The PDCA cycle — Plan, Do, Check, Act — is the simplest structured framework for continuous improvement. First described by Walter Shewhart and popularized by W. Edwards Deming in post-war Japan, it underlies most modern quality management systems: Lean, Six Sigma, ISO 9001, and Toyota’s kaizen approach. This guide explains each phase, shows how PDCA connects to root cause analysis tools like 5 Whys and fishbone diagrams, and walks through three real-world examples.

What Is the PDCA Cycle?

PDCA is a four-phase iterative loop. Each cycle produces an improvement or a learning that feeds the next cycle. The output of every Act phase becomes the new baseline for the next Plan phase — which is why teams that run PDCA continuously outperform those that run it once.

Phase 1

Plan

Define the problem. Identify root causes using 5 Whys, fishbone, or Pareto. Set a measurable improvement goal. Design a solution.

5 WhysFishbone diagramPareto chart
Phase 2

Do

Implement the solution — but on a small scale. A pilot limits risk and gives you clean data to evaluate before a full rollout.

PilotA/B testProcess change
Phase 3

Check

Compare actual results to your goal. Did the solution work? By how much? What unexpected effects appeared?

Control chartsKPI dashboardsBefore/after
Phase 4

Act

If the improvement worked — standardize it. Update procedures, train the team, document the change. If not — return to Plan with new learnings.

SOP updateTrainingNext cycle

The cycle then repeats. Each iteration either improves the process or adds knowledge that makes the next attempt smarter. This is why PDCA is called a cycle, not a project.

PDCA and Root Cause Analysis

The most critical part of the PDCA cycle is the Plan phase — specifically the root cause analysis step. Many PDCA cycles fail not in the Do phase, but in the Plan phase, because the team skips proper RCA and jumps straight to a solution. If the root cause is wrong, the solution won’t stick and the problem will recur.

The standard RCA tools slot directly into the Plan phase: use a Pareto chart to identify which problem category to tackle first, then use a 5 Whys or fishbone diagram to find the root cause of that top category.

Pareto chart → identify vital few  →  Fishbone diagram → map possible causes  →  5 Whys → drill to root cause  →  PDCA Do phase

How to Run a PDCA Cycle (Step by Step)

  1. Define the problem clearly — use data, not opinions. State what is wrong, where, and since when.
  2. Set a measurable goal — “reduce defect rate from 8% to under 2% within 60 days.” Without a measurable goal, the Check phase has nothing to evaluate against.
  3. Analyze root causes — use 5 Whys, a fishbone diagram, or a Pareto chart. Don’t skip this step. A solution without root cause analysis is just guessing.
  4. Design the solution — based on the root cause, not the symptom. Document what you will change, who is responsible, and how you will measure it.
  5. Run a small-scale pilot (Do) — test in one team, one line, one location. Limit scope to limit risk.
  6. Measure and compare (Check) — collect data during and after the pilot. Compare to your baseline and your goal. Note unexpected results.
  7. Standardize or restart (Act) — if the improvement worked, update procedures and roll out broadly. If not, return to step 1 with what you learned.

PDCA Cycle Examples

Manufacturing: Reducing Assembly Defects

PlanPareto shows 67% of defects are solder bridging. 5 Whys: solder paste applied at wrong temperature in morning shift.
DoPilot: add temperature check log to morning startup on Line 3 for 4 weeks.
CheckDefect rate on Line 3: 8.2% → 1.4%. Other lines unchanged (8.1–8.5%).
ActRoll out to all 6 lines. Update SOP. Add to new-hire training.

Software: Reducing Login Failures

Plan31% of support tickets are login/auth failures. 5 Whys: session token expires in 15 min — too short for slow connections.
DoExtend token expiry from 15 to 60 min for 10% of users via feature flag, 14 days.
CheckLogin failure tickets from pilot group drop 74%. No increase in security incidents.
ActRoll out to 100% of users. Update session management policy. Monitor 30 days.

Healthcare: Reducing Medication Errors

PlanPareto: 65% of errors are wrong/omitted dose. 5 Whys: no double-check step during handover for high-alert medications.
DoPilot double-check protocol for high-alert meds on one ward for 6 weeks.
CheckErrors on pilot ward drop 58%. Staff report +12 min per shift for handover.
ActRoll out to all wards. Add 15 min overlap to handover schedule. Update protocol.

PDCA vs. Other Improvement Frameworks

FrameworkBest ForRelationship to PDCA
PDCAIterative operational improvements, daily kaizenBaseline framework
DMAIC (Six Sigma)Complex, data-heavy process improvementsUses PDCA within the Improve phase for solution testing
A3 Problem SolvingStructured single-problem deep divesAn A3 report documents one full PDCA cycle
8D ReportCustomer complaint resolutionD1–D3 = Plan, D4–D5 = Do, D6 = Check, D7–D8 = Act
Kaizen eventsRapid team-based improvementsA compressed PDCA cycle (1–5 days)

Common PDCA Mistakes

Run Your Next PDCA with Free RCA Tools

Start the Plan phase with our free 5 Whys, Fishbone, and Pareto tools — no signup, no downloads.

See All Free Tools →

Related resources

Frequently asked questions

What is the PDCA cycle?

The PDCA cycle (Plan-Do-Check-Act) is a four-phase iterative model for continuous improvement. Developed by Walter Shewhart and popularized by W. Edwards Deming, it provides a structured approach: Plan (identify the problem and design a solution), Do (implement on a small scale), Check (measure results), Act (standardize if successful or restart). Also called the Deming Cycle or PDSA cycle.

What are the 4 stages of the PDCA cycle?

(1) Plan — define the problem, analyze root causes, set measurable goals, design a solution. (2) Do — implement on a small scale or pilot. (3) Check — compare actual results to expected, identify gaps. (4) Act — if the improvement worked, standardize it; if not, return to Plan with new information.

What is the difference between PDCA and PDSA?

PDCA and PDSA (Plan-Do-Study-Act) are essentially the same model. Deming later preferred “Study” over “Check” because Study implies deeper analysis and learning, not just verifying compliance. In practice both terms are used interchangeably. PDCA is more common in manufacturing and Lean; PDSA appears more often in healthcare improvement frameworks like IHI.

How does the 5 Whys fit into the PDCA cycle?

The 5 Whys is used in the Plan phase for root cause analysis before designing a solution. Without proper RCA in the Plan phase, teams treat symptoms instead of causes — leading to recurring problems. Running 5 Whys inside PDCA ensures the solution addresses the real root cause, not just the visible symptom.

What is a PDCA cycle example?

Manufacturing example: Plan — Pareto chart shows 67% of defects are solder bridging; 5 Whys finds root cause is wrong temperature in morning shift. Do — pilot temperature check log on one line for 4 weeks. Check — defect rate drops from 8.2% to 1.4%. Act — roll out to all lines, update SOP.

What is the difference between PDCA and Six Sigma DMAIC?

PDCA is simpler and faster for iterative incremental improvements in daily operations. DMAIC (Define-Measure-Analyze-Improve-Control) is more rigorous and data-intensive for complex problems requiring statistical analysis. Many Six Sigma projects use PDCA cycles within the Improve phase for rapid solution testing.

How long should a PDCA cycle take?

Anywhere from one day (a simple process tweak) to several months (a complex operational improvement). The key principle is that the Do phase should be a small-scale pilot — not a full rollout — to limit the cost of failure. Lean teams often run weekly PDCA cycles for operational improvements; quality improvement projects in healthcare or manufacturing typically run 4–12 week cycles.