If you have spent any time manufacturing furniture, home decor, or consumer goods—especially across international supply chains in Asia—you know that quality control (QC) is never a one-and-done checkbox. It is an ongoing, daily operational discipline.
Whether you are battling finish defects on a lacquer cabinet, warping on solid timber dining tables, or inconsistent weaving across hand-woven rattan baskets, quality issues demand systematic problem-solving. That is why so many operations teams turn to the classic PDSA Cycle (Plan, Do, Study, Act).
Stop Fixing Everything at Once: The Real Secret to Mastering PDSA in Manufacturing Quality Control
PDSA (Plan, Do, Study, Act). is originally pioneered by Walter Shewhart and championed by W. Edwards Deming, PDSA (Plan, Do, Study, Act) is taught in every business school and continuous improvement seminar worldwide. On paper, it looks straightforward:
- Plan: Spot an issue and design a test.
- Do: Run the test.
- Study: Look at the data.
- Act: Decide what to do next.
Yet, despite its apparent simplicity, brands and factory managers stumble constantly when applying it on the factory floor.
At Mondoro, we work alongside artisans, factory directors, and international logistics teams every single day. We frequently see teams fall into two massive traps:
- Confusing “Doing” with “Implementing.”
- Trying to fix three or four variables at the exact same time.
Let’s unpack why these mistakes happen, how to know when your improvement project has bitten off more than it can chew, and how to use reflective decision-making to lock in lasting manufacturing improvements.
1. The Critical Distinction: “Am I Actually Implementing Instead of Doing? And PDSA Quality Contro
In regular workplace conversation, words like “doing,” “executing,” and “implementing” are thrown around interchangeably. But in quality improvement, “Doing” and “Implementing” are opposite ends of the operational spectrum.

What “Doing” Means in a PDSA Cycle
The Do phase is an experiment, not a permanent roll-out. The primary deliverable of the “Do” phase is not perfection—it is data and learning.
When you “do” in PDSA, you are testing a small, temporary, low-risk hypothesis:
- Instead of overhauling the sanding protocol for an entire 10,000-unit production run, you test a new sandpaper grit progression on five sample panels.
- Instead of changing your packaging spec across all twenty product lines, you run drop tests on three cartons packaged with corner protectors.
- Instead of ordering an entire container of a new water-based polyurethane sealer, you apply it to two side tables to observe drying time, adhesion, and clarity under high-humidity factory conditions.
The hallmark of the “Do” phase is that it can fail without destroying your business, blowing your budget, or ruining a container shipment.
What “Implementing” Means
Implementation is the permanent adoption of a proven change into your standard operating procedures (SOPs). It means:
- Rewriting factory work instructions and technical specification sheets.
- Retraining factory operators, line leads, and QC inspectors.
- Updating purchase orders, raw material specifications, and vendor contracts.
- Hard-wiring the process across the entire production line.
Why Confusing the Two Causes Chaos
When teams ask, “Am I actually implementing instead of doing?” they are asking a crucial gut-check question: Have we locked ourselves into a massive, irreversible change before we even have the data to prove it works?
Imagine you notice minor micro-cracking in the lacquer of a new coffee table line during monsoonal humidity. A team that jumps straight to implementing immediately orders three tons of an expensive new chemical additive, updates all curing protocols, and alters their assembly line setup overnight.
If that additive turns out to cause yellowing after 30 days of sun exposure, they have just ruined an entire shipping container of finished inventory.
A disciplined QC team instead runs a true Do test: they take ten test boards, apply the formulation in a controlled curing booth, measure the cure rate, and study the adhesion. Only after studying the results do they decide whether to implement.
2. The Multi-Variable Dilemma: Can You Solve Multiple Problems in One Cycle? And PDSA Quality Control
A question we hear often from operations managers and product developers goes like this:
“In training examples, we see people tackling multiple issues at once—like getting in better shape by cutting calories, drinking more water, and running daily. If a problem has multiple moving parts, shouldn’t we tackle them simultaneously? Isn’t it impossible to fix complex systems by changing only one variable?”
It is a valid observation. In reality, industrial manufacturing and global supply chains are complex, interconnected webs. But there is a huge difference between having an overarching improvement goal and running an uncontrolled experiment.
The Rule of Variable Isolation
In pure Lean, Six Sigma, and scientific experimentation, iterative improvement with isolated variables is fundamentally superior.
Why? Because of attribution.
