KPIs, strategy & continuous improvement
The PDCA cycle: Plan, Do, Check, Act with practical examples
How to use the PDCA (Plan-Do-Check-Act) cycle for continuous improvement in offices, stores and plants, with examples, templates and common mistakes.
The PDCA cycle (Plan, Do, Check, Act), also called the Deming cycle, is the simplest structured way to improve a process. You plan a change based on a hypothesis, do it on a small scale, check whether it worked using data, and act by standardizing the change or trying something else. Repeated over time, small PDCA cycles add up to significant improvement.
The four steps
| Step | What happens | Output |
|---|---|---|
| Plan | Define the problem, current state, target, root cause hypothesis and the change to test | A clear test plan with a measure |
| Do | Run the change on a small scale (one branch, shift or product) | Data from the test |
| Check | Compare results with the target and baseline; learn what happened | Evidence: worked, partly worked or failed |
| Act | Standardize if it worked; adjust and repeat if not | New standard or next cycle |
Example 1: payroll corrections
Plan. Payroll corrections after release average 45 per cutoff. Most come from late overtime approvals. Hypothesis: reminders to supervisors two days before cutoff will reduce late approvals. Target: under 20 corrections.
Do. Pilot reminders in two departments for two cutoffs.
Check. Corrections in the pilot departments drop by 60%; other departments unchanged.
Act. Roll out reminders company-wide and add an exception list for HR. Next cycle: tackle missing time logs. See payroll cutoff best practices.
Example 2: store receiving errors
Plan. Branch receiving errors cause stock discrepancies. Hypothesis: blind counts at receiving will reduce errors.
Do. Pilot blind receiving in three branches for a month.
Check. Discrepancies in pilot branches fall from 6% to 2% of lines received.
Act. Update the receiving procedure for all branches and train staff (goods receiving).
A one-page PDCA template
| Section | Content |
|---|---|
| Problem | What is wrong, with numbers |
| Current state | Baseline data |
| Target | Measurable goal and date |
| Root cause hypothesis | Why we think it happens |
| Countermeasure | What we will change |
| Test plan | Where, when, who, how long |
| Results | Data after the test |
| Next action | Standardize, adjust or abandon |
PDCA vs. DMAIC
PDCA is lighter and faster, suited to many small improvements by frontline teams. DMAIC adds more rigor for complex, recurring problems with unclear causes. Many companies use PDCA as the everyday habit and DMAIC for a few major projects.
Common mistakes
- Skipping Check: declaring success without data
- Big-bang Do: rolling out to all branches before testing
- No Act: improvements work in the pilot but are never standardized
- Vague Plan: no baseline or measurable target
Building a PDCA habit
- Make PDCA the standard format for improvement ideas from teams
- Review PDCA boards in daily or weekly meetings
- Recognize completed cycles, including well-documented failures
- Store completed PDCAs so other branches can copy successful changes
See Kaizen and lean manufacturing for related approaches.
Where Cirquolus fits
The Cirquolus improvement portfolio includes PDCA, CAPA and Kaizen workspaces alongside DMAIC, DMADV, Lean, FMEA and project management, linked to scorecards and operational reviews.
Frequently asked questions
What does PDCA stand for?
Plan, Do, Check, Act.
Who created the PDCA cycle?
It is associated with W. Edwards Deming, building on earlier work by Walter Shewhart, and is often called the Deming cycle.
How long is a PDCA cycle?
It can be days or weeks. Short cycles allow faster learning.
When should PDCA be used instead of DMAIC?
For smaller, faster improvements where a simple test can show whether a change works.
Bring this workflow to a Cirquolus demo.
Show us your current process, companies and branches. We will map it to the modules you need and propose a realistic rollout.