Business Intelligence (BI) is the gathering, management and analysis of vast amounts of data in order to gain insights to drive strategic business decisions, and to support Operational processes with new functions.
BI is about the development of information that is conclusive, fact based, and actionable. It includes technology practices like data warehouses, data marts, data mining, text mining, and on-line analytical processing (OLAP). The objective of a BI solution is to transform data into useful information, such as customer profiles, buying habits, product profitability and competitive analysis. It may involve analyzing volumes of data for unsuspected, but valuable, associations and insight. It includes streamlining data into useful reports and sharing that information with people inside and outside the organization who need that information.
However, implementing a successful BI initiative is not as simple as just installing the required technology. It is imperative that the business objectives for the project be clearly defined at the outset and that the project has upper management's complete support. At this point, the technological solution can be developed, and the expected benefits of undertaking the project quantified. By predicting the return on investment expected from a project, management will have a means by which to measure the success of the project. Equally as important is the communication that must take place between a company's IT staff and the end users on the business side of the company. A data warehouse will not be a success if the end users are not fully aware of the many ways they can use it.
Assuming that the overall strategic business approach is sound, the next important factor is the use of reliable, consistent data. Too many business decisions are being made using data of dubious quality and consistency. The process of extracting data from disparate sources, transforming it into a consistent record -- also known as cleansing -- and then loading it into a data warehouse is a critical process. Another important aspect is the data model, the map of how the database is organized and how its fields relate to each other. Using tried and tested models specific to the needs of a given industry will greatly enhance the effectiveness of the solution.