Fielding this question to the community to see if you have run across this same debate with your team ...
What is the best practice for handling surveys when the respondent only answers the embedded NPS survey question in the email and nothing else?
Do you still record that score b/c any info is better than no info?
Or do you not keep their score b/c they hadn't bothered to give you any further input to help us understand their score?
It's hard to know whether they thought the survey was only the one question in the email, whether they didn't answer the other questions b/c the upload time took too long, whether they just didn't care to answer the other questions, etc.
We found that almost all our partial completes were people who only answered the embedded question in the survey.
Do you keep all your partial completes? Do you only keep the partial completes if they answered beyond the first embedded NPS question?
Any use cases would be very helpful to us.
Thank you!
Keep responses for those who only answer the embedded survey question in the email and nothing else?
+1
Userlevel 7
+27
hi @addyc_12
As you mentioned any info is better than no info , and given that it was an NPS question which would be probably rating some experience. so storing that data would make the most sense. Some companies even keep their NPS question mandatory and rest are shown but not kept mandatory. And as this questions were sent by email so this would have been sent to your targeted customers so it is not recommended you delete the data.
As you mentioned any info is better than no info , and given that it was an NPS question which would be probably rating some experience. so storing that data would make the most sense. Some companies even keep their NPS question mandatory and rest are shown but not kept mandatory. And as this questions were sent by email so this would have been sent to your targeted customers so it is not recommended you delete the data.
Since we're talking "Best Practices" then I think I'd like to re-framing the question a bit. I would keep the incomplete, but more over: I wouldn't embed the question to begin with.
Speaking frankly, If you are embedding a question in an email you are telling customers "This is the single question I want answered". If there is anything else after that- it is a betrayal of customer expectations.
If you expect them to answer more questions, rather than embedding a question inline, try communicating that "if you can take our survey, it has just 5 questions and will take less than 5 minutes of your time".
I think you'll find informed consent goes a long way to helping maintain healthy completion rates! Keep surveys short and ask only questions that REALLY matter to you. Customers are more wiling to help when they know that their feedback is important, and their time is VALUED. Sneaking in other questions without notice is going to make them feel a little betrayed- they will certainly loose trust in your research, and perhaps your brand.
Speaking frankly, If you are embedding a question in an email you are telling customers "This is the single question I want answered". If there is anything else after that- it is a betrayal of customer expectations.
If you expect them to answer more questions, rather than embedding a question inline, try communicating that "if you can take our survey, it has just 5 questions and will take less than 5 minutes of your time".
I think you'll find informed consent goes a long way to helping maintain healthy completion rates! Keep surveys short and ask only questions that REALLY matter to you. Customers are more wiling to help when they know that their feedback is important, and their time is VALUED. Sneaking in other questions without notice is going to make them feel a little betrayed- they will certainly loose trust in your research, and perhaps your brand.
+1
Well it turns out that an email server firewall may have been the culprit to causing a significant # of responses recording a 0 NPS score, fortunately easily identified as "Spam" in the Response Type column. The first alert to us was noticing that we had many more respondents for the first NPS question embedded in the email but then a drop in # respondents for the rest of the survey --> then we noticed almost all of our partial respondents only answered the 1st question and nothing else --> then we noticed the partial respondents (vs full respondents) were driving down our NPS score and almost all of the detractor scores were 0 --> then we noticed "Spam" in the Response Type column.
Qualtrics engineering team helped identify this as an email server firewall issue where they're checking links and then getting logged as a survey response however with no user input.
Have any of you run across the same issue where you found several respondents who only answered the embedded NPS question and nothing else, and they were all 0 NPS scores?
What kind of workaround did you do to survey these customers? Not embed the NPS survey question in the email? Remove images?
Qualtrics engineering team helped identify this as an email server firewall issue where they're checking links and then getting logged as a survey response however with no user input.
Have any of you run across the same issue where you found several respondents who only answered the embedded NPS question and nothing else, and they were all 0 NPS scores?
What kind of workaround did you do to survey these customers? Not embed the NPS survey question in the email? Remove images?
Leave a Reply
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.