1 of 1 people found this helpful
One option might be to run your SQL data through some data prep beforehand using Tableau Prep, Alteryx, or another tool. You might also look at joining your SQL data to Excel to create the structure needed. Another option might be to take a look at Ian Baldwin's brilliant post on creating a sankey without any data prep: https://www.theinformationlab.co.uk/2018/03/09/build-sankey-diagram-tableau-without-data-prep-beforehand/
I love Sankey diagrams, but remember they are graphic intensive. If you have 100k rows per user, and each line requires a sigmoid of 25 instances (to make for a smooth curve)-- that's a LOT of graphic detail.
Perhaps you should look into a different type of visualization?
Depending on how you wanted to use the Sankey, you might be able to replicate it using graphic elements:
This could be a lot less of a memory hog than using a Sankey.
You could also try Radar Chart/Spider Chart Color would be mapped to ID and size to count.
Your plot for Product ID 2 might look something like this:
(these are for illustrative purposes)
You can find details here:
These might be more work than it's worth, though, since many of your ID's have just one line.
Nevertheless, I hope this may give you some different ideas on how to conceptualize your data