We examined the practices and major problems that can easily occur in the use of balanced scorecards in the June 3 issue of InsideSixSigma. Now let’s look at why attention to causal event analyses can improve the future performance of any process. Often, this traditional method fails to make metric statements that describe process variability.

The E-DMAIC system within an integrated enterprise excellence (IEE) process provides a scorecard approach that more structurally integrates and orchestrates a company’s whole system.

The approach has five characteristics:

1. Part of the measure phase of E-DMAIC, the scorecard structurally assesses the enterprise system as a whole. In the value chain, business measures are tracked using a corporate level metric report format, while functional areas of the business area are tracked using an operational format to describe how this form of tracking can transition an organization from fire fighting to fire prevention. At this stage, everyone in an organization has all the information needed for an enterprise to move more closely in meeting the three Rs of business: Everyone is doing the right things and doing them right at the right time.
2. Every metric will have an owner where the measurement’s performance can be part of the manager’s plan and discussed at reviews. Sometimes the manager might learn later that no improvement efforts are needed. He or she needs only to maintain the current operational level of performance—a welcome relief from the activity-oriented conclusion drawn from traditional scorecards.
3. In the analyze phase of E-DMAIC, analytics help facilitate, create, and achieve smart financial improvement goals and strategies. By integrating analytics with innovation, an organization can identify and then give focus to areas that, when improved, can yield the most benefit to the business as a whole.
4. Through an enterprise improvement plan (EIP) system, goals are established for value chain performance metrics, which can be cascaded with alignment throughout an organization.
5. This system creates a measurement-improvement-need pull for an improvement or design project - in contrast to the traditional push-for-project creation system under most lean Six Sigma deployments; i.e., “Let’s vote on which projects the black and green belt practitioners should be working on.”

With traditional red-yellow-green scorecards, metrics can be established throughout an organization with corresponding goals. When a metric goal is being met, all is well and the color is green. When measurements are close to not being met, the color is yellow. The metric is colored red when a goal isn’t being met and corrective action needs to be taken.

One rule of thumb is that most scorecards should include 7 to 10 metrics. Any more than that and a person will struggle to monitor and act on them. How can you have a metric that is red for the entire reporting period? Is no one monitoring it? Is it based on an arbitrary target and just ignored? Who knows, but all are possible.

The IEE analysis process includes the following two steps:

1. Assess process predictability.
2. When the process is considered predictable, formulate a prediction statement for the latest region of stability. The usual reporting format for this statement is the following:
   
   a. When there is a specification requirement: nonconformance percentage or deffects per million opportunities
   b. When there is no specification requirement: median response and 80 percent frequency of occurrence rate

Applying the scorecard/dashboard metric reporting process to this data shows:

1. The purpose of red-yellow-green charting is to stimulate improvements. Figure 2 provides an assessment of how well this is accomplished. This figure contains a control chart, probability plot, and histogram. When there are no occurrences beyond the two horizontal lines (i.e., upper and lower control limits), no patterns, or data shifts, the process is said to be in control. Again, when this occurs, there’s no reason to not believe that the up-and-down monthly variability is the result of common-cause variability, i.e. the process is predictable. Because the process is predictable, we can consider past data from the region of stability to be a random sample of the future.
   
2. The histogram in figure 2 is a traditional tool that can be used to describe the distribution of random data from a population with continuous response. It’s difficult to determine from a histogram the expected percentage beyond a criterion. A probability plot is a better tool to determine the nonconformance percentage estimate since actual data values are plotted on a coordinate system where percentage less than is on the y-axis. Figure 2 provides an estimate that approximately 32.6 percent of future monthly reporting will be less than the lower-bound criterion—unless a fundamental process improvement is made or something else externally occurs to the process. The percentage value is consistent with an estimated proportion below the 2.2 reference line in the histogram graph. This percentage is similar also to the percentage of red occurrences; that is, 5 out of 13. If this nonconformance percentage of 32.6 percent is undesirable, this metric would pull for an improvement project creation.

To reemphasize, effective long-lasting improvements to processes aren’t made by firefighting individual time-line conditions beyond a desired objective. Process improvements can be made by collectively examining process data over the period of stability so that insight might be gained for the purpose of determining what could be done differently to improve the overall response. This can be accomplished through the execution of a project define-measure-analyze-improve-control (P-DMAIC) roadmap. (Figure 2 below illustrates IEE-improved finance metric B continuous-response reporting: Red-yellow-green scorecard versus IEE reporting. The histogram is included for illustrative purposes only.) figure 2

The future
As domestic and international competition continues to intensify, businesses will rely increasingly on scorecarding. Industry reports indicate that midsize businesses made reporting and dashboards their top business intelligence investment in 2007, spending more for this function than for data integration.

To achieve the true potential of a balanced scorecard, ways must always be examined on a continuing basis to be sure it is integrated with an enterprise’s operations as effectively and productively as possible. Balanced scorecards, properly structured, are essential in helping organizations achieve business strategies and bottom line goals. Otherwise, an unbalanced or incomplete scorecard can result in surprises with significant negatives for an enterprise.