic, vulnerable to competitors.
- Detractors (0 to 6): unhappy customers who can damage your brand through negative word of mouth.
The score is calculated as:
NPS = % Promoters minus % Detractors
Passives are counted in the total base but excluded from the subtraction. The result ranges from -100 to +100. Note that NPS is a whole number, not a percentage, despite being derived from percentages.
Why it matters
NPS is popular because it is simple, comparable across time and (with caution) across peers, and correlates in many industries with retention and growth. It gives executives a single directional signal that is easy to communicate to boards and teams.
Its limitations matter too: it is a lagging indicator, sensitive to survey timing and sample bias, and a raw score reveals little without the follow-up "why?" question that surfaces root causes.
How it is used in practice
- Relationship NPS: measured periodically (quarterly) to track overall brand loyalty.
- Transactional NPS: triggered after a specific interaction (purchase, support ticket) to diagnose touchpoints.
- Benchmarking: compared against your own trend line first, then against industry norms.
- Driver analysis: verbatim comments are coded (increasingly with LLMs) to explain the number.
Worked example
You survey 200 customers and receive 150 responses:
- 90 Promoters (9 to 10) = 60%
- 30 Passives (7 to 8) = 20%
- 30 Detractors (0 to 6) = 20%
Calculation:
NPS = 60% minus 20% = +40
A score of +40 is strong in most sectors. If next quarter Detractors rise to 40% and Promoters fall to 45%, NPS drops to +5, signaling an urgent problem to investigate through the verbatim feedback.
Cautions
- Always pair the score with the open-ended "why" question.
- Report the response rate and sample size alongside the score.
- Avoid gaming (staff nudging customers toward 10) which corrupts the signal.
See also
Frequently asked questions
What is NPS and how is it calculated?
NPS (Net Promoter Score) is a loyalty metric built on one question: how likely a customer is to recommend a company or product on a 0 to 10 scale. Responses split into Promoters (9-10), Passives (7-8) and Detractors (0-6), and the score equals % Promoters minus % Detractors. Passives count in the base but drop out of the subtraction, so the result lands anywhere between -100 and +100.
Is NPS a percentage?
No. NPS is a whole number on a scale from -100 to +100, even though it is derived from two percentages. Writing "our NPS is 40%" is incorrect; the right form is "+40".
What counts as a good NPS score?
A score of +40 is strong in most sectors, but the absolute number matters less than your own trend line. Compare each measurement to your previous ones first, then to industry norms, since response scales and survey habits differ widely between markets. A drop from +40 to +5 in one quarter is a far more useful signal than any external benchmark.
What is the difference between relationship NPS and transactional NPS?
Relationship NPS is measured periodically, typically quarterly, to track overall brand loyalty. Transactional NPS is triggered right after a specific interaction such as a purchase or a support ticket, to diagnose that touchpoint. The first tells you where you stand, the second tells you which moment of the journey is hurting.
Why is a raw NPS score not enough to act on?
Because the number tells you the direction, not the cause. NPS is a lagging indicator, sensitive to survey timing and sample bias, so it should always travel with the open-ended "why" question, the response rate and the sample size. Coding the verbatim comments, increasingly with LLMs, is what turns a score into a list of things to fix.