Statistical Business Analysis
sustain the gain my blogcontact us

Concepts:


To introduce the concepts used in our statistical approach, lets start with the idea that you need to evaluate Market Share performance. We have already exposed you to the Index versis Year Ago and its shortcomings, so we will consider some alternatives, beginning with a simple line graph. The vertical axis is for market share and the horizontal axis is time, in this case quarters.


For most company metrics, someone has an expectation of performance. On our graph, it shows as the Manager’s Goal. We will Now add the results to date for 5 years in the graph below.

Here is where the Manager expectation effects the analysis. In most companies, action would begin by asking “what happened in the last quarter since it is below the goal”! Teams would form, problems would be identified and solutions put forth. Finally after the agreed action was taken, the following quarter would be used to justify this course of action, as seen in the following graph.



If you ignore all the data points except the last 2, there is a 50/50 chance that the last result was going to rise. However, if you consider all the graphed history, there is a greater than 90% chance that the last quarter would rise, whether action was taken or not!! Yet, this “objective” data of the final quarter will be used to justify all the meetings and expenses associated with the action taken to “fix the problem” identified in the quarter below the goal.

So the question is: was there really a problem? The result below the goal was treated as being uniquely different from all the results above the goal. We must be VERY careful not to treat “undesirable” results as uniquely different without an objective way to determine the difference. This is where statistics comes in.

To determine if the most recent result is different, we need to establish what is “normal” for this business result. This is done by calculating the standard deviation of the actual data. Multiply this by 3 and add it to the average for the upper control limit (UCL). Take the standard deviation times 3 and subtract it from the average to get the lower control limit (LCL). Statistics tells us that 99.7% of all the results will fall between the UCL and the LCL if the business is stable and consistent without any changes or breakthroughs.


The sources of variation that make the results go up and down within the limits and cluster around the average are called the Common Cause. The Common Cause is made up a many interrelated sources. For Market Share, the common causes include growth of the total market, competitor product quality, marketing, pricing, distribution, etc., your product quality, marketing, pricing, distribution, etc., industry practices, retail practices, macro economics, micro economics, government policy, weather, and the list goes on! Unfortunately, you only control only a small percentage of these causes. Therefore, it is critical to understand, at least better than your competition, the complete set of common causes. Only then can you change the “rules of the game” to create and sustain a statistically relevant change or BREAKTHROUGH.

If a result goes beyond the 3 sigma limits, there is a very good likelihood that something out of the ordinary has occurred. This is a Special Cause caused by a singular event that is not typical of the common causes. If these changes are positive, we would hope it is due to actions that we took. But remember that it might also have been due to the actions of your competitors. In any case, it is critical that these positive breakthrough actions be identified so that you can make them part of your ongoing strategy. There are a couple of additional methods to identify a Special Cause. One is to look for 7 or more consecutive results that run above or below the average. Second is to find “non-normal” patterns in the data which is aided by creating one sigma boundaries around the average. This area is usually colored to help with interpretation.

Building a knowledge base of the actions which only sustain past business results and those actions which create sustainable breakthroughs, will improve the cost effectiveness of your business plans and strategies. For instance, one business discovered temporary price reductions would create a breakthrough in share. However, profit pressures would soon reverse this pricing causing share to drop back again. The business team realized this was only increasing the variation without creating share growth.

web services by adaptive convergence, llc