Counting Raw Unique Visitor Data Dynamically

[czmudzin]

Has anyone found a way to dynamically aggregate & count unique visitors using raw data (data feeds) + data viz tool, the way that Workspace does? One solution is importing every distinct ID, but that's too heavy of a dataset for me to use.

The obvious issue is that aggregating at different date grains or channels (or combination of both) overcounts because the same IDs are present multiple times, whereas Workspace accounts for duplicates. I haven't found a solution other than creating calculations in SQL for every date grain or channel I need. Just wondering if anyone has ever found a more dynamic solution downstream that doesn't involve importing every ID in the dataset?

[Jennifer_Dungan]

I know that our data team gets very close numbers (like usually within 10 of what I see in Workspace).

I don't know the SQL they use, but I do know they use the Visid High and Visid Low together, and get the unique count of those identifiers.

[czmudzin]

For anybody that's interested in dynamically counting UVs by any dimension in Tableau without importing millions/billions of rows of data (and getting the data to match Workspace), setting the data source connection to "Live" instead of creating an extract works for this use case. You do have to wait a few seconds for the data to process, but it's a viable solution.