PDCA as a Daily Habit: Common Mistakes and How to Avoid Them

Plan-Do-Check-Act (PDCA), also known as the Deming cycle or the Shewhart cycle, is a four-step improvement method: Plan (identify a problem and develop a hypothesis about its cause and a proposed solution), Do (implement the solution on a small scale), Check (measure whether the solution produced the expected result), Act (standardise the successful solution or return to Plan with new learning). PDCA is a foundational tool in Lean and continuous improvement, but most organisations that adopt it make predictable mistakes that prevent it from becoming a genuine improvement practice.

Mistake 1: Skipping the Plan Step

The most common PDCA mistake is jumping directly to Do -- implementing a solution before developing a clear hypothesis about the root cause and the expected effect of the solution. Without a clear hypothesis, the Check step has nothing to verify. If the solution does not produce the expected result, there is no learning about why, because the expected result was never defined. Every PDCA cycle should begin with a written hypothesis: 'We believe [the root cause is X]. If we implement [change Y], we expect [result Z]. We will measure [metric M] over [time period T] to verify this.'

Mistake 2: Doing at Scale Instead of Piloting

The Do step in PDCA is a pilot -- a small-scale test of the hypothesis, designed to be reversible if the hypothesis is wrong. Most organisations skip the pilot and implement the solution fully. If the full-scale implementation does not produce the expected result, rolling it back is expensive and disruptive. If the small-scale pilot does not produce the expected result, it costs little to return to the Plan step with the new learning.

Mistake 3: Checking the Wrong Things

The Check step requires measuring the right things, at the right time, to determine whether the hypothesis was correct. Common mistakes include: measuring outputs (the work was done) rather than outcomes (the problem was reduced); measuring too soon (before the change has had time to affect the process); and not establishing a pre-implementation baseline to compare against. Before implementing any PDCA change, record the current-state metric value. If you do not know what the metric was before the change, you cannot determine whether the change made it better.

Mistake 4: Act Means Standardise OR Learn -- Not Always Standardise

The Act step is often misunderstood as 'standardise the solution.' It should be understood as 'decide what to do with what you learned.' If the hypothesis was correct and the expected result was achieved, standardise the solution. If the hypothesis was wrong or the result was not achieved, return to the Plan step with the learning from the Check. PDCA is a learning cycle, not a project completion cycle. A PDCA cycle that ends in 'the result was not what we expected' is a successful cycle -- because the team learned something.

How to Make PDCA a Daily Habit

• Use visual management to make PDCA visible. A simple board with four columns (Plan, Do, Check, Act) where each active PDCA cycle is represented by a card makes improvement work visible and creates accountability for completing each step.

• Start with small, contained problems. A PDCA cycle that takes three months to complete does not build the habit. Start with problems that can be completed in a week or two, and build the muscle before applying PDCA to larger problems.

• Celebrate learning, not just success. If the Check step shows the hypothesis was wrong, that is learning. Teams that are punished for failed PDCA cycles will stop running PDCA cycles.

XNM applies PDCA and Lean continuous improvement methodology to public-sector and capital-project environments. Reach out to XNM's strategic advisory team to discuss continuous improvement and PDCA implementation for your organisation.