This article won’t answer the question of what is NPS. I’m writing it assuming the reader already knows what Net Promoter Score is. Thoughts expressed in this article are related to tactical, touchpoint NPS.

Since the introduction of the NPS there’s a constant debate is it good, or is it bad, even worth nothing? And to be honest, I understand both points of views, moreover, I completely understand why it has a bad rep.

Don’t put a target on the number!

Let me start by saying that I believe in fast metrics. Fast? Yes. From a customer point of view: you have only two questions, one of which is mandatory. Your customer is using you to do their job. Filling in a comprehensive questionnaire is out of the question. But, pressing on a number and maybe leaving a comment takes 20 seconds of your life – so you’re more inclined in filling it out. Yes, maybe a comment. (In my experience, getting 10% of respondents to leave a comment is considered good. That’s why you need to strive to survey as many as possible)

From a company point of view – it gets you a baseline very fast, and that allows you to have a good foundation for further deep-dive analysis.

So, here’s where the ones that are against NPS see a problem – for a reason. You have a score of 32, 500 responders and a low amount of comments. This does not tell you how to improve. It does not even tell you where you are. It just tells you that the score is OK because you have more promoters than detractors.

And that’s true. That’s why I believe that focusing on the score, on the pure number does not lead you anywhere. And when done like that, NPS is a worthless metric. Don’t focus on the score, define targets to improve detected problem areas. That said, NPS should be considered as an indicator, not as a finite number.

Why an indicator? Let’s look at the following example that should explain why I think that focusing solely on the number is bad.

We have company A. Their target for NPS the previous year was 15. And they overachieved it. Everyone got their bonuses, everyone is happy. Except for the customers.

NPS score development without category data

Everything looks legit, right? But what if we dig deeper and reveal the category distribution?

NPS score development with category data

Noticing anything suspicious? See why I wrote customers are not happy? The number of Detractors is rising. And, from Q1 to Q4 rose by 30 points. Do we know why? What to do? How to reduce the number of Detractors? We probably don’t know, because the target was met. (And, yes, I know. Likelihood of this turn of events to happen is small, but it’s here only to show why it’s important to go beyond the number and establish a feedback loop.)

Analyze the data behind the responders

So, how to avoid getting trapped inside the “everything above 50 is awesome” trap. Simple, don’t present the NPS score by itself. Always draw conclusions and correlate them with the score itself. Dig even deeper, and see how different customer segments, geographical regions, or industries score your company. And that’s the easy part, if you set up your survey right, you’ll be able to draw your customer data and analyze it further. (assuming you didn’t set up an anonymous survey)

And there’s the hard part. Feedback. Actual customer feedback besides the number. How to get that? Depending on the size and maturity of the company you work for, there are two ways of getting more responses from your customer base.

Implement a strong feedback loop

To be able to make NPS actionable, you need to implement a strong feedback loop. You need to show your customers their feedback matters. By nurturing a relationship where feedback is appreciated, with the time you’ll increase the number of respondents, as they’ll see the value of taking the time to leave feedback. A feedback loop can mean:

  • Collect score – in a product.
  • Ask for feedback (email, live interview)
  • Detect pain-points
  • Define action points
  • Implement
  • Score again
  • Ask for feedback...

Collecting the score in the product maximizes your response rate. While response rate over email was often no more than 10%, in my personal experience once started using in-app surveys we pushed the response rate up to 60%, which we can all agree on is an impressive number. (I've used Gainsight PX for in-product communication) And, with 10% of comments, it’s a good playing ground for the initial analysis. But that is not enough. 90% of the respondents still didn’t leave you a comment – and without acting upon it, it only leaves you room to speculate. In short, you need to get to the bottom of the “why”.

NPS Category distribution

Don’t be afraid to ask for additional feedback

But don’t spam your customers. (I know, it sounds contradictory, but that’s a whole different topic to deep dive into). Contact them in two to three iterations maximum, with a minimum pause of 24 hours between the communications. Automate your survey by setting up auto-responses to customers that didn’t leave you any comment. Divide them into three groups – Detractors, Passives, and Promoters. After they score you, send them an email where you thank them for taking the time and ask why is it they scored you with X.

Follow up email example


Make the email as personal as it gets. Many tools support sending “on behalf of”. Make that on behalf of their Customer Success manager. Show them you care for their experience and explain how their feedback will make it easier for us to make their experience better.

With the spot to leave written feedback, integrate a link to your calendar to allow the customer to book a 30-minute slot with you if they prefer to talk live. Sometimes, customers just want to be heard.

Besides the automation, feed the NPS data to your Customer Success. Make it a routine for them to talk about the score when they have their regular meetings with the customers. And, not only talk – they should provide you with the feedback so you can assign that to the responder.

These actions will help you detect pain-points and problem areas. Backed up by numbers and analytics. It definitely helps to have it up your sleeve when proposing action points to change and mitigate the problem areas.

Did you say analyze, again?

Yes. Perform periodical analysis of NPS results. Re-survey your customers every 3-6 months. Look at the trends, and always, always look for problem areas and define action points.

Again, even if the NPS is 50 (which is considered good!) we still have an X number of unhappy customers. And they are the ones that could predict a future problem, even churn if the problems are not addressed.

And yes, never hide the bad comments. The opposite - be thankful for them as they allow you to evolve your product, give your customers a better product experience that directly contributes to your company growth.

Would you agree, disagree? What is your experience with NPS? I'm interested in hearing your thoughts. You can leave them in the comments below.