The Ritual That Isn’t Working
You know the scene. Every Tuesday morning, a group of managers
gathers at the entrance of the production floor. Clipboards in hand.
Safety glasses on. They walk the line, nodding at operators, glancing at
machines, occasionally pausing to ask a question that sounds
suspiciously like “everything going okay?” Two hours later, they’re back
in the conference room, satisfied that they’ve “been to the gemba.”
Nothing changes.
This is the gemba walk in most organizations: a well-intentioned walk
that produces zero insight, zero improvement, and zero trust. The
Japanese word gemba (現場) means “the actual place” — where
real work happens, where value is created, where problems reveal
themselves. The concept is powerful. The execution, in most companies,
is not.
The gap between the philosophy and the practice is not a minor
inefficiency. It’s a structural failure that quietly undermines every
other quality initiative you’re trying to implement. Because if the
people who make decisions don’t understand what’s actually happening on
the floor, every decision they make is built on assumption rather than
reality.
Where Gemba Walks Go Wrong
Walking Without Purpose
Most gemba walks suffer from a fundamental design flaw: they have no
structure. A team shows up at the floor with a vague intention to
“observe” and “engage,” but without specific questions, themes, or focus
areas. The result is a walk that covers physical distance without
covering any intellectual ground.
A purposeful gemba walk starts before you set foot on the floor. What
are you looking for? Is it a recurring quality defect? A bottleneck that
showed up in yesterday’s OEE report? A new process change that’s been in
place for three weeks? Without a defined theme, your walk becomes
tourism — and operators can tell the difference instantly.
Questioning Without
Listening
Here’s a pattern I’ve watched play out in dozens of factories: a
manager walks up to an operator, asks a question, and immediately starts
talking about what should be done. The operator, who has been running
this machine for eleven years, barely gets a sentence out before the
manager is already formulating a solution based on what worked at their
previous employer.
This is not engagement. It’s a monologue disguised as a conversation.
The entire premise of a gemba walk is that the people closest to the
work know things you don’t. If you’re not creating space for them to
tell you, you’re wasting everyone’s time.
Documenting Without Acting
I’ve seen gemba walk logs that stretch back years — beautifully
organized spreadsheets tracking every walk, every observation, every
“action item.” When you cross-reference those action items with actual
changes made on the floor, the correlation is near zero. The walk became
a documentation exercise. The log became the deliverable instead of the
improvement.
This is particularly insidious because it feels productive.
You have records. You have data. You have a paper trail showing that
leadership is “engaged with the floor.” But paper doesn’t prevent
defects. Logs don’t improve cycle times. If your gemba walk doesn’t
result in at least one tangible change per week, it has become
bureaucratic theater.
What a Real Gemba Walk Looks
Like
Preparation: Defining the
Lens
Every effective gemba walk needs a theme — a specific lens through
which you’ll observe the floor. Not a generic theme like “quality” or
“safety,” but something precise:
- “We’re investigating why Line 3’s first-pass yield drops every shift
change.” - “We’re observing how operators handle the new torque specification
on Station 7.” - “We’re looking at material flow between the machining cell and
assembly.”
A focused theme changes what you see. When you walk the floor looking
for “anything unusual,” your brain filters out the abnormal because it’s
become normal. When you walk looking for a specific pattern, you notice
things you’ve walked past a hundred times.
The theme should come from data — not from a gut feeling. Pull your
last week’s quality reports, maintenance logs, production metrics, and
customer complaints. Where are the gaps? What question, if answered,
would unlock the most improvement? That’s your theme.
Observation: Seeing Without
Solving
The hardest part of a gemba walk is resisting the urge to solve.
You’ll see something that looks wrong — an operator reaching above
shoulder height, a bin overflowing with parts, a gauge reading outside
the expected range — and your instinct will be to correct it
immediately. Don’t.
Instead, ask. “I notice you’re reaching up for that fixture. Is there
a reason it’s positioned that high?” You might learn that the operator
asked maintenance to move it six months ago and nothing happened. You
might learn that the reaching is actually the least bad option given the
alternative constraint. You might learn something you couldn’t have
figured out by looking.
Observation means gathering information, not dispensing solutions.
The time for solutions comes later — after you’ve walked the entire
floor, talked to multiple operators, and cross-referenced what you’ve
seen with the data you brought with you.
Here is a framework I use for structuring observations during a
walk:
| Observation Category | What to Look For | What to Ask |
|---|---|---|
| Safety | Awkward motions, bypassed guards, ergonomic strain | “Have you reported this? What happened when you did?” |
| Quality | Rework stations in use, visual nonconformities, sort bins | “How often does this happen? What do you think causes it?” |
| Delivery | WIP accumulation, idle operators, expedited material | “Is this the normal flow, or is today unusual?” |
| Cost | Excess inventory, redundant steps, unnecessary handling | “If you could change one thing about this process, what would it be?” |
| People | Frustration signals, disengagement, workarounds | “What’s the hardest part of your job right now?” |
Notice that every category pairs observation with a question. Seeing
tells you what. Asking tells you why. You need both.
Engagement:
Building Trust One Conversation at a Time
Operators have a finely calibrated BS detector. They know when a
manager is walking the floor because they genuinely want to understand,
and when they’re there because their boss told them to “be more
visible.” You can’t fake it, and trying to fake it is worse than not
showing up at all.
