How to solve this Data Analysis problem ...? need guidance & Suggestions

@deva123 wrote:

I try to solve this using sqldf package from R but after 1st question i couldn’t figure it out how to solve next questions
I dont how to extract and separate data from columns ( events and properties, timestamp) …All suggestions are welcome

Problem

• eventdump - contains event triggered within the app for customers

• Info:

• Customer information contains install-uninstall-reinstall cycles in vertical format
  [so for a customer 1 lifetime cycle is a single install-uninstall event set or a reinstall-uninstall event set]
  [so if a customer install-uninstalls-reinstalls-uninstalls then he has 2 life cycles within the app which need to be treated separately]

• Event logs contain the events triggered within the app by these customers

• Generate the following from the data provided [Analysis questions below are along the lines of CLTV(Customer Lifetime Value) analysis]

1- Customer retention trends from their lifetime cycles [frequency chart or histogram plot]
[retention is defined as the duration of one install-uninstall cycle, so multiple re-installs have to be treated separately]

2- Find out the time of day when the customers are most active [use your own discretion for time of day bucketing] [activity is defined on the basis of events]

3 - Purchase value buckets [find purchase/checkout events from event logs and parse the ‘properties’ column to get total value associated and generate a simple bucketed frequency chart/histogram plot]

4- Behavior of purchasing and non-purchasing customers [something along the lines of their in-app event frequency in a given install-uninstall cycle]

5- Week over Week revenue trends for purchasing customers

6- How are their purchases distributed post install? [the number and value of purchases after installing the app in one retention cycle]

7- Do they perform purchases in the 2nd,3rd etc weeks post install? [if their retention cycle is greater than 1 week]

8- Is there a steady inflow of revenue for customers with high retention? [growth can decline but is it still a positive gradient?]

9- Any other actionable insights that can be drawn from the given data?

Data Sets :—

cycles

events

Posts: 1

Participants: 1

Read full topic