Forum Discussion

racx321's avatar
racx321
New Member
9 years ago

Calculating Retention with Limited Data

 Racking my mind on how to calculate retention. I am pretty sure I have enough data for this.

 

I have a table with 3 columns.

 

1st column is the timestamp of the event.

2nd column is the event, either a user_create or user_login.

3rd column is the user unique ID. 

 

As a sample, it could easily look like:

Date                                  Event                      ID
08/01/2017 12:12:12         user_create            10910AB
08/02/2017 15:10:19         user_login              10910AB

etc...

 

I want to be able to calculate aggregated user retention by any time frame (days/weeks/months). So I can see retention rate of our users within X time frame after being created. 

 

I'm new to powerbi and data engineering in general, so any help is appreciated.

 

Thanks!

 

 

Need Help

3 Replies

  • Anonymous's avatar
    Anonymous
    Not applicable
    9 years ago

    What is your general formula for calculating retention?

  • Anonymous's avatar
    Anonymous
    Not applicable
    9 years ago

    Hi racx321

     

    A simple measure like CountoftheUser = Countrows(yourfacttablename) will do the job.

     

    I am assuming you have a calendar table and this is linked to your fact table.

     

    Create a matrix table and use ID on the rows, year month on the columns and COuntoftheuser measure as values.

     

    If this works for you please accept this as a solution and also give KUDOS.

     

    Cheers

     

    CheenuSing

  • v-jiascu-msft's avatar
    v-jiascu-msft
    Microsoft Employee
    9 years ago

    Hi racx321,

     

    I have some questions.

    1. How to determine the retention? One user_create and one user_login?

    2. Are there any duplicates? Such as, one user has many logins.

    Date                        Event                      ID
08/01/2017 12:12:12         user_create             10910AB
08/02/2017 15:10:19         user_login              10910AB
08/03/2017 16:10:19         user_login              10910AB

    Best Regards!

    Dale