Organizations believe it is a good determinant of customer satisfaction. It not only tracks who among an organization’s audience are loyal customers, but also those who help the organization scale through positive word of mouth.

A low score might mean that there are too many detractors, which will hinder the business in scaling up with more customers. It can also track the expected rates of negative word of mouth, given that some people who do not recommend an organization’s products might instead tell people to not buy them.

Developed by Fred Reichheld, software provider Satmetrix and the business consultancy Bain & Company, Inc., NPS asks users who have experienced a touchpoint how likely they are to recommend the company to others. It was introduced in a well-known 2003 Harvard Business Review (link resides outside of IBM.com) article “The One Metric You Need to Grow” and has since become an important tool to understand customer relationships.¹

It originated when former Enterprise Car Rentals CEO Andy Taylor shared a case study with an audience. He said the company measured customer loyalty by asking two questions of its customers. Those questions involved how the customers ranked the quality of their experience and whether they would use the company’s services again.

Reichheld built upon that simplicity and created NPS, which asked people whether they would recommend a product or service to their social groups. Bain & Company has evolved the NPS to a Net Promoter System, which it uses to help companies better serve their customers across all aspects of the customer experience.

The NPS methodology is straightforward. Organizations ask customers to rank how likely they would recommend a company’s product or service in an NPS survey. Respondents reply by using a scaled answer, often 1–10.

Different responses mean different things. Companies consider customers who scored 9’s and 10’s to be promoters, 7’s and 8’s are often treated as passives, and 6 or under are seen as detractors.

The Net Promoter Score calculation formula involves subtracting the number of low scores (6 or under) from the number of “promoters” (9’s and 10’s) and the net is converted into a percentage. This percentage can range from -100 (all 6’s or less) to +100 (all 9’s and 10’s). The best performers receive an average NPS score of 80 or better.

Good net promoter scores often demonstrate a customer base with enthusiasm for the company. Companies can use an NPS template or develop their own. Some companies choose to include some open-ended questions to dig deeper into which product features they customers like. Sample survey questions of this nature include asking why people chose their specific NPS score and what a company can do to improve it.

Every industry is slightly different in baseline NPS scores, so it’s important to benchmark against industry norms. A 2024 Retently study (link resides outside of IBM.com)² found significant differences between NPS averages across industries. For example, the average NPS for an insurance company is 80, whereas it is 39 for cloud and hosting. Companies that fall below this average should focus on improving the customer experience to better meet customer needs and expectations.

NPS is one of several metrics companies use to gauge customer experience.

A Customer Satisfaction Score (CSAT) calculates the percentage of respondents who claim to be satisfied (4) or very satisfied (5) in surveys.

A Customer Effort Score (CES) determines how easy or difficult customers found accomplishing their goals as it relates to an interaction with a company. Examples might be making a purchase or having an issue resolved. Most CES surveys ask customers to rate the difficulty of completing the action from 1 (easy) to 5 or 7 (difficult).

While all three scores help gauge general customer happiness with a company, they do vary. CES is focused more on a specific action and can be less representative of the individual’s overall satisfaction with a company. CSAT helps understand a customer’s satisfaction, but does not determine whether the individual is likely to recommend the company’s products to their networks. The NPS score

Given the specificity of the questions, it does not tell the whole story. Companies might have plenty of happy customers who nonetheless do not intend to tell others about their products and services. That might be an impression that customers are not happy, but that would not be accurate.

It also might not provide actionable insights, unless the organization asks more open-ended follow-up questions. And, for those companies that ask follow-up questions, it can be challenging to analyze all that information. Many turn to sentiment analysis, but that historically struggled with understanding slang and euphemisms. Thankfully, artificial intelligence (AI)-enhanced sentiment analysis can more intelligently understand the full scope of a statement and determine whether it was positive or negative.

Finally, companies might need to send multiple requests to customers to get them to complete the NPS survey. As a result, they might influence more negative scores from respondents who don’t like the company’s aggressive survey tactics. As such, a company should study when and through which medium most responses occur, so they can prioritize that channel.