Opens in a new tab
3 4 5 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

What is Net Promoter Score – NPS

Net Promoter ScoreDefinition:

The Net Promoter Score (NPS) is a metric that summarizes a person’s stated intention to recommend a company, product, or service. It is obtained from a question rated from 0 to 10 and classifies responses as promoters, passives, or detractors.

NPS is used as a relationship indicator within customer experience programs. It does not directly measure overall satisfaction, actual loyalty, or future behavior: it records an answer to a specific question from a sample, through a channel, at a particular time.

Origin and purpose of NPS

Fred Reichheld introduced Net Promoter Score in the Harvard Business Review article “The One Number You Need to Grow”, published in 2003. The method was developed with Bain & Company and Satmetrix to provide a summary measure of customers’ willingness to recommend.

The metric was later incorporated into the Net Promoter System, a broader framework for collecting and following up on feedback. NPS can form part of a Voice of Customer program, but a score alone does not explain the reasons behind responses or replace qualitative analysis.

How NPS is calculated

The usual question is: “How likely are you to recommend this company, product, or service to a friend or colleague?”. The respondent answers on a 0 to 10 scale and is assigned to one of these categories:

  • Promoters: give a score of 9 or 10.
  • Passives: give a score of 7 or 8.
  • Detractors: give a score from 0 to 6.

The formula is % promoters minus % detractors. Passives remain part of the response total used to calculate both percentages, although they are not added or subtracted directly. The result is reported as a number from -100 to +100, not as a percentage.

For example, if 60% of responses come from promoters, 20% from passives, and 20% from detractors, NPS is 60 – 20 = 40. The calculation provides a summary distribution, but it hides differences between responses within the same category.

Types and applications

Relational NPS is collected periodically to observe the overall perception of a customer’s relationship with an organization. Transactional NPS is requested after a specific interaction, such as a purchase or support conversation. In the latter case, the immediate context can strongly influence the answer.

The metric can be analyzed by product, market, segment, or stage of the Customer Journey, provided there are enough responses and the method remains comparable. It can also be added to a customer’s CRM record to connect the response with context, subject to the stated purpose and applicable privacy criteria.

A survey often adds an open-ended question about the reason for the score. These comments help identify themes and form hypotheses, while the numeric score supports aggregate tracking. Neither component demonstrates on its own which change will produce an improvement.

Interpretation and comparison

A positive result means that promoters make up a larger percentage than detractors; a negative result indicates the opposite. However, there is no universal value that is good for every industry, country, product, and type of relationship.

The most consistent comparison uses the same question, population, channel, timing, and calculation rules over time. Comparing figures from different organizations can be misleading when respondent profiles, response rates, or the survey design differ.

NPS concerns recommendation intention, whereas CSAT asks about satisfaction with an experience and other metrics observe effort, retention, or churn. Customer loyalty is a broader phenomenon that includes behavior, duration, and relationship value. These measures are complementary, not interchangeable.

Limitations and data quality

The simplicity of the calculation does not remove the problems inherent in survey research. Several quality factors must be considered when interpreting the result:

  • Sample: respondents may not represent the full customer population.
  • Nonresponse: people who ignore the survey may have different experiences from those who participate.
  • Timing and channel: a recent incident or the delivery method can change the score.
  • Cultural context: the use of the ends of a rating scale can vary between markets.
  • Sample size: a small number of cases produces large swings and unstable estimates.

Recommendation intention does not necessarily equal an actual recommendation, repeat purchase, or revenue increase. A correlation between NPS and growth does not show that the score caused that growth or guarantee predictive power in every context. Its interpretation therefore requires additional evidence, open-ended feedback, and behavioral metrics.