Q. 196  The rating in Net Promoter Score is obtained as Ordinal data (rating scale from 1 to 10), but this data is later modified to three discrete categories (categorized as promoter, neutral or detractor). When Ordinal is a better type of data, why is Ordinal data converted to Categorical data?

 

Note for website visitors - Two questions are asked every week on this platform. One on Tuesday and the other on Friday.

• All questions so far can be seen here - https://www.benchmarksixsigma.com/forum/lean-six-sigma-business-excellence-questions/
• Please visit the forum home page at https://www.benchmarksixsigma.com/forum/ to respond to the latest question open till the next Tuesday/ Friday evening as per Indian Standard Time.  
• The best answer is always shown at the top among responses and the author finds honorable mention in our Business Excellence dictionary at https://www.benchmarksixsigma.com/forum/business-excellence-dictionary-glossary/ along with the related term.

Net Promoter Score is an index of customer loyalty & not customer satisfaction.

True loyalty affects the profitability of an organization by reducing customer acquisition cost.no company can grow if their customer bucket is leaky, Loyal customers will arrest the leak and add more customers to the bucket. 

 

Loyal customers act as references and will risk their reputation by recommending a product or service from a company. Truly loyal customers bring in new customers at no cost to the company. The path for profitable growth for any organization lies on their ability to get its loyal customer to become its brand ambassadors.

 

Right Measure of loyalty

 

Because loyalty is so important to profitable growth, measuring and managing it make good sense.

Frederick F. Reichheld, Bain & Company & Satmetrix after a research of 6-12 months came up with one question which will gauge the customer loyalty after framing the question it is important to arrive at a method to analyse the response which is easy and simple to understand without statistical knowledge.

 

After finalizing the right question “How likely is it that you would recommend [company X] to a friend or colleague?” Fred & team worked upon a scale which is easily understandable by investors, regulators, employees and grasp the basic message without needing a handbook or session.

 

Initially the scale was framed from 0-10 where 10 means extremely likely to recommend, 5 means neutral and zero means not likely to recommend. The customer responses and repurchase behaviours were examined along the scale and the three clusters were defined.

 

Promoter - the customers with the highest rates of repurchase and referral, who gives a rating of nine or ten to the question.

Passively satisfied who gives a seven or an eight classified as Passives

Detractors- scored from zero to six.

Upon more surveys and analysis it was concluded that a strong correlation existed between net-promoter figures and a company’s average growth rate.

 

To have a common measure of customer loyalty which can be understood easily Fred and team Came up with a methodology which converts the ordinal scale (1-10) to a Categorical data and provides the customer loyalty index which is measured as NPS.

Net Promoter Score is a metric commonly used to measure the loyalty of the Customer.

NPS scoring was created by Fred Reichheld from Bain & Company.

 

NPS originally stood for Net Promoter Score. However, it has evolved to stand for Net Promoter System.

 

Most of the Fortune companies like Apple, GE, Amex, Allstate, Walmart and other companies use NPS. NPS could be focal point for Organizational learning’s.

 

It is calculated by simply asking the customer just one question.

On a scale of 0 to 10, how likely are you to recommend this Product / Service to family, friend or Colleague?

 

11 Point scale,
0 being Not likely to recommend and
10 being Most likely to recommend the Service / Product

 

This again is further categorized into 3 segments

• Promoters (9,10 Ratings) 
• Passives (7,8 Ratings)
• Detractors (0 to 6 Ratings)
   

Promoters are delighted by the Service / Product, Loyal and most likely to recommend
Passives have Neutral opinion, they neither Promote nor Demote
Detractors are dissatisfied and will most likely to switch to competitors, or where the service is Excellent (E.g.: From SBI to AmEx)

 

Objective of NPS is to listen to detractors, fix the dissatisfaction and move them to promote.

NPS = % of Promoters - % of Detractors

NPS score ranges from -100 to 100 and simple calculated by the below formula,

-100 signifying, all are detractors &
100 signifies that all are promoters

Score >0, implies, promoters are more than detractors

 

Below are the NPS Leaders by Industry:

Source: NICE Satmetrix – US Consumer Report 2018

Allstate finished 4.9 points higher in 2018 compared to year-end 2017

 

Some of the Quick Benefits Include:

Simplicity

Ease of Use

Easy and Quick Follow-up

Learning and Experience

Adaptability

 

Categorical values (Qualitative): observations clubbed in groups or categories (Promoters, Passives, Detractors)  
Ordinal values: observations have rating scale (0 - 10 rating). This has implied order.

