Strategic information systems planning (SISP) remains a critical issue of many organizations and also the top IS concern of chief executives. Therefore, researchers have investigated SISP practices and tried to improve a methodology. Among the various issues of SISP, systematically determining subject database groupings and fully automating the processes are important aspects. This study presents an alternate methodology using a statistical technique, a variable clustering approach, and systematic rules for determining database groupings, which can be fully automated. This methodology provides a strong theoritical justification as well as systematic and simple criteria for database groupings, enhanced interpretability of the output, and would be easy to include in CASE software application.