Using ChatGPT to gaugue app reviews
Published Date
Identifying Trends In App User Comments
Like the rest of the world, we've been enjoying playing around with ChatGPT for the last few months. One nice use case we've found is to point ChatGPT's skills in analysis of natural lagage to use in shifting through app reviews for some of our our clients.
What We Did?
As a very basic proof of concept, we created a small phyton scrpit to scrape the reviews out the last 1000 app reviews from the app stores of one of our apps. This gives us a neat CSV files upon which we can run all sorts of analysis.
For our purposes, we simply fed this file into the ChatGPT browser and ran an analysis to uncover the most common user problems people were facing with one of our apps. This confirmed a few of the main issues we're already working on but also flagged that a few others that we knew about needed more immediate attention.
Where Could We Go Next?
The next logical step for this tool would be to integrate the scrapping element and ChatGPT's API into a single application so that we could get fully formed analysis out without a two step process. Regardless, we've found that this is a helpful tool for keeping us focused on the main issues users are facing at a scale that's bigger than we had before.
Some Prompts To Try On Your Own Data
-
