Knowledge Base · Solutions
What is the PDCA cycle, and how do I run one?
PDCA is a repeating four-step loop for testing change: Plan a change and predict its result, Do it on a small scale, Check what actually happened against the prediction, then Act to standardise it or Adjust and run the loop again. It turns improvement into small, honest experiments instead of one big launch.
Chapter 4 calls PDCA (Plan, Do, Check, Act) the engine of continuous improvement: a simple, repeatable loop that keeps work grounded in reality, and the antidote to a plan, do, done habit that leaves results unexamined.
- Plan: name the problem, form a hypothesis, and plan the test: what you will do, who will do it, and how success will be measured.
- Do: carry out the plan, on a small scale first where the change is large: a pilot or limited trial.
- Check: measure the result honestly against what you predicted, and capture what the team learned.
- Act: if it worked, standardise it and roll it out wider. If it partly worked or failed, adjust the plan and cycle again.
The book's example is an order entry checklist, planned to cut errors by 25 percent. After two weeks errors had fallen by 15 percent, and the team found one section of the checklist confusing. That is not failure in PDCA terms. The team adjusted the checklist and ran the test again toward a 20 percent goal, which is exactly how the loop is meant to learn.
“It’s the scientific method applied to our daily work, and it’s the engine that powers real, sustainable change.”
Continuous Improvement in the Workplace, Ch. 4
Track a cycle in the free PDCA Tracker
Related questions
- How do I pilot a solution before rolling it out?
- How do I run a small improvement experiment this week?
- What is DMAIC, and when should I use it?
Source: Continuous Improvement in the Workplace, Ch. 4 (Solution Creation and Implementation) and the CI glossary in the book's Appendix.