Let’s translate the personal health analogy into a real factory scenario:
Suppose you are manufacturing dining chairs, and you are experiencing a high defect rate:
- Joint wobble (structural weakness).
- Finish orange-peel (rough, uneven surface).
- Moisture-related dimensional expansion.
If you decide in a single PDSA cycle to simultaneously:
- Change the tenon adhesive formula,
- Adjust the spray-gun pressure, and
- Increase kiln-drying dwell time from 10 to 14 days…
…and your defect rate drops by 60%, which change actually fixed the problem?
- Was it the kiln-drying?
- Was it the glue?
- Did the spray-gun adjustment actually do nothing—or worse, did it increase spray cycle time unnecessarily without adding value?
Because you changed three critical variables at once, you have no clear data attribution. You cannot determine your return on investment (ROI), nor do you know which adjustment was the true root-cause corrective action. If costs rise or defects creep back in six months later, you will have no idea which lever to pull.
| Approach | Advantages | Disadvantages | Best Used When |
| Single-Variable Cycles | • Crystal-clear data attribution • Pinpoints exact root causes • Low cognitive load for factory floor | • Requires multiple iterative cycles • Feels slower upfront | The variables operate independently (e.g., sanding grit vs. carton tape strength). |
| Coupled Multi-Variable Cycles | • Accounts for direct physical interdependencies • Avoids optimizing one step while breaking another | • High risk of confused metrics • Harder to diagnose secondary failures | Variables are physically bonded in the system (e.g., kiln temperature and timber moisture equilibrium). |
3. When Systems Demand Simultaneous Action (Coupled Variables) PDSA Quality Control
Your intuition is correct on one front: some systems cannot be improved by moving one dial in isolation.
In systems engineering, these are called coupled variables. In furniture and home decor manufacturing, you see them frequently:
- Kiln Drying & Airflow: You cannot raise drying temperature without simultaneously modulating relative humidity and internal airflow. If you adjust temperature alone, you case-harden and split the timber.
- Packaging & Palletization: You cannot change carton wall thickness from single-wall to double-wall without also adjusting outer dimensions, pallet stacking patterns, and container cube utilization.
- Finishing & Drying Times: If you switch from solvent-based to water-based stain to meet environmental standards, you must also adjust airflow, booth temperature, and curing flash-off time. The chemical switch cannot happen in isolation.
The Professional Solution: The “Nested Cycle” Approach
Experienced quality leaders resolve this tension by distinguishing between the Aim and the Cycle.

By nesting sequential, single-variable (or tightly coupled) PDSA cycles under one broad quality initiative, you maintain the momentum of solving a large, multi-faceted problem without drowning in muddy data.
4. How Do You Know You Are Taking on Too Much?
Whether you are improving a workshop’s daily output or redesigning your incoming quality inspection protocol, here is a practical diagnostic checklist to tell if your PDSA scope has expanded too far.
1. The Attribution Test
Ask yourself: “If this experiment produces positive or negative results, can our team point directly to the exact change that caused it?”
If your honest answer is “Not really, because we adjusted three things at once,” your scope is too wide. Strip it back.
2. The Floor Burden Test (Cognitive Load)
Manufacturing happens on the floor, not in conference rooms. If your proposed test requires line operators, carpenters, or QC inspectors to memorize a 12-step mental checklist of newly altered behaviors all at once, compliance will drop. When compliance drops, your data becomes worthless because you won’t know if the protocol failed or if people simply stopped following it.
3. The Measurement Overload Test
Every change variable requires verification data. If testing your project demands logging six brand-new data points every hour across three different departments, your team will suffer data fatigue. A great PDSA experiment usually hinges on one primary outcome metric (e.g., delamination rate) and one balance metric (e.g., line cycle time) to ensure the fix doesn’t cause downstream bottlenecks.
5. From Study to Act: Using Borton’s Framework and the “Adopt, Adapt, Abandon” Decision
Once your small-scale experiment has concluded, you enter the Study and Act phases. This is where teams often stumble into ambiguity. They review the numbers, hold a brief meeting, and walk away without a concrete, structured path forward.
To turn experimental data into decisive manufacturing policy, pair Borton’s Model of Reflection with Deming’s three standard options: Adopt, Adapt, or Abandon.

Phase 1: “What?” (State the Facts)
Strip away emotion, opinion, and factory politics. What actually occurred during your test cycle?
