NPS Best Practices: The most effective way to send a Net Promoter Score℠ survey

Have more questions? Submit a request


  • Chris

    I agree with Dawkins. Not just the NPS survey but the Satisfaction survey. It really should be customizable. You already proved you can make a data model that handles binary responses, and responses from 1-10. The question text on both surveys should be editable. Ideally you would be able to expand to at least 2-3 questions as well (non open-ended). This would increase the value of both surveys significantly for many companies.

    Needless to say putting all this in Insights so it can be sliced by issue, office, agent, and a host of other measures for analysis would be invaluable.

  • Pallavi Barnwal

    In the article it is been assumed that you receive a 15% response rate in your NPS mails sent through Zendesk. Now we have 1%. How to improve that?

  • Seama Rezai


    I would recommend to continue sending the surveys, but to also make sure to send them to all your customers so you are more likely to get responses. Really, it's up to your customers to fill them out, so if you give them the opportunity hopefully they will take it :)

  • Andrei Kamarouski

    Hi all!

    We are stoked to inform you that the first version of our Loyalty Radar app is out and available for installation in your Zendesk as a private app. You can download it from this article. Before you proceed to the installation, we recommend to take a look at a couple of other articles that we created to guide you through the installation and setup processes.

  • James Green

    Nice series!

    I saw one little typo:

    "We recommend using one of the various randomizer tools available on the Interne."

  • Jennifer Rowe

    Oops! Thanks for letting us know, James. It's fixed now.

    And glad you like the series!

  • Brad Rhees

    I'm most interested in measuring NPS scores of different customer types and comparing them to each other. Specifically, I'm interested in knowing how the feedback of a customer who is in-trial differs from a customer who has been with us for 2-3 months, or a customer who has been with us for 6-months, or a customer who has been with us for over a year. 

    Do I need to launch a different NPS campaign for each of these groups? If so, is there a good way for me to compare the data? 

  • Dan Kondzela

    Thank you for your question, Brad.

    You are correct, you will need to construct a few unique NPS Surveys to capture the information you're looking for here. You can do so either by uploading a custom, external CSV File or using Customer Lists, should you have that particular function. (

    Once you close out these NPS surveys you will be able to compare past surveys by selecting a date range. Your options are described here (


  • Brad Rhees
    Thank you, Dan. 
    Is there a recommended time for how long each NSP survey should last? Because if I've got 5 different groups to measure, I'm worried that enough time will pass from the first survey to the last survey that other factors will impact the scores (bug changes, new feature rollouts, etc.) and i won't be a very accurate comparison. 

    I would really like to measure these responses by customer-type simultaneous to each other. 
  • Dan Kondzela

    I can understand the desire there, Brad, unfortunately there is not a best time frame to suggest. It is dependent upon your own workflow's needs. If you are looking for relatively contemporaneous results I would recommend about a week's lifecycle for each of these, if you feel that would be long enough.


Please sign in to leave a comment.

Powered by Zendesk