Visual Six Sigma: Making Data Analysis Lean

7.7. Conclusion

Using visual techniques, Rick has been able to construct a good understanding of the operation of his sales force and to answer the two questions posed to him by his manager in a complete and compelling manner. Rick's work in the section "Validating and Scoping the Data" was largely directed toward his personal goal of obtaining better knowledge of his sales force. In the "Investigating Promotional Activity" and "A Deeper Understanding of Regional Differences" sections, he addressed the business questions posed by his manager. The analysis that led to the required answers in these two sections was clean, quick, and compelling, and much more efficient than anything Rick could have done without JMP.

This case study is a good example of where force-fitting important business questions either into a project framework or into the traditional Six Sigma DMAIC methodology would be ill-advised and counterproductive. Rick Fincham's analysis raises two major points:

1. Sales data are likely to be more, not less, complex than the data presented here, and there are likely to be more records. Questions relating to data quality, the distribution of missing values, and the balance of the data (namely, which levels of descriptive variables occur jointly) become crucial. As this case study seeks to show, understanding these issues is a necessary step in any kind of sensible analysis. Once these patterns of variation are understood, the analysis itself is often relatively straightforward ...