Kissmetrics Blog

A blog about analytics, marketing and testing

Built to optimize growth. Track, analyze and engage to get more customers.

Why Customer Satisfaction Surveys Aren’t Useful and What To Do About It.

Definition: Cus-tom-er Sat-is-fac-tion Sur-vey  (n): A long, complicated and arduous task for both a company and their customers that can often yield little to no useful results.

Do you really think your customers WANT to fill out your customer satisfaction survey? How many times have you blasted through a survey just to get the freebie that comes with it? Well guess what? Most of your customers are probably doing the same thing and ruining all your “precious” data.

Furthermore, it can be overwhelming and difficult to make any sense out of the survey responses. Wouldn’t it be nice if there was just one question you could ask your customers that revealed how healthy your business was doing? More importantly, one question that resulted in data that correlated to profitability?

Well, actually, there is. It’s called the Net Promoter Score or “NPS”. NPS is used by some of the biggest brands to measure customer satisfaction and business performance. Who uses NPS?

  • General Electric
  • Procter and Gamble
  • Verizon Wireless
  • eBay
  • Intuit

How NPS Originated

the one number you need to grow

NPS was originally introduced by Fred Reichheld in a Harvard Business Review article back in December of 2003. The article entitled “The One Number You Need to Grow” was based on a two-year study that Reichheld decided to conduct, all stemming from inspiration gained at an afternoon seminar.

It all started by Reichheld listening to a talk by Andy Taylor, CEO of Enterprise Rent-A-Car. Enterprise would conduct quick two day surveys across all their locations by only asking their customers two questions. On top of that, the only responses counted and internalized were those from the most enthusiastic customers (these are the customers that drive growth). By doing so, Enterprise could quickly keep tabs on the best performing locations and spread the lessons learned to other branches.

Inspired by this talk, Reichheld went to work trying to develop a simpler and more direct survey that would become a metric for profitability. After two years, Reichheld boiled down that one magic question:

“How likely is it that you would recommend [enter the name of your company] to a friend or colleague?”

How NPS Works

net promoter score survey

The NPS system states that every company has three types of customers:

  1. Promoters – Evangelists. Customers that preach and spread the word about your product or service. And of course, they continue to be loyal customers.
  2. Passives – Customers who are blasé about your company. They may have been satisfied with your offering but can easily go with one of your competitors.
  3. Detractors – Former customers who loathe your company. They love to preach the “bad news.”

Customers respond to your net promoter score survey by choosing a number between 0 and 10 — “0” being the lowest possible rating (not likely to recommend your company) and “10” being the highest possible rating (most likely to recommend your company). The following scale determines which of the three types of customer buckets they fall into based on their numerical choice:

  • Promoters (score 9-10)
  • Passives (score 7-8)
  • Detractors (score 0-6)

To calculate your company’s NPS, take the percentage of customers who are promoters and subtract the percentage who are detractors. To give you an idea of some (very high) NPS scores, here are a few scores from some familiar brands:

  • USAA – Banking 87%
  • Costco – Dept Store and Wholesale 77%
  • Apple – Computer Hardware 72%

Average companies usually score between 5 and 10%. Stellar companies operate at NPS efficiency ratings of 50 to 80%.

KISSinsights and Net Promoter Score

Update: We have sold our tool, KISSinsights, to CatchFree. The tool is now called Qualaroo. Qualaroo provides all of the support for the product.

By simply enabling Qualaroo on your company website you can continually gather NPS data. This will give you year round feedback on how your company is doing and provide a great compass for profitability.

You can easily configure Qualaroo to only show the net promoter survey on certain webpages, or to registered users only. Also, many timing and triggering options are available, so that the survey is only shown to customers at appropriate times.

nps sign up

If you’re interested in trying our NPS solution on your company website, please click on the button above. Feel free to contact Qualaroo if you have any questions with installation or understanding your NPS results.


The One Number You Need to Grow (PDF file)

About the Author: Sean Work is the Marketing Director at KISSmetrics.

  1. NPS makes a lot of sense, I have had no success in pulling useful data out of past customer surveys.

    The question I have is: Once I collect the data, what are some best practices that I can use to improve the score?

    NPS should be an easy sell to the E-suite if I can show that there are actionable best practices that we can implement based on the score.

    • Hmmm. I would actually think of it in another way. I wouldn’t focus too much on just the number. But ask yourself, how can you improve your customers satisfaction as a whole.

      Of course every company has its own intricacies so you would have to look into the details of what makes your customers happy and go from there.

    • Alistair Nicholls Aug 14, 2011 at 11:26 am

      We love extracting useful information from data, explaining it (including statistical errors), and producing resulting logical ways to improve customer satisfaction.

