The Deming cycle: plan–do–check–act

Plan–do–check–act (PDCA), often called the Deming cycle, structures improvement as repeated experiments. A team plans a change and its expected result, carries it out, studies the evidence and then standardises, adapts or abandons the change before beginning the next cycle.

When to use it

Use PDCA to bring discipline to continuous improvement, problem-solving, training, objectives and control processes. The method distinguishes active management from repeated firefighting: each cycle makes the theory, action and learning explicit.

Teach the method to everyone expected to lead improvement. Visible results from small cycles build confidence and encourage further work, while the repeated loop prevents a successful change from becoming a one-off event.

Origins

The cycle grew from Walter A. Shewhart’s statistical approach to quality control, which treated specification, production and inspection as an iterative learning process. W. Edwards Deming taught and popularised that experimental logic in post-war Japan. Deming later preferred plan–do–study–act, because ‘study’ stresses learning from evidence rather than merely checking whether a task was completed.

What it is

How to use it

Work through every stage and keep the scale of the test proportionate to the uncertainty and risk.

1. Plan

Describe the current condition, the problem and the higher-purpose objective. Develop a theory about what change may improve the result and predict what should happen. Define measures before acting, identify affected people and locations, specify the procedure, timing and responsibilities, and establish any safety or stopping rules.

Useful questions include:

• What are we trying to achieve, and why does it matter?
• Who will be affected, where and when?
• What exact sequence will be tested?
• What result do we predict?
• Which evidence would show improvement or harm?

2. Do

Run the test on a controlled, suitably small scale. Follow the plan closely enough to interpret the result, record deviations and observations, and protect participants from avoidable risk.

3. Check

Compare the observed result with the prediction. Determine not only whether the target was reached but also what explains the difference, what unintended effects appeared and what the evidence says about the original theory.

4. Act

If the change works reliably, incorporate it into standard practice and plan how to scale it. If it does not, preserve the learning, revise the theory and begin another cycle. Remove practices shown not to contribute and update objectives when the evidence justifies doing so.

Final analysis

Many organisations struggle to state objectives, activities and expected results clearly. PDCA cannot supply that discipline automatically. Teams often stop after plan and do, then move immediately to another action without studying the outcome.

Adaptations may divide planning into goals and methods, or doing into education and implementation. The labels matter less than preserving the experimental loop. PDCA is closely aligned with kaizen: repeated, evidence-based learning turns local changes into cumulative improvement.

Top practical tip

Test a clear prediction through a small, controlled change so the observed improvement or failure can be attributed and learned from.

Top pitfall

Do not reduce ‘check’ to final inspection or PDCA to paperwork. Study the evidence and revise the theory before beginning again.

Further reading

Deming, W.E. (1986) Out of the Crisis. Cambridge, MA: MIT Press.