Who Was W. Edwards Deming? The Creator of the PDCA Method
When discussing the PDCA cycle, it is impossible not to briefly mention the man to whom we owe this simple yet incredibly effective tool. The author of the method described here is W. Edwards Deming, a statistician and graduate of institutions such as the University of Wyoming, Colorado State University, and Yale University. He was a distinguished reformer who advocated the idea of continuous improvement and was one of the greatest pioneers in the field of quality management known worldwide. During World War II, Deming worked as a statistical consultant for the U.S. government, helping to improve the quality of war production. After the war in 1950, he was invited to Japan by the Union of Scientists and Engineers (JUSE) to give lectures on quality control. Deming conducted seminars for engineers and managers, promoting the principles of statistical quality control and continuous improvement. His methods were widely adopted by the local companies, contributing to a significant increase in product quality and their competitiveness in the global market. Through his lectures and writings, Deming promoted the PDCA cycle as a tool for continuous process improvement. Thanks to his influence, the cycle has become widely used around the world. The Deming Cycle has broad applications across various fields, not just in manufacturing. It is employed in quality management, project management, and broadly defined problem-solving. We will be exploring this method today, in relation to this last aspect. PDCA is a key element of a professional and well-designed problem-solving session. Nevertheless, it is important to remember that it will not yield the desired results in this context alone if we strip it of the several other complementary tools used in the preceding steps of the problem-solving cycle.
Pitfalls of PDCA
Before we proceed further, it is worth highlighting some of the most significant pitfalls associated with using PDCA for problem-solving:
- Lack of a detailed understanding of the problem: Failing to fully comprehend the problem’s characteristics, structure, and impact on the area or process in question.
- Inadequate selection of the most important causes: Choosing the most crucial causes of the problem in an unstructured or intuitive manner.
- Lack of proper root cause identification: This can lead to taking further steps in the wrong direction, addressing superficial symptoms rather than deeper root cause, resulting in no results or in best case scenario a short-term effect.
- Skipping the comparison of effect vs. effort: The precise analysis of the expected effect of the implemented changes with the amount of resources needed for its implementation.
- Incomplete analysis of the environment surrounding the problem: This may lead to unexpected impacts on other processes or operational aspects after implementing a potential solution. In other words, a common issue for beginners: solving one problem may unintentionally create two new ones. #trap of unintended consequences
- Lack of specific criteria for assessing the solution’s effectiveness: Poor preparation, including the failure to identify and quantify the damage generated by our problem, will make it difficult to define what we want to achieve and what will be the measure of success that we can refer to when summarizing the experiment.
Problem solving vs. Continuous improvement
As previously mentioned, the PDCA cycle is a method that applies to various contexts related to improving quality and performance through effective problem-solving. It is worth emphasizing, that PDCA can be applied in two distinct scenarios:
One: using it as a component of a comprehensive problem-solving session aimed at intense, moderated short-term resolution of a specific issue. Second: is connected to the broader initiative regarding #Kaizen philosophy and the culture of #Continuous improvement, which focuses on the ongoing search of both minor and major problems and resolving them endlessly, with a non-project-based approach. This approach relies on the cyclical observation of processes and the assessment of their characteristics using well-designed indicators that reflect both quality and efficiency. The range of indicators is vast, and their proper selection should be closely aligned with strategic goals and critical success factors. However, a detailed discussion of this topic would be more appropriate for a separate article.Returning to the main topic: Objectives that reflect the condition of the process should be the trigger for further improvements. We evaluate the indicators, and if we are unable to achieve them, it should initiate a further cycle of analysis, often summarized as the: Objectives – Results – Improvements cycle. The purpose of the PDCA method is to provide a structured approach to managing corrective actions in a controlled environment. As you might expect, making decisions about changes that have a direct or indirect impact on the business requires a thoughtful and analytical approach. Determining the problem itself is not enough to discover an infallible solution. It requires much more research than that. With only process maps and SME (Subject Matter Expert) knowledge, it is impossible to develop a perfect theory within the confines of a meeting room where the analysis takes place.The key issue is not the lack of data or appropriate knowledge. The challenge lies in the immense complexity of problems and the potential impact of solutions. Even seemingly simple processes or their components are directly or indirectly interconnected and affect other processes and subprocesses that coexist around them. Therefore, the safest approach is to identify potential solutions, minimize the margin of error through thorough analysis, and proceed with the experiment in a controlled environment.
