One thing I learned at Toyota is that the most important part of problem-solving happens before anyone proposes a solution.
Yet I still see companies choosing improvement methodologies before fully understanding what’s going wrong.
If you’re unsure where to start, my free Process Optimization Audit Checklist can help you identify where your processes need attention.
This week’s article explores why Six Sigma and PDCA were developed to address different problems and what that means for how we approach improvement today.
Six Sigma and PDCA Were Never Designed to Solve the Same Problems
Six Sigma and PDCA are often put in the same category. Both are associated with process improvement. Both use data. Both aim to improve quality and performance. And both are frequently taught as methodologies a company can adopt to solve operational problems.
That makes it easy to assume they’re simply two different ways of accomplishing the same thing. They’re not.
Six Sigma and PDCA emerged at different times, from different circumstances, with different problems at the center of their development.
That doesn’t make one better than the other. But it does raise a question I think more companies should ask before choosing any improvement methodology:
What kind of problem are we actually trying to solve?
Because one of the easiest mistakes to make in operational improvement is choosing the methodology before understanding the problem.
PDCA Came Long Before Six Sigma
PDCA is frequently called the “Deming Cycle,” which leads many people to assume W. Edwards Deming invented it.
The history is more complicated.
Walter Shewhart was developing the thinking behind an iterative improvement cycle decades before Six Sigma existed. ASQ traces the concept to Shewhart’s work and his 1939 book Statistical Method From the Viewpoint of Quality Control. Deming later brought Shewhart’s ideas to Japan, where they became highly influential in the development of Japanese quality management.
Here’s the interesting part: Deming apparently didn’t call it PDCA.
An account published by the American Society for Quality notes that Deming never used the phrase “PDCA” during his lectures in Japan, even though he introduced Shewhart’s methods to Japanese audiences. The familiar Plan-Do-Check-Act formulation developed through the evolution of those ideas.
The important part isn’t who gets credit for the acronym. It’s the thinking underneath it.
PDCA is fundamentally iterative. You establish a plan based on what you understand about the problem, implement the change, examine what happened, and use what you learned to determine what happens next. Then you repeat the cycle.
That’s why I have always viewed PDCA as much more than a four-step diagram.
When I worked at Toyota, the emphasis wasn’t simply on getting to a solution quickly.
We spent significant time understanding the current condition before changing it. I’ve written before about why the planning portion of PDCA matters so much and how we approached PDCA in service-based process transformation.
The purpose is not to prove that your original idea was correct. It’s to learn.
Six Sigma Started With a Different Problem
Six Sigma came decades later. Motorola developed Six Sigma in 1986 in response to quality challenges and the need for a common way to measure and improve process performance. Motorola describes Six Sigma as a quality-improvement process that became a common language for measuring quality.
Its development placed much greater emphasis on variation, defects, process capability, measurement, and statistical analysis.
One number became particularly associated with Six Sigma: 3.4 defects per million opportunities. ASQ’s history of Motorola’s Six Sigma program describes how statistical methods were used to pursue that level of quality by controlling variation.
That’s a very specific way of looking at a problem.
Suppose a company processes thousands of transactions every month. The process is established and measurable, but the output varies. Some transactions are correct. Others contain errors. Some take two days. Others take seven. The organization has enough data to examine that variation statistically.
That’s exactly the kind of environment where Six Sigma thinking becomes powerful.
It asks you to quantify performance, understand variation, identify the factors contributing to defects, improve the process, and control it so the gains hold.
PDCA can certainly use data, and Six Sigma involves experimentation and learning. The two aren’t opposites.
But they don’t start from exactly the same place.
Not Every Operational Problem Is a Variation Problem
This is where I see companies get into trouble. Once an organization becomes trained in a methodology, the methodology can start determining how it sees problems.
A Six Sigma organization sees a Six Sigma project.
A Lean organization sees waste.
A company trained in root-cause analysis starts looking for root causes.
A PDCA organization sees something to cycle through.
Eventually, every problem starts looking like the framework the organization already knows.
I’ve written before about the differences between Kaizen, Lean and Six Sigma because these approaches are often grouped together even though they focus attention on different aspects of improvement.
But there’s an even more basic question that should come first:
What do we actually know about the problem?
Imagine customer onboarding takes too long.
Before choosing a methodology, I want to know what “too long” actually means.
