Net Promoter Score
Also known as: NPS, customer loyalty score, customer satisfaction score (related), recommend score
Net Promoter Score (NPS) is a widely used customer loyalty measurement system developed by Fred Reichheld and Bain & Company in 2003.
It is based on asking customers one question: 'On a scale of 0-10, how likely are you to recommend this company/product to a friend or colleague?' Responses categorize customers as Promoters (9-10), Passives (7-8), or Detractors (0-6). NPS = % Promoters - % Detractors. Scores range from -100 to +100. NPS is used across industries as a leading indicator of customer retention and revenue growth, though its predictive value and methodology are debated among researchers.
Net Promoter Score, explained at three levels
At home
NPS is a simple customer satisfaction score based on one question: 'How likely are you to recommend us to a friend, on a scale of 0 to 10?' People who answer 9 or 10 are called Promoters (they love you and spread the word). People who answer 0 to 6 are called Detractors (they're unhappy and might warn others away). The NPS is calculated by subtracting the percentage of Detractors from the percentage of Promoters: the result is a score between -100 and +100.
At work
Companies use NPS to track customer loyalty over time and benchmark against industry standards. A score above 0 means more customers would recommend you than warn others away. A score above 50 is considered excellent in most industries. More importantly, companies use the follow-up open text question ('What's the main reason for your score?') to understand what's driving satisfaction or dissatisfaction, making NPS both a measurement tool and a qualitative feedback system.
At home
The NPS calculation: survey 100 customers, 40 score 9-10 (Promoters = 40%), 30 score 7-8 (Passives: excluded from NPS calculation), 30 score 0-6 (Detractors = 30%). NPS = 40 - 30 = 10. Industry benchmarks vary widely: consumer apps average NPS of 30-40; banking averages 20-35; B2B software ranges from 20-60. Comparing your NPS against your own historical trend and against industry benchmarks provides more context than the absolute number alone.
At work
NPS implementation variations affect data quality significantly: transactional NPS (surveys triggered after a specific event like a support ticket closure or a purchase) measures satisfaction with a specific interaction; relationship NPS (surveyed periodically: quarterly or annually) measures overall relationship health. Transactional NPS tends to be more actionable (you can close the loop with dissatisfied customers quickly) but reflects one event rather than the overall relationship. Survey timing, channel, and frequency all affect response rates and score accuracy.
At home
The deeper trade-off is that NPS compresses an 11-point scale into three buckets, so a lot of information disappears. Moving a customer from 0 to 6 changes nothing, while moving one from 8 to 9 shifts the score. Small samples swing wildly, because each response moves the result by several points, and researchers have found other satisfaction measures predict growth just as well. Its real strength is the follow-up question about why. In my view, NPS is useful as a conversation starter and weak as a scoreboard.
At work
Statistical validity of NPS surveys is a common concern: low response rates (often 10-20%) create selection bias: respondents are typically either very satisfied or very dissatisfied, inflating both Promoter and Detractor segments. Sample size requirements for statistically significant NPS are large: a 10-point NPS change needs hundreds of responses to be statistically significant. Enterprise NPS programs use statistical significance thresholds and confidence intervals in their reporting. NPS correlates with revenue retention metrics (net revenue retention, gross revenue retention) in B2B SaaS, which can be used to validate that your NPS is predictive of business outcomes in your specific context.
Real-world examples
At home
Sophie ran a yoga studio and sent a one-question survey via email after each new member's first month: 'On a scale of 0-10, how likely are you to recommend our studio to a friend?' Members who answered 9 or 10 received a follow-up offering a free referral card for a friend's first class. Members who answered 0-6 received a personal email from Sophie asking what could be improved. Over six months, she turned three unhappy members into loyal regulars through this follow-up approach.
At work
Thomas ran NPS as a quarterly measurement across his B2B SaaS company's 800 customers. He tracked NPS by customer segment, tenure, product tier, and industry. Detractors who didn't receive a call had a 40% higher churn rate than those who did receive a personal follow-up within 48 hours. He used NPS data to prioritize customer success team outreach and to surface product feedback themes to the product roadmap process.
Common confusions
- Many companies treat NPS as a target to maximize rather than a diagnostic tool — Gaming NPS (surveying only happy customers, timing surveys right after a good experience, pressuring customers for high scores) produces a meaningless number that looks good but provides no actionable insight. NPS is valuable as a diagnostic tool: understanding why customers score as they do and using that to improve. A company with a genuine NPS of 45 that understands why is more valuable than one with a surveyed NPS of 65 that doesn't know what drives it.
- Some believe NPS alone is sufficient to predict customer retention and growth — Research has consistently shown NPS correlates with but doesn't perfectly predict revenue outcomes. Usage data (product engagement), renewal rate, expansion revenue, and customer health scores (in B2B) are more directly predictive of retention than NPS alone. NPS is best used as one component of a broader customer health model, particularly valuable for capturing qualitative feedback on what customers value and what frustrates them.
For analysts
When a client asks for help improving their NPS, the first question is whether they have a closed-loop process for following up with detractors. Companies that collect NPS but don't act on it are spending survey respondents' goodwill without generating value. The highest-ROI NPS improvement comes from operationalizing follow-up: a workflow that notifies a customer success manager immediately when a detractor response arrives, with a 48-hour follow-up target.
History & origin
NPS was introduced by Fred Reichheld in a 2003 Harvard Business Review article titled 'The One Number You Need to Grow,' and expanded in his 2006 book 'The Ultimate Question.' Bain & Company and Satmetrix commercialized NPS as a methodology. The simplicity of a single-question survey drove rapid adoption: companies replaced lengthy satisfaction surveys with one question. By the 2010s, NPS had become a standard board-level metric in enterprise companies, tied to executive compensation in some organizations. Academic research has challenged NPS's predictive validity, showing other customer satisfaction metrics correlate equally or better with growth in many contexts. Fred Reichheld introduced NPS 3.0 (Earned Growth Rate) in 2021, focusing on revenue earned from existing customers vs. new customer acquisition, as a more rigorous companion metric.