Our environment is VA 6.3 and does not support a star schema.  Additionally, our IT group will not allow data builder to be used in the VA environment as they don't want any users creating new (calculated data) such that the volume grows too large and impacts other users.   Question: We have ~10mm customers worth of data.  There are 1000s of products.  Each customer makes multiple purchases with each purchase containing up to ~35 products of varying quantities.  Each customer may have purchased hundreds of products throughout the year and purchased the same product multiple times.   One of the key querry types that will be run on the data is to understand how many unique customers have purchased any one of ~60 select products.  We don't want to know how many products were purchased, just how many unique customers purchased any one of ~60 products.  This same querry will be run over and over with different mixes of up to 60 products to identify the unique customer count that purchased any one of those 60 products.   Once we run a querry to identify the unique count of customers, we will then want to apply all sorts of filters(age, gender, income, geo, etc. etc.....) to this customer segment.   Is there a recommended data layout/format for the data to load into VA?  We only get one chance at loading the data due to internal costs by another group of extracting/appending the data for load into VA.   Thanks for any guidance/recommendations! ... View more