A star schema is designed to efficiently answer metric business questions, such as those questions that return amounts, quantities, and counts. For example, “What are our total sales by customer and product over the last six months.” Stars are generally small in scope with a handful of dimensions surrounding a single fact table. But a colleague one day stops by your desk with a printout of a star schema that makes you jump right out of your seat. It is a Sales super star! Over 1000 data elements from the Sales area are on this model. You learn that this star was produced by placing every metric data element from Sales in the fact table and every non-metric data element in a dimension. The fact table contains over 100 different metrics and there are over 30 dimensions!
You are even more shocked to learn that this star schema is in production! It is experiencing some performance and space issues, and your colleague would like advice on how best to restructure this model. What is your response?