Tổng quan

With the current booming of technology development, creating mobile applications has become a trend. Did you know that in 2022 alone, there were approximately 110 billion app downloads on Google Play (Google Play annual app downloads 2022 | Statista ). In this project, we will analyze the application market on the famous app marketplace worldwide (Google Play). From there, we can use our findings to inform our decisions if we want to create our own app.

Owner: Google Play
Industry: Mobile application and digital content distribution industry on mobile devices.
Services: App Distribution, Digital Content, In-App Purchases, Subscription Services, Developer Tools.

Challenge

With a vast market like app design, it is challenging to identify valuable features. This requires not only understanding the positive values within the data but also gaining insights into the industry (such as understanding app download habits, what users genuinely dislike, and sometimes their irrational behaviors driven purely by temporary emotions [like network lag, someone triggering their anger, or a heavy app]).

After much analysis, we should ultimately decide on the type of app we want to create, among hundreds of options

Solution

Data cleaning and validation

Luckily, we have a guy who loves downloading apps and especially enjoys playing and exploring mobile games :D. We already have someone who is a user and understands the app market.

Practicing data cleaning skills and exploring the distribution of different types of apps. Comparing features with each other to discover correlations. This will be a long EDA journey.

We also explore user preferences for applications. We have two bases in the data to do this. One is app ratings, and the other is sentiment analysis.

Results

Understanding the current trends of apps on Google Play.

We gain insights into the relationship associated with the price and size of the app.

We gain many insights from the data, especially seeing the hidden differences behind ratings and sentiment analysis. We can conclude whether users prefer paid or free apps (downloading many free apps does not necessarily mean they prefer free apps).

We can conclude that if you want to design your own app, according to the market, you should design the most advantageous type of app.