The Net Promoter Score: Does the Single-Question Scoring System Work?
The Net Promoter Score (NPS), since its development in 2003, is championed by many as one of the most reliable indicators of company growth and customer loyalty. This paper aimed to answer three important questions regarding the NPS to determine this theory:
- What does the NPS tell us about company performance?
- Can it be used universally?
- Can the NPS successfully indicate customer loyalty?
Research examining the reliability of the NPS reveals flaws in predicting both company performance and growth. Also, the NPS does not prove relevant for all industries, and can therefore not be used universally. Lastly, it was noted that it the NPS is not a reliable indicator of customer loyalty.
It was concluded that the NPS does not paint the full picture with regards to company performance or growth. It was therefore recommended that several additional questions be linked to the NPS, allowing companies to dig deeper into why customers may choose not to recommend them to friends or family. Furthermore, questions regarding the repurchase intentions of the customer could be used, as they may provide a clearer indication of customer loyalty than the Net Promoter Score.
In 2003, Fred Reichheld introduced the Net Promoter Score, or NPS as a key measure of customer loyalty, and claimed that it could accurately measure the growth capacity of any company, more so than any other metric . Reichheld designed the NPS as a single question with an 11 point scale, asked in either isolation or as part of a larger customer satisfaction survey.
1. How likely are you to recommend Company X? Where 0 is not at all and 10 is extremely likely.
Those who score between 0 and 6 are considered “detractors”, those who score between 7 and 8 are “passives” and those scoring 9 or above are “promoters”. The eventual NPS is then calculated by subtracting the percentage of detractors from the percentage of promoters. For example, if 20% of Company X’s customers are detractors, and 40% are promoters, then Company X has scored an NPS of 20%.
The scores can then be used to compare either against previous years scores, or with set individual industry NPS benchmarks – these range from banks, insurers to wholesale manufacturers.
Reichheld advocated this scoring system as ideal for top level executives, who could use the benchmarks provided to assess company performance. For many, however, this fails on several counts, including:
- Its ability to measure company performance
- Its ability to predict company growth
- Its application
- Its design
This paper will take a closer look at each of these points by asking three important questions:
- What does the NPS tell us about the company?
- Is the NPS for everyone?
- Can the NPS successfully indicate customer loyalty?
The answers to these questions will allow us to form a greater understanding of the Net Promoter Score, and its ability to aid in the gathering of customer insight .
1. What does the Net promoter Score tell us about the company?
The concept behind the single score measure was to create a level of simplicity, and allow those at the head of businesses to grasp the nature of their performance in the simplest way – a high NPS can be interpreted as a strong company performance and growth, and the inverse indeed true. But many have claimed that the NPS is not a reliable predictor of either company performance or growth.
The NPS as an indicator of Company Performance
“I’m the CEO of Company X, and we have an NPS of 20%, which shows me that we are performing well. This is excellent news. But it tells me nothing about what areas we are performing well at. Are levels of service across the business consistent, or is one section outperforming another?”
Many may feel that these are unnecessary questions. After all, performance is good and our customers appear satisfied. But if service levels are to be maintained in the long term, these questions need to be answered. Similarly, if the NPS score was -20%, poor performing areas would need to be identified immediately so they can be improved.
A study into indicators of customer performance by Keiningham and colleagues in 2007 noted that a single question is not efficient in providing enough information on the performance of a company, and must be supplemented with addition questions. For example,
1a. How likely are you to recommend Company X? Where 0 is not at all and 10 is extremely likely.
1b. Why would you recommend/not recommend?
By asking the customer why he or she has scored the company as such, we can begin to understand more about the company’s current true performance, and the areas which may need improving. Keiningham wrote that in order to maintain high levels of customer service, we first must understand which areas of the business are performing well, and more importantly, which aren’t.
For many, however, the principal benefit of the Net Promoter Score is its simplicity, and adding to the single question could prove to be an unnecessary complication. But what if the single score itself is misleading?
Company X has an NPS of 20%, as does Company Y. Does this mean that both companies have the same number of detractors as they do promoters? Not necessarily. Figure 1 highlights two possible permutations of promoters and detractors, both resulting in an NPS of 20%. Therefore, the presentation of a 20% NPS alone does not indicate how many promoters a company actually has, a figure necessary if they hope to understand the customer experience they provide. Similarly, if a score of 20% was gained for two consecutive years, how would the company know if the number of promoters has increased, decreased or stayed the same?
Figure 1. Breakdown of detractors, passives and promoters for Companies X and Y
The NPS as a measure of Company Growth
The Listening Company and the London School of Economics tested Reichheld’s theory in 2005, and found support for his claim . Over 1000 customer telephone surveys and an analysis of over 50 companies’ yearly revenue highlighted a clear correlation between NPS and growth of companies such as Honda, O2 and HSBC over a one year period.
