"Run the PDCA cycle" — you hear it all the time, but how exactly do you do it in practice? PDCA is one of the most widely known business frameworks, yet many teams fall into the trap of running PDCA in name only, with little actual impact on results.
This article walks you through everything from the basic meaning of the PDCA cycle, to how each step works in practice, concrete marketing examples, and tips for avoiding the most common pitfalls.
The PDCA cycle is a framework for continuous improvement that repeats four steps: Plan, Do, Check, and Act. Originally developed in the field of quality management, it is now used across virtually every business domain — from marketing and sales to HR and corporate planning.
The defining feature of PDCA is that it is a cycle — it is meant to be repeated, not done once. Lessons learned in one round feed into the next plan, and with each iteration the precision of your initiatives improves, gradually lifting results.
PDCA is important for three reasons. First, it enables data-driven improvement rather than gut-feel decisions. Second, it gives the whole team a shared process, preventing knowledge silos. Third, small, cumulative improvements compound into significant results over time.
Let’s break down what you should do at each stage of the cycle.
In the Plan phase, you define your target numerically and design hypotheses and initiatives to achieve it. The critical part is making your hypothesis explicit: "Why do we believe this initiative will hit the target?" Without a clear hypothesis, you will not know what to verify in the Check phase.
Key points for the Plan phase: quantify the current state, set a numerical goal, identify the gap between the two, design hypotheses and initiatives to close the gap, and assign a schedule and owners. Defining "what counts as success" at the planning stage improves the quality of the entire cycle.
In the Do phase, you carry out the plan. The key here is not just executing but also keeping records. Document what was done, when, how, and whether anything unexpected occurred. Good records dramatically improve the accuracy of your analysis in the next Check phase.
A "small start" approach is also effective — test on a limited scale before rolling out broadly. This minimizes risk while giving you early feedback.
In the Check phase, you evaluate the results of Do against the goals and hypotheses set in Plan. The essence of Check goes beyond simply asking "Did we hit the target?" — it requires digging into "Why did we get this result?"
Specifically, this means: quantifying the gap between target and actual, verifying whether your hypothesis was correct, isolating success factors from failure factors, and checking for external influences beyond the initiative itself. A shallow Check leads to misguided actions in Act, so invest sufficient time here.
In the Act phase, you decide on improvements based on Check findings and feed them into the next Plan. Actions generally fall into three categories: continue or scale what worked, modify and re-execute what did not work, and stop initiatives with no prospect of impact.
The crucial point is translating improvements into a concrete "next Plan." If Act ends with just "we need to improve," the cycle breaks. Keep in mind that the output of Act is the input for the next Plan.
To deepen understanding, here are three concrete PDCA examples from the marketing field.
Plan: Analyze Search Console data to identify 10 keywords ranking 11–20. Set a goal to move 5 of them into the top 10 within 3 months. Hypothesis: content lacks comprehensive coverage of search intent.
Do: Analyze top-ranking articles for target keywords, add missing information, revise heading structures, and add internal links. Record implementation dates and changes.
Check: One month after rewrites, review ranking changes in Search Console. 4 of 10 rose into the top 10, 2 showed no change, 4 improved slightly. Identify that "adding concrete examples" was the common factor in articles that improved.
Act: Since adding examples proved effective, apply the same approach to remaining articles. For the 2 with no change, determine that search intent mismatch is the root cause and plan a structural overhaul for the next cycle.
Plan: Newsletter open rate is stuck at 15%. Set a goal to raise it to 25% within 3 months. Hypothesis: subject lines lack personalization.
Do: Segment the mailing list and customize subject lines per segment. Run an A/B test comparing "name-included" vs. "no-name" subject lines.
Check: Name-included subject lines achieved a 22% open rate vs. 16% for no-name. Among segments, "site visitors in the past 3 months" had the highest open rate.
Act: Standardize name-included subject lines for all sends. Plan to test optimal send timing in the next cycle.
Plan: CPA (cost per acquisition) on search ads exceeds $100. Set a target of $80 or less. Hypothesis: messaging mismatch between ad copy and landing page.
Do: Create a new ad-and-landing-page pattern with unified messaging. Run a 2-week A/B test against the existing pattern.
Check: New pattern CPA dropped to $72. CVR rose from 1.8% to 2.5%, while click-through rate showed little difference — indicating the improvement was driven by the landing page.
Act: Deploy the new pattern as the default. Next cycle: optimize the form UI on the landing page to push CVR even higher.
While the concept is simple, running PDCA in practice often falls into several typical failure patterns.
Pursuing the perfect plan can consume weeks. But plans can only be validated through execution. Move to Do at about 80% confidence and course-correct during Check. Speed matters, especially in fast-moving digital marketing.
Without records of what was done and when, accurate analysis in Check becomes impossible. Build a habit of briefly noting implementation dates, changes made, and unexpected events. No elaborate system is needed — a simple spreadsheet log will suffice.
The most common Check failure is ending at the impression level — "it worked" or "it didn’t." Check’s real job is to quantify results and dig into root causes. Go beyond qualitative reflections and perform data-backed analysis.
Even with good analysis, meetings sometimes end with a vague "so what do we do next?" In Act, define "who does what by when" and translate it into concrete tasks for the next Plan. Making Act and Plan seamless is the secret to keeping the cycle alive.
If your PDCA cycle only turns every six months or a year, the pace of improvement cannot keep up. While it depends on the nature of the initiative, aim for weekly to monthly cycles for marketing activities. Shorter cycles generate more learning and faster improvement.
The OODA loop — Observe, Orient, Decide, Act — is a framework frequently compared with PDCA. Originating from military strategy, OODA is an observation-driven approach that prioritizes real-time sensing and rapid action, as opposed to PDCA’s plan-driven approach.
OODA is better suited to volatile situations where there is no time to plan, while PDCA excels where conditions are reasonably predictable. The two are not mutually exclusive. In marketing, you can use PDCA for medium-to-long-term strategy and OODA for day-to-day operational decisions, achieving both planning rigor and agility.
Finally, here are three tips to prevent your PDCA cycle from becoming an empty ritual and ensure it drives real outcomes.
Stating your hypothesis in the Plan phase — "why we believe this will work" — is the single biggest quality lever for the entire cycle. A hypothesis makes the Check validation point clear and the Act improvements more targeted. Without a hypothesis, PDCA degrades into a "try and reflect" loop.
To ensure Check and Act actually happen, set up recurring review meetings. Embedding PDCA reviews into weekly or biweekly standing meetings creates a mechanism that keeps the cycle turning naturally. In busy day-to-day operations, Check and Act are easily deprioritized without a built-in structure.
Changing multiple variables at once makes it impossible to determine which change drove the result. Limit each cycle to 1–2 changes so you can establish clear cause-and-effect relationships. This approach builds a library of reproducible learnings over time.
The PDCA cycle is a framework for continuous improvement through repeated rounds of Plan, Do, Check, and Act. While the concept is simple, it only delivers results when you raise the quality of each step — articulating hypotheses, logging execution details, evaluating with data, and translating insights into concrete next actions.
Start by picking one initiative on your team and running a small PDCA cycle. When you reliably carry each cycle’s learnings into the next round, both the speed and accuracy of your improvements will steadily increase.
Learn how to calculate LTV (Customer Lifetime Value) and CAC (Customer Acquisition Cost), why the ideal LTV/CAC ratio is...
A practical guide to SWOT analysis templates and how to apply them to marketing strategy. Covers how to identify strengt...
A practical guide to RFM analysis from setup to segment-specific strategy execution. Learn how to segment customers by R...
