Week 25: The Floor Operator Who Changed a $50M Decision With One Sentence
The Floor Operator Who Changed a $50M Decision With One Sentence
This is the story of a Gemba walk that almost didn't happen, a floor operator who said five words that nobody wanted to hear, and a twelve-thousand-dollar project that saved fifty million.
It's also a story about the most expensive words in business.
I was at Shell. This was during a period when a major capital investment was being evaluated for one of the production facilities. The project was a capacity expansion — new equipment, new infrastructure, significant capital outlay. The business case had been through multiple rounds of review. The engineering team had done their analysis. The finance team had done theirs. The number on the table was approximately fifty million dollars.
The logic was straightforward. The facility was hitting capacity constraints. Production was being limited by a bottleneck in one of the key processing units. The demand was there. The margins justified the investment. The engineering solution was technically sound. Multiple levels of management had reviewed and approved the progression of the business case.
It was, by every conventional measure, a solid investment proposal.
Around the same time, I was running a CI programme at the facility. We had Gemba walks scheduled as part of the regular operational cadence — leadership walking the floor, observing processes, talking to the people doing the work. These walks were non-negotiable. Every week. Rain or shine. Regardless of what else was on the calendar.
Except this particular week, there was pressure to cancel. The capital investment review was consuming leadership attention. The operations director was preparing for a steering committee presentation. His calendar was full. He asked if we could skip the Gemba walk.
I pushed back. I always push back on this, because the moment you make Gemba walks optional, they become rare. And once they become rare, they become extinct.
He agreed to keep it. Reluctantly.
We walked the process. The operations director, two of his team leads, a CI practitioner, and me. We followed the production flow from intake through the processing unit that was identified as the bottleneck — the same unit that was the justification for the fifty-million-dollar capacity expansion.
We stopped at the processing unit. Observed the operation. Watched the cycle. Noted the changeover happening between product runs.
The changeover was long. Not a little long. Visibly, obviously, painfully long. The equipment was idle for an extended period while the crew went through the changeover process — cleaning, reconfiguring, testing, restarting. The crew was working steadily. They weren't standing around. But the process itself was consuming a disproportionate amount of available production time.
I knew this, because I'd looked at the OEE data. The availability losses from changeovers were significant. But data on a screen is abstract. Watching it happen in real time, standing next to the equipment, feeling the minutes pass while a fifty-million-dollar capital project is being justified by the capacity constraint you're watching — that's concrete.
The operations director saw it too. He turned to the team lead and asked: "How long is this changeover?"
The team lead said it depended on the product transition but typically ran about ninety minutes.
The operations director frowned. Then he looked at the operator — a woman who had been running this equipment for years. She was wiping down a component, waiting for the reconfiguration to complete.
He asked her: "What would help you most?"
She didn't hesitate.
"Just fix the changeover."
Five words. No analysis. No PowerPoint. No business case. Just the direct, unfiltered perspective of someone who lived with the problem every day.
Those five words changed everything.
After the walk, I sat down with the operations director and the CI team. I showed them the data — the changeover frequency, the time per changeover, the total production hours lost. Then I showed them what a SMED analysis could potentially deliver.
SMED — Single Minute Exchange of Die — is a methodology for reducing changeover times. It was developed by Shigeo Shingo and has been a core lean tool for decades. The principle is straightforward: separate internal activities (which can only happen when the equipment is stopped) from external activities (which can be done while the equipment is still running), and then streamline both.
We ran the SMED project. The team included operators, maintenance technicians, and engineers. We videotaped the changeover. We analysed every step. We separated internal from external. We redesigned the sequence. We built pre-staging protocols. We modified a few fixtures.
Total project cost: approximately twelve thousand dollars. Most of that was the fixture modifications. The rest was labour time for the analysis and implementation.
The result: changeover time dropped from ninety minutes to under thirty-five minutes. More than a fifty percent reduction.
The impact on capacity was immediate and dramatic. The bottleneck opened up. Production hours increased significantly. The facility's throughput rose by enough to close the majority of the capacity gap that had justified the fifty-million-dollar expansion.
The capital project wasn't cancelled outright. But it was deferred. Significantly deferred. Because the constraint that justified it had been largely eliminated by a twelve-thousand-dollar SMED project that started because a floor operator said five words during a Gemba walk that almost got cancelled.
