Inventory the Fields and Groupings to be Supported
The first step in planning a Data Mart is to take as complete an inventory as possible of all the fields and groupings you plan to support in your information delivery project. The more complete the inventory, the better you can create a logical and economical design of models, model extensions and groupings. This does not mean that use of the Mart is an all or nothing or big bang step, only that within the scope of the data to be supported in the Mart, a complete inventory of it makes for a better plan.
This inventory process should include an effort to standardize where possible on a minimal set of definitions of common fields like market value and security price. Standardization makes possible a compact Mart where data elements are widely shared among consumers of the data.
Start Small and Build on Your Successes
While it is best practice to create a complete inventory of fields you will need for a specific information delivery goal, that goal may be a modest one, especially if it is your first Data Mart project. The mart is architected for future growth. So doing a complete analysis of a present known need does not require foreknowledge of all future reporting requirements.
Identify and Create any Missing Meta Data Elements
Every Data Mart field requires a field attribute, and every Data Mart grouping requires some set of field attributes, range rules and dictionaries (regular and performance). The Mart implementation plan should identify these meta data elements quite specifically, and indicate where something must be created.
Use the Data Mart Planning Template
Eagle has a template that is very useful for storing all of the Data Mart and regular meta data elements involved in a Mart, and can serve as the basis of an implementation plan.