---
name: plan-do-check-act
description: Use when asked to explain or apply the PDCA (Plan-Do-Check-Act) improvement cycle, its PDSA (Plan-Do-Study-Act) variant, or how it compares to other iterative-improvement cycles (OODA, DMAIC) — sourced from Wikipedia's PDCA article.
---

# Plan-Do-Check-Act

PDCA (Plan-Do-Check-Act) is an iterative four-stage cycle for continuous
improvement of a process, product, or system — each cycle closer to the
goal than the last, rather than a one-time fix.

## History

PDCA originated with physicist **Walter Shewhart** at Bell Telephone
Laboratories in the 1920s as a three-step quality-control process.
**W. Edwards Deming** modified it in the 1940s and popularized it in Japan
during lectures in the early 1950s, where Japanese participants shortened
the steps to the now-traditional Plan-Do-Check-Act naming.

## The four stages

- **Plan** — establish the objectives and the process changes needed to
  achieve the desired result.
- **Do** — execute the plan, ideally first on a small scale so a flawed
  plan doesn't cause large-scale damage.
- **Check** — evaluate the data and results against the expected outcome,
  looking for both similarities and differences, and for trends across
  repeated cycles.
- **Act** — improve the process based on what Check revealed: investigate
  root causes, adjust the process, and set a new (better) standard for the
  next cycle.

Then the cycle repeats — Act's output becomes the next cycle's Plan input,
which is what makes PDCA a *spiral* toward the goal rather than a single
loop.

## PDSA and other variants

- **PDSA (Plan-Do-Study-Act)** — Deming's own preferred later terminology;
  he considered "Study" a better description than "Check" of what that
  stage should actually involve — genuine analysis, not a pass/fail box
  check.
- **Shewhart cycle** / **control circle** — earlier names for the same
  underlying idea.
- **OPDCA** — adds an explicit "Observe" stage before Plan, emphasizing
  assessing current conditions before planning a change.

## Where it's used

PDCA applies across manufacturing, healthcare, education, and service
industries — reducing defects, optimizing efficiency, improving clinical
processes, refining teaching methods, and raising customer satisfaction.
It supports both large, breakthrough improvements and small, incremental
**kaizen**-style refinements, and is a foundational element of most
formal quality-management systems (ISO 9001 among them).

## Comparison to other improvement/decision cycles

PDCA is often placed alongside other iterative cycles that share the same
underlying shape (assess, plan, execute, review) but originate in
different domains:

- **DMAIC (Define, Measure, Analyze, Improve, Control)** — Six Sigma's
  data-driven improvement cycle, more heavily quantitative and typically
  applied to a defined, bounded project rather than an ongoing spiral.
- **OODA (Observe, Orient, Decide, Act)** — a military decision cycle
  (John Boyd) emphasizing speed and adapting faster than an adversary,
  applied to business strategy and competitive situations rather than
  process quality specifically; see the reasoning-under-uncertainty
  emphasis contrasted with PDCA's steadier improvement-cycle framing.

Roughly: PDCA's "Plan" maps to identifying and measuring a problem, "Do"
to creating and executing action plans, and "Check"/"Act" to verifying and
then controlling/institutionalizing the fix — the same underlying shape
DMAIC and OODA each formalize with their own vocabulary and emphasis.

## Common pitfalls

- **Treating "Check" as a pass/fail gate** rather than genuine analysis —
  Deming's push toward "Study" exists precisely because a shallow check
  misses the trends and root causes the next Act stage needs.
- **Skipping the small-scale "Do"** and rolling a change out fully before
  any Check — defeats the purpose of an iterative cycle, since a flawed
  plan now causes full-scale damage before anyone learns it was flawed.
- **Treating one pass through the cycle as "done"** — PDCA's value comes
  from repeating it; a single Plan-Do-Check-Act with no follow-up cycle is
  a one-off fix, not continuous improvement.
- **Not institutionalizing the improvement in "Act"** — identifying a fix
  but not updating the standard process to reflect it means the same
  problem likely recurs, and the next team has to rediscover the same
  fix.

## Learn more

- [Wikipedia: PDCA](https://en.wikipedia.org/wiki/PDCA) — the source for this skill.
- W. Edwards Deming, *Out of the Crisis* — Deming's own account of the cycle and quality management more broadly.