Twelve thousand dollars versus fifty million. That's not an ROI. That's an argument for never cancelling a Gemba walk again.
Now, here's the part of this story that keeps me up at night.
This almost didn't happen.
If the Gemba walk had been cancelled — which it very nearly was — nobody would have stood next to that changeover and watched it happen. The OEE data existed. The availability losses were documented somewhere in a system that people reviewed periodically. But the data hadn't triggered action. The data had been there for months. Maybe years.
Data alone doesn't create urgency. Standing next to the problem does.
If the operations director hadn't asked the operator a genuine question — not "what are your KPIs" or "how's the shift going" but "what would help you most" — the conversation wouldn't have happened. The operator had the answer. She'd had it for years. Nobody had asked.
And if the words "just fix the changeover" had been dismissed — "operators don't understand strategy," "the capital project is already in the pipeline," "we've done the analysis" — the fifty-million-dollar project would have proceeded. Nobody would have known they were solving the wrong problem. Or rather, solving the right problem with the wrong solution.
The most expensive words in business are not in any financial report. They're the words that prevent leaders from hearing the people who are closest to the work.
"Operators don't understand strategy." I've heard this in boardrooms. I've heard it from directors, from VPs, from consulting partners. And every time I hear it, I think about the floor operator who understood the strategy better than anyone in the capital planning team — because she understood the process, and the process is where strategy lives or dies.
"We've already done the analysis." The analysis was correct. The facility did have a capacity constraint. The capital project was technically sound. But the analysis didn't include a twelve-thousand-dollar alternative that an operator could have identified, because nobody asked her.
"That's not how investment decisions work." True. Investment decisions work through business cases, engineering studies, financial modelling, and governance approvals. But sometimes the most important input to an investment decision is a five-word sentence from the person who operates the equipment.
I've told this story in dozens of leadership sessions. The reaction is always the same.
First, people are impressed by the numbers. Twelve thousand versus fifty million. That gets attention.
Then they get uncomfortable. Because they start thinking about the capital projects, the strategic investments, the big decisions in their own organisations. And they start wondering: is there a floor operator in my facility who has the answer to a question we haven't asked?
The honest answer is almost certainly yes.
In my experience, the people closest to the work almost always understand the constraints better than the people making the investment decisions. Not because the decision-makers are incompetent. But because the decision-makers are operating at a level of abstraction that smooths over the operational details where the real leverage lives.
A changeover that takes ninety minutes doesn't show up as "changeover problem" in a capital planning review. It shows up as "capacity constraint requiring infrastructure investment." The translation from operational detail to strategic summary loses the information that would have prevented a fifty-million-dollar misallocation.
This is why Gemba walks are not optional. This is why asking genuine questions matters. This is why the most valuable voice in many rooms is the one that doesn't get invited to the meeting.
At Shell, we built this into the operating rhythm. Gemba walks were not tours. They were structured conversations between leaders and frontline teams, with the explicit purpose of surfacing operational reality that might be invisible at higher levels of abstraction.
The format was simple. Walk the process. Observe. Ask three questions: What's working? What's not working? What would help?
And then — critically — act on what you hear.
The acting is what makes it real. If the operations director had listened to the operator, nodded thoughtfully, and then gone back to his office and continued with the capital project, the Gemba walk would have been theatre. But he didn't. He listened. He investigated. He acted. And the result was a twelve-thousand-dollar project that changed a fifty-million-dollar decision.
So here's my challenge.
When was the last time you asked a floor operator — or a service desk agent, or a warehouse worker, or a call centre representative, or whoever does the actual work in your organisation — what would help them most?
Not what their KPIs are. Not what their targets are. Not what their compliance score is.
What would help them most.
Because somewhere in your organisation, right now, there is someone who has an answer that could change a decision worth more than you imagine. They have it because they live with the problem every day. They see it. They feel it. They work around it.
They're waiting for someone to ask.
Are you willing to walk the floor and listen?
If you want to build the performance management and Gemba discipline that surfaces these insights systematically — not by accident, but by design — explore the Performance and Process Management Mastery programme at stormholt.org/products/performance-and-process-management-mastery.
New to Stormholt? Start here — it locates your stage in minutes.
Building capability? Stage III — Skill — the tools, the courses, the simulations.