App user reviews are like gold. To hit paydirt, you have to know where to dig.
So let’s dig into how you can go about surfacing the valuable insights running through all those app store reviews.
Map Comments Within Reviews to the App User Experience Flow
Start by considering how users move through your app. The goal is to get them using the app completely. So any hiccup or speedbump they incur along the way is worth looking into.
If you’re conducting entity sentiment analysis on those reviews, the process will first identify all the entities in the reviews, and then generate sentiment ratings for each entity (‘entity’ is essentially a $10 word for ‘noun’). Those entities/nouns, when pulled from reviews, are typically user journey touchpoints.
So, by selecting a given entity along with the related text in the review, you’re getting user commentary on precisely what they think about that given touchpoint.
Add that up across all your reviews, and you have extensive input on what users think about an entire series of app touchpoints.
Entity Analysis for Healthcare Apps
Let’s take a look at how this applies to a healthcare app.
In this case, we’re digging into the Aetna Healthcare app reviews. After running entity sentiment analysis for reviews from 2021 on, we get a list of over 10,000 entities. Remember, these are authentic user comments on specific sections of the app user journey.
After looking at the entities, we can see which offer the best opportunity for optimization. Those with both high volume and with lower sentiment ratings. The high volume means this is a sizable problem impacting a number of the apps users. The low sentiment ratings mean there’s a lot of room for improvement, and fixing the problems described should improve the apps user ratings.
Password issues are at a critical point in the user journey. They often prevent people from onboarding or using the app. Or at least they make the process far less pleasant, prompting reviews with lower ratings.
Viewing all user comments on a given topic
After identifying a critical user touchpoint like ‘password’, we can select the various versions of that entity to get the full picture of what’s being said about that touchpoint.
That helps us focus on a very specific area along the user journey.
To further isolate the most negative comments, we can focus on only the comments where the negativity is at a specific level. Let’s say a negative sentiment of -.2 or lower.
And shazam! Like that, we’ve just poured those 10k+ reviews, revealing key user feedback on one of the most troubling issues being reported.
And of course, you can do this for any touchpoint up and down the user experience flow. This not only informs your product design team, but also provides the ‘why’ behind decisions. Ultimately, it answers the big question, “What should we do to improve our app?”
Interested in reviews analysis for app design projects? Let’s chat.