Controlled experimentation environment: Sandbox concept
In IT terminology, a controlled environment for conducting experiments is often referred to as a “sandbox,” and I will be using this term in the context of problem-solving as well. A sandbox is typically a copy of the operational environment that is isolated from the overall infrastructure, allowing changes to affect only the designated copy without impacting the surrounding systems. Our task when creating a space for PDCA experiments is to, as much as possible, organize similar conditions—a sort of testing ground for our solutions. The difference is that we are often unable to arrange a synthetic process or area solely for testing purposes, as this would be inadequate compared to the actual #Gemba experience, leading to potentially false results. Additionally, setting up such an environment would consume disproportionate resources and time. The solution in this case is to test the changes in an isolated fragment of the real process or a separated process variation.
For example, let’s assume there is a problem in a client invoicing process. We cannot afford to change the entire process overnight and expose ourselves to unexpected external and internal consequences. As previously mentioned, it is impossible to assess with complete accuracy whether a solution will work and what additional effects it might have on concurrent tasks and processes. Instead, we can isolate a small segment and decide that 1 out of our 100 clients will be serviced based on the new process variation for a testing period of time. While all the other clients will be continuously and secured with the former process. This approach offers two key benefits:
- Minimized impact of unsuccessful changes: If the changes are ineffective or produce unexpected negative effects on the direct process or other concurrent processes, the impact is limited and can be quickly corrected, returning to the pre-experiment standard.
- Comparative analysis of new vs. old processes: By having results from the area managed by the old process, we can compare them with results from the area managed by the new process. This allows us to easily evaluate whether the new solution is effective by comparing key KPIs such as timeliness, error rates, process efficiency, lead time etc.
Achieving the best results with the deming cycle
Now that we understand the context, know what to avoid, and recognize the benefits of correctly applying the method, it is crucial to define the conditions under which the PDCA cycle can yield the best results. It is important to remember that the Deming Cycle on its own will not resolve the problem. Paradoxically, it might even escalate the problem by addressing non-existent issues in a highly professional manner, thereby creating new problems. Let me explain: The problem-solving method is just one of the final stages in a multi-step process. To maximize the effectiveness and precision of PDCA, it should ideally be the last step in an eight-step process for analyzing problems and finding solutions.
The preceding stages include:
In-depth understanding of the problem and its characteristics
Identifying and selecting key causes
Analyzing the cost of resources required for implementing proposed solutions and comparing that with the potential benefits #effort vs. effect matrix
Without delving too deeply into the specifics, I will move directly to the final, eighth step. For those interested in professionally conducting problem-solving sessions and establishing an effective culture of continuous improvement, I invite you to our training sessions on this topic, click button below to explore.
The Deming’s PDCA: Summary
PDCA is an acronym representing four stages:
PLAN – At this stage, we must establish five key elements:
- What is the problem – and what quantified damage does it cause
- What we intend to change and in which scope
- What are the key criteria by which we will evaluate the experiment
- Who is responsible for particular change cycle
- When the experiment will start and end
DO – The acronym has evolved from PDCA to PTCA nowadays, where the T stands for TRY. This stage represents the core of the method: to implement ideas and test solutions that have the potential to improve the current state before deciding whether the solution is effective and will ensure the desired outcome.
CHECK – In this stage, we analyze the impact of the implemented solution and compare it to the expectations defined at the PLAN stage through the use of indicators that reflect the condition of the process or issue. We can also compare the results from two parallel standards—the old and the new—to clearly determine if our new initiative is effective. It is crucial to have established criteria beforehand to base the evaluation on facts rather than opinions.
ACT – The final stage is a decision-making step where we must choose between two options. If we have clearly established that the solution is effective and there are no unforeseen repercussions on other concurrent processes, we can declare the experiment a success and proceed to standardize the new solution, which means implementing it in a broader perspective outside the isolated environment. However, if the solution fails to meet expectations or negatively impacts the surrounding process environment, we must return to the beginning of the cycle, start planning, and methodically work through all the stages of PDCA again until the desired effect is achieved.
Thank you for reading this article.
For more insights on effective management, explore other articles in our Free Learning Hub. Have questions or want to develop your team’s skills? Contact us via the Contact Form or reach out to me directly on Linkedin.
All quotes used in this article are from the following books:
Out of the Crisis by William Edwards Deming
