final report analysis on google play stores potential apps and categories

VIETNAM GENERAL CONFEDERATION OF LABOUR TON DUC THANG UNIVERSITY FACULTY OF BUSINESS ADMINISTRATION

DAI HOC TON BUC THANG

TON DUC THANG UNIVERSITY

CONTRIBUTION OF TEAM MEMBERS

CONTENTS

CONTRIBUTION OF TEAM MEMBERS ccccccecceeteteecteeeecnectenesnseeretesneetieeenesnneeenes il IISIE.S90)000I6)0).405cciiiiiiaiaiaiẢi.5a-:ä-:.- Ắ.ẮẮÝ IV [.\0).9)i9À)4M09.6)0)/10ìk:aiiiaiaiaiddđdiiáiááäá Vv

2 DATA PREPARATHION nent ein tenes cie teen eneeteeeneeteeneeseestesneneestesiesieesiesiesteesiees 1 2.1 CLEAN AND CONVERT THE VARIABLES 0 enenteeeecieeenenieeeer etnies 1 2.2 FILTER THE DATA SET IN ORDER TO PERFORM EXPLORATION AND ANALYSIS G11 T41 t1 111111 11 11111 1111111 1111111116111 1111116111111 6 111 11111 11111 111 1191111 1111191111111 11 151 112111 119112112 3 E'4S10/.1972.9000)0.)0090.0)/.16 51 3 199) 0900639) 4 8

1H

IV

ACKNOWLEDGEMENT

We are thankful to all of the lecturers at the Faculty of Business Administration at Ton

Duc Thang University in general for making it possible for us to complete the final report and this course

We are particularly grateful for Mr.Doan Xuan Huy Minh, who directly taught and guided us so that we could complete this course in the best possible way Although it has not been a long time, we have received passionate advice and support from the lecturer Thank you for the lessons, which not only benefit our specialist expertise but also have a wide range of implications for our future development

There are many imperfections in the process of creating this report because we do not have much real-life experience and have little expertise We hope that after finishing the report, we will be able to gain more knowledge for ourselves and receive comments and opinions from lecturers to improve in the future

Once again, we appreciate all of the support and are hopeful that all of you have wellness and success.

1, INTRODUCTION

Types of Data in statistics can be classified in:

Categorical/Qualitative Data: is non-numerical data and involves categorical variables

that describe the features such as gender, home town, etc

Numerical/Quantitive Data: is numerical data that represents the numerical value such

as size, price, height

In Google Play Store data, we identify 9 qualitative variables including App, Category,

Type, Installs, Content rating, Genres, Current ver, Android ver, Last updated, and 3 quantitative variables: Ratings, Size, Price, Reviews

Objective: This final project aims to evaluate and provide how paid and free categories and potential apps affect customer usage experience by analyzing the relationship between Reviews, Content Rating, Rating points, and the number of installs to see what categories or which apps will get a boost in search results visibility, bring more attention to these potential apps and categories, which help to promote Google’s revenue and its development These

variables are collected from Google’s users, so they will be indicators to clearly evaluate

customer experience as well as the performance level of apps

consistent in Excel

- Pull together the data table in Excel:

Import data into Excel to create a table according to the following steps:

On the toolbar select Data => get data => from text/CSV => Import => Transform data => The Power Query window will open => Changed Type: This is the step where the Power

Query Editor automatically recognizes the format type for each information field (column) when

inputting data We will see the Price column automatically change the format, which losing units => Delete the changed type to keep the Price column format => close and load

- Clean and Transform the data

Using mathematical transformations is crucial for data analysis and mining programs to better understand the raw data using the following operations: Normalize, Value mapping, Discretization, Aggregation

Because the data set of Google Play Store is in text format, quantitative variables need to be converted to numbers for analysis using the Value Mapping method Specifically, my group will use the “number value formula” to translate text values into numbers in Excel:

¢ Ratings: Create an additional Rating Points column in the form of numbers by applying the formula =NUMBERVALUE([@rating])

¢ Review: Create an additional Reviews Number column as a number by applying

© Adda column named Size in Mb

" Delete the letter M: =numbervalue(left([@size],len([@size])-1)) " Delete the letter K and change to Mb unit: =left([@size],len([@size])-

for example, 100000+, so if you delete the "+" character, then data means the

lowest number of downloads for each app We use the formula:

