In my last posting I laid out three principles, with the first one being that time is the ultimate constraint. The authors propose to us that if "time is the ultimate constraint and the promotion of flow is the best way to manage it, then we need to understand what is most often impeding our ability to promote flow." In other words, what is the primary element that destroys flow in systems? The obvious answer is variability within the system. The APICS dictionary (12th edition (Blackstone, 2008 page 71) tells us that the law of variability states, "The more that variability exists in a process, the less productive that process will be." The authors rightfully have a concern with this definition in that it may not adequately highlight the impact of variability at the system level. To this end, the authors have proposed a new law as follows:
"The law of system variability: The more that variability is passed between discrete areas, steps, or processes in a system, the less productive that system will be. The more areas, steps, or processes and connections in the system, the more erosive the effect to system productivity will be."
I absolutely love this concept and the authors have created the following graphic to illustrate their law of system variability. The lower half of the graphic depicts a network of connections, The authors tell us that this could represent a project network, a bill of materials, or even a routing. Their point is that it depicts a set of relationships between discrete events, areas, or entities that culminate in some form of completed product, project, or end state.
The large squiggly line represents a variability wave that accumulates and amplifies through the system. The authors point out a very important fact that delays frequently accumulate, whereas gains rarely accumulate. The authors have added a graph above the network section to show the impact of the variability wave on system lead time and output. In short, they tell us, lead time expands while output decays. This is such an important concept to understand and realize! Think about it.....a potentially small amount of variation at the beginning of the system continues to amplify and accumulate as it works its way through the system.
Let's now turn our attention to a more in-depth discussion of variability, including the various types that the authors have so eloquently written about in their book.
The authors propose to us that if flow is paramount to protect and system variability kills flow, then we should explore the nature of system variability encountered by companies. The authors are quick to point out that much of their discussion on variability is largely taken from the third edition of Orlicky's Material Requirements Planning by Carol Ptak and Chad Smith (McGraw-Hill, 2011, pages 16-18).
The authors explain that variability can be systematically minimized and managed, but not eliminated. They rightfully point out that the Six Sigma toolset provides an excellent approach to reducing variability at the discrete process level, but even the best master black belt cannot totally eliminate variability (at least not without a magic wand). The authors further explain that there are four distinct sources (two internal and two external) of variability within any system. Furthermore, these sources come together to create total system variation. So what are these four sources? In the following graphic, the authors describe these four sources of system variability.
The types of variation
As the graphic states, the four sources of variability are:
- Demand variability - an external form of variation characterized by fluctuations and deviations experienced in demand patterns and plans. The authors explain that this type of variability is driven by system nervousness of the major players near the top of the supply chain.
- Supply variability - an external form of variation that is seen as a disruption in the supply network or deviation from the requested and/or promised dates for supply order receipts. The authors rightfully point out that something as simple as a five-cent fastener can block the delivery of a multi-million dollar assembly.
- Operational variability or "Murphy" - an internal source of variability which Dr. Deming referred to as common cause or natural cause variability. This is the ever-present variation that normally exists within all processes. Even if a process is in a state of "statistical control" variation is still present.
- Management variability - an internal source of variability that is associated with the human element within a system. Dr. Deming referred to this as special or assignable cause and told us that this form of variability should be the first target for improvement. The authors explain that this form of variation is associated with the human element and since humans make and interpret the rules of the system, this type of variation is associated with the decisions made and actions taken. Management makes the policies within the system and many times these policies motivate opposing behaviors within the system.
The authors ask one simple question. Of the fours sources of variation, which is under our direct control? Let's look at each type.
- Customer and market behavior (Demand variability)? Hardly.
- Supplier performance (Supply variability)? Only indirectly.
- Random events (Operational variability)? The definition of random says it all.
- How we decide to run the business (Management variability) Bingo!!
The authors make a very important point that I hope everyone sees right away. "Could it be that a significant portion of the variability that we experience as an organization is largely self-inflicted?"
In closing this posting, I want to encourage everyone to go purchase Debra and Chad's new book. It is full of amazing content and is sure to help everyone who reads it. I personally and publically want to thank Debra and Chad for their outstanding contribution to system's theory and for showing all of us a better way!! My next posting will be the last in this series and will be about becoming what the authors refer to as, Becoming Demand Driven.