I am newly using PowerBI to connect to my Adobe Analytics website data and while pulling some basic data (unique visitors, visits, page views with one dimension called site section) from AA I noticed big differences in the unique visitor numbers that I get in PowerBI versus what I see within my AA workspace reporting. After investigating, I believe it may have to do with aggregation of the unique visitor numbers because if I remove the Date Granularity fields from the data being loaded to PowerBI, the total values I am getting are correct, but if I include the Date Granularity fields the total values are incorrect. Also interestingly, if I use a line chart to show trend of unique visitors by site section, if I look at it from a monthly level it is also wrong, but if I drill down to daily values then it is correct / matches AA workspace.
The problem is that I need the date fields for my reporting to be able to show the data by year, month, etc., so is there any way to have accurate unique visitor numbers with the date granularity?
Hi @TheScottR24 ,
Based on my understanding, the unique visitor number can vary in Analysis Workspace & Power BI due to the following reasons.
1. Due to segments, if you have currently applied any segment then I would request you to try to compare the number by removing the segments.
2. Analysis Workspace & Power BI both use different Analytics API version. Analysis Workspace is based on Analytics API 2.0 whereas the Power BI integration uses the Analytics API 1.4. If you need to compare the Power BI numbers with Analytics then my suggestion will be to try comparing these numbers in either Reports section of Analytics user interface or try creating a similar request in Adobe Report Builder to see if the number matches because Reports section of Analytics & Report Builder are both based on Analytics API 1.4.
Hope this helps. In case even after trying the above suggestion, you notice any discrepancies then I will suggest you to log this issue with the Client Care team. So, that we can investigate further on the issue if needed.
Thank you for the reply.
I think the main problem as I mentioned in my post is the aggregation of the unique visitor numbers as it is pulled into PowerBI.
We used to have an export set up that pulled the data from Data Warehouse and was loaded onto a SQL database. That data contained a field called Unique Visitor Identifier which could be used in tools like Tableau or PowerBI to do Count Distinct functions, which provided accurate unique visitor numbers for whatever date granularity was desired.
Whereas with the Adobe Analytics connector, there is no such Unique Visitor Identifier field available. There is only the default unique visitors field/value that can be used and that seems to only provide a sum for the date granularity selected, and when aggregating to higher levels it just adds up those values, it doesn't actually account for the number of true unique visitors because I don't think it can.
For example: Say on Aug 1, 2020 we had 10 unique visitors at 1 AM, 10 at 2 AM, 10 at 3 AM, and so on, the way the data is aggregating would actually add all the hourly totals up and give a total of 240 unique visitors for Aug 1, 2020 even though in reality many of those unique visitors may actually be the same unique visitor across different hours of the day. The same applies across different days and then months.