The PDCA cycle is a four-step loop for improving anything: Plan a change, Do it on a small scale, Check whether it worked, and Act on what you learned. Then you go round again. It is the method underneath continuous improvement, kaizen and lean, and the reason it lasts is that it is hard to argue with: try something small, look at what happened, then decide.
Most teams know the four words. Fewer use them well. The usual problem is not the cycle itself, but skipping the parts that feel slow: a clear problem in Plan, and an honest look at the numbers in Check.
What is the PDCA cycle?
PDCA stands for Plan, Do, Check, Act. It is a method for testing a change before you roll it out, so that decisions rest on what actually happened rather than on what people expected to happen. It is also called the Deming cycle, the Deming wheel or the Shewhart cycle, after the two quality pioneers most linked with it.
The cycle is meant to repeat. Each round either confirms a change, which becomes the new standard, or teaches you something that shapes the next plan.
What are the four steps of PDCA?
| Step | What you do | Questions to answer | Typical tools |
|---|---|---|---|
| Plan | Describe the problem, find the cause, choose a change and decide how you will measure it | What is the problem, in numbers? What causes it? What will we change? What result do we expect? | Problem statement, 5 whys, fishbone diagram, baseline data |
| Do | Try the change on a small scale: one line, one shift, one team, one week | Did we do what we planned? What surprised us? | A short trial, a log of what happened |
| Check | Compare the result with the baseline and with what you expected | Did the numbers move? Why, or why not? What did we learn? | Before-and-after chart, run chart, a short review |
| Act | Decide: make it the standard, adjust and try again, or drop it | Do we roll it out, change it, or stop? What is the next problem? | Updated standard work, training, a new plan |
Two things make the difference between PDCA and simply trying things. First, Plan includes a prediction: you write down what you expect to happen before you start. Second, Check compares the result with that prediction. Without both, there is nothing to learn from.
What does PDCA look like in practice?
Here is a typical example from a factory floor, simplified.
Plan. Changeovers on packing line 2 take 48 minutes on average over the last month, against 30 on line 1. An operator suggests that most of the time goes on looking for the right tools and settings. The team watches three changeovers and confirms it: about 15 minutes each time is spent fetching parts and checking settings. The change: a shadow board with the tools for line 2, and a one-page settings card for each product. The prediction: changeovers under 35 minutes.
Do. The team tries it on the day shift for two weeks and logs every changeover, including what went wrong.
Check. Day-shift changeovers average 34 minutes. Two still took over 45, both on the product with no settings card yet. The idea works; the cards need to cover every product.
Act. The team finishes the missing cards, rolls the board and cards out to all shifts, and adds them to the line’s standard work. The next round looks at the 34 minutes that are left.
Notice that the idea came from an operator. PDCA works best when the people closest to the work bring the problems and the first ideas, which is why it sits so well with a system for employee ideas.
Where does the PDCA cycle come from?
The cycle grew out of the scientific method applied to quality.
- Walter Shewhart, a statistician at Bell Telephone Laboratories, described a cycle of specification, production and inspection in his 1939 book Statistical Method from the Viewpoint of Quality Control.
- W. Edwards Deming took the idea to Japan. In his 1950 lectures to Japanese engineers and managers, organised by JUSE, the Union of Japanese Scientists and Engineers, he presented it as a wheel of design, production, sales and research.
- Japanese managers turned Deming’s wheel into Plan, Do, Check, Act. According to Ronald Moen and Clifford Norman, who traced its history, no single person claims to have written that version.
- Deming himself later preferred PDSA, Plan, Do, Study, Act. He felt “check” sounded like inspection, when the point is to study what happened and learn from it.
Today PDCA is part of many management standards. ISO 9001, the quality management standard, describes its approach in terms of the PDCA cycle, and so do other ISO management system standards such as ISO 14001 for environmental management.
What is the difference between PDCA, PDSA, A3, DMAIC and kaizen?
| Method | What it is | When to use it |
|---|---|---|
| PDCA | The basic four-step improvement loop | Any improvement, small or large |
| PDSA | The same loop, with Study instead of Check to stress learning | Common in healthcare and in teams that follow Deming closely |
| A3 | A one-page report that walks through a PDCA cycle: background, current state, cause, countermeasures, plan, follow-up | Problems worth writing down and sharing, often with a coach |
| DMAIC | Define, Measure, Analyse, Improve, Control: the Six Sigma version, heavier on statistics | Complex problems with lots of data and variation |
| Kaizen | The philosophy of small, continuous improvement by everyone, every day | The culture PDCA runs inside; see what kaizen is |
In short: kaizen is the attitude, PDCA is the method, A3 is a way to write it down, and DMAIC is the heavier version for problems that need it.
