Net Promoter Score (NPS) is a simple but powerful tool for measuring customer loyalty and predicting business growth. Introduced in 2003 by Fred Reichheld of Bain & Company, NPS has been the standard in both B2B and B2C for years — used by Apple, Amazon, Microsoft and thousands of companies worldwide.
When customers have hundreds of options and a single negative review on Google can kill a deal, NPS is not just a survey — it is your early warning system against churn and a roadmap for generating more referrals.
History and origin of NPS
NPS was created in 2003 as a response to overly complicated satisfaction surveys (hundreds of questions, low response rates). Fred Reichheld published his findings in the Harvard Business Review ("The One Number You Need to Grow"), showing that one question correlated with revenue growth better than 242 other survey questions.
That question: "How likely are you to recommend our company/product/service to a friend or colleague?" (scale 0–10)
Since then, NPS has been used by two-thirds of Fortune 500 companies. It has evolved significantly — in 2026, it integrates with AI (churn prediction) and real-time feedback triggered after every customer interaction.
How NPS works — categories and formula
The survey asks one question on a 0–10 scale. Based on responses, customers fall into three categories:
- Promoters (9–10) — loyal fans who return, buy more, and refer others; they generate 2–5× more revenue than average customers
- Passives (7–8) — satisfied but not enthusiastic; vulnerable to switching to competitors
- Detractors (0–6) — dissatisfied customers who won't return and may leave negative reviews; one negative review deters an average of 10 potential customers
NPS formula:
NPS = % Promoters − % Detractors
Range: −100 to +100.
Example (100 responses): 60 promoters (60%), 30 passives (30%), 10 detractors (10%) → NPS = 60 − 10 = +50
Why NPS matters
NPS matters because it predicts business outcomes, not just current satisfaction:
- Churn prediction — declining NPS is an early warning sign that customer retention will worsen in the coming months
- Revenue growth correlation — companies with NPS in the top quartile of their industry grow at more than twice the rate of competitors
- Word-of-mouth measurement — NPS captures the likelihood of organic referrals, which are the highest-quality, lowest-cost acquisition channel
- Benchmarking — NPS allows comparison against industry peers and historical performance over time
- Employee alignment — sharing NPS data with teams creates customer-centric culture and accountability
Companies in the top NPS quartile grow at more than twice the rate of the average in their industry. — Bain & Company research
How to measure NPS step by step
- Choose your survey method — email surveys (post-purchase or periodic), in-app surveys, SMS, or live interviews
- Add a follow-up question — always follow the rating with "What is the main reason for your score?" This qualitative data tells you what to fix
- Define your sample — minimum 30 responses for statistical significance; 100+ for reliable trend analysis
- Time it right — send surveys shortly after a key interaction (delivery, onboarding, support resolution), not months later when the experience has faded
- Calculate your NPS — apply the formula: % Promoters minus % Detractors
- Segment results — break down NPS by customer segment, product line, geography, or account manager to identify patterns
- Close the loop — follow up with detractors personally to resolve issues; thank promoters and ask for referrals
Interpreting your NPS results
| NPS Score | Interpretation | Typical examples |
|---|---|---|
| +70 and above | World class | Apple, Amazon, Tesla |
| +50 to +70 | Excellent — industry leader | Top SaaS companies |
| +0 to +50 | Good — B2B average | Most healthy businesses |
| 0 to −10 | Warning — action required | Companies with CX issues |
| Below −10 | Crisis — losing customers faster than gaining | Companies at risk of churn spiral |
Note: NPS benchmarks vary by industry. A +30 NPS in financial services (competitive market) is excellent. A +30 in software may be below average. Always compare against your industry peers, not just a generic scale.
