The Data Warehouse Data Model: The Floor Plan for Your BI Environment

  Business Intelligence   Strategy  

I've used the analogy between a home and a data warehouse before, and in what I've been seeing lately, I know that the analogy continues to hold true. Like a house, a data warehouse has many diverse components that need to be planned, architected and built properly in order for it to function optimally. In this column, I want to stress the importance of a robust data warehouse model as a part of the overall data warehouse architecture.

The data warehouse data model is like the floor plan of your house: Each entity is like a room and each attribute is like a characteristic of that room. If the analogy holds true, getting the model right seems pretty significant, doesn't it? Imagine drawing up a floor plan for your house and forgetting the sinks in the kitchen or the tub in the bathroom. Most people would be pretty ticked off. Data warehouse users are much the same. If they can't find the data that they need or the data they need exists in multiple places, they will look upon the system unfavorably. Getting the data model right and designing it to scale is key in building a successful data warehouse.

The key driver in building a sound data model is a solid understanding of the business requirements. All the potential users of the data warehouse need to participate in the business requirements gathering process. The system designers need to have a detailed understanding of …

Please log in or sign up below to read the rest of the article.