Large improvements often begin as small tests. The Plan–Do–Check–Act cycle describes a loop: specify a problem and an expected result, test the change on a manageable scale, examine the evidence and decide whether to adopt, revise or abandon the approach.
The core idea
ASQ describes PDCA as a four-step model for continuous improvement. It is most useful when teams are explicit about what they expect to improve and what they will measure. Without a comparison or a realistic feedback period, “check” becomes an impression rather than evidence.
How to apply it
- Plan: Define the problem, baseline, success measure and hypothesis for one change.
- Do: Test the intervention in a small, controlled setting where risks are acceptable.
- Check: Compare actual results with the baseline; include unintended effects and feedback from people affected.
- Act: Standardize successful practice, revise the approach or stop an ineffective change.
- Repeat: Set a new question for the next cycle and keep a short record of the decisions made.
A workplace example
A customer service team sees repeat complaints about incomplete responses. Rather than redesigning all operations, it tests a one-page response checklist with one team for two weeks. It reviews customer follow-ups and response time, corrects weak checklist items and decides whether wider adoption is justified.
What to watch out for
PDCA is not a substitute for formal engineering validation, safety approvals or controlled experiments when these are necessary. Avoid using only positive examples in the Check stage. Record unintended costs and make sure changes do not shift a burden to another team.
Practice and reflect
Choose a recurring irritation and design a two-week test with one measurable outcome.