Thanks, Gimwell

Dimensional modelling proved to be useful particularly designing the entire data model from scratch. Have a look at Kimball's "Enterprise bus" and it makes it easy to design anything you might need in a few hours. The difference between Snowflake and Star schemas lies in materialisation strategy. Snowflake tends to be more transactional and that's why we need "fresh" data from multiple fact and dims. STAR on the other hand is more "data obsolete" and looks like a Snowflake's snapshot.