- What was the hypothesis? (e.g., “Adding a foam corner guard to our standard dining chair shipping carton will eliminate transit edge-denting.”)
- What did the data reveal? (e.g., “We shipped 20 test cartons via domestic freight. 18 arrived pristine; 2 showed impact tearing where the carton tape split.”)
- What unexpected events occurred? (e.g., “Packing line time increased by 45 seconds per chair.”)
Phase 2: “So What?” (Analyze the Meaning)
This is your diagnostic reflection. Why did the results look like this, and what does it mean for production cost, lead time, and customer satisfaction?
- Why did the tape split on those two units? (The corner protector shifted under compression because the tape width was insufficient to bridge the seam.)
- Is the 45-second packaging delay acceptable? (At our daily volume of 200 chairs, that adds 2.5 labor hours per day. Is that cost lower than the cost of replacing damaged goods?)
Phase 3: “Now What?” (The Decision: Adopt, Adapt, or Abandon)
Based on your reflection, you must select one of three operational pathways:

Option A: Adopt (Standardize and Scale)
You choose Adopt when the data decisively validates your hypothesis without creating unmanageable secondary problems.
- The Action Plan: You shift immediately from “Doing” (testing) to “Implementing” (scaling).
- Execution: Update the master technical files, revise quality control manuals, conduct mandatory floor training, and update purchase orders for all incoming raw materials.
Option B: Adapt (Tweak and Re-Test)
You choose Adapt when the fundamental concept demonstrates merit, but the test data exposed a flaw, variance, or unintended consequence. This is the heart of continuous improvement.
- The Action Plan: You do not roll this out factory-wide yet. Instead, you modify a specific parameter and immediately initiate a follow-up PDSA cycle.
- Execution: In the packaging example above, you adapt by switching from 2-inch tape to 3-inch fiber-reinforced tape to resolve the seam split, then test another small batch of 20 cartons.
Option C: Abandon (Cut Your Losses and Re-evaluate)
You choose Abandon when the data shows that the idea either failed to resolve the defect, introduced intolerable costs, or disrupted adjacent manufacturing operations.
- The Action Plan: You scrap the trial procedure cleanly.
- Execution: Reset the process back to the original baseline, document the failure thoroughly in your QC logbook so future teams do not repeat the exact same experiment, and return to the root-cause analysis stage to brainstorm a completely new direction.

6. Real-World Case Study: Applying the Model to Supply Chain Inventory Control
To see how these concepts function together, let’s look at a common operations challenge: monitoring and managing consumable production hardware (screws, dowels, brackets, sandpaper discs, or safety equipment) on a busy workshop floor.
The Problem
Assembly workers constantly run out of critical M6 connector bolts mid-shift. This causes assembly lines to stall, forces workers to search adjacent storage areas, and leads to unauthorized substitutions of non-spec fasteners just to finish the day’s quota.
The Flawed Instinct: Premature Implementation
A frustrated factory manager might instantly spend $15,000 to purchase and install automated badge-scanning vending lockers across five production halls, mandatory for all 300 employees.
Within two weeks:
- Lines back up at the start of shifts as workers wait to badge into lockers.
- Fasteners run out anyway because replenishment lead times from local hardware distributors weren’t adjusted.
- The system is hated by operators and bypassed within a month.
Why did it fail?The manager implemented before doing. They skipped small-scale testing and ignored variable isolation.
The Disciplined PDSA Approach
Cycle 1: Plan & Do (Small Scale)
- Plan: The team formulates a hypothesis: “Placing a simple two-bin visual Kanban system at Assembly Station #3 will prevent stockouts of M6 bolts without interrupting worker workflow.”
- Do: They don’t touch the whole factory. They place two identical plastic bins at Station #3 containing a calculated two-day supply of M6 bolts. When Bin 1 empties, the worker slides it to the top shelf (a visual cue for the floor runner to restock it within 24 hours) and begins using Bin 2.
Cycle 2: Study (Borton’s “What? So What?”)
- What? Over a two-week trial, Station #3 never stopped assembly due to missing bolts. However, on two occasions, the floor runner forgot to log the restock order in the central purchasing software, meaning the central warehouse ran dangerously low.
- So What? The physical two-bin visual card system works perfectly for floor operators (zero assembly stoppage). The breakdown occurred at the handoff between the floor runner and central warehouse inventory planning.
Cycle 3: Act (The Decision: Adapt)
- Now What? The team decides to ADAPT.
