how to design a "random dimension"???

[alfredjp]

to be honest, i really dont know how to state the subject, thus id begin with the situation at hand

i am to design a cube for a supply chain-based company with the following dimensions = time (granuality of days), product (granuality of sku), organization (sales group/team), and customers (granuality of store level of the supply chain)

for basic reporting purposes, i dont encounter much design problems

the problem is that i want to create a cube wherein it would aggregate not only based on the dimensions but ALSO on the attributes. ie., each product has a different selling price based on which customer (as seen from the operational data).

in which case, the report would be:

time (slicer), organization (slicer), customer (slicer)
----
sku (dimension) sku price (attribute) total sales (cube measure)

* in this scenario, a single SKU product in the dimension would have TWO entries based on its selling price

thanks!!! i would really appreciate some insight into this problem - i may have overlooked something...

ps: im using MS SQL Server 2000's Analysis Services, and im using Excel's PTS for the reports