Forum Discussion
Google Analytics # users vs # users PowerBI
Here is an explaination AND a potential solution
The solution outlined makes everything a little more complicated (you'll have to create a DAILY fact, a WEEKLY fact, a MONTHLY fact etc.) *and* does not still resolve all possible use cases (if you want to report from Wednesday of week 1 to Tuesday of week 2, you still won't have the correct data)
That is perfectly normal, you can't avoid it. Users are Unique Users.
The reason behind it is due to the way GA counts "unique users".
Imagine John Doe browsing your website at 23:30 of Day 1 and at 00:30 of Day 2.
When you ask GA to count your users from day 1 to day 2, it will go into its database and search for UNIQUE users across those days. Looks like John Doe is the same person on day 1 and day 2, so it will report 1 Unique user.
However, PowerBI does not download the entire dataset of GA (it would be massive). What it does is querying GA for the data for every single dimension. So if you ask PBI to connect to GA and getting users From day 1 to Day 2 it will make two queries
- give me the users of day 1. GA answers "here you are, we have John Doe as unique on day 1"
- now give me the users of day 2. GA answers "Here you are, we have John Doe as unique on day 2"
And then PBI will sum users on day 1 + day 2 and will report 2 unique users, which is "wrong". Technically PBI will show the sum of unique users on the selected days, while GA will count exact unique users on day 1 and day 2.
This limitation CANNOT be avoided (*) and in fact means that any measure that is not summable on a dimension to be reported wrongly (for example Sessions and Users are wrong, while Page Views are correct).
This is the reason why, for our final customers, opted to use Google Data Studio instead of PowerBI to show GA data.
*= There is one way to avoid this, and it's to use a Direct Query connector to GA. Direct Query will re-query GA each time you make a new filter and GA will report correct number. But it has a cost AND when you switch to DQ everything gets a little more complicated.
(you can see this behaviour if you choose just one day in PowerBI, data will match GA)
Here is an explaination AND a potential solution
The solution outlined makes everything a little more complicated (you'll have to create a DAILY fact, a WEEKLY fact, a MONTHLY fact etc.) *and* does not still resolve all possible use cases (if you want to report from Wednesday of week 1 to Tuesday of week 2, you still won't have the correct data)or you can spend 499/599$ /year
https://www.cdata.com/drivers/ganalytics/powerbi/
(note: cdata is quite well known, but i NEVER used this connector)
Hi Anonymous , how I connect to GA using Direct Query connector? Tks