 

With Ordinal data it is easy to detect responses, even when there is change in the distribution.

 

For Instance:

When 30% of the respondents have rated between 0 - 3 &

40% of respondents have rated between 4 - 6;  

With regards to Categorical classification, entire 70% is considered as Detractors.

 

Any movement between the detractors are also untraceable when it is categorical.

 

So why Ordinal Data converted to Categorical Data?

By moving does statistical power and precision is lost?

 

We know, Standard error is derived from Variance.

So applying variance, we get

 

Var [NPS] = Var [% of Promoters - % of Detractors]

Var [NPS] = Var [% of Promoters] + Var [% of Detractors] - 2 Cov [% of Promoters, % of Detractors]

 

Note: Cov [% of Promoters, % of Detractors] is going to be negative, as the customer cannot be both Promoter and Detractor at the same time.

therefore,

Var [NPS] = Var [% of Promoters] + Var [% of Detractors] - 2 (- number)

Var [NPS] = Var [% of Promoters] + Var [% of Detractors]  + number

 

Here, Variance of NPS is > sum of variance of parts.

So Categorical is better over Ordinal.

 

Categorizing also influence the customer. Customer originally planned to give 5 or 6 would give better category when labeled.

Categorization is not symmetrical.

 

Converting from Ordinal to Categorical will lose few information. But the extreme responses (Most Likely and Least Likely) are good predictors. In the sense what is lost in scale conversion is ok to lose.

 

We will not be able to start categorically first with 3 points, as we need at least 4 points to understand the intensity of the agreement.  

11 Point is used to capture satisfaction at granular levels, macro and micro levels and then categorize to understand the group better.

Categorizing / segmenting is a better way to notice patterns / movements between the customers. This will help in improving the experience and on certain touchpoints.

Benchmark Six Sigma Expert View by Venugopal R

 

As most of you must be aware, the Net Promoter Score (NPS) gives an indication of how well the customers of a business are likely to recommend a product or service to their friends, colleagues or relatives. Those customers, who are highly likely to continue to avail the product or services and also recommend to their friends positively are called “Promoters”. Those customers who are unlikely to avail the product or services and would spread negative publicity are called  “Detractors”. We may have some customers who fall in between i.e. they are neither Promoter or Detractor, but are people who behave neutrally. They are neither very excited about the product or Service, nor are they very unhappy.

 

The actual NPS score is calculated by subtracting the % of detractors from the % of promoters.

NPS score = (% of Promoters) – (% of detractors)

 

Obviously, once we know the % of the discrete categories viz. Promoters and Detractors, we can calculate the NPS scores. However, the process for obtaining this discrete data is through a survey of a sample of customers. During the survey, only one question is asked “On a scale of 0 to 10, how likely are you to recommend this company’s product or service to your friends, colleagues or relatives?”

 

Those who give a score of 9 or 10 are classified as “Promoters”. Those who give a score of 6 or less are classified as “Detractors”.  Those who give a score of 7 or 8 are classified as “Passives” or “Neutral”.

We can see that the purpose of using the ordinal scale is to gather the data objectively from a pool of customers. However, the way the NPS is defined, we need to identify the “Promoters” and “Detractors”. Hence the need to classify the ordinal data into discrete categories. The NPS score will not be influenced by the individual scores within Promoters or Detractors.

 

One may also wonder how the line is drawn on the ordinal scale to define the Promoters, Detractors and Passives. It may so happen that even customers who gave score of 7 and 8 could be promoters. Also, when we classify one who gave a score of 6 as detractors, isn’t there a possibility that someone who gave 7 could also be a detractor rather than being passive. Or, looking at it the other way, even a person who gave a score of 6, need not be a detractor.

 

Bain and Company have done several researches on this topic and the customer behaviors and it has been decided that the scores of 9 and 10 represent customers who are extremely likely to recommend whereas the scores of 6 and below represents extreme unlikeliness - and these guidelines are being used as a standard for deciding the Promoters and Detractors - and to be applied on the defined calculation for NPS.

Definition:

As the wiki definition says "Net Promoter score(NPS) is a management tool that can be used to gauge the loyalty of a firm's customer relationships." 

 

Objective: The tool is used to predict customer loyalty to a product, service,brand or company.  

