When you create an OLAP model, your goal is to create a multidimensional Essbase database. This topic assumes you understand design principles for multidimensional databases. See OLAP and the Oracle Essbase Database Administrator's Guide.
Before designing OLAP models, answer these questions:
What data do users want to see in the Essbase database?
Consider your business environment and the available data. Decide whether to create one or multiple Essbase databases.
What is the level of detail users need in the Essbase database?
Consider your business environment and the performance requirements of the Essbase database. (In general, the less detail stored in the Essbase database, the faster the performance.)
You can use drill-through reports or Hybrid Analysis to offer users at the spreadsheet level alternative views of data and direct access to the source data.
Which dimensions apply to each fact table row?
A dimension is a data category. Typical dimensions are Product, Market, and Time. Consider how users want to view data.
Each dimension has a number of members (characteristic values in data warehouses). For example, the Market dimension can include members representing cities. Each row in the fact table represents a combination of members, one from each dimension. For example, a row in the fact table stores the sales for Product A in New York in February.
Which measures do you want to represent in the fact table?
Measures are numeric quantities that vary over time. Examples of measures are quantity sold, cost of goods sold, and profit. Not all numeric values are measures, but all measures are numeric values.