      Your comment about not pulling useful data from past surveys is a challenge I’d love to take on! Send me an email at alistair@perrynichollscouk if you’d be interested in talking. We will tell you straight if there is nothing we could add.

  2. Absolutely beautiful. Love this NPS widget. You guys are amazing.


  3. Siddharth Mangharam Aug 13, 2011 at 4:14 am

    Very timely article. We’ve been struggling with one page surveys for two months now – our members hated them and the data had no correlation with their actual behavior. We’re going to switch to NPS and will also check out the NPS widget.

  4. @bobbleheadguru Aug 13, 2011 at 9:26 am

    With all due respect, I have seen comprehensive Customer Satisfaction Surveys be extremely useful for literally 100s of organizations. Perhaps more useful than any other analytics data that practitioners collect and analyze.

    I know that you are trying to sell a new product but to imply that the work of ForeSee or iPerceptions or Opinion Lab is not useful is probably not the best way to improve your own credibility.

    I would venture to guess that many more top companies use comprehensive satisfaction surveys than use NPS, and for good reason.

    NPS is not statistically valid, sensitive to change or reliable.

    The idea of asking people to rate one question (which is not a satisfaction question), then have them subtract detractors from promoters reduces the precision and actionability of the data to the point of almost randomness.

    There is no peer reviewed academic study that I am aware of that links NPS score with financial performance.

    Think about it logically. If the “likelihood to recommend” question were asked as the only question on a survey, then was graphed as a histogram, the distribution could be bipolar, a bell curve or flat.

    However, they would all potentially have the same NPS score, because the Detractors and Promoters are negated and the Passives are ignored.

    The fact that is it not credible or actionable, does not mean it is not popular, however. It key benefit is that it is it is easy and Reichheld is a master marketer. [His book is a best seller despite the fact that the correlations he implies are spurious.] The key drawback is that it does not work.

    Please feel free to contact me to discuss further. I am not trying to put rain on your parade, but NPS does not follow the basics of statistical validity, reliability or precision. Ask any student of a basic freshman level statistics to look at NPS and they will tell you you the same thing. I hope that I would have some influence on your thoughts with respect to that concept.

    As mentioned before, I would love to discuss further. I do not want to come across as putting a dampener on your new product, but I do think it could be better if it truly measured satisfaction in an actionable way.

    • Thank you for your long thought out comment!
      We should make something clear first. KISSinsights does a lot more than just Net Promoter Score. As a matter of fact, watch this video to understand what its primary usage is for:

      What I was looking forward to saying in the comments is: the real takeaway is to do what works best for your customers and your company. To put it simply: be careful about creating incredibly long surveys.

      A lot of new companies (start-ups) probably dive into customer satisfaction with long surveys and are not really prepared to make sense out of the responses. It can be very expensive and time consuming. NPS is a very simple way to get started and it may work wonders. Besides, why not start with something simple?

      Bobbleheadguru – I will definitely be contacting you soon! :)

  5. Alistair Nicholls Aug 14, 2011 at 11:16 am

    We do a lot of customer satisfaction surveys & include Net Promoter Score question.

    There are often very large statistical errors in NPS %ages (typically 10%-15%) so take results with a pinch of salt, and don’t worry about minor fluctuations.

    Also, results vary dependant upon industry sector; some products / services are much easier to deliver well than others.

    30% is a good score. Many organisations run at 5%-10%, and some whole industries can be negative scores!

    And customer satisfaction surveys if done well, and analysed thoroughly and robustly can be very powerful. We have helped organisations improve their NPS from 5% to 30%!

    If anyone has any data they can’t extract useful information from, I regard that as a challenge! Contact me by email ( and lets talk. It will cost you nothing.

  6. Do you think this is something that bloggers or small online businesses could use?

    • I definitely think it’s worth a shot. It’s really easy to start using. If you find that this metric helps with achieving your goals then what do you have to lose?

      Running a small business is extremely time consuming and difficult. You ideally want your tasks to be quick and easy because your time is VERY expensive. If you end up creating a monumental survey task for yourself, I could see it never getting done. I know so many small businesses that attempt to try different activities, and they end up on the eternal back burner…

  7. Why is Enterprise Rent A Car involved in the NPS conversation? The question they ask is “How Satified”. And they count top two box scores, they don’t take out detractors. I see their methodolgy closer to CSI.

  8. An interesting article and great read on NPS.



Please use your real name and a corresponding social media profile when commenting. Otherwise, your comment may be deleted.

← Previous ArticleNext Article →