- Instead of rolling the two-bin system across all 30 stations immediately (premature implementation) or scrapping it (abandonment), they adapt the protocol: they attach a laminated re-order card with a QR code directly to the bottom of each bin. When the floor runner picks up the empty bin, they scan the QR code with their inventory tablet in two seconds, instantly triggering a replenishment request in the central purchasing system.
- They run Cycle 2 on Stations #3, #4, and #5. Only after that adapted test proves completely reliable across all three stations do they formally Adopt and roll the visual replenishment protocol across the entire factory floor.
7. The Core Takeaways for Your Quality Control Playbook
- Keep “Doing” and “Implementing” in Their Proper Lanes:
Treat every new quality idea as an experiment first. “Do” is a low-stakes, short-duration lab test aimed at gathering actionable intelligence. “Implementation” is the long-term, institutional rollout of a vetted solution. Never implement until your data forces you to. - Respect the Power of Single-Variable Testing:
When tracking down defects or redesigning workflows, isolate your changes whenever possible. If you change everything at once, you will never know which action saved your margins or which one added hidden waste. - Master Coupled Variables through Nested Cycles:
If your production reality prevents you from changing one element in total isolation, maintain your broad project objective, but design structured, sequential sub-cycles. Track dedicated, distinct metrics for each facet. - Close Every Loop with a Decisive “Now What?”:
Don’t let quality meetings end in vague discussions. Anchor your conclusions in the empirical facts of Borton’s framework, and force a clear, conscious verdict: Adopt the new baseline, Adapt the parameters for another quick test cycle, or Abandon the concept and protect your team’s valuable bandwidth.
Quality in global manufacturing is never achieved through massive, sweeping, overnight transformations. It is built one disciplined, well-designed, meticulously measured PDSA cycle at a time.
Summary of What this Post Covers:
- “Implementing vs. Doing”: Breaks down the essential difference between running a low-risk, small-scale learning experiment (the “Do” phase) versus permanently updating SOPs and infrastructure (implementation) before validation.
- Single vs. Multi-Variable Iteration: Explains the scientific and lean rationale for variable isolation (data attribution) and provides a practical framework for when and how to manage “coupled variables” via nested PDSA cycles.
- Scope Management Check: Features a diagnostic test (The Attribution Test, The Floor Burden Test, and The Measurement Overload Test) to recognize when a quality improvement project is biting off too much at once.
- Borton’s Model + “Adopt, Adapt, Abandon”: Contextualizes “What? So What? Now What?” into a manufacturing decision matrix, complete with a practical workshop inventory/Kanban case study.
Find out more about how Mondoro can help you create, develop, and manufacture excellent home decor and furniture products – don’t hesitate to contact me, Anita. Check out my email by clicking here or become a part of our community and join our newsletter by clicking here.
Mondoro gives out a FREE Lookbook to anyone interested. You can receive a copy of our latest Lookbook by clicking here.
Listen to our Podcast called Global Trade Gal and Deep Dive by Global Trade Gal. You can find it on all major podcast platforms. Try out listening to one of our podcasts by clicking here.
Subscribe to our Mondoro Company Limited YouTube Channel with great videos and information by clicking here.
Related Content
Pantone’s Color of the Year: Cloud Dancer – A Breath of Fresh Air for 2026
In true Pantone fashion, they surprised us all with a choice that goes beyond expectations: Cloud Dancer. This serene, versatile shade has captured my heart, and I believe it perfectly encapsulates the emotions, aspirations, and mindset of 2026.
Read our blog Pantone’s Color of the Year: Cloud Dancer – A Breath of Fresh Air for 2026 by clicking here.
Using Shapr3D and Google Gemini: A Powerful Design Duo
Read on as we will explore how I use Shapr3D, a cutting-edge CAD program, alongside Google Gemini, an AI-powered visualization tool, to enhance my design process. These tools are not just software—they’re design companions that help me bring my creative visions to life in ways that were previously unimaginable.
Read our blog Using Shapr3D and Google Gemini: A Powerful Design Duo by clicking here.
Why AI is the New Design Companion
Design is a deeply personal and creative process, driven by human imagination and ingenuity. At its core, design is about expressing ideas, solving problems, and creating something that resonates with others. But as the world evolves, so do the tools we use to bring our ideas to life. Enter artificial intelligence (AI), a technology that, when used correctly, can become one of the most valuable companions for designers, product developers, and creatives alike.
Read our blog Why AI is the New Design Companion by clicking here

