Aggregations can be classified into two types: identified and ad hoc. Identified aggregations are those that the data warehouse already has stated requirements for. Ad hoc aggregations are those that are created to solve particular performance problems. These will be created and deleted as necessary, and are impossible to test.

Each identified aggregation should be tested thoroughly. There are often a number of ways in which a given aggregation can be created or maintained. Each method should be checked and the best compromise of performance and resource usage chosen. These tests are important, because they will often show up the fact that a particular aggregation is impractical.