In the context of a Business Intelligence solution, a data warehouse is the central data collection point (one or several databases, in most cases), the contents of which are comprised of data from various sources.
The reasons for setting up a data warehouse are:
- unsuitable organisation and quality of data in the ERP system; there is no facility for evaluating functions which span multiple ERP systems, for example in a group
- insufficient ability to include outside data, e.g. from competitors and research institutes
- impact of analytical calculations on the ERP system
- constant changes to the database in an ERP system
Data transfer into the data warehouse takes place during an ETL process, whereby data are extracted from the previous system, transformed (in the sense of being prepared) and loaded into the data warehouse.
What is set up is usually relational or multidimensional databases, depending on the intended use. Optimised data storage facilities, used for collection and preparation of accounting data, are called financial data warehouses.