Question about survey changes on active survey | XM Community
Skip to main content

Hello!

I've been tasked with standardizing a few CES questions. These Customer Effort Score questions are being asked in various surveys in our customer journey (we are a SaaS organization). All of these surveys are active and we’re actively leveraging the data.

Old -

How east was it to work with COMPANY on your tech support case?

Answers: Very Easy - Easy - Neutral - Difficult - Very Difficult

New -

Tell us how you feel about the following statement:

Company made it easy to verb my noun.

Answers - Strongly Disagree to Strongly Agree

 

After some research I understand that I need to keep the old and “hide” it then display the new question. The main question I am trying to figure out is how I can use the old data and the new. I’ve looked at bucketing and combining fields but am not sure which (or if there is another alternative). Ultimately, I feel like the question differences are small enough that we can use both together. How would I go about combining the previous response and the new responses (all centered around Customer Effort Score)?

Could you ask both questions in the same survey for a short time to test if the two questions get similar responses?

In the similar scenario, we had used filter and bulk edit option in data and analysis tab i.e. trasferred the old data in the relative new data format.

Could you ask both questions in the same survey for a short time to test if the two questions get similar responses?

@MikeW Great question, the reason for this change is to standardize across similar surveys so, that we can compare this metric across the customer journey. I’m trying to make this change without losing the impact of the previous 17k responses that have already answered. 

I might have misunderstood, but if you’re happy that this metric is directly comparable, then you could just update the wording of the original question and the individual labels. That way you don’t need to move/ recode any data - it will be in the same variable. You’ll just need to be mindful when analysing that the data is both question versions combined. If you have a version variable in your script then you’ll know when the change was made and which respondents saw the old wording and which respondents saw the new wording.

Also, take care when changing the labels (best to change each one at a time, so you don’t accidently create new codes/ don’t accidently delete old codes).

Perhaps do a test on a small dummy project to check it’s doing what you want.

We had similar, but different scenario recently.  Our question was more around what happens when survey distributions overlap (old & new).  Those who got the original survey, will continue to see the original survey even if you change/update the survey questions whilst it is in field.  So as MikeW says, you should be able to update the current question and labels for the 5 point scale if the data values are remaining the same and your happy that it’s comparable.  Keeping the data variable the same is key to not separating the data.  

Leave a Reply


Terms & Conditions