 

What is it all about? 
As succintly put in the wiki for Net Promoter, the tool measures the loyalty between a provider and a consumer .  The Net Promoter Score (NPS) tool is used by a provider who can be a service provider or an employer or who can be any other entity .  The Provider will ask question as a survey, which need to be answered by the consumer. The consumer is the one who can be a customer (to the service providing organisation), or an employee(in case this is measured by an organisation for its employees) or a respondent to the survey for this NPS.

 

How it works ?
NPS is based on the calculation obtained from a question posed to the respondent. Often the question could be "How likely is it that you would recommend our company/product/service to a friend or colleague?" NPS score can range from -100 to +100 when measured across multiple people, where negatives indicate that the respondents are deemed as detractors and positive scores indicate that the respondents are called as Promoters.  In a 11-point scale (0-10), the rating is classified such that respondents who provide values in the range of 0-6 are branded as "Detractors", respondents who provide values between 7 and 8 are branded as "Passives" and respondents who provide values between 9 and 10 are branded as "Promoters". 

 

Formula to find NPS score
NPS score = % of Promoters - % of Detractors

 

Promoters are loyal and would provide support to the provider of service/product/service providing organisation.  Passives are neutral support people who are susceptible to become detractors quite quickly at any point. Detractors are unhappy customers who can damage the service provider's brand and can portray a bad picture about the organisation, through word of mouth.

 

Why Ordinal data is converted into categorical data ?

Before we talk about that let us see what categorical data and Ordinal data do. Categorical data will have data classified into multiple categories and has no intrinsic ordering. Whereas in Ordinal data, the ordering happens.  Now let us take a look at the conversion part.

Imagine the 0-10  point scale.   Assume (in a hypothetic way) that each ordinal point is representing a categorical value as mentioned below : 
0- worst, 1-very poor, 2-poor, 3-Not bad, 4-Average, 5- Fair, 6-OK,7-good, 8-Better,9-Very Good, 10- Excellent

Now let us compare the difference in categories between 0 and 2  and the difference in categories between 4 and 6. The difference in these 2 set of categories would be quite different from being worst to somewhat palatable (but sill not good enough from the customer perspective-which is a different issue altogether).  Here the ordering happens but we can see the size of the difference between the categories is inconsistent as the spacing between the categories in each set is varying.

 

Also is the fact that upto the 7th point of the scale(0-6), it is deemed as "Detractors" and 8th and 9th point of the scale(7-8) called as "Passives" and 10th and 11th point (9-10)  as "Promoters" . Therefore there is no need to focus on the Ordinality. Hence  it is branded as Categorical data.

 

Conclusion
Effective use of NPS tool can bring in the right perspective to an organisation about its customer and can potentially make an organisation succeed in understanding the customer's expectations and also ensure a satisfied customer which can bring more additional customers and revenues for the organisation
 

Converting the ordinal data into categorical data helps in identifying the more loyal customers, customers who are really unhappy and neutral customers in net promoter score
Data can be made in relevant groups to bring more usefulness or better understanding of the current state
Also the categorization helps in understanding the data and take a useful decision i.e. where we need to focus, where we need to be supportive etc.,

There are 02 reasons of using categorical data instead of ordinal data in Net promoter measurement :

 

1) ordinal data has more values specified while categorical data groups the parameters in a category thus reducing the values into a single category. This helps in quick comparison of the categories with one process to another process and also helps in taking actions on the bottom quartile measurement which is detractor easily.

 

2) apart from this, the categories are defined in regards to customer experience depending on the function.Thus, instead of working on 0 to 6 counts and taking action, the team works on Detractors or Red alerts(0-3) and frame actionables.

Net promoter score or NPS, as we call shortly, is one of the analytical tool to measure the customer satisfaction index or loyalty of the customer to a particular manufacturer or service provider. It is an important metric to gauge the performance of producers against their stated brand or value product. The Metric uses scale from -X to +X and places some questions to the customers. List of questions varies widely from basic features of the product/services , customers likelihood of using the same product/services again, their probability to refer the similar brand/product/services to other potential customers. Based on the score, all customers and buyers are categorized into Detractors, neutrals and promoters. Former being negative or poor response and the later being positive or worthy response.

 

The purpose of converting ordinal data (-X to +X) to discrete data (Categorical) are :

 

a) To have a holistic view point on the categories of customers being detractors, Neutral or promoter at the larger level in the organisation which can help the organisation for data based decision making in expedited manner.

b) Categorization of data helps to have a focused  attention of a particular type of customers based on the questionnaire type and response. This helps the management team and staff to brainstorm on proposed mitigation plan against those challenges.

c)Logical interpretation of data can be better executed through categorical data than ordinal data which can be helpful in complex problem solving and decision making. This is because the ordinal data are spread in unorganized and unsystematic manner. Categorical data takes into cognizance various independent variables to make a distinction between different zones for further analysis.

