fernando,
Great question.
In demand planning each value stored in a key figure is described by a CVC characteristic value combination. This is a set of "characteristics" that completely describe the values in the key figures. They would be things like product, customer, plant, sales office, etc. The characteristics act very much like filters do in Excel. They allow you to summarize the data in many different ways. This is a concept called multidimensionality.
A feature of Characteristics in demand planning is that once they are created with the transaction /n/sapapo/mc62 they remain the same unless modified with realignment.
Attributes are descriptive values associated with a characteristic that help to more completely describe that characteristic. As an example: If the characteristic was social security number one of the values of that characteristic would be your social security number. The characteristic social security number could have many different attributes that would more completely describe you as an individual (e.g., age, eye color, address). One of the nice features about attributes is that they can be changed without the use of realignment. So, for example, if you moved and address was an attribute you could change your address without changing any characteristics in realignment.
Technically in DP characteristics are stored as a part of a CVC in the Planning object structure of your planning area, and Attributes are stored as a part of the master data table of the infobject of the characteristic.
In actual use in a demand planning planning area, interactive planning attributes and characteristics appear to the user the same.
Some people say that you have a performance hit using attributes. I have never been able to measure that difference in practice.
Attributes certainly allow you to simplify your planning object structure and are an essential tool in DP.
George
