Text IQ and impact of topics on NPS and other metrics | XM Community

Text IQ and impact of topics on NPS and other metrics

  • 11 February 2019
  • 6 replies
  • 53 views

Userlevel 7
Badge +23
  • Level 5 ●●●●●
  • 449 replies
How do you guys calculate impact of open comments topics on such metrics as NPS? We would like to show our top management that there are some topics that need to be addressed and if we fix this or that, our NPS potentially improves and reaches some particular range? We would really want to have it in a more visual way than just circles and hopes.

6 replies

Userlevel 7
Badge +33
Good Question, there are number of ways to do this like one below.
Create topics and then show these topics in widget across NPS, like topics % as a bar and NPS as a line in same graph.
You can use filters like NPS Groups(Promoters, Passives, detractors) and Sentiments (+ve, -ve) when you filter page using Detractos, -ve sentiments you come to know the pain areas or areas needs improvement.

I would love to learn other ways you guys are using to do this task...
Userlevel 1
Badge
We also utilize the "% topics as bars overlaid with OSAT (or NPS) as a line". This allows us to see particular topics where NPS is lower, implying that these are pain points. It's been pretty useful, so I wanted to echo that suggestion!
Userlevel 7
Badge +61
Are you in general happy with the Text IQ functionality?

If you're looking to identify areas of opportunity that will have the most gain for the company, I would say that the number of negative mentions of that topic is a good start, but to get the best bang for your buck I'd look at Customer Segment to see what the "most important" customers say.
No use in implementing change because 10 single users complain. Better focus on solving the issues 5 of your largest Accounts complain about.

I've also tried to investigate the "value of increasing the NPS", meaning the actual kickback to the company that comes with a higher NPS score. So far, the internet has not provided me with any good answers 🙂
Userlevel 7
Badge +11
This may or may not be helpful, but I use TextIQ to categorize my data into "buckets", then verify that the assigned sentiment score and polarity make sense. Then I give all of that to my statistician to do a key driver analysis.

I know that you're supposed to be able to do that within Q, but every time I do, it tells me that _everything_ is correlated so my statistician gives me a definitive answer on what would improve NPS the most.
Userlevel 7
Badge +11
> @MattM said:
> I've also tried to investigate the "value of increasing the NPS", meaning the actual kickback to the company that comes with a higher NPS score. So far, the internet has not provided me with any good answers :)

Agreed! We plan to do a statistical study to see the relationship between NPS and churn or upgrades later in the year. I can't wait to see how strongly they're correlated since I've long championed that we should focus on several CX metrics and not just NPS.
Userlevel 1
Badge +4

https://community.qualtrics.com/XMcommunity/discussion/comment/11388#Comment_11388I would love some details on how you built this widget please! 🙂 I've tried a few different way and have been unsuccessful.

Leave a Reply