0 Replies Latest reply: Aug 1, 2017 9:32 AM by Luca Bambin RSS

    Qliksense - Identifying Repeated Customers between Week N & Week (N-1)

    Luca Bambin

      Hi members of the Qlik Community!

       

      I have just joined a tech startup as a business analyst, and while I have no problems performing routine tasks right within Qliksense, I am starting to run into more complex problems, some of which I have little to no clue to go about solving it

       

       

      For instance, I am trying to generate a list of customers who fulfills the below conditions:

      [A] Orders at least once in 2016/10 and at least once in 2016/17 AND

      [B] Delivery fails to reach them due to 'Error' at least once in both times

       

      ...and I have the following key fields available:

      [1] customer_id (e.g. alphanumeric string)

      [2] package_id (e.g. alphanumeric string)

      [3] delivery_failed_reason (e.g. 'Error', 'Rejected', 'Cancelled', 'N/A')

      [4] delivered_date (e.g. 2016/10, 2016/11, 2016/12, ...)

      [5] country (e.g. RU, DE, AU, SK)




      In my feeble attempt I have tried the following in a straight table:

       

      DIM:

      if(delivered_date = '2016/10' or delivered_date = '2016/11', customer_id)


      MEASURE:

      count (distinct if(wildmatch('*Error*, delivery_failed_reason), package_id))


       

      I then apply a count function to the measure column, which would ideally provide the count of customers fitting conditions [A] & [B]. However numbers dont look right so far..

       

      Any advice would be much appreciated!

       

       

      Thank you!