Here, we will describe the customer feedback loop that you need to build and execute in order to successfully launch and grow your app. It doesn’t matter if you work for a big brand with apps that generate millions of downloads, or if you are a solo developer. The process would remain the same, and you would just need more tools to handle the dataset.
If you enjoy this type of videos, subscribe to my channel in YouTube
So why it is so important to analyze app reviews?
Because reviews drive downloads, ranking on the app stores, and consequently…revenue. It sounds like a paradox, but what mostly drives downloads, is downloads. In fact, the more your app is downloaded, the more Google or Apple will push it up on the ranking, generating more visibility, and more downloads. So, we should shift our focus to what makes people download your app instead of the others. The answer is “reviews”.
More than 80% of customers read reviews before downloading an app, and almost 95% of the top 100 apps on the app store have a minimum of 4 stars.
It’s interesting to know that about 75% of users leave a review after a negative experience, while only 60% leave a review after a positive one. This means that you will need many more happy clients to generate a good score and a sufficient amount of reviews.
Let’s summarise what we said so far: to increase the revenue of your app, you need more downloads. Downloads are mostly determined by two factors: the ranking on the app stores, and the average review score. To hit the top 100 you need a minimum of 4 stars.
At this point, the question that comes next is how do I get a high star rating? The number one reason for people to leave a negative review is to file a bug report. In fact, 40 percent of reviews talk about bugs. This percentage increases even more if we consider only negative reviews. If you look at 1-star reviews only, you will notice that almost every review talks about bugs, stability or other functioning issues, such as too many popups, or permission requests.
Consequently, the answer to get a high star rating is…fixing the bugs. And you know what is interesting? Customers use the app store reviews not just to complain, but they often give a detailed description of what bugs they find, or when is the app crashing. 25% of customers even suggest new features to improve their experience. In short, they use reviews to interact with the creator of the app. Why they do that? Because customers expect you to read their comments, and release their desired version update.
Now that we have fully understood the importance of reviews in the app stores, we need to find a way to analyze them, and seamlessly design the improvements that make customers happy. How to do that? You can do it manually if you don’t have many reviews, but you need a tool if you work for a bigger company.
If you do it manually, you would have to:
1. Read the reviews: you can do it directly on the app stores or download an extract to your computer. I recommend doing so because some reviews may be deleted from the app stores, and if you have them downloaded on your computer, you would not lose them.
2. Classify reviews using a dictionary that suits your product. Create a list of topics and use them to classify the records. Be consistent during the analysis, and don’t change the dictionary during the analysis, otherwise, it would be hard to trust the results.
3. Understand the correlation between negative reviews and negative topics. This would tell you which topics are more relevant for your customers, and determine the priorities of your development team. Even if impact, relevance or prediction are usually performed by complex algorithms, you can still try to derive some conclusions yourself.
It’s good to do this type of analysis also for competitors’ apps. You can do it pre-launch to study customers’ expectations, or continuously after-launch, to understand how to win the competition.
To wrap this up, we can say that reviews play a crucial role in app stores. Having a good score is the only solution to generate organic growth and increase sales. How to have a good score? Analyze what customers say in reviews and determine which factors influence their opinion the most. Release updated versions with bugs fixed, and keep this loop going.
If this ends to be too much work, you can use some automated solutions that use scraping, data mining and prediction models. Here, you will find a link to a use case of one of the brands we have been working with.