Net Promoter Score: The Metric That Flatters and the One That Fixes

Net Promoter Score is a single-question customer loyalty metric that asks respondents how likely they are to recommend your company to a friend or colleague, on a scale of zero to ten. Scores of nine or ten make someone a Promoter, seven or eight a Passive, and zero to six a Detractor. Your NPS is the percentage of Promoters minus the percentage of Detractors, giving you a number between -100 and +100.

It is one of the most widely used metrics in business. It is also one of the most misunderstood, most gamed, and most frequently mistaken for something it is not.

Where NPS Came From and Why It Spread So Fast

Fred Reichheld introduced Net Promoter Score in a 2003 Harvard Business Review article. The core argument was that customer loyalty, measured through a single question about likelihood to recommend, correlated strongly with revenue growth. The simplicity was the point. One question, one number, one benchmark you could track over time and compare across competitors.

It spread because it gave executives something clean to put in a board deck. Customer satisfaction data had always been messy, multi-dimensional, and hard to summarise. NPS collapsed all of that into a single integer. Boards loved it. Consultants sold it. Software platforms built dashboards around it. Within a decade, it had become the default customer metric for companies across virtually every sector.

I have sat in enough boardrooms to know that the appeal of a single number is almost irresistible. When I was running agencies and presenting to C-suite clients, the questions that got real attention were always the ones with clean answers. NPS fit that brief perfectly. Whether it answered the right question is a different matter.

If you are building out a fuller picture of how customers experience your brand, the Customer Experience Hub covers the broader framework, from satisfaction measurement to retention strategy, in one place.

How NPS Is Actually Calculated

The mechanics are straightforward. You ask: “On a scale of 0 to 10, how likely are you to recommend [company/product/service] to a friend or colleague?” Then you segment responses into three groups.

Promoters score nine or ten. These are your most loyal customers, the ones most likely to generate word-of-mouth referrals and least likely to churn. Passives score seven or eight. They are satisfied but not enthusiastic, and are vulnerable to competitive offers. Detractors score zero to six. They are unhappy customers who can actively damage your reputation through negative word of mouth.

Your NPS equals the percentage of Promoters minus the percentage of Detractors. Passives are excluded from the calculation entirely, which is worth noting because it means a shift from Passive to Promoter moves your score even if nothing about your Detractor population has changed.

Scores above zero are considered positive. Scores above 50 are considered excellent. Scores above 70 are world-class. But these benchmarks are heavily industry-dependent. A score of 40 in financial services is exceptional. A score of 40 in consumer software might be mediocre. Comparing your NPS to a competitor in a different sector is largely meaningless.

The Difference Between Transactional and Relational NPS

There are two distinct ways to deploy NPS, and conflating them is a common mistake that produces unreliable data.

Relational NPS is a periodic survey sent to your customer base, typically quarterly or annually, to get a read on overall loyalty and sentiment. It is designed to reflect the cumulative experience of the relationship, not any single interaction. This is the version most useful for strategic benchmarking and board reporting.

Transactional NPS is triggered by a specific event: a purchase, a support interaction, a product onboarding. It captures how a customer felt about that particular touchpoint. It is more granular and more actionable in the short term, but it is also more susceptible to recency bias. A customer who just had a brilliant experience with your help desk team will score you higher than their overall relationship with you might warrant. A customer who just had a billing dispute resolved badly will score you lower.

Neither version is wrong. But they answer different questions, and treating transactional NPS data as a proxy for overall loyalty will skew your understanding of where you actually stand.

Why the Score Alone Tells You Almost Nothing

Here is the part that gets glossed over in most NPS implementations. The number is not the insight. The number is a prompt to go find the insight.

I spent a period working with a retail client whose NPS had been hovering in the mid-30s for two consecutive years. The leadership team treated this as a stable, acceptable position. When we actually dug into the verbatim comments attached to the scores, a clear pattern emerged. A specific part of the post-purchase experience, the delivery communication process, was generating a disproportionate share of Detractor responses. The fix was operational, not marketing. Within two quarters of addressing it, the score moved to the low 50s. The number had been there all along. Nobody had read the comments carefully enough to act on them.

This is why well-designed customer feedback surveys treat the NPS question as the opening of a conversation, not the end of one. The follow-up open text field asking “what is the main reason for your score?” is where the commercial value actually sits. Without it, you are tracking a number without understanding what is driving it.

HubSpot’s breakdown of how to measure customer satisfaction makes this point well. Quantitative scores need qualitative context to be actionable. A score without a reason is just noise with a decimal point.

The Gaming Problem Nobody Talks About Honestly

NPS is one of the most gamed metrics in business, and the gaming often happens with the best intentions.

When customer-facing teams are measured and compensated on NPS, they find ways to influence the score. Staff ask customers directly for a “ten out of ten.” Survey invitations are timed to follow positive interactions. Detractor responses are excluded from reporting on the basis of being “outliers.” None of this is necessarily malicious. It is what happens when you tie incentives to a metric that is easy to manipulate.

I have judged the Effie Awards, where effectiveness is scrutinised hard, and the thing that strikes me about companies that genuinely perform well on customer metrics is that they are not optimising the measurement. They are optimising the experience. The score follows. When you reverse that, optimising the score rather than the experience, you end up with flattering data and a business that is quietly deteriorating underneath it.

The broader point is one I have come back to repeatedly across two decades of agency work. Marketing is frequently used as a blunt instrument to prop up companies with more fundamental problems. If a business genuinely delighted customers at every touchpoint, it would not need to chase its NPS. Growth would be a natural consequence of the experience. When NPS becomes a target rather than a signal, it stops functioning as either.

What NPS Does Not Measure

Understanding the boundaries of any metric is as important as understanding what it captures. NPS has several significant blind spots.

