"Operational Excellence" (OE) is a term now in common usage in the business world. But what exactly is OE? Before defining OE, we must put it into context. OE involves efforts at the level of individual processes within an organization. In practice, it demands that any process be designed, executed and improved according to Deming's Plan/Do/Check/Act (PDCA) loop  (Figure 1).
In this framework, the "Plan" phase contains the process design and how it should be executed to achieve optimal performance, which often is called the control plan or control strategy. Once the plan is in place, the process is run (the "Do" phase). In an ideal world, the plan would be perfect and its execution would be flawless — the process would deliver optimal performance from day one. However, in reality, this never is true and process performance isn't optimal. So, we use the "Check" phase to compare actual with desired performance, and decide what actions must be taken to close the performance gap. These actions then are implemented in the "Act" phase.
OE generally is thought of as applying to the Do/Check/Act phases. However, because the Plan phase is the output of other processes that have their own PDCA loops, in a general way OE applies to all phases of the loop. So, OE can be viewed from two perspectives — an end in mind (the goal of optimal process performance) and a way to achieve it (flawless execution and, when issues and opportunities occur, the ability to recognize them and make improvements).
Given the above, the following definition seems reasonable: For any organization or part of that organization, OE is being able to reliably produce great quality at low cost while complying with all regulatory requirements.
This definition focuses on all the results of the process. It doesn't involve choosing one desired performance measure over another. Rather, it requires that all goals be met to truly achieve OE. The term "quality" reflects the degree to which the output of the process satisfies internal and external customer expectations. The word "reliably" is perhaps the most important word in the definition simply because many times it's the part of OE that's not recognized. However, performance reliability is a hallmark of OE. It's not only about optimal results but also about achieving them consistently over time. Such reliability is fundamental to a good customer/supplier relationship. Indeed, it's why process stability (statistical control) must be established in a process before process capability can be quantified [2,3].
OE represents an ideal that never can be fully reached in practice. Nevertheless, an organization should strive for OE. As legendary football coach Vince Lombardi said, "Perfection is not attainable. But if we chase perfection, we can catch excellence." Customer expectations always are changing, as are those of society (as reflected in evolving regulatory requirements), so at best any performance goal is temporary. Given the time it takes for an organization to make significant improvement, by the time the change has occurred the goal already may be out of date. As an example, in the past some companies considered three-sigma performance as world class but now strive for six sigma — and, for high-consequence situations, even six sigma may not suffice.
Given that goals are changing and reliability is so important, the second view of OE becomes critical. That is to say, as well as OE representing an end in mind, its never-attainable nature begs for an investigation of "how" it is attained. Having a solid approach to achieving the goals of OE means that reliability can be built into a process and the organization can respond quicker to changing requirements. However, actual performance is the only way to judge the success of efforts. If an organization is using what it considers to be best-in-class approaches to improve OE but defect rates still are very high and not improving, then clearly its approaches are flawed and must change.
People execute and manage processes — so, the core of the approach to OE must focus on their behaviors. Such behaviors are the easiest way to see OE when it's being done right (and when it's being done wrong). In addition, we must recognize that behaviors don't just happen but are caused. Therefore, we also must look at the elements that must be in place to promote the right behaviors to support OE and minimize the risk of the wrong behaviors that undermine OE.
One way to think about this is to consider how organizations react to human error issues. One approach is simply to consider such errors to be root causes and punish the individuals involved. A more-enlightened approach is to ask why such errors are occurring and whether organizational elements that support people — e.g., training, timeliness of important process data, suppression of nuisance alarms, etc. — are adequate. This view treats improper behaviors as symptoms of gaps in the underlying systems.