Example: Optimizing Performance 
Suppose you have the following characteristics:
In this example, the product master group is dependent on the product group, and the product group is dependent on the product. Furthermore, the customer group is dependent on the customer. No dependence exists between any other characteristics. Each customer purchases each product. This results in a million records (1000 products * 1000 customers) in the transaction database.
In your report, you wish to navigate from the list of product master groups over the characteristics product group, product and customer group to the customer. If, for this purpose, you define a report with all characteristics, this would lead to long response times when you execute the report online, as the data would be loaded at the lowest level of detail in the memory.
You can solve this problem as follows:
When you execute the first report, the system reads 1,000 records from the summarization level, as the product group and product master group are dependent on the product, and there are 1,000 products.
To navigate to the customer, navigate to a product in the first report. Now call up the second report using the report/report interface. As the three required entry fields are transferred from the first report, no more than 1,000 records can be read from the transaction database. For this product, you can now navigate to the customer.
In contrast to the single report with all characteristics, where the system reads a million records, in this case the system only reads 1,000 records twice. You can therefore execute the reports online.