The Net Promoter Score is so popular because it is quick to gather, easy to understand and is a single metric that can be benchmarked with other companies from your industry. Especially the last topic, seems to be of great interest for many companies that want to implement the NPS. There are places on the internet where you can find NPS Benchmarks to compare yourself to competitors and to companies that have similar processes to your company.
People are always baffled when we tell them that the actual NPS score is just the second most important information we get from the Net Promoter System and the customer feedback is the most valuable data that we gather. For us the NPS is just a number to see if we are progressing into the right direction. Even though we still benchmark ourselves against other companies this is not our primary goal with the NPS and we are very careful to which companies we compare ourselves to. The simplicity of the NPS eliminates many nuances and abnormalities in the customer sentiment and reduces it into a single number, which is usually why people find the NPS appealing. Thus when selecting your benchmarking partners, you will need to break down this number and see if the NPS scores are comparable.
Selecting the right Industry
Having a look at some recent industry benchmarks reports for the NPS we can see that the range of NPS across industries varies greatly. With the highest average for “Auto dealers” being an NPS of +40 and the average for “TV service providers” is -1 in 2015 you can already see that it depends greatly in what industry you position yourself in. Sometimes it is obvious but for a larger TV/Cabel/Phone/Internet provider it makes a huge difference if they see themselves in the “Wireless Carriers”, “Internet Service Providers” or even “TV Service Providers” bracket. If your company specializes in different industries, you will need to dig deep into the companies that are included in the industry survey and find out which companies have a similar product line and are comparable to you.
The other option is to do it like some bigger companies who are facing this problem and go into a much more granular NPS view. Apple for example asks the NPS question not for the entire company but for each of their products that they offer. For Apple this makes total sense, since no one would recommend the company Apple to a friend or colleague but their products like IPad, IPhone or ITunes. That is why Apple asks the NPS for their individual products with varying results: For instance, Ipad (+65), Iphone (+70) and Laptop (+76) [*1]. By doing this they can compare each product with the industry average for tablets, phones and laptops and see where they stand.
Depending on the country of origin of your customers, NPS and all other feedback ratings that you would receive can vary greatly.
Geert Hofsteade classified countries and societies into different dimensions, one of them being characterized by more masculine or feminine traits. More Masculine societies are characterized by traits like: High ego, large gender wage gap, girls cry and boys don’t, boys fight and girls don’t, failing is a disaster. While feminine cultures are characterized by: Quality of life, importance of people, more women in management, lower wage gap, failing is a minor accident and a flexible family structure.
Customers of different nationality give different feedback.
Depending on your cultural background the NPS feedback will thus be very different. When comparing Hoffstedes model based on NPS Feedback for travel services, John Crotts and Ron Erdmann (2000) found that feedback and the intention of recommending the service varied greatly by cultural groups.
On Hofstede’s masculinity index (lower rank, higher masculinity) the countries are ranked as followed
While some passengers threatened to sue others where quietly waiting for 3 hoursIn the experiment by Crotts and Erdmann where they had flight passengers wait for three hours to enter the airplane, the feedback of high masculinity cultures (America, Great Britain and Germany) where very different compared to the ones of low masculinity (feminine) cultures. While high masculinity cultures demanded change of flights, refunds or even threatened to sue the airline, low masculinity cultures accepted the delay and where very polite and appreciative to the flight attendants and pilots during the flight. This shows how different the perception of situations is due to cultural backgrounds and obviously this can affect the NPS score of your company. Always keep this in mind when benchmarking your NPS against companies that might have customers with a different cultural background.
NPS Collection Channel
With the rising popularity of the Net Promoter System many different models of feedback collection were introduced. We have seen web services, telephone and even in-person NPS surveys. Even in a single collection method there can be different approaches on how to collect the feedback. Let’s take for example the use of web services to collect NPS data. The main methodical difference which also contributes to a significant difference in the final score is between in-service (inside app or ticket system) NPS collection and transactional NPS collection through email.
Sam Klaidman and Frederick C. Van Bennekom [*3] analyzed the response rates and NPS results for a large US company that use both a web service survey for customers with an email on file and a phone survey for the rest.
The data showed two main outcomes:
The response rate for telephone surveys is much higher. The response rate for telephone surveys is around 44% and for web services at around 13% for customers of this company [*4]. This shows that on telephone surveys a larger pool of customer feedback is collected and there is less self-selection of users who give feedback because it is harder to withdraw from a telephone survey than one you get by email. A self-selection bias occurs when people can freely choose to participate in a survey. People who are engaged enough to participate will on average give a different score than people who do not care if their voice is heard.
In personal conversations customers would rarely give extremely negative scores, through more anonymous channels like email this would change significantly
Having a personal conversation makes it much harder to give an extremely negative feedback compared to an anonymous environment like email voting. This second finding by Klaidman and Van Bennekom is also connected to the self-selection bias and are especially true to the personal one on one conversation while conducting the NPS survey by phone. While through a phone survey only 9.7% of the respondents were detractors, when asked the same question through a web form 26.4% gave the company a score lower than 7.
That is the reason why the average NPS score for a phone survey is 58,1% and the NPS score for an email survey only 18,0% for the same company [*3].
Looking at the scale distribution confirms the theory that people who give the vote in a personal conversation often shy away from the extreme scores. This finding is another argument for using web services to collect your NPS as it seems to produce a more honest NPS. Accounting for lower response rates through email feedback collection can be accounted for grading all non-respondents as detractors.
As you can see the conclusion with whom you should benchmark yourself is not an easy one. As mentioned before you should concentrate on using the comment feedback part of the NPS question (the second question) to reduce problems or frustrations of passives and detractors as well as improve your own score. When all these limitations that alter the NPS score do not apply or are accounted for in your NPS competitor analysis you can and should benchmark yourself against other companies but always remember that the score is not the most important thing what makes the NPS such a strong metric.
 Tempkin Group
 Sam Klaidman and Frederick C. Van Bennekom
 Dillman et al. 2009