 

One example can be taken from service industry : Let us assume, company XYZ has implemented a process to reward first 50 customers with exciting gift pack in order to promote the newly launched brand. In order to determine the NPS, list of questions are asked from existing customers in sample (n) basis. once the survey results are out, the categories are made for the customers to distinguish them based on questions whether they are detractors, promoters or neutrals. For an  example, if one of the Question has got 70 % of the sample population are detractors, than it is evident for the management team to focus on the ways to mitigate the challenges pertaining to that particular question.  In case of categorical data, if this would have been a ordinal data, it would bring ambiguity to identify areas for continuous improvement and make critical business decisions.

Net promoter score or NPS, as we call shortly, is one of the analytical tool to measure the customer satisfaction index or loyalty of the customer to a particular manufacturer or service provider. It is an important metric to gauge the performance of producers against their stated brand or value product. The Metric uses scale from -X to +X and places some questions to the customers. List of questions varies widely from basic features of the product/services , customers likelihood of using the same product/services again, their probability to refer the similar brand/product/services to other potential customers. Based on the score, all customers and buyers are categorized into Detractors, neutrals and promoters. Former being negative or poor response and the later being positive or worthy response.

 

The purpose of converting ordinal data (-X to +X) to discrete data (Categorical) are :

 

a) To have a holistic view point on the categories of customers being detractors, Neutral or promoter at the larger level in the organisation which can help the organisation for data based decision making in expedited manner.

b) Categorization of data helps to have a focused  attention of a particular type of customers based on the questionnaire type and response. This helps the management team and staff to brainstorm on proposed mitigation plan against those challenges.

c)Logical interpretation of data can be better executed through categorical data than ordinal data which can be helpful in complex problem solving and decision making. This is because the ordinal data are spread in unorganized and unsystematic manner. Categorical data takes into cognizance various independent variables to make a distinction between different zones for further analysis.

 

One example can be taken from service industry : Let us assume, company XYZ has implemented a process to reward first 50 customers with exciting gift pack in order to promote the newly launched brand. In order to determine the NPS, list of questions are asked from existing customers in sample (n) basis. once the survey results are out, the categories are made for the customers to distinguish them based on questions whether they are detractors, promoters or neutrals. For an  example, if one of the Question has got 70 % of the sample population are detractors, than it is evident for the management team to focus on the ways to mitigate the challenges pertaining to that particular question.  In case of categorical data, if this would have been a ordinal data, it would bring ambiguity to identify areas for continuous improvement and make critical business decisions.

The Net Promoter Score is an index ranging from -100 to 100. It reflects the tendency or likelihood of customers to recommend an organization’s products or services to others. It is used to measure the customer's overall satisfaction of the product or service and brand loyalty. ‘Word of mouth’ being an integral part for any buyer community NPS is a critical scoring that every organization has embraced to make their brand larger and memorable for customers.

All of this is calculated with ordinal data as the scoring is between 1-10 however this data is later modified to three discrete categories (categorized as promoter, neutral or detractor).

NPS survey score of 9 or 10 are categorized as  ‘Promoters’ which means they are very satisfied with the product or service and are very likely to recommend the product or services to other organisations. NPS of 7 or 8 are categorized as ‘ Neutral’ or ‘Passives’. They are not too satisfied with the product or services and will not be sharing any negative opinion though but are less likely to recommend to other organisations. NPS score less than 7 are termed as ‘Detractors’ which means these customers neither will repurchase or avail any further services of the company and could most likely negatively impact the reputation of the company.

This is because categorization of the type of customers gives us a qualitative view of the customer experience and what kind of measures one must take to ensure maximum protection to the company’s reputation, higher growth and market share and better customer experience.

Hence the judgement to promote a brand or not cannot just depend on the ordinal data which are independent variables. There has to be an accumulation of the customer experiences which comes through the categorical data only.  

 

 

While many respondents have provided very good responses. I agree categorization makes things simpler and easier to use. 

 

The critical aspect is provided by Jayaram T who clearly mentioned that research showed that the NPS% (the way it is calculated) showed good correlation with growth rates of Organizations. The winner for this question is therefore Jayaram T. 

 

I liked the responses from Asif, Shreyas, Rajesh, Indrani, Praveen, Bhagyashree, and Indrani. Do go through the expert view from Venugopal as well.