Mistakes Leaders Make: Not Appreciating And Managing Variation
Mistakes Leaders Make: Not Appreciating And Managing Variation

Mistakes Leaders Make: Not Appreciating And Managing Variation

This final blog in the series of seven explains one of the most critical issues facing managers and organizations every day, namely a manager’s inability to understand, appreciate, and manage variation.

Dr. W. Edwards Deming had a unique way of thinking about the roles and responsibilities for managers.  For example, Deming wanted us to use the phrase “the ability to predict” to describe one of the most important skills a manager must develop. An effective manager will, if he/she does their job, be able to make predictions about processes. Predictions are valuable to an organization. Unpredictability is costly. Predictability improves the ability to deliver customer value. Unpredictability damages customer loyalty.

The first six mistakes typical managers make that damage employee engagement are:

Mistake #1: Not appreciating the concept of a system.

Mistake #2:  Believing you must have all the answers.

Mistake #3: Believing an improvement in employee performance improves organizational performance.

Mistake #4: Misunderstanding motivation – Manipulation is NOT Motivation.

Mistake #5: Being unaware of your most important characteristic.

Mistake #6:  Misunderstanding feedback and its importance.

Mistake #7 is about the inability or unwillingness of managers to appreciate and manage variation. Typical managers use blame language to explain variation. This is one of the worst mistakes they can make because it not only damages employee engagement but prevents innovation and improvement.

Mistake #7: Not appreciating and managing variation

Medical malpractice is rife with misdiagnosis. Leadership malpractice is as well.   Below are data for 8 workers making the same product, all working at about the same rate for 12 weeks.  Imagine you are the leader in this plant (doctor) of these workers, what do you make of these data?  What would you prescribe?  What is your diagnosis? (Tribus, 1990)


What would you do?  How would you go about reducing errors?  What would be your approach?  Myron Tribus explains how he has presented this table to audiences across the USA, in Mexico, in Canada, in Australia, in the UK and he always got about the same reaction.  People almost always suggest one of the following regarding Eva: a “good” stern talk; have Eva work alongside Mary and learn; to fire her; give her more training.  In this true-life example those typical answers are consistent with the blaming an individual(s) for the inferior quality.  This is an enormous mistake.  The inferior quality comes from variation in the production system and not the people.  All variation comes from the processes and policies as designed by management. The root cause is not with the people.

Dr. W. Edwards Deming describes two types of mistakes that leaders can make, 1) to act when they should NOT act and 2) NOT act when they should. We can know when to act if we collect data and create a control chart.  A control chart is a method to display data to help managers predict when to act and when not to act.  By collecting data and plotting it onto a “control chart” a leader can begin to manage the variation in a process because a control chart is used to monitor the variation in a process.  Data points that fall within the limits (upper and lower control limits) indicate common cause variation or random variation.  Any points that fall outside these limits indicate a major change to the system.

Think about how long it takes you to travel to your work every day (drive, walk, ride your bike etc.).  What is the average time?  Is there too much variation in this process such that it is chaotic or can you predict (with accuracy) how long the trip will take?

Let’s assume it takes a person 41 minutes to travel from home to his/her work in the morning.  It is not always exactly 41 minutes, 41 minutes is the average.  There is always going to be some variation. This variation comes from various factors that occur in the route that person takes.  These common cause factors might be, traffic, the time he/she leaves the house, the number of lights he/she catches as green, the weather, accidents, the day of week, holidays celebrated, etc.  These common causes create expected variation.  If all times fall between the upper control limit and the lower control limit the process is referred to as “stable.”  Being stable doesn’t necessarily equal “good.”  It just means predictable (within the limits).  Here is 30 days of travel data going to work in the graph “Travel Time to Work.”


Here the upper control limit is 63 minutes and the lower limit is 24 minutes.   If the system was stable we could expect all days of travel to take between 24 minutes and 63 minutes.  But, on the third Monday the time was nearly 100 minutes.  Ouch! This is a “special cause.” Something must have influenced the time, something outside the system, something special. Perhaps it was an accident on the highway and the State Police shut down the highway to clean up the oil spill caused by the overturned truck.

It’s the leader’s job to create a control chart to identify special causes and common causes especially with the management of people. Deming explains that the most important contribution control charts can make is with the management of people. (Deming, The New Economics – Second Edition, 1994)  To avoid this special cause in the future requires special action. Perhaps the worker might want to check with the State Police each morning in order to avoid any future road closures.

Assuming there are no special causes (overturned trucks) the only way to create a consistent shift in travel time is to change something about the system. Any attempt by the driver to reduce the individual daily times will NOT produce significant improvement (value will remain within the control limits) and it may create a safety issue.

Now, let’s discuss Eva and the “Flaws per Worker/Week” data. Remember, most viewing this data suggested Eva was the problem. However, when we review the control charts we see a different perspective. There are NO special causes. Eva is NOT a special cause.  The errors are all common causes which mean the factors creating the errors all came from the system and not the people.



All data points are within the control limits. Any attempt to help Eva to change is probably a waste of time because all the factors that cause the errors come from the overall system within which all the people work, including Eva. The system influenced Eva’s performance.

This is a transformational insight for most leaders and managers. The typical manager wants to create a linear connection between the errors and Eva. They want to pair Mary with Eva in an effort to re-train Eva. These actions reveal an inability to understand variation.  When we use the typical performance appraisal we are also demonstrating our inability to understand a system and our inability to understand variation.

Common cause variation is the “voice” of the system. It is the way the system is currently operating and the way the leaders designed it. If a leader wants to improve common cause variation he/she must study the process using quality improvement tools and make decisions based on knowledge of the system. A change in one process can cause an impact (unintended negative consequences) in another process.

Managing variation improves quality, profitability, and, shareholder value. Leaders who manage variation well can bring new products to market more quickly with minimal errors.  Their teams can produce high-quality products with minimal rework.  Their teams can grow customer loyalty faster.

Leaders often talk a good game about the importance of producing high-quality products and services at a low cost, but few use control charts effectively and very few use control charts to help manage people.

Too often leaders are reactive by fixing problems after they have already been created instead of using control charts to make predictions and thereby knowing when to act and when not to act.  The typical managers pretend to be “omniscient problem solvers” deciding quickly what he/she thinks are the answers to the variation. This approach often makes things worse. Without an understanding of variation and without the use of a control charts managers can are committing leadership malpractice.

Wally HauckWally Hauck, PhD, has a cure for the “deadly disease” known as the typical performance appraisal.  Wally holds a doctorate in organizational leadership from Warren National University, a Master’s of Business Administration in finance from Iona College, and a Bachelor’s degree in philosophy from the University of Pennsylvania.  

Wally is a Certified Speaking Professional or CSP.  Wally has a passion for helping leaders let go of the old and embrace new thinking to improve leadership skills, employee engagement, and performance.

Share this post

Start typing and press Enter to search

Shopping Cart