One question, a score from 0 to 10, a result between minus one hundred and plus one hundred. NPS's simplicity explains both its massive adoption and its status as one of marketing's most criticised metrics.
It is neither truth nor a scam. It is a thermometer: useful if you know what it measures, misleading if you make it say what it does not.
Definition
NPS, Net Promoter Score, rests on a single question: "how likely are you to recommend this product to a friend or colleague?", scored 0 to 10.
Answers fall into three groups:
| Score | Group | Reading |
|---|---|---|
| 9 and 10 | Promoters | Actively recommend |
| 7 and 8 | Passives | Satisfied but indifferent |
| 0 to 6 | Detractors | Risk of leaving and bad word of mouth |
NPS = percentage of promoters minus percentage of detractors
Passives count in the total but not in the subtraction, which is why a product with decent satisfaction often shows a modest score.
Its three real limits
It does not say why
A score of 20 teaches you nothing actionable. The open question asked immediately after, "what explains your score?", holds all the value. Without it you are collecting a decorative number.
It measures intent, not action
Saying you would recommend and actually recommending are different things. Real Churn rate and the share of customers arriving through referral measure the same intent, but after the fact and without self-reporting.
Cross-sector comparisons are worthless
Rating habits vary sharply between countries and sectors. Comparing your score to a published average is meaningless; comparing it to your own last quarter is not.
Using it without fooling yourself
- Ask at the right moment. After real usage, never right after purchase where you measure hope rather than experience.
- Follow the trend. The absolute value means almost nothing, its movement means a lot.
- Handle detractors individually. A customer scoring 3 who explains why is worth ten anonymous surveys.
- Cross it with Retention. A good NPS with poor retention signals a product people like but do not need.
Frequently asked questions
What score should I aim for?
Positive to start with, then improving. Chasing a score published by a sector giant wastes time: customer bases and rating habits are nothing alike.
How often should I send it?
Twice a year is plenty for a small business. Beyond that you tire your customers and mostly measure their weariness at being asked.
Can I use it as Social proof?
A score alone convinces few people, because it is unverifiable and abstract. Promoter comments, on the other hand, make excellent testimonials, provided you get the author's permission before publishing.
What should I do about a falling score?
Read the comments before touching the product. A drop almost always follows an identifiable change: a price rise, a removed feature, a longer response time. The cause is usually in the written answers.