Business Intelligence (BI) is best understood by dividing the concept in distinct areas that represent things people actually do as part of their normal routine to run manage and operate organizations. After years of working
in the fields of data warehousing and BI we have developed this overview to help our customers better understand how and where BI can be beneficial. The five pillars of Business Intelligence represent the five things that are "done" by either systems or people to provide a better understanding of the information available about
an organization.
Pillar 1 – Collection: This represents activities primarily executed by the systems representing BI. It includes the extractioin of data from
source systems such as ERP or CRM, the transformation and integration of such data, the loading of this data into a data
warehouse or data store and the creation of semantics and meta data to make the data secure and business friendly to
end users.
Pillar 2 – Reporting: Most organizations have some form of this capability already. BI extends typical database reporting to make electronic
versions of reports interactive and available online. It also provides a mechansim for automating the distribution of
reports in a standardized format such as PDF or Excel.
Pillar 3 – Analysis: Sometime referred to as OLAP (online analytical processing) many organizations need to provide certain people with the
ability to define their own views of the data and to summarize the data on the fly. Ad hoc analysis and query tools
provide such capabilities and often reduce the number and complexity of reports.
Pillar 4 – Visualization: In many cases executives and others wish to understand now things compare over time or against a predetermined goal
without having to review details. Visualization provides graphically rich, interactive dashboards and scorecards, which
help users quickly identify exceptions and problems. This content can also serve as a launching point for more detailed
analysis with other tools.
Pillar 5 – Discovery: An often overlooked aspect of BI is the ability to identify trends and relationships in the organization's data. This
"discovery" process can often be based on a non-structured approach to investigating an area of interest. BI offers
tools and techniques for supporting such activities.