On the other hand, Keiningham and colleagues in 2007 showed no link between the two. After a series of customer interviews and revenue data from 21 companies, the authors were unable to support Reichheld’s theory, and claimed that the Net Promoter Score was not predictive of a company’s growth.
Could this lack of correlation be related to the nature of the question itself? Many, including Keiningham, believe that the question is a measure of customer attitude (I am likely to recommend) rather than behaviour (I have recommended), and therefore does not accurately show a company’s capacity for growth . After all, a customer may state that they are likely to recommend, but how can the company be sure that they will? Some authors have gone as far as suggesting that an additional question, highlighting past recommendation behaviours be included, and taken into account when scoring the overall NPS.
2. Can it be used universally?
Another potential flaw with the NPS is in its application, with some authors claiming that the question is not suited to all industries.
A study by The Loyalty Research Centre last year highlighted that “likelihood to recommend” does not fit every industry and in some cases may prove to be an inaccurate measure of customer loyalty .
For example, a customer is perhaps more likely to recommend a car manufacturer than they are a brand of soap. This does not mean however, that the customer is less loyal to their soap brand than the car brand; the predictor of loyalty is simply irrelevant. To gauging loyalty in these industries, questions regarding the likelihood that the customer will switch brands or repurchase could be used.
3. Can the NPS successfully indicate customer loyalty?
Thus far, flaws in the ability of the NPS to predict company performance and growth, but what about the prediction of customer loyalty? In other words, does a high NPS score mean a high level of customer repurchase and can the NPS be a true measure of customer loyalty, if a link between likelihood to recommend and propensity to repurchase cannot be made?
Morgan and Rego sought to explore this relationship further in 2006, and compared the results from a variety of metrics on customer loyalty. They found that an average score gained from a large customer satisfaction survey (measuring satisfaction versus dissatisfaction on a 7 point Likert scale) measured customer loyalty more accurately than a standalone NPS. They added that the inclusion of a “likelihood to repurchase” question would increase the firms capability of calculating customer loyalty and company growth. However, one could argue that this form of question would incur the same problems as the NPS – is it a measure of attitude or behaviour?)
Fred Reichhard’s intention when he devised the Net Promoter Score was that it be used to successfully predict the growth of any given company, and indicate how loyal its customers are. However, the NPS now seems to fail on predictive quality, application and design.
While many have supported its use in a customer satisfaction survey, others have highlighted problems in:
- Predicting company growth
- Measuring company performance
- Measuring customer loyalty
- Its application on a wider scale
Should we use the NPS to measure customer loyalty?
The satisfaction survey is key to understanding the customer experience, and pinpointing where problems in the journey may lie, but should the NPS be included in such a survey as a measure of loyalty?
Morgan and Rego (2006) argued that the customer satisfaction survey is best presented without the NPS, and should include questions specific to loyalty and repurchase intentions (although they failed to highlight the possible expense or inconvenience of creating large customer satisfaction surveys).
While it is clear that the NPS could provide insight, the potential flaws of the NPS noted throughout this paper are hard to ignore.
It can be recommended therefore that if the NPS score is used, it be supplemented with additional questions which would allow companies to further understand their customers and their reasons for recommending to friends and family. In doing so, the single question score once championed by Reichhard may become obsolete.
It can also be recommended that companies select questions which highlight customer intention to repurchase to more accurately measure customer loyalty. Companies must strive to select relevant and appropriate questions based on the customer insight they hope to gain.
Regardless of whether companies continue to include the NPS in their satisfaction surveys, one thing is clear; insight is more important than ever in gaining a true understanding of your customer. Percepta strive to create value in the customer insight we provide for our clients. Through tailored customer satisfaction surveys, our clients can better understand the behaviour of their customers and can begin to comprehend what the future holds for their business.
To discuss this paper, to receive in a different format or to find out more
Please call us on +44 (0)141 438 2474 or email
You can also join our mailing list to receive future communications directly to your inbox. Simply send us an email.
References: Neil Morgan and Lopo Rego “The Value of Different Customer Satisfaction and Loyalty Metrics in predicting Business Performance” (Marketing Science, 2006, 25:5 pp426-439) Satmetrix www.satmetrix.com Tim Keiningham , Bruce Cooil, Tor Wallin Andreassen and Lerzan Aksoy “A Longitudinal Examination of Net promoter and Firm Revenue Growth” (Journal of Marketing, 2007, 71 p39-51) Neville Upton, Alain Samson and Dr Paul Marsden, “Advocacy Drives Growth” (Research paper by The Listening Company and The London School of Economics, 2005) Alex Bollen, “Measuring and Managing Loyalty”, (Ipsos Research paper, 2008) The Loyalty Research Center www.loyaltyresearch.com