When should you use PDCA?
PDCA is the right tool when:
- you want to test a change before rolling it out. A trial on one line costs much less than a failed rollout on ten.
- the cause of a problem is not obvious. Each round narrows it down.
- you want to improve the same process step by step. Each Act becomes the next Plan.
- an idea from an employee needs testing. PDCA gives a fair, quick way to try it and decide.
It is less useful for a decision that cannot be tested small, like choosing a new site, or for an emergency, where you fix first and analyse afterwards.
What are the most common PDCA mistakes?
- Starting at Do. Teams jump to a solution without describing the problem or measuring the starting point. Then Check has nothing to compare with.
- No prediction. If you did not write down what you expected, any result can be called a success.
- Testing too big. A trial across the whole plant is a rollout, not a test. Start with one line, one shift, one week.
- Skipping Check. The trial ends, people are busy, and nobody looks at the numbers. This is the step most often dropped.
- No Act. A successful trial that never becomes the standard fades within weeks. Write it into standard work and training.
- One round only. PDCA is a cycle. The end of one round is the start of the next.
A simple PDCA template
A PDCA cycle fits on one page. These are the fields most teams use:
| Field | What to write |
|---|---|
| Problem | What is wrong, where, and since when, with a number |
| Baseline | The current figure, measured the same way you will measure the result |
| Cause | What you found when you asked why, with evidence |
| Change | What you will try, where, and for how long |
| Prediction | The result you expect, in numbers |
| Trial log | What you did and what happened, day by day |
| Result | The new figure, compared with the baseline and the prediction |
| What we learned | Why it worked or did not |
| Decision | Standardise, adjust and repeat, or stop |
| Next step | The next problem, or the next round |
Our idea scoring scorecard helps you decide which ideas are worth a PDCA cycle in the first place.
How do employee ideas fit into PDCA?
Most PDCA cycles start with someone noticing a problem or suggesting a fix. In many organisations those ideas get lost: written on a whiteboard, mentioned in a meeting, never tested.
In Hives, each idea moves through steps you choose, so a collection for improvement ideas can follow the cycle: a step to plan and measure, a step for the trial, a step to check the result and a step to decide. Everyone who submitted or followed the idea sees where it is and what was decided, which keeps people suggesting the next one. See what a workflow is in our help centre.
Frequently asked questions
What does PDCA stand for?
PDCA stands for Plan, Do, Check, Act. You plan a change, try it on a small scale, check the result against what you expected, and act on what you learned by standardising, adjusting or dropping the change.
Who invented the PDCA cycle?
The cycle grew from Walter Shewhart’s work at Bell Labs in the 1930s and W. Edwards Deming’s lectures in Japan in 1950. Japanese managers turned Deming’s wheel into Plan, Do, Check, Act, and no single person claims that version. Deming himself later preferred PDSA, Plan, Do, Study, Act.
What is the difference between PDCA and PDSA?
They are the same four-step loop. PDSA replaces Check with Study, which Deming preferred because the point is to understand why a result happened, not just to inspect it. PDSA is common in healthcare; PDCA is more common in manufacturing and in ISO standards.
Is PDCA the same as kaizen?
No. Kaizen is the philosophy of continuous, small improvements by everyone. PDCA is the method many teams use to carry out each improvement. Kaizen describes the culture, and PDCA describes the steps.
How long does a PDCA cycle take?
As long as the problem needs. A simple fix on one workstation can go round in a few days. A change to a process across several shifts often takes four to eight weeks, mostly because the trial needs enough time to give reliable numbers.
Can PDCA be used outside manufacturing?
Yes. It is used in healthcare, offices, schools, software teams and public services, anywhere a team wants to test a change before rolling it out. The steps are the same; only the measures change.
Sources
- Walter A. Shewhart, Statistical Method from the Viewpoint of Quality Control, Graduate School of the Department of Agriculture, Washington, 1939
- Ronald D. Moen and Clifford L. Norman, “Circling Back: Clearing up myths about the Deming cycle and seeing how it keeps evolving”, Quality Progress, November 2010
- W. Edwards Deming, The New Economics for Industry, Government, Education, MIT, 1993
- ISO, ISO 9001:2015 Quality management systems, Requirements