- Is the process consistently taking 14 days when the business needs it to take seven?
- Is the average seven days, but some customers take two days and others take 21?
- Do we even know how long it takes because nobody has mapped the process from beginning to end?
- Is the delay caused by variation inside one step, or does the work spend most of its time waiting between departments?
- Does Sales send incomplete information to Operations?
- Are people following different processes?
- Or are we assuming onboarding is the problem when the actual constraint occurs somewhere else?
Those are very different operational conditions. They shouldn’t automatically lead to the same improvement approach.
This Is What PDCA Changed for Me at Toyota
One of the most important things I learned at Toyota was not to rush from problem to solution. That sounds simple. In practice, most organizations do it constantly.
A customer complains, so we add a checkpoint.
A deadline gets missed, so we add a meeting.
Someone makes an error, so we add an approval.
A process is slow, so we add another person.
We experience a symptom and immediately start discussing solutions.
PDCA forces a different discipline.
What is happening now? What should be happening? Where is the gap? What do we know? What are we assuming? What countermeasure are we testing? And what happened after we tested it?
The Plan stage is particularly important. In my work on the process improvement cycle, I’ve explained why so much attention goes into understanding and planning before implementation. I’ve also written about how we used the PDCA process in service-based businesses to identify problems and develop countermeasures.
The value isn’t the acronym.
It’s the discipline of learning before standardizing a solution.
Six Sigma Brings a Different Kind of Discipline
Six Sigma asks for discipline too, but often in a different form.
If the problem is measurable variation in a repeatable process, intuition alone isn’t enough. You need to understand what the data is telling you.
That is why Six Sigma developed such a strong statistical foundation.
The methodology eventually expanded far beyond Motorola. Companies including AlliedSignal and General Electric adopted and developed it further, and Six Sigma moved beyond manufacturing into service and other business processes. ASQ describes this evolution as Six Sigma moving from its initial quality focus toward broader organizational and financial improvement.
This matters because neither methodology should be frozen in its original historical context.
PDCA isn’t only for Toyota or manufacturing. Six Sigma isn’t only for factories measuring physical defects. Both have been adapted extensively.
But understanding where they came from helps us understand the kind of thinking each one brings to a problem.
The Mistake Is Falling in Love With the Methodology
I saw something similar happen with Lean.
As I explained in my article about the origin of Lean manufacturing, Toyota developed the management system that inspired what the world eventually called Lean, but inside Toyota we weren’t walking around saying, “Let’s do Lean.” We were solving problems. That distinction matters here too.
Companies love methodologies because methodologies create structure. They give us language, tools, certifications, steps, and something tangible to implement.
But the methodology should serve the problem.
The problem should not be forced to serve the methodology. Sometimes you need rigorous statistical analysis to understand why a process is producing unacceptable variation.
Sometimes you need to test a countermeasure quickly and learn from what happens.
Sometimes you need to map the work because nobody actually understands the end-to-end process.
Sometimes you don’t have an improvement problem at all. You have an ownership problem, an information problem, or a lack of clarity about what the standard is supposed to be.
And sometimes the problem is complex enough that you will draw from more than one approach.
Start With the Problem
I don’t think leaders need to become experts in every improvement methodology.
They do need to become better at understanding problems before deciding how to solve them.
That’s the lesson I would take from the history of Six Sigma and PDCA.
Six Sigma emerged from Motorola’s pursuit of dramatically better quality through understanding and controlling variation. PDCA evolved from a much older tradition of iterative experimentation, learning, and improvement.
Both became much broader than their origins. Both can be powerful. And both can be misused when the framework becomes more important than the problem.
So the next time someone asks whether your organization should use Six Sigma, PDCA, Lean, Kaizen, or another methodology, I wouldn’t start by debating the tools.
I’d start somewhere simpler.
What problem are we actually trying to solve?
Then ask what you need to learn about it.
Because the discipline isn’t knowing which improvement methodology your company uses.
It’s understanding the problem well enough to know what kind of problem-solving it requires.
Want to Work With Us?
You can’t improve a process you don’t fully understand.
In my one-day virtual FlowState Workshop, we’ll map how work actually moves across your business, uncover hidden bottlenecks, and identify where improvement will have the greatest impact.
Before investing in another solution, make sure you’re solving the right problem.
Explore the FlowState Workshop →






