It is another day from the plan life. The plant manager has scheduled a gemba walk for 9 AM.
The management team assembles in the corridor — operations director, quality manager, maintenance lead, HR director, a visiting consultant. Eight people in total. Everyone has a clipboard or a phone. Someone has printed a checklist.
They walk the floor for forty-five minutes. They stop at three workstations. They ask a few questions. The operators answer carefully. Nothing unusual is flagged. The team returns to the conference room, discusses "observations" for twenty minutes, and agrees to follow up on two minor housekeeping issues.
By Friday, the notes are filed. By the following Thursday, another gemba walk is scheduled.
Nothing changes. Nobody expected it to.
This is not a gemba walk. This is a factory tour with a Japanese name.
And in my experience across manufacturing plants in Central Europe, it is by far the most common version of gemba that exists in organizations that have adopted the terminology without understanding the purpose.
What Gemba Actually Means
Gemba is a Japanese word that means "the actual place" — the place where value is created, where the work happens, where reality lives.
The principle behind gemba is simple: you cannot manage what you do not see, and you cannot see what you do not go to. Data, reports, and dashboards tell you what happened. The gemba tells you why.
But the principle only works if you go to the gemba with a purpose. Not to inspect. Not to audit. Not to be seen walking the floor. To observe — specifically, against a defined standard, with a clear question in mind.
The question that separates a gemba walk from a factory tour is not "what do I see?" It is "what did I expect to see — and what is the gap?"
Without that gap, there is no observation. There is only looking.
Why Most Gemba Walks Fail
The root cause of ineffective gemba walks is almost never a lack of discipline or commitment. It is a lack of system — specifically, the absence of three things that give a gemba walk its purpose.
No True North
If the organization has not defined its single most important goal for the year — the one outcome that, if achieved by December 31st, makes everything else secondary — the gemba walk has no anchor.
The management team walks the floor and sees many things. A machine that vibrates slightly. An operator who seems slow. A board that has not been updated. A pile of material in the wrong location.
All of these observations may be valid. But without a True North, there is no criterion for deciding which of them matters. The team discusses everything. They act on whatever felt most urgent in the room. The improvement energy dissipates in ten directions.
When the True North is clear — say, a -25 TEuro reduction in scrap on Line 2 — the gemba walk becomes a focused observation. You walk to Line 2. You look at what is generating scrap. You ask why. You connect what you see to the data you already have. The observation has a purpose before you leave the building.
No Annual Improvement Plan
The Annual Improvement Plan translates the True North into a set of prioritized projects — each with a defined scope, expected impact, responsible owner, and key milestones.
Without an AIP, the gemba walk produces observations that have nowhere to go. Something is flagged. A corrective action is assigned. It gets completed — or it does not — with no connection to the plant's improvement trajectory.
With an AIP, every gemba walk observation is evaluated against a simple question: does this connect to a project already in the plan? If yes — it becomes data that informs the project. If no — is it significant enough to open a new project, or is it a daily management issue to be handled at team leader level?
That distinction — between AIP-level observations and DMS-level observations — is what prevents the management team from being overwhelmed by everything they see on the floor.
No Defined KRI, KPI, and PI per Project
This is the failure mode that is most invisible — and most expensive.
Every project in the Annual Improvement Plan should have a defined cascade of indicators: the KRI it is expected to improve, the KPIs that measure the area performance contributing to that KRI, and the PIs that track the specific process variables the project is targeting.
When this cascade exists, the gemba walk has a measurement framework. You are not just looking — you are verifying. You check whether the PI on the board matches what you observe on the floor. You ask whether the KPI trend is moving in the right direction. You evaluate whether the project is on track to deliver the KRI impact that justified it in the first place.
Without this cascade, the gemba walk is disconnected from the measurement system. Observations are subjective. Actions are ad hoc. Progress is unmeasurable.
What a Connected Gemba Walk Looks Like
A gemba walk that is connected to the improvement system looks completely different from a factory tour.
Before leaving the office, the plant manager reviews three things: the True North status, the AIP project dashboard, and the KPI trend for the area being visited. She is not looking for surprises. She is building the context that will make her observations meaningful.
On the floor, the observation is structured around the PI. Not "how is production going?" but "what is the actual cycle time on this machine right now, and how does it compare to the standard?" Not "any problems today?" but "what is the scrap count since the last shift change, and where are the pieces going?"
The questions are specific because the indicators are specific. The operator is not asked to evaluate her own performance — she is asked to describe what is happening at the process level, in the language of the PI: units, kilograms, minutes.
The management team is not looking for blame. They are looking for the gap between the expected PI and the actual PI — and asking why the gap exists. That question — why — is where the gemba walk generates value. Not in the observation itself, but in the conversation that the observation triggers.
When the conversation produces a root cause hypothesis, it gets connected to the relevant AIP project. When it reveals a deviation that is not yet captured in the measurement system, it triggers a review of whether the PI cascade is complete.
The gemba walk ends not with a list of corrective actions, but with a set of updated observations that feed back into the improvement system.
