Most companies build separate systems for business intelligence applications. This is because data typically needs to be organized differently for decision-making purposes, and because companies want to protect their operational systems from performance impacts. Operational systems require data to be organized according to the requirements of a given application, such as order entry. Business intelligence systems require the consolidation of data across multiple systems, that the data be summarized, and that it be maintained over time to identify trends.

To build these systems, tools are needed to extract, cleanse, and transform data from source systems that can be on a variety of hardware and software platforms. When the data is prepared for business intelligence applications, it is stored in a database system and must be refreshed and managed. These systems are called data warehouses, and the process of building and maintaining them is called data warehousing.