I was recently asked by a client to produce a high performing cube that included measures that should be calculated differently depending on which of their customers was being viewed, and allow them to define a set of other customers to be used in comparisons.
During this session we will look at how to construct a datamart to support this flexibility whilst allowing the OLAP layer to take full advantage of aggregations that would not be possible with cube based calculated measures.
catch a preview at
http://throughadarkglassclearly.blogspot.com/2011/01/dimensional-modelling-for-flexible-cube.html