Friday 13 May 2016

Limits of The Net Promoter Score (NPS) System


What is the Net Promoter Score?

The Net Promoter or Net Promoter Score (NPS) is a management tool that can be used to gauge the loyalty of a company’s customers. NPS was first introduced by Fred Reichheld in his 2003 Harvard Business Review article "One Number You Need to Grow"[1].

The NPS is based on a single question:
"How likely are you to recommend [product or company name]
 to a friend or colleague?"

Customers are asked to rate their answers on an 11 point scale ranging from (0 = not at all likely to recommend and 10 = extremely likely to recommend).

These scores are then split into three categories “Detractors,” “Passives,” and “Promoters.”
  •         Score of 0 – 6: Detractors
  •        Score of 7 – 8: Passives
  •        Score of 9-10: Promoters
The "Net" in Net Promoter Score comes from subtracting the percentage of detractors from the percentage of promoters. A negative score means you have more detractors than promoters and a positive score means there are more promoters (that is, more positive word of mouth than negative word of mouth). An NPS can be as low as −100 (everybody is a detractor) or as high as +100 (everybody is a promoter). An NPS that is positive (i.e., higher than zero) is felt to be good, and an NPS of +50 is considered excellent.

Criticisms of NPS

A key benefit of NPS is that it provides a single score that is easy to understand. However, in our view it has only limited practical application for companies who want to monitor, understand and improve their overall customer experience as it measures only one aspect of the overall experience - loyalty. Below you will find some of the most common criticisms of the NPS.

1.      NPS correlates with customer satisfaction but should not replace it

Customer loyalty and customer satisfaction are related concepts and they are correlated (people usually provide similar ratings for satisfaction and loyalty). At the same time customer experience questions and satisfaction questions are also correlated. However, where a measure is correlated with another measure, that does not mean that one is a replacement for the other, which NPS attempts to do.

It's unlikely to have customers that are loyal but unsatisfied, but you can have satisfied customers who aren't loyal.  However, a key concept behind NPS and loyalty measures in general is that customer satisfaction and experience were excluded to focus only on loyalty.

In Fred Reichheld’s, the creater of NPS, book “The Loyalty Effect” he makes the case that 60% to 80% of customers who defected or didn't repurchase a product were in fact satisfied or very satisfied. In the Auto Industry, for example, he cited figures which showed a 90% satisfaction rate and yet, on average, only 40% of customers repurchased the same brand of car.  This is an important issue as satisfaction does not mean loyalty (see our previous blog posting about satisfaction Do you really need satisfied customers.

So, while satisfaction relates to loyalty, they are not the same thing. Satisfaction measures the customer’s experience at any given time whilst NPS measures long term attachment (loyalty.) It is therefore best to measure both and not rely solely on one. Having multiple measures requires minimal additional effort to record and analyse and is significantly more valuable and actionable for you to monitor and manage – this is our approach.

2.      NPS is not actionable

NPS is also not actionable. What if your company has a good score What does that mean? How can you change the score? We simply don’t know the “So what?”. Similarly what if you have a bad NPS? What do you do? Is it linked to a group of underperforming or unprofitable gyms or a general company wide issue with the customer experience?

If a metric is just an “it is what it is” number, as NPS is, it has no context and no predictive power, can’t be used alone, and doesn’t give you clues about what to do. Its usefulness must therefore be questioned, especially if used in isolation.

At CJM Research we believe that you should measure not only overall satisfaction and loyalty but measure Key Performance Indicators of the customer experience too. This is then reported as a composite experience index score made up of all the KPI scores. This approach has been reported as best practice in other research[2] [3] [4].

Combining the experience scores provides a more robust and stable overall experience score that is directly linked to its component KPI scores. This not only provides a truer overview of the customer experience but also makes it easier to identify and act upon areas for improvement and measure the impact of any changes or improvements to service.

For example, if you have a low overall score for a part of your business you can directly link this to a particular KPI and act to improve this specific aspect of service. With NPS you would know that a part of your business has a lower score but not necessarily why. Our scoring is designed to help management identify and act on the information to improve service. This in turn makes our scoring easier to use for you as it is not only a single experience score but one that is linked to the actual customer experience and and this helps you improve overall satisfaction and loyalty for customers in a faster and more targeted way.

Furthermore, by tracking specific aspects of the customer experience it allows further detailed analysis that can tell you which parts of the experience are most important in driving satisfaction and loyalty. NPS on its own cannot do this.

3.      NPS only works in the context of your wider industry

The Net Promoter score has no real meaning without the context of competitor scores. Whilst a score of +50 is considered excellent this actually differs significantly between industries. For example, a utility company with an NPS of +1 may be excellent in the context of an industry where the average NPS could be -20.

