october 11, 2019

NPS: Not Particularly Sufficient

CX 2.0 2020 platform

How can you write a book on Customer Experience without discussing the ideas and systems that brought us here? Over the past three years, we have put the NPS® question to the test and found it is no longer the only number you need to grow your business.

NPS® is an old standard that is still highly adopted

The Net Promoter System® (NPS®) was created in 1993—before Google, before Netflix and before Microsoft Windows—and has become, today’s most widely adopted measurement standard for customer experience was first created. I am neither a promoter, nor a detractor of the NPS questions.

It is a good method for benchmarking performance, assuming that the groups being benchmarked are using the same methods, wording, graphics, and requesting timing to collect the NPS® feedback. To drive that point home, using the same question doesn’t always yield the same results.

  1. Colorizing the questions with red, yellow and green will yield better results than the same question without the colors.

  2. The timing of the survey will yield different results. For instance, if a car dealer asks a customer the “Likely to Refer” question as they are driving their new car off the lot, they will get better results than asking the same question a week later, when the new car smell is wearing off and the customer still can’t figure out how to connect their phone with Bluetooth.

  3. The front-line person interacting with the customer can manipulate the results. If an employee discusses the survey with the customer and persuades them to complete it, this level of engagement will yield both higher completion rates and higher scores.


The biggest decline was the sixes. Just over 10 percent of the test group selected that they had an OK experience, but answered the NPS® question with a six. While we offer the NPS® question to customers, we are not passionate about it either way. It only fits into the plan if it is important to the business for benchmarking or to measure against a previous data set. It can be fun to watch your NPS® scores improve month over month as you evolve your CX strategy. 

As a standalone, it definitely fits into a Data-at-Rest bucket. We will dig into this more later, but think about it like this—the NPS® question doesn’t tell Jill, your counter agent in Denver, anything about her personal service in real-time. You may be able to use your point of sale and figure out who was working that day at that time and create an employee level report, but the question doesn’t ask if the customer would refer based on their interaction with Jil. Instead, itasks if the customer is likely to refer. By the time Jill sees her scores (if ever), the data has aged and usually has been sitting on a report for days, weeks or months. In this case, it will have little to no impact on Jill’s behavior.

NPS® is said to be a great predictor of growth. This means that  if you have bad NPS® scores then your customers are not happy and you probably have products, people and processes to fix. If you have great scores, your customers are happy and they’ll be back—and may bring their friends. Additionally, you can parse their comments and likely figure out where to focus your improvement. I agree that NPS® is definitely a predictor of growth. But, the word “predict” is where I get twisted up—to predict is to estimate what will happen in the future. I predict Alabama, Clemson or Ohio State will win the next College Football National Championship. Like my brilliant observation about the 2019 football season, NPS® feedback can be pretty good at telling us what we already know.

Whether or not NPS® is a good predictor of growth is not the right question. The right question is, “Is NPS® a great DRIVER of growth?” The answer is, not really. If you are measuring the NPS® of a product, then the comments will offer some good insights for improvement. If the measurement is about a service that is impacted by an employee, then the employees knowing that the survey is being sent may cause improved behavior and customer experiences. Also, seeing monthly reporting can make service levels more top-of-mind for employees. But NPS® isn’t intended to measure individual interactions or experiences. The magic happens when survey questions are specifically about a client’s interaction with an individual employee, and the responses are put into motion to drive improvement in real time.

Below is a brief history of the NPS® question and the NPS® scoring model. It is worth mentioning that in part, I owe my job to Fred and this question. In the evolution of customer experience management, NPS® played and continues to play a big role.

The Origin of NPS®

In 2003, Fred Reichheld of Bain & Company launched a project to determine to find a better way of measuring how well an organization treats customers and generates loyal relationships, working with data supplied by Satmetrix. The result was Net Promoter Score (NPS®): a metric said to predict customer purchase and referral behavior. Reichheld shared the methodology so anyone could apply it, and NPS® became widely adopted by companies to gather customer feedback. 

It was determined that the strongest sign of customer loyalty is measured by customers who are happy enough to recommend a product/ service to a friend. Reichheld said that when customers act as references, they put their own reputations on the line, but will do so only if they feel intense loyalty. 