Trust is built through consistency and follow-through. If an operator
tells you something on Tuesday’s walk and you follow up on Thursday —
not with a solution, but with a genuine “I’ve been thinking about what
you said, and I have a couple more questions” — you’ve just created more
engagement value than a year of suggestion boxes.
Here’s what doesn’t work: showing up once a quarter, acting alarmed
at conditions operators have been reporting for months, making promises
you don’t keep, and then disappearing. That pattern doesn’t just fail to
improve things — it actively makes them worse by teaching operators that
leadership engagement is performative and temporary.
The Cadence Question
How often should you do gemba walks? The honest answer depends
entirely on your role and your organization’s maturity with the
practice. But here are rough guidelines that work in most manufacturing
environments:
For plant managers and operations directors: Daily.
Not a full two-hour walk — even fifteen minutes on the floor every day,
focused on one area, creates a rhythm that operators come to expect and
value. The daily frequency matters less for what you see and more for
the signal it sends: this is important enough that I make time for it
every single day.
For quality managers and engineering leaders: Two to
three structured walks per week, each with a specific technical theme.
These walks go deeper — they involve spending real time at a station,
watching entire cycles, and talking with operators about process
details.
For executives and corporate staff: Weekly or
biweekly, always hosted by someone from the floor who can provide
context. Corporate visitors should be the exception, not the norm — but
when they do visit, the visit should be substantive enough to inform
decision-making at the enterprise level.
The wrong answer is “whenever we have time.” You will never have
time. You will always have meetings that feel more urgent, reports that
need review, and fires that need fighting. Gemba walks are the kind of
activity that gets displaced by urgency because their impact is deferred
— you don’t see the payoff immediately, so it feels expendable. It
isn’t.
Connecting Gemba
Walks to Your Quality System
A gemba walk is not a standalone activity. It’s an input into your
quality management system, and it should connect to your other tools and
processes:
FMEA: Observations from the floor should feed back
into your FMEA reviews. If you’re seeing failure modes during walks that
aren’t in your FMEA, your risk analysis is incomplete.
CAPA: When you identify an issue during a walk, it
should enter your corrective action system with the same rigor as a
customer complaint or audit finding. No special treatment because “we
found it ourselves.”
Layered Process Audits: Gemba walks and LPA share
DNA but serve different purposes. LPA verifies compliance with
established standards. Gemba walks question whether the standards
themselves are correct. Both are needed; they are not substitutes.
Management Review: Your monthly or quarterly
management reviews should include a summary of gemba walk findings. If
leadership never hears what’s being observed on the floor, the
information has no pathway to strategic decision-making.
Measuring
Whether Your Gemba Walks Are Working
This is the question that makes people uncomfortable, because it
forces them to confront the possibility that their walks aren’t
producing value. But if you’re investing hours of leadership time every
week, you should be able to answer the question: “What changed because
of our gemba walks?”
Here are metrics worth tracking:
-
Issues identified per walk — A leading indicator
of observation quality. If you’re consistently finding zero issues,
you’re not observing effectively. -
Issues resolved within 30 days — A measure of
follow-through. Identification without resolution breeds
cynicism. -
Operator feedback trends — Survey or pulse-check
your floor operators quarterly. Are they seeing more leadership
presence? Do they feel heard? -
Repeat findings — If the same issue appears on
walk after walk without resolution, you have a systemic problem with
your action system, not your observation system. -
Time to resolution — How long does it take from
“observed on the floor” to “change implemented”? If it’s longer than
your quality issue resolution time for customer complaints, you’re
treating internal findings as lower priority.
The Cultural Dimension
Tools and frameworks are necessary but insufficient. The real power
of gemba walks emerges when they become part of a culture where going to
the source is instinctive — not a scheduled event, but a reflex.
In organizations where this culture exists, a quality engineer who
hears about a defect doesn’t open a spreadsheet first. They walk to the
station where the defect was found. A production manager who gets a call
about a delivery risk doesn’t send an email asking for an explanation.
They go to the cell and watch the process.
This cultural shift doesn’t happen because of a memo or a training
session. It happens when leaders model the behavior consistently, over
months and years, until it becomes the norm rather than the exception.
It happens when operators trust that speaking up during a walk leads to
real change, not to blame. It happens when the organization stops
treating the floor as a place to visit and starts treating it as the
place where truth lives.
A Personal Note
In twenty-five years of quality engineering, I’ve never solved a
significant manufacturing problem from behind a desk. Every meaningful
improvement I’ve been part of started with standing on the floor,
watching the process, and listening to the people who run it every day.
The gemba walk isn’t a technique. It’s a mindset — the radical idea that
to understand what’s happening, you have to go where it’s happening.
If your gemba walks have become routine, comfortable, and
consequence-free, they’ve lost their purpose. The floor should make you
slightly uncomfortable. You should see things that concern you. You
should hear things that challenge your assumptions. If you don’t, you’re
not walking deeply enough — or you’ve stopped being honest about what
you see.
Go to the gemba. Go with questions, not answers. Go consistently. And
when you find something, fix it. That’s the entire practice. It’s not
complicated. It’s just hard.
About the Author: Peter Stasko is a Quality
Architect with over 25 years of experience in manufacturing quality,
process improvement, and Lean methodologies. He has helped organizations
across automotive, electronics, and industrial sectors transform their
quality systems from compliance-driven paperwork into genuine drivers of
operational excellence. Peter writes about the gap between quality
theory and shop-floor reality — because that’s where most improvement
opportunities live.
