This is a high level data design question but knowing the expertise out there I am hoping for some guidance as to how best to design this. Our company is trying to track turnover - sounds simple however the turnover calculation is terminations/headcount. Sounds simple enough except headcount is a rolling 12 month average. I can calculate this in SQL and create a table of those numbers however the dashboard has a large number of filters and dimensions. Given that I can not store every possible combination in my data warehouse so think I need to either perform the calculation in Tableau after filtering has been applied or use a cube. The analyst wants to be able to filter on management/non management, part-time/full time they need to drill to departments or particular managers etc. Terminations are easy as an associate is fully defined at the time of termination. For active associates for the first 6 months of the rolling 12 months you may be non management and the last 6 you are management etc etc.
Anyone handled something similar?
Anyone using Tableau against a SQL Server cube - any gotchas I should know about.