Information technology continues to change and evolve, and it is the responsibility of IT managers to continually provide their organizations with the best architectural solutions by identifying the opportunities and validating and proving out new solutions. This case study is about just that, a proof of concept initiative supporting the development of an architectural pattern for analytics within large manufacturing organizations referred to as the analytics workbench.

This research project, conducted using action research methodologies, is a study intended to be a proof of concept in assessing the effectiveness of a relatively lower cost commercial database based business intelligence platform and the more sophisticated analytical tools it provides in helping to understand complex business problems in very large, complex organizations. The business problem addressed (related to reducing warranty costs) is of a complex type common to most large manufacturing organizations, with the following characteristics: (1) It involves very large amounts of disparate data from multiple sources; (2) Efforts to investigate the problem have been only marginally successful; (3) Desktop tools and corporate applications have proven in-effective.

This study sought to discover whether business intelligence tools have evolved to the point that a major commercial database based analytics platform can provide value quickly and inexpensively to help address this type of complex problem. Further, the project develops a recommendation for supporting the product and the business users, from an IT perspective; this includes both the operational support for the associated hardware and software as well as the actual implementation and use by business (non-IT) personnel.

The research's conclusions support the use of the platform as a viable architectural pattern for specific, defined applications, and confirms the business value of using the BI tools. It identifies some apparent potential weaknesses and the need for some improvements by vendors, particularly in the area of usability for end users, and provides evidence that the tools still require strong IT support to be effective. Finally, a case is made for the need to improve IT competencies in BI tools in order to properly leverage the tools within an organization.