The PI Flexibility Principle
One of the most underappreciated aspects of the KRI/KPI/PI framework is that not all Process Indicators carry the same weight at all times.
Some PIs are consistent drivers of performance throughout the year. Others become critical only in specific periods — seasonal, cyclical, or tied to particular phases of a project.
Treating all PIs as equally important at all times is one of the most common causes of measurement noise in manufacturing plants. The management team tracks thirty indicators simultaneously, none of them with the focus and frequency that the critical ones require.
The principle: For any given period, identify the 3-5 PIs that have the highest influence on the KPI you are trying to move. Focus your gemba observation on those. Let the others run on standard monitoring.
The Absenteeism Example
Consider a plant that is running a workforce availability project as part of its Annual Improvement Plan. The standard PI might be: number of absent employees per day.
This PI is valid. But in autumn and spring — when illness rates spike — it becomes a lagging indicator. By the time the absence is recorded, the line is already running short.
A more useful PI for this period is: number of cross-trained employees available per line per shift.
This PI measures capacity to absorb absence rather than the absence itself. A team leader who knows she has three cross-trained operators available can manage a two-person absence without escalation. A team leader who only tracks the absence count knows there is a problem after the fact.
The shift from absence count to cross-training availability is not a permanent change to the measurement system. It is a seasonal adjustment — a PI that becomes critical in high-risk periods and returns to standard monitoring when the risk passes.
This is what PI flexibility means in practice: the measurement system adapts to where the risk is, not just to where the standard says to look.
Other Examples of Flexible PIs
Customer delivery periods: When a major customer launches a new program or runs an end-of-quarter push, on-time delivery PI frequency may need to increase from daily to shift-by-shift monitoring.
New equipment ramp-up: In the first 90 days after a new machine is installed, cycle time variation and tool wear PIs require closer tracking than at steady state.
Summer shutdown re-start: The first two weeks after a plant restart are high-risk for quality deviations. First-pass yield PI monitoring should be elevated for this period regardless of the baseline performance.
Supplier disruption periods: When a key supplier is under stress, incoming material quality PI becomes a priority observation point on every gemba walk until the risk is resolved.
In each case, the PI is not new — it was always part of the system. What changes is the attention and frequency applied to it, based on where the current risk to the True North is located.
A Structured Gemba Walk — The Questions That Matter
Use this framework before, during, and after your next gemba walk.
Before you leave the office (5 minutes)
- What is the current status of our True North goal?
- Which AIP project is most at risk this week, and what does its PI trend look like?
- What am I expecting to see on the floor — and what would a deviation look like?
On the floor — for each stop (10 minutes maximum)
- What is the actual PI right now? (Ask for the number, do not estimate)
- What is the target PI for this shift?
- What is the gap, and does the team leader know the reason?
- Is the gap connected to an open AIP project — or is this new?
- Is this a PI that should be on elevated monitoring given the current period?
After the walk (10 minutes)
- Which observations connect to existing AIP projects? (Update project status)
- Which observations are DMS-level issues? (Assign to daily management)
- Are there any PIs that should change frequency or focus based on what we saw?
- Does the True North trajectory need to be reviewed?
What Changes When the System Is Connected
In a Tier-1 automotive plant where I supported the implementation of a connected improvement system — True North, AIP, KRI/KPI/PI cascade, and structured gemba walks — the shift in management behavior was visible within eight weeks.
Before the system was in place, the weekly gemba walk produced an average of twelve corrective actions. Six months later, those actions had not been closed. New ones were being added faster than old ones were resolved.
After the system was connected, the gemba walk produced an average of three observations per session. Each one was connected to an AIP project or escalated to daily management. Closure rate went from below 40% to above 85% within three months.
The difference was not discipline. It was focus. The management team stopped seeing everything and started seeing what mattered — because the system had defined what mattered before they walked out the door.
Where to Start
If your gemba walks are producing observations that go nowhere, the problem is almost certainly upstream of the walk itself.
Start with the True North question: does your leadership team have a single, specific, measurable goal for this year? If not — define it before your next gemba walk. Without it, the walk has no anchor.
Then check the AIP: do you have a prioritized set of improvement projects, each with a defined PI cascade? If not — build it. The gemba walk is the verification tool for the improvement system. It cannot replace the system it is meant to verify.
And before every gemba walk, ask one question: what am I looking for today, and why does it matter for our True North?
If you cannot answer that question in one sentence — the walk is not ready to happen.
At AdaptiveOps, we support manufacturing plants in building the connected improvement system that makes gemba walks genuinely useful — from True North definition through AIP development to KRI/KPI/PI cascade design. If you want to discuss how this applies to your plant, book a free 30-minute diagnostic call at /contact.
Related: Why Having 7 Priorities Means Having None — the True North question that anchors every gemba walk.
Related: The €1M Mistake Most Factories Make Every January — how the Annual Improvement Plan gives gemba walks their direction.
Related: Why Your Operators Should Never See a Percentage — the KRI/KPI/PI framework that gives gemba walks their measurement language.