What Is NPS? A Complete Guide to Net Promoter Score

Imagine you run two companies. Both have an NPS of 35. But Company A has 50% Promoters and 15% Detractors, while Company B has 60% Promoters and 25% Detractors. Same score, completely different situations — and only one is growing sustainably. The number alone tells you almost nothing.

Net Promoter Score was created by Fred Reichheld and Bain & Company in 2003. Today, two-thirds of Fortune 1000 companies use it. This guide will show you not just how to calculate NPS, but how to use it to actually drive growth.

How NPS Works

NPS is built on a single question: "On a scale of 0-10, how likely are you to recommend us to a friend or colleague?" Simple to ask, powerful to analyze. Based on their score, every respondent falls into one of three groups:

• Promoters (9-10): Loyal enthusiasts who keep buying and actively refer others. They spend more, churn less, and do your marketing for you.
• Passives (7-8): Satisfied but unenthusiastic. They won't badmouth you, but they're one competitor offer away from leaving.
• Detractors (0-6): Unhappy customers who can damage your brand through negative word-of-mouth. One viral complaint can undo thousands of dollars of marketing spend.

How to Calculate NPS

The NPS formula looks simple, but the math trips people up because Passives disappear from the equation entirely. Here's exactly how it works:

What's a Good NPS Score?

Context is everything. An NPS of 30 is exceptional in telecom and mediocre in software. Before benchmarking against the world, benchmark against your own industry first.

• Technology / SaaS: 35–50 is good; 50+ is excellent
• E-commerce: 45–60 is good; 60+ is excellent
• Financial Services: 25–40 is good; 40+ is excellent
• Healthcare: 50–70 is good; 70+ is excellent
• Telecommunications: −5 to 15 is average; 25+ is excellent
• Retail: 40–55 is good; 60+ is excellent

Why the Number Alone Isn't Enough

Here's the fundamental problem with traditional NPS surveys: a score of 4 from a long-term enterprise customer who just had a terrible support experience means something completely different than a 4 from a new user still figuring out your product. But if all you collect is the number, you can't tell them apart.

Going Beyond the Number with AI

This is where AI-powered NPS surveys transform the metric from a thermometer into a diagnostic tool. Instead of a generic "Why did you give that score?" text box (which most people skip), AI asks personalized follow-ups based on the specific score each customer gave.

• Score 0–3: "What's the primary issue you've experienced with us?"
• Score 4–6: "What's the main thing preventing you from rating us higher?"
• Score 7–8: "What would need to change to make you a strong promoter?"
• Score 9–10: "What do you love most about our product? Can we share your experience?"

Traditional NPS surveys stop at the number. AI-powered NPS surveys uncover the story behind every score, giving you 3-5x more actionable insights from the same number of respondents.

Implementing NPS Effectively

NPS is only as useful as the systems you build around it. Collecting the score is step one — here's what the rest looks like:

• Survey at the right moments: Post-purchase, post-support, post-onboarding, and quarterly relationship surveys each capture a different slice of the experience.
• Close the loop with Detractors: Contact every Detractor within 48 hours. One saved customer is worth more than ten new ones — and this alone can move your NPS 5-10 points.
• Track trends, not one-offs: Month-to-month NPS can be noisy. Look at 3-month rolling averages and compare quarter-over-quarter.
• Segment your scores: Break NPS down by customer type, product line, or tenure. Your average score hides the real story.