[Sarah Vonderberg]
[DAFT, MAR21, Remote Campus]
The purpose of this data analysis was to find distinct similarities between the users of an astrology app as well as painting a picture of who is the apps main user and if the horoscope sign of the user plays a role in user behavior. From this analysis the founders of the app wanted recommendations on ad-targeting. The astrology app runs on a freemium model that offers different subscriptions in order to get a personalised horoscope as well as the option to purchase several questions to ask a real life astrologer.
Hypotheses: Female users are the main-paying users as astrology seems to be a mostly female interest.
Questions:
- Which platform generates on average the most amount of money?
- Do star-signs affect purchase behavior?
- Who is the main user of the app and how does it differ from the most preferred user?
- Who prefers which subscription type or question purchase in terms of gender, sign and age?
The data was provided to me by the founders of the app. The data contained user data of the astrology app, which is collected upon signup (eg. email, name, registration date, birthday etc.) as well as later upon purchase of different purchase options. It came as one excel file, separated by platforms via tabs in the document.
- Organisation and scope of the project on trello
- github repo setup
- Dividing the datafile and first data checks
- Starting the cleaning process platform by platform
- Data Analysis and visualisations
- Technical report
- Presentation
- update readme
I organised my project with trello into tasks which are either in progress, done or in review. I also assigned deadlines to some tasks in order to be reminded.
I tried to keep my github repo as clean and lean as possible, so it contains:
- a gitignore
- a readme
- a Graph folder with all graphs used in my presentation
- NO Data folder for privacy purposes as I was working with real user data
- NO Code folder for privacy purposes as I was working with real user data