Content area
Full Text
photo, Kent Bauer
Once again we will tackle the challenge of reducing the 1,000+ metrics shipped "out-of-the box" with business intelligence and performance management applications to a manageable 10 to 15 key performance indicators (KPIs). Last month we discussed affinity analysis as one potential approach. This month the focus will be on a Six Sigma technique called correlation analysis, which will allow us to understand the relationships between individual metrics while simultaneously identifying potential KPI candidates. The e-commerce case study in this column will illustrate how correlation analysis can highlight the relationships between Web metrics and bottom-line financials. The approach is to first determine which metrics are related to the business financial drivers. The second step is to discover the interrelationships that exist among the Web metrics themselves. This prevents the inclusion of any two Web metrics that might be measuring the same process or outcome. The result of the analyses will be a reduced set of KPI metrics that are uncorrelated...