Net Promoter Feedback Loop
Measure loyalty, learn why, and ask your strongest advocates what to improve
- Difficulty
- Easy
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 99%
The Net Promoter Feedback Loop starts with one behavioural proxy: how likely a customer is to recommend the organisation to a friend on a zero-to-ten scale. Scores separate promoters from passives and detractors, but the number is only the entry point. A verbatim follow-up reveals why the experience enriched or diminished the customer's life, and a third question asks what would have made it better. The responsible team reads those words directly, preserving the emotion and context that automated aggregation can erase. The result is a compact operating loop: identify the relationship state, understand its cause, and turn customer language into recovery or innovation work. The survey stays capped at three questions so the measurement itself does not become excessive friction.
Origin
Fred Reichheld developed the system after seeking a simple predictor of loyalty behaviour. Enterprise Rent-A-Car's two-question branch survey helped demonstrate the power of a short feedback loop, and his daughter Jenny persuaded him to add the third improvement question.
Core principles
- 01Keep feedback requests short enough to respect customers' time
- 02Treat recommendation intent as a proxy for loyalty behaviour
- 03Read customers' own words rather than reducing emotion to statistics
- 04Use promoters as partners in innovation
How to run it
- 1
Ask the loyalty question
Ask how likely the customer is to recommend the organisation to a friend, using a zero-to-ten scale. Keep the request focused rather than surrounding it with a long market-research survey.
Pro tip Tie the question to a recent, identifiable experience so the response can reach the right team.
Watch out Do not pressure customers to award a ten.
- 2
Classify the relationship
Treat nines and tens as promoters, sevens and eights as passives, and zeroes through sixes as detractors. Use the groups as a simple signal of enrichment, adequacy, or failure.
Watch out The category is a signal, not the root cause.
- 3
Capture the reason
Ask an open-text follow-up that lets the customer explain the score in their own words. Have people responsible for the experience read the response directly.
Pro tip Look for the emotion behind an enriched or diminished experience.
Watch out Do not delegate all interpretation to statistics or automation.
- 4
Invite a better idea
Ask what else the organisation could have done to improve the experience. Promoters are especially useful here because they already want the organisation to succeed.
Pro tip Mine promoter answers for ideas that could create even more enthusiastic referrals.
Watch out Do not inflate the survey beyond three questions.
- 5
Act where the experience happened
Send recovery and improvement work to the branch, product, or team that created the experience. Track whether changes produce more promoters, repeat purchases, and referrals.
Pro tip Discuss verbatims in regular team huddles.
Watch out A headquarters-only response prevents learning at the point of service.
In the wild
Andy Taylor used a two-question survey across Enterprise Rent-A-Car branches. Reichheld saw how a short survey could make customer loyalty visible at the local level and used that insight in the development of Net Promoter.
→ A simple branch-level practice helped inspire a system later adopted by thousands of organisations.
A software customer scores an experience ten, explains that onboarding was unusually clear, and suggests adding a reusable setup checklist. The product team tests the checklist with new users rather than adding more generic survey questions.
→ The team converts advocate goodwill into a focused product experiment.
Common mistakes
Begging for tens
Pressuring customers corrupts the signal and turns the score into theatre rather than evidence of loyalty.
Automating away the verbatim
Reducing open-text feedback to aggregate statistics keeps responsible people from feeling and understanding the customer's experience.
Allowing question inflation
Adding five, ten, or dozens of questions creates friction and lowers the quality of participation.
Is it for you?
Best for
It is best for organisations that can connect customer feedback to a specific product, branch, team, or interaction.
Not ideal for
It is not ideal for teams seeking a perfectly comparable market benchmark from a self-selected customer survey.
From the transcript
“How likely would you recommend us to a friend? So, zero through 10, we asked that question.”
“Then, we have an open text verbatim follow-up to tell us why.”
“is there anything else we could have done to make your experience better?”
From the episode
Fred Reichheld: Love Your Customers
Fred Reichheld