How to improve NPS — 10 concrete steps
- Respond to every detractor within 24–48 hours — personal outreach to resolve issues shows customers they matter and can convert some detractors into passives or even promoters
- Identify root causes — use follow-up question responses to identify the top 3 reasons for low scores; prioritize fixing these systematically
- Improve onboarding — most churn happens early; a stronger onboarding process often yields the fastest NPS improvement
- Reduce response times — slow support is consistently among the top reasons for low NPS scores; set and enforce response time SLAs
- Proactively communicate issues — don't wait for customers to discover problems; inform them proactively; this dramatically reduces detractor formation
- Create a promoter program — incentivize promoters to leave reviews and refer others through referral programs, case studies and testimonial campaigns
- Train your team on NPS data — share NPS results with everyone who touches the customer; connect team bonuses to NPS improvement
- Segment and prioritize — focus improvement efforts on your highest-value customer segments first; a detractor with $500K ARR deserves more attention than one with $5K
- Measure NPS quarterly — too frequent measurement dilutes signal; quarterly is sufficient for trend analysis
- Benchmark against competitors — if your NPS is +40 but the industry average is +55, you have a competitive disadvantage that needs addressing
NPS by industry
NPS benchmarks and improvement tactics vary significantly by sector. Here is how NPS applies across the most common industries:
- SaaS B2B — measured after demo and onboarding; average +35–55. Key improvement lever: faster time-to-value in the first 30 days.
- E-commerce — measured after purchase and delivery; average +20–45. Key improvement lever: reducing cart abandonment and improving post-purchase communication.
- IT agencies and consulting — measured after project completion; average +25–50. Key improvement lever: better quarterly business reviews (QBRs) and proactive communication.
- Manufacturing and machinery — measured after implementation and service events; average +15–40. Key improvement lever: service training and faster response to technical issues.
- Real estate developers — measured after property handover; average +30–55. Key improvement lever: consistent and transparent communication throughout the buying journey.
NPS vs other metrics (CSAT, CES, Churn, LTV)
| Metric | Measures | Best used for | Limitation |
|---|---|---|---|
| NPS | Overall loyalty & referral intent | Business health, growth prediction | Doesn't explain why |
| CSAT | Satisfaction with specific interaction | Support tickets, transactions | Short-term, not predictive |
| CES | Ease of customer effort | Self-service, checkout, support | Narrow scope |
| Churn rate | Customers lost in a period | SaaS, subscriptions | Lagging indicator |
| LTV/CLV | Total revenue per customer | Revenue forecasting | Complex to calculate accurately |
The best practice is to use NPS as your primary loyalty metric, complemented by CSAT for transactional moments and CES for self-service interactions. Together they give a complete picture of customer experience. Note: NPS below +20 correlates with churn rates above 10%, while promoters generate 2.6× higher LTV than average customers.
See also: Most important indicators in the sales process
The future of NPS — AI and real-time
- AI-powered prediction — systems such as Gong and HubSpot already predict NPS based on call recordings and email sentiment, with 80–90% accuracy. You can identify detractors before they complete the survey.
- Real-time NPS — measurement triggered after every customer touchpoint (call, email, chat) enables response within less than one hour, turning potential detractors into neutrals or promoters before churn occurs.
- Multichannel integration — NPS signals collected from LinkedIn, WhatsApp and other channels give a richer picture of customer sentiment beyond traditional email surveys.
- AI benchmarking — automatic industry comparison in real time (e.g. your NPS +45 vs SaaS industry average +52) helps prioritize improvement efforts with competitive context.
In 2026, NPS is not just a number — it is your customer loyalty radar. Measure it, analyze it, improve it — and you will see growth in referrals and lifetime value.
FAQ
How often should I run NPS surveys?
Quarterly relationship NPS surveys work best for most businesses. Additionally, run transactional NPS after key moments (purchase completion, support interaction, onboarding milestone). Avoid surveying customers more than once every 90 days to prevent survey fatigue.
What is a good NPS score?
Any score above 0 means you have more promoters than detractors. A score above +50 is generally considered excellent. However, what matters most is the trend over time and your performance relative to industry benchmarks.
Why is my NPS low even though customers seem happy?
There is often a gap between stated satisfaction and actual loyalty. Customers may rate service as "good" but not feel strongly enough to recommend you. Focus on creating exceptional, memorable experiences — not just avoiding dissatisfaction.
Should I use NPS for B2B?
Yes — NPS is highly effective for B2B, where relationship quality directly drives renewal and expansion revenue. In B2B, survey key stakeholders and decision-makers separately, as perceptions often differ within the same account.
Can I tie NPS to employee compensation?
Yes, and it can be highly motivating. However, avoid tying individual NPS scores to individual pay — this creates incentives to game the system. Instead, tie team or company-wide NPS improvement to bonuses.