2 Replies Latest reply: Jan 4, 2012 2:52 PM by jaredhislop RSS

    Calculated Metrics (ie - Conversion, Avg Order Value) and Time Comparison Differences?

      Hello --

       

      I am new to the QlikCommunity.  I have been working with a client to build an online marketing dashboard - and unfortunately I've been using a competitive product that cannot do one key thing.... and I need to know if QlikView can do the following:

       

      1.  Perform calculated metrics (ie - conversion, AOV, etc.) on aggregated data after it has been summed by traffic source type and week.

      2.  Provide week over week % differences (ie - last week was 20% higher than the previous week) after all metrics have been summed by traffic source and by week.

       

      Example below:

       

      They want to see the following in a table

       

      1.  Standard metrics - visits, orders, revenue, etc.

      2.  Calculated metrics - order conversion (orders/visits), AOV (revenue/orders), etc.

      3.  Broken down by traffic source (paid search, email, online banner, etc)

      4.  Results for most recent week, prior week, same week last year, quarter to date

      5.  % difference calculation - w/w, y/y, etc.

       

      ---

       

      The data that builds this table comes from results at a 'tracking code' level on a daily basis.  Each tracking code is then mapped to a traffic source type then the data is a added to other tracking codes that are grouped to the same type then added together to give me a weekly summary.  The data to generate this report comprises approximately 2.2 million rows of data.

       

      Thanks for any insight into the capabilities of QlikView.

       

      Darren

        • Calculated Metrics (ie - Conversion, Avg Order Value) and Time Comparison Differences?
          Karl Pover

          I don't see anything you've mentioned that you couldn't do in QlikView.  You mention that you are doing data aggregations and I think it goes without saying that you can't ask QlikView what is the average for Mondays if the data is pre-aggregated by week.  No product would give you that information.

           

          Also, 2.2 million rows should run well on your laptop unless you have more than, say, 200 columns.

           

          Karl

            • Calculated Metrics (ie - Conversion, Avg Order Value) and Time Comparison Differences?

              This question is in conjunction with Darren's earlier question.

              I am looking for some instructions on how to get a view within Qlikview that looks like the following:

               

              Desired View.JPG

              I have been using a competitor's visualization tool and am wondering if Qlikview can do what I need in a more fashionable way.

              My data is broken down by day over the past 3 years and can be in either format:

              Date, Revenue, Orders, Visits ...

              Jan 1, 54434,    56,        783432

               

              OR

               

              Date,  Metric,  Value

              Jan 1, Revenue, 54434

              Jan 1, Orders,   56

              Jan 1, Visits,    783432

               

              We operate on a Monday to Sunday basis so if today is Jan 4, 2012 "Last week" would mean December 26, 2011 through January 1, 2012.

              Last week Previous Year would be same week as last week one year ago.

              Week Previous would be December 19, 2011 through December 25, 2011.

              % Diff mean "(last week - last week prev year) / last week"

               

              I am somewhat new to Qlikview and am unfamiliar with the scripting process or the funcions within the expression box so any and all help is much appreciated. Detail will help significantly as well.

               

              Please let me know what you can come up with because I am kind of at a loss right now.

              I am using Personal Edition so I can't open any other documents either, so screenshot or something of the like would be best.

              Thank you tons in advance

              Jared