Visual Six Sigma: Making Data Analysis Lean
by Ian Cox, Marie A. Gaudard, Philip J. Ramsey, Mia L. Stephens, Leo Wright
5.3. Collecting Baseline Data
Now Jane and her team embark on the process of obtaining the historical data and verifying its integrity. Once this is done, they will compute some baseline measures.
5.3.1. Obtaining the Data
Recall that Jane has decided to study two periods during which 5 percent price increases were targeted. She has identified 20 products and, for each product, 12 customers representing various sizes and regional locations (3 customers per region). Each of these product and customer combinations had sales in each of two periods: oversupply and shortage. This results in 480 different product-by-customer-by-period combinations.
Having identified the data required, the corresponding invoice figures are directly downloaded from Polymat's data warehouse into a standard database. After entering the Product Categorization Matrix categories determined earlier, along with the sales representative experience and buyer sophistication rankings, Jane loads the data from the two ill-fated pricing interventions into JMP for further investigation. The raw data are given in the first ten columns in the 480-row data table BaselinePricing.jmp (Exhibit 5.4).
Figure 5.4. Partial View of Baseline Pricing Data Table
Jane is thinking more and more about the failure of a sale to achieve a 5 percent price increase as a defect; she has even heard some members of the management team refer ...