Forum Discussion

raymond's avatar
raymond
Post Patron
8 years ago

Google Analytics Connector: User Count doesnt match Google Analytics

Hello together,

I have just encounterd something strange that I cannot explain. The user data pulled from PowerBI through the Google Analytics Connector doesn't match the user count provided by the google analytics dashboard. That is strange because count of sessions are exatcly the same. 

 

Within a specific date range google analytics displays 56 K users and with the exact same settings powerbi gets 73 K users. 

 

Anyone here experienced something similar? Is there a known solution or explanaition?

Need Help

4 Replies

  • CitySparkInc's avatar
    CitySparkInc
    New Member
    7 years ago

    It's a quirk with how the User data from Google Analytics is binned according to the Date. Rather than selecting 'Date' in the Google Analytics Connector, select only the time period that you're interested in and it will display it correctly. For example, if you'd like to see the number of users for a particular month, then select 'Month of Year' (not to be confused w/ 'Month of the year') in the Connector. Doing it this way puts serious constraints on the visual because you're unable to use Date hierarchy, but at the very least the numbers will be correct.

    • raymond's avatar
      raymond
      Post Patron
      7 years ago

      Hi CitySparkInc,

      thanks for getting back to me :)

       

      I found there is also another quirk while working with the google analytics api. There are restrictions and due to the high volume of data GA works with a sampling as well. Hence the numbers will almost never be on point. They will rather show a very well calculated trend. 

       

      For my clients I keep extracting the data from GA in various forms. One report only comprises date, another date and device, yet another date and user type. Most of the time there are slight variations in user, sessions and other metrics as well. However the variations are slim if your reports consist of a few dimensions only. 

       

      I found the more dimensions you are going to use, the more data GA needs to provide. Hence the api limits kicks in and the deviations become larger.

    • raymond's avatar
      raymond
      Post Patron
      8 years ago

      Interesting. Would think it might also have to do something with the sampling of the google data?