=NUMBERVALUE(left([@install],len([@install])-1)) - Categorize the data

After cleaning the data set, our team converted the Installs variable into a quantitative variable named Installs Min to serve analysis Therefore we will reclassify the data as follows”

¢* Qualitative data includes: App, Category, Type, Content rating, Genres, Current

ver, Android ver, Last updated

* Quantitive data includes: Rating Points, Size in Mb, Price App, Installs Min,

Review Number - Remove unnecessary data

Remove variables that will not contribute to any analysis

On the toolbar, select Data => Remove duplicate => select the App column to remove duplicate app names

2.2 FILTER THE DATA SET IN ORDER TO PERFORM EXPLORATION AND ANALYSIS

We chose the Type variable to filter the data to see if there is any difference or impact on

the users’s demand for using apps and their experience between free and paid apps At the same time, the Type variable has a relationship with other factors such as reviews, ratings, and Installs to classify potential apps or categories that can bring high revenue to Google

In addition, Category is used to filter only for the App Reviews chart to identify which categories the top apps with the most user reviews belong to

3 VISUALIZATION AND ANALYSIS

Dataset: The Google Play Store team has launched a new feature that will boost the visibility of select potential apps It is represented by the priority level in the software suggestions section, which includes phrases such as "for you", "top charts", "new and updated games", etc Therefore, it will help enhance the popularity of search results and attract users’ attention to new potential applications

Analysis to be performed: The major issue is identifying excellent apps for Google to promote Customer reviews will demonstrate the application's quality (a high ranking indicates that the application has been favored by users) The remaining issue will be anticipating which apps will receive high ratings

Review Card Install Card Avg Rating Card

) Google Play Store

Google Play Dashboard 2.086.948.465 74.333.679.130 117

Figure I Statistics on number of downloads, number of reviews and average rating Google Play Store is one of the largest platforms in the world, it allows users to download any applications such as game apps, entertainment apps, social networking apps, etc It can be clearly seen from the dashboard that there are more than 2 billion reviews for all apps on the Google Play Store including both paid and free apps In addition, the number of application installations on this platform has a huge number of up to 75 billion downloads Users often tend to look at the average rating when deciding to install and use any application Through statistics on Google Play Store, 4.17 out of 5 point is the average rating for all applications available on this platform and this is a fairly reputable rating.

Avg Rating Type

4,17

Free

4,27 Paid

Figure 2 Average Rating Type

In this pie chart, our team used Average Rating Points data to determine the overall quality and user satisfaction when using apps of both free and paid types App ratings are scores assigned to a mobile app, apps with high ratings are much more likely to be downloaded by potential users browsing the Google Play Store Rating points are calculated by the formula ratings points is equal to the total number of points (points) divided by the total number of reviewers (users)

The Average Rating Point of the apps in the paid type is 4.27, while the remaining apps in the free type have a lower Average Rating Point of 4.17 This shows that paid apps have better overall quality than free apps, the reason may come from paid app developers using revenue from app sales to continuously develop and update new features, refine old functions to improve the quality of user experience.

Review by Content Rating Adults only 18+

Figure 3 Review by Content Rating

According to the chart, the highest number of user satisfaction ratings belongs to apps without age restrictions, about 1.295.211K Meanwhile, age-restricted apps such as Adults only

18+ have the lowest number of reviews Normally, at this age, users tend to read comics and

enjoy sports programs, most of these applications are free to use In 3rd place, there are more than 500,000K reviews of apps including free apps and paid apps belonging to Teen age apps In particular, the number of reviews for free apps is higher than that of apps that download and pay to use and most of the needs for using this type of application at this age are communication and social Finally, there are more than 200,000K reviews for apps with an age limit of over 10+ and less than half of the figures belong to apps for mature 17+ To sum up, users tend to use free apps more than paid apps, because free apps meet the basic needs of users Nowadays, there are some apps that have quite outstanding features and no need to pay any additional fees

According to statistics on the number of reviews for free apps, it tends to be similar to the chart of total reviews (free and paid apps), apps without age restrictions will have the highest reviews However, it has fluctuations in the chart of the number of people rating the paid apps, the most reviews belong to everyone 10+ and the adults only 18+ is completely missing in this chart, because according to statistics, this age group only uses apps that do not require payment such as Manga Master - Best manga & comic reader; DraftKings - Daily Fantasy Sports and Manga Books.

