PDCA – Framework Thinking

PDCA

Plan-Do-Check-Act — The Universal Cycle of Continuous Improvement

1. Overview of PDCA: A Methodology That Achieves Continuous Improvement Through a 4-Stage Cycle

    flowchart LR
    A["Ad hoc improvement<br/>relying on experience<br/>repeated one-off fixes"] --"Systematic approach<br/>via a 4-stage cycle"--> B["Plan, Do, Check, Act —<br/>data-driven improvement"] --"Standardize, repeat,<br/>continuously improve"--> C["A culture of<br/>continuous improvement<br/>spiraling performance gains"]

    style A fill:#FFEBEE,stroke:#D32F2F,color:#000
    style B fill:#E3F2FD,stroke:#1976D2,color:#000
    style C fill:#E8F5E9,stroke:#388E3C,color:#000

Definition: A quality management methodology devised by Walter Shewhart and popularized by W. Edwards Deming, in which the four stages Plan → Do → Check → Act are repeated in a cycle to continuously improve the quality of a process, product, or service — a universal improvement cycle.

Characteristics: (Iterative, spiral improvement) After one cycle completes, it starts again at a higher level — continuous improvement (Kaizen). (Small-scale pilot first) In the Do stage, risk is minimized by experimenting and verifying at small scale before full rollout. (A common foundation for standards) The common foundation of virtually every major quality/management standard, including TQM, ISO 9001, ITIL, Lean, Six Sigma, and ISO 27001.

2. Core Structure of PDCA

a. The 4-Stage Improvement Cycle

    flowchart TD
    PLAN["Plan<br/>Define the problem, analyze current state<br/>set goals & metrics<br/>establish an improvement plan"]
    DO["Do<br/>Run a small-scale pilot<br/>collect data<br/>execute against the plan"]
    CHECK["Check<br/>Measure & analyze results<br/>assess goal achievement<br/>identify causes"]
    ACT["Act<br/>Success: standardize & roll out org-wide<br/>Failure: eliminate cause & replan<br/>raise the goal for the next cycle"]

    PLAN --> DO --> CHECK --> ACT --> PLAN

    style PLAN  fill:#E3F2FD,stroke:#1976D2,color:#000
    style DO    fill:#E8F5E9,stroke:#388E3C,color:#000
    style CHECK fill:#FFF3E0,stroke:#F57C00,color:#000
    style ACT   fill:#1E3A5F,stroke:#1E3A5F,color:#fff

Key Activities and Tools by Stage

Stage	Key Question	Key Activities	Tools Used
Plan	What is the problem? How will we improve it?	Collect and analyze current-state data, set goals/KPIs, establish an improvement plan	Pareto chart, Fishbone, 5-Why, hypothesis formulation
Do	Is execution proceeding as planned?	Run a small-scale pilot, collect data, control variables	Check sheets, experiment logs, observation records
Check	Do the results match the plan?	Compare actual results against goals, analyze deviations, identify success/failure factors	Control charts, histograms, scatter plots, dashboards
Act	How will we make the improvement stick?	Standardize and roll out if successful; replan and repeat if not	SOP documentation, training, setting the next Plan’s goals

PDCA vs. DMAIC Comparison

Comparison	PDCA	DMAIC (Six Sigma)
Origin	Shewhart / Deming (1950s)	Motorola / GE (1980s–90s)
Approach	Simple, fast iterative improvement	Deep statistical analysis of complex problems
Data	Basic data analysis focused	Uses precise statistical tools
Best fit	Everyday, repetitive process improvement	Large-scale, complex quality problems
Duration	Short to medium term (weeks to months)	Medium to long term (months to quarters)
Complementarity	DMAIC methodology can be applied within a PDCA cycle

b. Application to IT Service and Process Improvement

    flowchart LR
    subgraph R1[""]
        direction LR
        A1["DevOps CI/CD<br/>Plan: sprint feature planning<br/>Do: code development & build<br/>Check: testing & monitoring<br/>Act: deploy & retrospective"]
        A2["ITSM/ITIL<br/>Plan: set SLA targets<br/>Do: operate the service<br/>Check: measure metrics<br/>Act: improve the process"]
    end
    subgraph R2[""]
        direction LR
        A3["Security management<br/>Plan: threat analysis & countermeasures<br/>Do: apply security controls<br/>Check: vulnerability scanning<br/>Act: update policy"]
        A4["Data quality<br/>Plan: set quality criteria<br/>Do: cleansing pipeline<br/>Check: measure quality<br/>Act: refine rules"]
    end

    style A1 fill:#E3F2FD,stroke:#1976D2,color:#000
    style A2 fill:#F3E5F5,stroke:#7B1FA2,color:#000
    style A3 fill:#FFEBEE,stroke:#D32F2F,color:#000
    style A4 fill:#E8F5E9,stroke:#388E3C,color:#000
    style R1 fill:none,stroke:none
    style R2 fill:none,stroke:none

Example of PDCA Applied to an IT Project — Reducing System Outages

Stage	Activity	Result
Plan	Plan to reduce average monthly outages from 12 to a target of 6 or fewer	Pareto analysis of outage types: DB connection errors account for 40%
Do	Optimize DB connection pool settings, deploy an automatic restart script	Ran a 2-week pilot, collected outage logs
Check	After 2 weeks, DB connection errors dropped 75%, total outages fell to 9	Goal not fully met (short of the target of 6); remaining cause: memory shortage
Act	Standardize the DB configuration change across all servers; carry the memory issue into the next Plan	New goal: establish a memory management policy → start the next PDCA cycle

3. Expected Benefits and Application of PDCA

Category	Expected Benefits	Application and Practical Use
Continuous improvement	After solving a problem, performance rises in a spiral toward higher goals	Run sprint retrospectives as a PDCA structure for continuous team process improvement
Risk minimization	Small-scale Do-stage experiments validate risk before organization-wide rollout	Apply a pilot → verify → org-wide rollout sequence when adopting new technology
Standardization	Successful improvements are immediately documented as SOPs in the Act stage	Immediately standardize incident resolution actions as a runbook
Cultural adoption	Repeated application embeds a data-driven improvement culture org-wide	Integrate the continuous improvement requirements of ISO 9001, ITIL, and ISMS-P through PDCA

TQM ISO/IEC 25010