What it means
Improvement efforts fail most often for a simple reason: someone rolled out a big change on a hunch and nobody measured what happened. PDCA exists to break that habit.
Plan means defining the problem, analysing causes, and designing a specific change with a predicted result, and vague plans produce vague outcomes, so the prediction matters. Do means trying the change small: one line, one branch, one team.
Small trials keep the cost of being wrong tiny and produce real data rather than conference-room opinions. Check means comparing the results against the prediction, and this step is the discipline's core, because skipping it converts PDCA into ordinary trial and error with extra paperwork.
Act means deciding: if the change worked, standardise it and train everyone; if it failed, adjust the plan and run the loop again. Either outcome feeds the next cycle.
The loop also scales, since a shop-floor team may cycle in days on a tooling tweak while a division cycles quarterly on a pricing process, and the discipline is identical at both speeds. The method descends from Walter Shewhart's work at Bell Labs and was carried to Japan by W.
Edwards Deming, which is why it is also called the Deming cycle or the Shewhart cycle. The American Society for Quality describes PDCA as the operating loop behind continuous improvement programmes from manufacturing floors to hospitals, and it underpins quality systems like ISO 9001.
Software teams will recognise the pattern in agile iterations and hypothesis-driven experiments, since modern product development is PDCA wearing new clothes, with the same insistence on small bets and measured outcomes. Its power is cultural as much as technical.
Teams that run honest cycles learn to distrust unmeasured assertions, and the loop turns improvement from an annual project into a daily habit. For a non-finance manager, PDCA is the antidote to the grand rollout: pick one metric, change one thing, measure honestly, then scale what actually worked.
In practice that means writing the prediction down before the trial starts and comparing the result with it afterwards. Without that comparison, a team has activity, not evidence.
In practice
Real-world examples.
Example
A hospital ward trials a new handover checklist on one shift for a week and measures medication errors. It standardises the checklist across all shifts after errors drop by half.
Example
A factory tests a modified jig on a single production cell, confirms defect rates fall, then rolls it across all lines with updated training. The written standard and retraining are what make the gain permanent.
Example
A call centre tests a new script with five agents and finds resolution rates unchanged. It discards the script in the Act step, saving the cost of a full rollout and keeping the team's attention for a better idea.
Formula
Calculation
There is no formula, only the loop: Plan (define and predict), Do (small trial), Check (compare results with prediction), Act (standardise or adjust), then repeat. Each full rotation should leave the process measurably better or the plan measurably wiser.
The Check step is where simple arithmetic belongs. In a fictional trial, a route produced 200 late parcels a week before the change and 90 after it, so the fall is (200 - 90) / 200 x 100 = 55%. If weekly overtime on that route was $10,000 before the change and rose 12%, the extra cost is $10,000 x 12% = $1,200 a week, which the Act step must weigh against the improvement.Case study
Seen in the real world.
This case study is fictional and illustrative. The made-up operations manager of a courier company in Warsaw faces rising customer complaints about late evening deliveries. Instead of overhauling the whole network, she runs PDCA on one depot. Plan: data shows 40 percent of late parcels come from a single overloaded route, so she predicts splitting it will cut late deliveries by half.
Do: the route is split for two weeks at that one depot. Check: late deliveries on the route fall 55 percent, but driver overtime rises 12 percent. Act: she keeps the split, adjusts start times to absorb the overtime, and schedules the next cycle to tackle the second-worst route. Six months of rolling cycles cut network-wide complaints by a third, with no heroic transformation programme and no consultants.
The manager also records each cycle's prediction and result on a single page that any depot can read. Other depots copy the pattern for their own worst routes, and the page becomes the company's evidence of what works. The courier, the figures and the outcome are invented for illustration.
Watch out
Common mistakes.
- Skipping the Check step, which turns PDCA into plan-do-plan-do with no learning and no evidence the changes help.
- Running the Do step at full scale, which converts a controlled experiment into an uncontrolled bet on the whole operation. The trial's small scale is the feature, not a compromise.
- Treating Act as the end; the cycle is meant to repeat, and teams that stop after one loop miss most of the available improvement.
Questions
People also ask.
What does PDCA stand for?
Plan, Do, Check, Act: design a change with a prediction, trial it small, compare results with the prediction, then standardise or adjust and repeat.
Who created it?
It derives from Walter Shewhart's cycle at Bell Labs and was popularised in Japan by W. Edwards Deming, hence the names Shewhart cycle and Deming cycle.
Where is PDCA used?
Across manufacturing, healthcare, services, and software, and it underlies quality management standards such as ISO 9001's process approach.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%Related