ceu_product-analytics

The project involves answering questions around Product analytics topics such as Acquisition, Activity and Retention. Done as a requisite of the Mastering Product Analytics course at Central European University in Hungary.

Article can be found on Medium.com

Datasets

Based on a real SaaS product subsampled and simplified for the task

Two datasets

• registrations.csv: unique users with basic demographics

• activity.csv: which users have been active in which month

Registration data

• Columns

  • id: unique identifier of a registered user(e.g id_5)

  • registration_month: number of the month from 1 to 21

    • Month1 and Month13 are both January in consecutive years

    • No need to deal with date type

  • region: America, EMEA, or ROW (every other country)

  • operating_system: Windows, Mac, Linux, Unknown

• About 40K records

• Each record is a unique registered user

Activity data

• Columns

  • id: unique identifier of a registered user(e.g id_5)

  • activity_month: number of the month from 1 to 21

    • Month1 and Month13 are both January in consecutive years

    • No need to deal with date type

  • Additional Columns:

    • Registration data joined to the activity records as convenience

• About 79K activity events

  • One record represents an active user in a specific month

  • It means a user is active in two months on average (long tail distribution)