LEAN SIX SIGMA WORLD CONFERENCE
Improved Analysis Technique for the Solution Selection Matrix
Presenter: Robert Rhyder, Contractor, GuideWell, Jacksonville, FL, USA
LSS, SSM, Analysis
Improved Analysis Technique for the Solution Selection Matrix The Solution Selection Matrix (SSM) has long been one of the mainstays of problem solving and process improvement activities within Six Sigma. It facilitates the development of unique solutions designed to address and remove specific root causes and then prioritizes solution implementation by assigning numerical ratings to factors such as individual solution effectiveness, cost and risk. The product of all three numeric ratings is then used to determine the desirability of implementing the solution. This is accomplished in a well-defined horizontal analysis schema that has been the standard analytical method for SSMs for many years.
Despite the benefit that is generally obtained by following the standard horizontal analysis method it has the potential to generate less than optimal results. Consider for a moment solutions classified as “low cost” and “low risk.” These have the highest “cost” and “risk” multipliers assigned and the product of these is often 25 or more. A multiplier that high may cause these solutions to have an overall desirability that is greater than most, if not all, other solutions on the SSM. These low cost and low risk solutions may or may not be applicable to the root causes the Subject Matter Experts (SMEs) think are the most likely causes of the problem at hand. Additional analysis techniques can be applied to the SSM to ensure the root causes chosen as the most “likely” causes of the problem are not ignored because only solutions with least cost and risk are promoted. This is accomplished by a vertical examination of each individual root cause column. In this analysis, the contributions that individual solutions make toward eliminating any one root cause are simply summed and recorded at the bottom of each column. Since unique solutions are selected for each root cause when the SSM is created, their combined effects on any individual root cause are frequently independent of one another and additive. If a scale of 0-10 is used when solution effectiveness is initially estimated, the vertical analysis goal is to select a set of solutions whose sum adds to at least 10 in every column of the SSM, since a “10” corresponds to 100% elimination of the root cause. Any score above “10” indicates a potential condition where the combined effects of multiple solutions yield complete problem resolution with an additional factor of safety, a concept that is exceptionally valuable in the event SMEs over- estimated the effectiveness of any individual solutions. A column score of 14 (140%) then, is more certain to eliminate a specific root cause as a contributor to a problem than a column score of 11 (110%). It represents multiple solutions working to eliminate a root cause. It also represents a larger factor of safety from an engineering perspective. In other words, when one considers variability introduced in solution implementations, the column with a total of 14 (140%) will prove to be a solution set that is likely more robust over time when compared to the column that has an overall sum of 11 (110%). Maximum versatility and control of the solution selection process is obtained when the horizontal and vertical analysis techniques are employed concurrently, and both are tied to a stacked bar graph. In this scenario, all the rank-ordered root causes are placed on the graph’s horizontal axis and a “combined effectiveness score” is along the vertical axis. Choosing desired solutions first from the horizontal analysis of the SSM will generate unique color-coded vertical bars to stack above each root cause the selected solutions impact. The height of each bar segment will coincide the magnitude of each solution’s effect on each root cause as determined by the SMEs’ estimates of the solution’s effectiveness. For each solution chosen, the vertical analysis will determine the cumulative effect of multiple solutions for each individual root cause. The resulting graph will illustrate with extreme clarity the degree to which each root cause will be addressed. It will also enable practitioners to visualize the effect of mixing and matching a variety of solution combinations to minimize costs as much as possible while ensuring that critical root causes are adequately addressed and completely eliminated.