A measure is a member of a special type of dimension called a measure dimension. A measure dimension and its measures belong to only one cube. An example of a measure dimension is the Demand dimension, which belongs to the Demand cube and contains measures such as Projected Sales and Booked Units. A measure dimension is not organized into hierarchies and contains only one namespace with the same name as the cube to which it belongs.
A measure, for example, Projected Sales, has numerical values located at intersections of cube dimensions, with coordinates given by one member from each dimension. All measures have a data type of double.
Loaded—Measure values are numerical data loaded from external systems or files into the Integrated Operational Planning database; for example, CSV files. Examples of loaded measures are Booked Sales and Shipped Units. When a loaded value does not exist, a formula can be used to derive it. The formula never overrides a loaded value.
For example, consider a formula in which the Inventory value for the current week is given by the value for the previous week. If loaded values for Inventory include a value for the current week, the loaded value is used. If a loaded value does not exist for the current week, the value for the previous week is used.
Input—Measure values are numerical data entered directly into an Integrated Operational Planning planning worksheet. For example, one source of input measure values can be data in existing Excel spreadsheets used for forecasting.
Derived—Measure values are calculated from formulas using Integrated Operational Planning built-in functions. Derived values can be based on other measures, including loaded and input measures. Examples of derived measures: Actual Average Selling Price and Excess Material Cost.
Cross-cube—Measure values in one cube are calculated from formulas that depend on measure values in other cubes. Examples of cross-cube measures: Forecast Projected Sales and Forecast Booked Sales, whose values depend on measures in the Demand and Forecast cubes.
A key metric corresponds to a measure value of interest; for example, a value of Projected Sales, that is located at a specified intersection of cube dimensions.
For example, the measure Projected Sales belongs to the Demand cube, with dimensions Demand, Product, and Fiscal. A key metric can be defined as the value of Projected Sales (in the Demand dimension) for product KFB (in the Product dimension) during the time period Q2 '05 (in the Fiscal dimension). A different key metric can be defined as the value of Projected Sales for All Products during the time period Fiscal 2005.
A scenario includes default key metrics whose values are calculated and displayed in the Key Metrics panel on the Perform Analysis screen or the Impact window in Excel during an analysis. Key metrics are chosen based on the measure values that are most important in an analysis type.