The promise of business intelligence, or BI, has always been to empower business users with better information than they could obtain from operational systems. While traditional BI continues to fulfill certain needs, self-service BI addresses the problem of ever-shifting demands for information.
In any organization, there exist multiple data sources that are constantly capturing information about that business. In addition to these sources (which are often in different formats), there are users with unanswered questions about their business. These users are seeking answers, but first need access to the data. Traditional BI relies on a centralized, cleansed and transformed store of data that users can access through standardized reports and perhaps an ad-hoc query tool. However, the development and maintenance of this type of system takes time and imposes constraints on the types of analysis users can perform – the data must be ‘in the system’ in order to be analyzed. On a day-to-day basis, the types of decisions that business users must make frequently require information that is not yet (or will never be) in the data warehouse. At best, a diligent user will spend the time to manually reformat and integrate the necessary data. At worst, users will make decisions without having all the information relevant to the problem at hand. Neither of these outcomes is best for the business or any of the typical endpoints along the spectrum.
Self-service BI shifts the emphasis away from the processes required to manage data in a centralized store and toward a process for finding, accessing and integrating pertinent information as needed. With self-service BI, decision makers are better able to respond to changes in business conditions quickly.
The focus of this paper is to:
- empower business users with better information in regards to business intelligence
- examine the obstacles and solutions self-service BI provides
- and propose some best practices for deployment and use