We have an example of this where one of our other clients had an NPS of +69 which they thought was an excellent score…until they conducted some competitor customer research and found they were only average compared to their main competitors.

This is because you have to interpret NPS scores relative to other NPS scores in the same industry. However, most NPS centred research does not include measures of competitors leaving the results without context i.e. you do not know if your score is good, average or poor compared to your competitors.

You also cannot directly compare NPS across different industries or product categories because some products simply don’t lend themselves to word-of-mouth (e.g., toilet paper or car oil), while others can evoke passionate positive or negative views (e.g., retailers or restaurants).

4.      The same NPS score can come from very different customer experiences

NPS is also problematic because there are many possible ways to arrive at the same number, for detractors, passive, and promoters respectively, you could achieve an NPS of +30, for example, in dozens of ways.

A company with a +30 NPS could have 30% promoters, 70% passives, and 0% detractors, while another company with a +30 NPS could 60% promoters, 10% passives, and 30% detractors. It is likely that a company with 30% promoters and 0% percent detractors is very different than the one with a polarized customer base with 60% promoters Vs 30% detractors — even though their NPS is seemingly the same.

NPS also ignores the fact that the voices and reach of promoters or detractors can be drastically different depending on their motivation and the digital channels they have at their disposal. There may be a small amount of very vocal detractors who go online and write negative reviews. That would outweigh a large non-vocal group of promoters, even if the company had a positive and high NPS.

5.      11 point scales are not as good as others

Research by Stanford University and others[5]. found that the 11-point scale advocated by NPS has the lowest predictive validity of the scales tested. We typically recommend a 10-point scale for overall satisfaction and often mirror NPS scoring for recommendation (for consistency with NPS rather than by preference). This provides a more detailed scoring which is useful for tracking research.

For experience KPI’s we suggest either a five or 10-point scale depending on our clients legacy research and/or objectives. Five point scales are simpler to interpret and easier for respondents to complete, especially on smart phones (50%+ of online surveys are completed using phones) but smaller changes in average ratings are more difficult to identify.

6.      The scoring inflates the margin of error

By converting an 11 point scale into essentially a 2 point scale made up of detractors and promoters (NPS ignores passives in the scoring), information is lost. Furthermore, this new categorisation increases the margin of error around the net score (promoters minus detractors). Unfortunately, this means that if you want to show an improvement in Net Promoter Scores over time, it can take a much larger sample size to calculate, otherwise the difference won't be distinguishable from the acceptable range of error. 

7.      The relationship between NPS and growth remains unproven

One final, but important criticism of NPS calls into question the strength and reliability of the link between NPS and measures of business growth and profitability of those with higher NPS. Research that has attempted to replicate the link between business performance and NPS has not find statistically significant relationships between the two[6]. Similarly, the relationship between growth and NPS is reported to be less consistent or strong when using tracking data[7].

Even Fred Reicheld, the inventor of NPS admits that the findings of his initial research for NPS was flawed stating[8]:
"A number of perspicacious readers have noted that the statistical evidence provided in my book The Ultimate Question is imperfect. It does not provide proof of a causal connection between NPS and growth. Nor are some of the timeframes ideal.”

We are not suggesting the NPS is a useless score but that it is not the one score you should only measure. Instead it has its place together with satisfaction and more detailed experience questions in a balanced customer experience research programme.




[1] https://hbr.org/2003/12/the-one-number-you-need-to-grow/ar/1
[2] For example: Hill, Nigel; Roche, Greg; Allen, Rachel (2007). Customer Satisfaction: The customer experience through the customer's eyes. London, England: Cogent Publishing. p. 7.
[3] Satisfaction as a Predictor of Future Performance: A Replication. Jenny van Doorn , Peter S.H. Leeflang, Marleen Tijs International Journal of Research in Marketing (Impact Factor: 1.71). 12/2013
[4] Szwarc (2005) “Researching Customer Satisfaction and Loyalty: How to Find Out What People Really Think” Kogan Page.
[5] Schneider, Daniel; Berent, Matt; Thomas, Randall; Krosnick, Jon (June 2008). "Measuring Customer Satisfaction and Loyalty: Improving the 'Net-Promoter' Score"
[6] Lawrie, Jock, Matta, Alsono, Roberts, Ken (2006). Value speaks louder than words. The management folly of adopting the Net Promoter Score as the ‘one measure’
[7] Keiningham, Timothy L., Cooil, Bruce, Aksoz, Lerzan, Andreassen, Tor Wallin, Weiner, Jay (2007). The value of different customer satisfaction and loyalty metrics in predicting customer retention, recommendation and share-of-wallet, in: Managing Service Quality, 17(4), 361-384.