Google Play Store Market Analysis: What App to Create?

Introduction

Choosing the right app to create is not an easy task. Using the Play Store apps data has enormous potential to make the task easier and drive your app-making businesses to success. Today, our team will introduce you to actionable insights that can be drawn to work on and capture the Android apps market!

Let’s dive into a  journey of using data science for figuring out the next steps for a profitable business strategy.

Method of Data Acquisition

First, we scraped the data using permutations:

Basically extracts id’s of an app in that order: a, aa, ab, ac...z,za,zb.

Then, used repository google_play_scraper to turn IDs into usable information and wrote it into a csv:

This has yielded 19772 pieces of information including the following types of data:

Data Analysis

1. Strategy Used

To analyze the data, our team first scraped useful data from the Google Play market (https://play.google.com/store). 

For each app on the Play Market, the following information was obtained: Installs, Score, Size, Price, In-AppProductPrice, OffersIAP, Genre.

• Installs - number of times users downloaded the app

• Score - apps’ rating (out of 5 stars)

• Size - how much storage the app takes (helps see how much time was spent on producing the app, as well)

• Price - self-explanatory :) 0 means freemium, other numbers - premium !

• InAppProductPrice - how much money people spend on average for in app additional stuff, or when the free trial ends

• OffersIAP - to see if the app offers inApp purchases

• Genre - category of the app (Communication, Photography, Sports, Education, etc.)

Here is a raw csv data frame that our team used for analyzing. 

Not readable? No worries!

Lucky for you, our team made the data easy to understand by searching for the correlations between given variables through the magical power of the seaborn module. Here is what was found from heatmaps and boxplots...

1. Correlation Heatmap for Google Play Market Apps

Through heatmaps, our team detected things to either ignore or focus on.

Look at installs. It is the main thing. If people don’t upload your apps, the opportunity to monetize declines. The only correlations here that matter are between offersIAP and installs, score and installs. Price and installs (-0.0049) and size and installs (-0.011) have insignificant correlations. OffersinAP and installs have a positive correlation(0.022). Score and installs - positive (0.034). 

Three insights:

1. Price and the size of the app does not significantly affect the installs.

2. People still install apps if they offer in-app purchases, so it is not a stop-factor.

3. If the app has a good score, people install it.

2. Popular Genres: What apps do people install more?

Top 6: Communication, Photography, Social, Video, Tools, Games.

3. Size of the market, Proportion of apps per category.

Here is a distribution that clearly shows what apps take the most proportion of the market. It helps to know which genre/category is the most profitable and easiest to enter. Let’s go through the top 6 installed apps.

1. Communication (416 apps). 4.5e7 installs, but a very low number of apps. It means that there are certain apps (Instagram, WhatsApp, Facebook) that account for the majority of the installs in this category. Also, this is a monopoly category mainly, as all the leading Communication apps are owned by Facebook. Thus, they already have a concrete stand there, and it is hard to take their place.
2. Photography(360 apps). 1.9e7 installs for such a small number of apps. Clearly, apps, such as Adobe Photoshop, PicsArt will be hard to shift, as they have all the basic features that are hard to replace with something new. It is hard to win against the companies who have all the possible filters and stickers for your pictures!
3. Social(114 apps - 1.8e7 installs). Communication and Social genres have similar apps in them, as for example, Facebook is both a place to share and socialize and to communicate via messages. Again, a monopoly that is hard to break…
4. Video (581 apps - 1.7e7 installs). Less of a monopoly category, but still not the best category option. VideoLeap are still hard to replace because they have been around for several years and are staples.
5. Tools(1665 apps - 1.2e7 installs). FileMaster and QR Code Scanners. Average monopoly and possible to enter. However…
6. Games(3475 apps - 0.9e7 installs). This is our top choice because the game genre has the most number of apps and a high mean number of installs, which means no staples and monopoly players, who make it hard for others to get their apps installed. There is no monopoly, as gamers want to try new things to play and compete in different skills. 

4. Monetization: Create a Paid App

As the goal of the business is to make profit, after understanding what genre is the easiest to enter (Game), our team needs to understand what apps bring the most profit too.

Three ways of monetizing apps:

1. Make users pay for the installation

2. Make users pay through in-app purchases

3. Make users pay for a subscription

Here is the total market share of apps(installs*price=430144784.65651715) for paid apps.

Top 3 are: Tools, Music, Games. Tools and Games, which are the markets possible to enter.

Percentage of Apps with In-app purchases in the Datase: 25.84%

Current Total Money Yield of Surveyed Apps: 618,378,154$

Top 3 Grossing Genres (Counting that 0.49$ for every app with in-app purchase):

1. Tools - 228,341,065.9532143.
2. Games - 207,209,149.50953418.
3. Education - 78,643,000.67937888.           

Counting that 0.49$ for every app with in app-purchase:

1. Tools - 228,341,065.9532143.
2. Games - 207,209,149.50953418.
3. Education - 78,643,000.67937888.               

5. Monetization: In-App Purchases

This is the area that we need to discover more about for the following reason:

1. InAppPurchase has a positive correlation with installs. It comes intuitively if something is free, people install it more, and the business has more potential clients. After people use the app, they are more inclined to buy it, because they use it and get the experience that they recognize is worth driving up their bills a bit! 
2. In the future, in-app purchase revenue is set to be the number one source of mobile app revenue, accounting for 48.2% of earnings compared to 14% from ads-based revenue and 37.8% from paid app downloads.
3. The average in-app purchase per user is $0.43 for Android users, according to an AppsFlyer study.

According to the boxplot above, Games is in the TOP Three here too. 

6. Bonus: Optimal in-app product Price

Now, we know that one of the profitable ideas to reach and spur action from a limited but potentially lucrative audience of paying users is to create an app that is free at the start, but gives such a unique experience that users will find worth paying. But.. What is the optimal in-app product Price? The above graph shows that users are willing to pay between 1-8$ for the premium features. 8$ seems to be a threshold, as the install rate declines sharply after this number. So, 1-8$ is a good range, for sure! A great idea is to have various options, such as premium(1-3$), premium+(3-5$), other(5-8$) packages for subscriptions, for example.

Conclusion

After some simple data exploration, our team strongly believes that your company will benefit the most from creating a free app with multiple in App purchase options within the 1-8$ range of the Games genre.

All this being said, remember this:

Choosing what app to create and creating it is not the only success factor. What will truly set a company apart is attention to learning what makes customers unique, and using that intel to market across devices. You can learn what motivates your customers not just to browse or add items to their carts, but actually complete their transactions and keep coming back for more. With the right data exploration - everything is possible.