Forum Discussion
Users from Google Analytics off
Hi,
I don't know what I am doing wrong...this is driving crazy.
Basically I am getting users, sessions, and PVs from Google Analytics. Sessions and PVs are spot on, yet Users are off. I created a new query to just get Users and the date to see if that would make a difference and it doesn't.
I even checked https://ga-dev-tools.appspot.com/query-explorer/ to see if it is sampled and it isn't.
Anybody know what I am doing wrong?
6 Replies
Hi procyon82,
Regarding the Users, do you mean the user container under the get data Window?
I encountered error when loading user data with “XX day active users” selected, leave them out and only select the other items would load the data OK from my side
In addition, please also take a look at the article below:
Third-party service: Google Analytics connector for Power BI Desktop
Check the Google Analyze changelog, or contact them to see if any further helpful information could be offered.
Regards
Well, turns out GA is wrong. The following adds up to 115,504 (the number from Power BI and the GA exports), yet visually in GA it shows 101,661.
2,319 2,636 2,874 3,630 5,440 4,427 4,551 3,971 2,984 3,540 4,428 4,105 4,101 3,892 3,475 2,924 3,311 4,375 4,266 3,984 3,925 3,658 2,794 3,181 4,108 4,081 4,035 4,253 3,861 2,963 3,412 Hi Procyon82,
GA is most likely not wrong if it shows you a table like the one above which sums up to 115,504, but with a total of only 101,661. The reason is that a user can figure in more that one cell, whereas the total is (and should be) the result of a distinct count of users. Imagine that you visit a hotel and that you are the first and only person to visit this hotel. You visit the lobby, the restroom and the bar. If you were to make a report about visitors in the hotel, you would see 1 person in the lobby, 1 person in the restroom and 1 person in the bar. If you sum across rooms, you end up with 3 persons, but in fact this is wrong, because all three rooms were visited by only 1 person, hence 1 in the total (distinct count).
The problem with pulling GA data into Power BI is that you create a data model, which allows you to aggregate your GA data on the fly. While this is a great feature, you risk doing all sorts of wrong things. In general, you can aggregate pageviews in any way you want without doing anything wrong, but sessions and users cannot.
I hope this helps :-)
