Net Promoter Score (NPS)

A summary and review of NPS as a measure of customer satisfaction and advocacy

“There are three kinds of lies: Lies, Damned Lies and Statistics.”

— Benjamin Disraeli (quote popularized by Mark Twain)

I don’t think NPS (Net Promoter Score) is particularly useful. It’s a bad quality metric and everyone has got too carried away using it.

As the business community has widely adopted the use of NPS to measure customer satisfaction, the commodization of this measurement technique amongst businesses has also brought with it a range of different interpretations and applications of the same.

Let me elaborate on this in a bit more detail and break things down.

---

What is NPS?

For those not familiar with this concept, NPS is a customer satisfaction methodology that measures the likelihood of a customer recommending your brand.

To provide further detail, there is a distinction between T-NPS (Transactional NPS) and R-NPS (Relationship NPS).

The difference between these metrics is that they measure the likelihood of the customer recommending your brand, based on the last interaction with your company or in relation to the overall business-customer relationship.

These customer advocacy ratings are measured on a scale between 0-10.

0-6 = Detractors

7-8 = Passives

9-10 = Promoters

Then, we work out the percentage allocation of each of the three categories.

NPS = % Promoter score - % Detractor score

Example: if 50% of your customers are Promoters and 20% Detractors, your NPS is 30.

---

What is a good NPS score?

Your company’s NPS score can range from -100 to +100.

A higher score reflects a stronger customer advocacy for your company.

Indicative ranges:

• Needs improvement: -100 to 0

• Good: 0-30

• Strong: 30-70

• Outstanding: 70-100

---

---

What is wrong with NPS?

One of the main issues has to do with the data range itself which can cause confusion.

For example, let’s say that a company with an NPS score of 10. Note that this is not the same as a score of 1 out of 10.

Rather, it means that there are twice as many Promoters as Detractors. For example, 20% Promoters less 10% Detractors equals an NPS of 10. Logically, it also means that the remainder must be 70% passives.

This is a fair score! But somehow, the result can be interpreted like being close to zero (which is not the actual bottom of the measuring scale) and some may instinctively feel that a good score should be much higher than this.

In addition, I would highlight a selection of other important drawbacks with NPS. Here are the five factors which I consider most relevant:

1. Narrow perspective

   An NPS score focuses on a single aspect of customer advocacy and customer satisfaction, but provide limited insights into the customer experience.

2. Is the likelihood to recommend even likely

   The top result to the question “How likely are you to recommend…?” doesn’t necessarily mean that the customer is going to recommend the brand to anyone. Also, the response rates to these type of surveys can be low and may generally not be statistically significant on their own.

3. It’s not a direct indicator of customer value

   The dimensions of NPS are either transactional or relationship-based, whereas customer value is a broader concept. However, it seems reasonable that if a customer finds value in the product, they might recommend it to others, but it’s not a causal data relationship as such.

4. Limited usefulness for senior management

   Ask yourself what strategic or tactical application the NPS score might have for a company. E.g. 30% of our company’s customers (net of the detractors) would be likely to recommend our brand. What should a senior leadership team do with this information?

5. Massaging the numbers

   Every company will want to get the highest score possible on NPS. Due to the nature of the NPS survey, it is subject to potential selection bias and results can be easily manipulated. See an example of this further down in this post.

---

Guidelines for carrying out an NPS survey

In case you still want to stick with NPS despite the flaws that have been discussed so far, here are some basic guidelines of how to run an NPS survey.

The key aspects I would highlight would be:

• Ensure that you use standard data range discussed above (0-10)

• Use a short, clear and standardised questions to your customers

• Use neutral language in the survey request

• Remove selection bias by:

  • Offering the survey to all customers

  • Offering the survey to randomly selected customers

  • Using an independent person or third-party to handle the survey process

• Do not offer customer incentives or otherwise try to influence the outcome

• Send out the survey requests in a timely manner that makes sense

• Collect verbatim comments where possible

  (e.g. “Why did you give this specific rating?)

• Automate and digitise the processes

It should be a fairly straightforward process to run an NPS survey but a lot of things that can go wrong along the way.

Here is a real-life example of how NOT to carry out such a survey:

---

What are the alternatives to NPS?

Let’s take a step back: What are we trying to do and achieve here by measuring NPS in the first place.

Remember — we trying to gauge the level of customer advocacy and customer engagement in our business, with a specific interest in those who are our strongest brand advocates.

The first question I would ask is whether this information might already be captured elsewhere in your business performance KPIs?

If you have a sub-set of customers who are transacting with you frequently, are high-spenders, have a long customer tenure, file no complaints etc — is it not likely that this customer profile also represents a brand advocate?

I’m alluding to data segmentation and customer profiling here, and as such, I think that as a first step it is advisable to double-down on ensuring that you have a really good and holistic KPI performance tracking in your business.

Here I am referring to:

• Revenue metrics

• Customer behavioural metrics

• Customer health scores

• Operational metrics

To my mind, there isn’t one single substitute for NPS at the moment but on a concluding note I would argue the case for going back to basics.

Rather than using a 10-step scale to measure customer satisfaction or similar, it would be helpful to simplify matters to use closed-end questions with only two alternatives to remove statistical ambiguity and provide clear and actionable results.

Example questions:

How would you rate the service you’ve received today?

• Good, I’m satisfied

• Bad, I’m dissatisfied

Do our products meet your customer needs and expectations?

• Yes

• No

In addition to the above, you can also use open-ended questions to capture customer feedback and verbatim in writing, such as in a Voice of Customer programme and I would also warmly recommend keeping a very close eye on complaint volumes and what customers are saying in these.

“Your most unhappy customers are your greatest source of learning.”

— Bill Gates

---