After NPS® was refined, Reichheld released an article in the Harvard Business Review that laid out the value of NPS® to business. It explained how NPS® frames the question around how likely customers are to recommend a company in general, as well as how to calculate a score indicating potential for growth through retention and word-of-mouth. 

The Question 

NPS® is different from other metrics because it does not measure a customer’s satisfaction with a specific interaction. Rather, NPS® is designed to measure a customer’s overall loyalty to the brand regarding repeat purchases and referrals. 

NPS® does this by asking a single question, determined to be applicable for most industries: “How likely are you to recommend [This Company] to a friend or colleague?” The customer is then prompted to give a score between 0 and 10, used to calculate the company’s Net Promoter Score.

According to Reichheld’s scale, 10 means “extremely likely” to recommend and zero means “not at all likely.” When examining the customer behaviors along this scale, three clusters were found:

“Promoters” gave ratings of 9 or 10, considered a powerful asset. They’re customers with the highest rates of repurchases and referrals, likely to actively recommend your company to others. In most cases, the respondent is also asked to leave comments about their answer.\

“Passives” logged a 7 or 8, meaning they will likely not hurt or help the business. Because they are only somewhat satisfied but not enthusiastically loyal, they are not likely to actively recommend your company to others.

“Detractors” scored from 0 to 6, indicating dissatisfaction with your company. They are someone who will not only discourage or abstain from referring friends, family, and colleagues to a business, but may even seek to do damage to a company’s reputation due to their unhappiness. In a sense, every detractor represents a missed opportunity to add a promoter to the customer population—one more unpaid salesperson to market and generate growth on your behalf. 

How to Calculate

The actual NPS® score is then calculated by taking the percentage of promoters and subtracting the percentage of detractors. Passive responders are not used in this final calculation, because they don’t influence a potential new customer’s decisions one way or the other. 

This will generate a score ranging from -100 to 100. A positive NPS® means that you have more people recommending your company or product organically than discouraging others from it. A negative score would mean the opposite. Scores higher than 0 are typically considered to be good and scores above 50 are considered to be excellent.

For example, if you surveyed 100 customers and 70% were promoters, 20% were passives and 10% were detractors, your NPS® score would be 60 (70-10 = 60). 

How do Companies Use NPS®?

Facebook: NPS® of -21

Facebook’s score is just ugly. While NPS® is considered an indicator of satisfaction and loyalty, it may not be the best tool for a product with over 2 billion users. 

According to Gibson Biddle, a former executive at Netflix and Chegg®, Facebook is considered a communication utility, and folks don’t tend to rave about utilities as much as other products /services. He also notes that some reports suggest Facebook has plenty of detractors due to privacy issues.  

Zappos: NPS® of 57

Online retail company Zappos is well-known for their culture and company values, one of which is “Build Open and Honest Relationships with Communication.” Zappos uses the Net Promoter Score drawn from customer surveys to measure customer service performance, and also came up with the Happiness Experience Form based on how well the agent fulfilled the following:

  • Tried to make a personal emotional connection
  • Maintained their rapport and if the customer responded positively 
  • Addressed the unstated needs of the customer 
  • Gave a “WOW” experience or went above and beyond what is expected of them


Starbucks®: NPS® of 77

Starbucks® survey receipts used to offer one dollar off your next purchase or be worth one free tall hand-crafted beverage. Today, in-store survey invitations are a thing of the past, as Starbucks® now sends emails to rewards members asking them to rate recent visits. This used to be in exchange for bonus stars, but changed in 2019 to offering no incentive at all. 

NPS® doesn’t measure specific interactions

NPS® is designed to measure a customer’s overall loyalty to the brand regarding repeat purchases and referrals. NPS® does not measure a customer’s satisfaction with a specific interaction.  For executives in the experience space, we are mapping journeys and interactions within journeys. We are searching for meaningful improvement as we go. The interactions are where we see the biggest lift in business outcomes and that is where the CX 2.0 magic happens.

To recommend or not to recommend? That is the question. 

The CX 2.0 Audit is coming soon! In the meantime, understand how to Get Better, Get Found, and Get Chosen.

Share this article on: