The goal of a data warehouse/business intelligence (DW/BI) solution is to publish the “right” data and make it easily accessible to decision-makers. Successful DW/BI implementations rest on the foundation of a dimensional model to deliver both ease-of-use and query performance.
As described in our book, The Data Warehouse Toolkit, Second Edition (Wiley 2002), effective dimensional models reflect the natural performance measurements and descriptive attributes of a business. They survive reorganizations, acquisitions, and other business changes, if designed properly. Most importantly, business partners will find a well-designed dimensional model intuitive because it mirrors the way they think about the business, not the way a data modeler has normalized the organization’s data. Dimensional models are critical to enabling the business to leverage the organization’s valuable information assets.
The Kimball Group often engages in a Dimensional Model Design project immediately following a Business Requirements effort. In other cases, the client may have independently completed the requirements analysis process. Kimball Group consultants work closely with key business liaisons and IT professionals to facilitate the design of flexible dimensional models through a series of joint design workshops. The process results in a design specification describing the fact and dimension tables and associated columns, along with sample values, slowly changing dimension policies, and required business rules transformations, where appropriate. The resulting logical data model is the starting point for the physical design, source to target mapping, and ETL efforts that will follow. The model will be independent of database platform and/or tool idiosyncrasies.
Contact Bob Becker to discuss how the Kimball Group can help with your Dimensional Modeling initiatives.