Top 10 Installed Categories

Figure 4 Top 10 Installed Categories

In this chart, our team will analyze the top 10 installed categories of users today Accounting for nearly 14,000M people tend to download game applications for entertainment purposes (including free and paid game applications to download) Today's users tend to download many app game genres to enjoy, not necessarily playing just one game app, which is the main reason the App game genre has the most installs Ranked 2nd belongs to communication category apps, about 11.038M, users tend to use free applications for this app because they provide basic communication needs through social networks

The tool application category ranked 3rd in the rankings (all apps), however there was a change in position in this category for paid apps and dropped to 4th place in the top 10 Productivity and Social have a similar trend as Tools, they fall in the rankings of paid apps because people do not need to pay for these types of apps In 4th place belongs to Photography apps, 4,639M installations in this app category including free and paid apps In particular, the need to pay for installing special photography apps increased and climbed to 3rd place in the ranking of paid apps In addition, there is a special payment to install family genre apps, from the top 7 in the general rankings, it has increased dramatically to rank 2nd in the rankings of paid apps Finally, the bottom 3 in the rankings are Video players, Travel_And Local and New_And_ Magazine, most users use these apps for free and do not pay at all to download apps of this category.

Top 5 Rated Categories

E

5 a

Ca =

f œ

1 9

EVENTS EDUCATION RT_AND_DESIGN BOOKS_AND_REFE PERSONALIZATION

Figure 5 Top 5 Rated Categories

According to the chart, when we apply in both filter free and paid, the event received the greatest score, accounting for 4.44 This demonstrates that users rate event-related applications over other categories of Google Play Store apps that are available in both free and paid versions in terms of overall quality The reason why people adore apps that relate to events because they can use them to set reminders for important occasions or to purchase festival tickets online

However, apps in the event category will not be rated when just paid filters are applied Although the event is among the top 5 most highly rated, none of the apps in this category are paid, thus if the paid filter is used, only the remaining 4 categories can be rated Instead, educational apps get the most favorable review, reaching 4.75 out of 5 points Education applications are primarily aimed at users who are concerned with learning new languages and abilities, and they are rated highest overall in the paid filter, which illustrates the desire for learning and development.

App Reviews

App =

Fecebook ES 73,16M WhetsApp Messenger es 69,121

Messenger - Text and Vid rs <6; 641 Clash of Clans iS 44 9m Clean Master- Space Clean EE 42,921

Subway Surfers i 27,72 YouTube MXMMNWa2:6sw Security Master-Antiviru i 24,90

Figure 6 App Reviews

According to the chart, in five different categories, 11 different applications including Facebook, WhatsApp Messenger, Instagram and Messenger - Free Video and Text Chat, Clash

of Clans, Clean Master - Space Clean and Antivirus, Subway Surfers, Youtube, Security Master, Clash Royale and Candy Crush Saga that have achieved reviews from 20 million upwards There

are 11 total apps which 4 are games, 2 are communication, and the remaining 7 are tools, social, and video players The number of users rating the Facebook app is the highest, totaling 78.16 million reviews, while WhatsApp Messenger takes second place in the number of reviews, accounting for 69.12 million Regarding to data, there are around 2 million fewer reviews for Instagram compared with WhatsApp Moreover, the number of users who rated Messenger - Free Video and Text Chat was about 56.64 million These apps acquire the most downloads and are considered the most popular by prospective users, which is why they get so many ratings and are related to social interaction and connection

Following that, the Clash and Clans app and the Clean Master - Space Clean and Antivirus app received around 44 million reviews, and the remaining 5 apps received less than 30 million reviews These apps have fewer reviews and fewer downloads than social media apps because they are about protecting personal devices and entertainment Hence, they do not receive the attention of many age groups compared to applications related to social interaction 4, CONCLUSION

It can be seen that there is a large difference between the number of downloads of free applications and paid applications, thus also leading to a difference between the number of reviews of these two types of applications However, there is a contrast in the rating variable when paid applications have higher ratings, perhaps because paid applications use the money collected from users to upgrade and maintain their applications that application.