The recent Global Business Process Excellence (BPE) benchmarking survey indicated that BPE professionals are shifting their focus from cost reductions to revenue generation, and that the most successful of these initiatives are explicitly linked to their competitive business strategies. This shift from an internal to an external focus requires an understanding of an extended set of statistical tools — tools associated with collecting and analyzing the Voice of the Market (VOM). Most Lean and Six Sigma certification programs, however, fail to address some of these tools altogether, or address them in ways that are not clearly linked to customer feedback.
Lean and Six Sigma practitioners don’t need to actually conduct the market research that’s required for setting market-focused, value-driven priorities — that’s why we advocate developing a partnership with the marketing team in your company. But they should understand how to use and interpret the results of five key statistical tools for identifying those priorities, and to be sure the tools are used properly and consistently. The critical tools for a comprehensive certification include:
- Factor Analysis
- Multiple Regression
- Analysis of Variance (ANOVA)
- Cluster Analysis
- Cause and Effect Matrix (modified)
1. Factor analysis
Factor analysis is the statistical tool used to identify and codify critical-to-quality factors (CTQs) from a market perspective. Essentially a data reduction tool, factor analysis is used to examine the relationships among a large number of questionnaire attributes in order to retain the actionability of those attributes, while simultaneously grouping them into a limited set of CTQs.
For example, following is a partial list of attributes randomly selected from a survey of auto body repair shops. Shop owners were asked to rate the performance of their supplier on a 10-point scale.
|
|
|
|
|
|
|
|
|
|
|
|
This level of detail on a questionnaire is necessary if you plan to actually do something with the results. But a list of 70 such questions must be reduced to a more manageable set of factors for interpretability. And that’s what factor analysis is for. In this case, the 12 performance attributes (survey questions) can be reduced to three critical-to-quality factors (CTQs):
- Business Support Processes
- Order and Delivery Processes
- Training and Problem Solving Processes
The numbers to the right of the individual attributes are factor loadings and, essentially, represent the correlation of each individual attribute with the corresponding CTQ. As such, they can be interpreted as importance weights, and are very useful when applied to the "Cause and Effect Matrix." (See below.)
2. Multiple Regression
Multiple regression is a statistical tool familiar to most Six Sigma practitioners. The tool is used to describe the relationship between a dependent, or criterion, variable and a group of independent, or predictor, variables. When used to understand the Voice of the Market, the criterion variable that is most important relevant is "value," because value is the best predictor of revenue growth and market share. We know that value represents the trade-off between the quality that the customer receives, relative to the price she or he must pay for that quality. Multiple regression will tell you what that trade-off is, and how important each component of quality (each CTQ) is for creating and delivering superior value. The result is a model of market value that provides quantitative direction for business process improvements. (Click on image to enlarge.)
4. Analysis of Variance (ANOVA)
Another tool that is already widely used by Lean and Six Sigma practitioners, ANOVA is used to evaluate the differences among multiple variables. When applied to the voice of the market, the differences we are interested in evaluating are differences in performance on the CTQs among the several competitors. Competitive performance ratings can be plotted on a matrix representing the interaction of quality and price, and differences in performance on those variables, as well as on the individual CTQs themselves, can be statistically evaluated in order to determine whether your company enjoys a competitive advantage, suffers a competitive disadvantage, or is at parity with key competitors. (Click on images to enlarge.)
4. Cluster Analysis
Cluster analysis is used to group things with similar characteristics. When applied to the Voice of the Market, cluster analysis can be used to evaluate the loyalty of your customer base, or the vulnerability of your competitors. Customers are grouped on the basis of the value they receive (quality of products and services relative to the price paid) in order to identify the potential for future defections, and how those defections might be prevented. Companies that have adopted the Net Promoter Score (NPS) metric can profile the loyalty groups on the basis of their NPS scores in order to identify specific actions for improvement. (Click on image to enlarge.)
5. Cause and Effect Matrix
When applied to the Voice of the Market, the effects that you are trying to achieve are improvements in performance on the CTQs that are having the greatest impact on value. Because you do business in a competitive environment, you want to "out value" your competition. This means:
- Identifying competitive performance gaps and evaluating the importance of each
- Identifying the value streams having the greatest impact on those performance gaps
- Evaluating the specific impact of each value stream process on those performance gaps you choose to address
You’ll find a more detailed explanation of this tool in a previous column on this site.
The BPE Benchmarking Survey clearly indicated that, when focusing process improvements on revenue growth or market share, those improvements are much more effective when driven by the Voice of the Market (VOM). In order to maximize that effectiveness, then, BPE certification programs need to provide instruction on the use of these five statistical tools and their application to the Voice of the Market.