It does not measure actual behaviour. A customer who says they are a nine out of ten may never refer a single person to you. A customer who scores you a six may have been with you for fifteen years and have no intention of leaving. The correlation between stated likelihood to recommend and actual recommendation behaviour is real but imperfect, and it varies considerably by category.

It does not capture the full customer experience. A single score cannot tell you whether a customer’s experience deteriorated at acquisition, onboarding, ongoing service, or renewal. You need experience-level data to diagnose where problems are occurring, not a single aggregate number.

It does not tell you why customers leave. Churn is often driven by customers who were Passives, not Detractors. They were not unhappy enough to complain or score you badly. They simply found something better and left quietly. NPS does not capture that population’s motivations effectively.

And it does not account for structural market factors. If your entire industry has poor NPS scores, a mediocre score in your category might still represent a competitive advantage. Context matters enormously, and a single number strips it out.

How NPS Fits Into a Broader Measurement Framework

The question is not whether to use NPS. It is how to use it alongside other metrics so that you are getting a genuinely useful picture of customer health rather than a comforting headline figure.

There are several metrics that complement NPS well. Customer Satisfaction Score (CSAT) measures satisfaction with a specific interaction and is better suited to transactional feedback. Customer Effort Score (CES) measures how easy it was for a customer to accomplish something, which is a strong predictor of loyalty in service-heavy businesses. Churn rate and retention rate are the behavioural measures that NPS is supposed to predict, so tracking them in parallel tells you whether the relationship between stated loyalty and actual loyalty holds in your specific context.

If you are working through which of these metrics belongs in your reporting stack, this breakdown of customer satisfaction metrics covers the trade-offs clearly. The short answer is that no single metric is sufficient, and the companies that make the best decisions around customer experience are the ones tracking a small, coherent set of complementary measures rather than optimising one number in isolation.

HubSpot’s data on customer service benchmarks is worth reviewing in this context. The relationship between customer experience quality and commercial outcomes is well-documented, but the mechanisms are more nuanced than a single loyalty score can capture.

Closing the Loop: What to Do With NPS Data

The most important operational step in any NPS programme is what happens after the score is collected. This is where most programmes fail.

For Detractors, the priority is a rapid, personal response. Not an automated email acknowledging their score. An actual conversation, either by phone or a personalised message, from someone with the authority to resolve the issue. The goal is not to change their score. It is to understand what went wrong and fix it where possible. Done well, this turns a proportion of Detractors into advocates precisely because the recovery was handled so well.

For Promoters, the goal is activation. These are customers who have already indicated they would recommend you. Give them a reason to actually do it. A referral programme, a review request, an invitation to a customer community. The willingness is there. Most companies do nothing with it.

For Passives, the question is what would move them to a nine or ten. This group is often underinvested in because they are not generating complaints. But they represent the largest pool of potential Promoters, and understanding what is holding them back is often the highest-leverage insight your NPS programme can produce.

A well-configured customer engagement platform can automate much of this follow-up workflow, routing Detractor responses to the right team, triggering Promoter activation sequences, and segmenting Passive responses for deeper analysis. The technology is not the hard part. The hard part is building the internal process and accountability to act on what the data tells you.

NPS in a Paid Media Context

One area where NPS data is consistently underused is in paid acquisition strategy. If you have a clear picture of which customer segments produce the highest Promoter rates, you have a targeting brief. You know the profile of customers who become advocates, and you can use that to sharpen your acquisition targeting, your messaging, and your offer.

I have worked with clients managing substantial Google Ads budgets where the targeting strategy was built almost entirely around demographic and behavioural signals from the platform itself. When we overlaid NPS data segmented by customer type, the picture changed considerably. Certain segments that looked efficient on a cost-per-acquisition basis were generating disproportionate numbers of Detractors. Others that looked expensive to acquire were producing Promoters at a far higher rate. The total cost of acquisition, when you factor in lifetime value and referral behaviour, looked very different from the platform metrics alone.

If you are running paid search and want to think about how customer quality data can inform your Google Ads strategy, the connection between acquisition targeting and downstream loyalty metrics is worth building into your planning process from the start, not as an afterthought once the campaigns are already running.

Buffer’s perspective on customer experience as a business driver is useful here. The companies that connect acquisition data to retention data tend to make better decisions at both ends of the funnel.

The Honest Assessment: When NPS Is Worth Using

NPS is worth using when it is part of a properly designed feedback programme, when the verbatim data is read and acted on, when it sits alongside complementary metrics rather than replacing them, and when the organisation has the operational capacity to close the loop with customers.

It is not worth much when it is tracked quarterly, reported to the board, and then filed away until the next quarter. It is not worth much when the survey timing is manipulated to inflate scores. It is not worth much when nobody has ownership of acting on Detractor responses. And it is not worth much when it becomes a substitute for the harder work of actually improving the customer experience at a structural level.

The companies I have seen use NPS most effectively are the ones that treat it as a diagnostic tool with genuine consequences. When a score drops, someone is accountable. When a pattern emerges in the verbatim responses, it gets escalated to operations or product, not just marketing. When a Promoter segment is identified, there is a plan to grow it. That is NPS working as intended.

Personalisation in how you follow up with different NPS segments also matters more than most programmes acknowledge. Personalised follow-up communication to Detractors and Promoters performs significantly better than generic automated responses, both in terms of engagement and in terms of the quality of qualitative feedback you receive back.

The broader customer experience discipline is worth investing in properly, not just as a measurement exercise but as a strategic priority. Everything on the Customer Experience Hub is built around that principle: that experience quality is a commercial lever, not a soft metric, and that the companies which take it seriously tend to outperform those that treat it as a reporting obligation.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions