Week 19: The One Metric That Predicts Whether Your CI Programme Will Survive
The One Metric That Predicts Whether Your CI Programme Will Survive
I've assessed CI programmes in more than forty organisations across twenty years. Energy. Manufacturing. Drilling. Financial services. Logistics. Government. Four continents. Programmes ranging from five people to five thousand.
And after all of that, I've found one metric that predicts programme health more reliably than anything else I've ever measured.
It's not the number of belts trained. It's not the number of projects completed. It's not the savings documented, the governance score, the maturity assessment rating, or the employee engagement survey results.
It's this: days since the last leadership Gemba walk.
That's it. One number. And it tells me almost everything I need to know.
Let me explain what I mean, and then let me show you the data.
A Gemba walk — for anyone not fluent in lean vocabulary — is when a leader goes to where the work happens and observes. Not inspects. Not audits. Observes. Asks questions. Listens. Learns. Sees the process as it actually runs, not as it appears in a PowerPoint.
The purpose is not to check up on people. The purpose is to understand reality. Because reality on the floor and reality in the boardroom are almost never the same thing, and the gap between them is where improvement programmes go to die.
When I assess a CI programme, one of the first questions I ask is: "When was the last time a senior leader — director level or above — walked the process and had a conversation with the people doing the work?"
The answer predicts the health of the programme with an accuracy that still surprises me, even after all these years.
Under seven days? The programme is almost certainly thriving. Teams are engaged. Projects are moving. People believe their work matters because they have evidence — in the form of a senior leader standing next to them — that it does.
Seven to fourteen days? Mixed signals. The programme is functional but fragile. There's enough leadership visibility to keep things moving, but not enough to build real momentum.
Fourteen to thirty days? Declining. The programme is still alive on paper, but the energy is draining. Teams are starting to question whether leadership actually cares.
Over thirty days? The programme is dying. It may still have governance structures, project trackers, and monthly reports. But the life has gone out of it. People are going through the motions. The best practitioners are quietly looking for other assignments. The improvement boards haven't been updated.
I've tracked this informally across dozens of organisations, and the correlation is almost perfect.
Let me tell you the story that convinced me.
At Shell, I was part of a deployment that spanned multiple geographies. One of the sites I worked with was an operations facility in Southeast Asia — large, complex, hundreds of people. The CI programme there was one of the strongest I'd been involved with. Projects were delivering real results. Teams were engaged. The Gemba boards were alive — updated daily, discussed daily, acted upon weekly.
The operations director at that site was a man who understood one thing instinctively that many leaders never learn: his presence was his most powerful leadership tool.
Every Tuesday and Thursday morning, he walked the floor. Not a quick tour. Not a photo opportunity. A genuine walk. He'd spend forty-five minutes to an hour. He'd stop at the boards. He'd ask the team lead what was stuck. He'd ask operators what was frustrating them. He'd ask the CI practitioner what support they needed. And — this is the critical part — he followed up. If someone told him on Tuesday that they needed maintenance to prioritise a repair, by Thursday he'd come back and ask whether it had happened.
His Gemba cadence was never more than three days. Usually two.
The result? That site consistently outperformed others in the same geography. Not by a small margin. The project completion rate was above ninety percent. The improvement sustainability rate — the percentage of improvements still in place twelve months later — was eighty-five percent. At other sites in the same region, with the same methodology and the same training, sustainability was typically forty to fifty percent.
The difference wasn't methodology. The difference was standing in the room.
Then something changed.
The operations director was promoted. Moved to a corporate role. His replacement was capable — strong technical background, experienced leader, genuinely committed to operational performance. But he had a different style. He trusted the data. He reviewed dashboards. He held monthly reviews. He was present in the governance sense — attending steering meetings, approving project charters, reviewing results.
But he didn't walk.
He believed — sincerely — that the data told him what he needed to know. He could see the KPIs. He could read the project updates. He didn't need to be on the floor because the information came to him.
Within four weeks, I noticed the first signs. The Gemba boards were being updated less frequently. Not because anyone had stopped caring, but because the audience had disappeared. When the director walked the floor, the boards mattered — they were the medium of conversation. When he stopped walking, the boards became paperwork.
By week six, two project sponsors had started missing steering meetings. Not because they were disengaged, but because the visible signal from the top had shifted. The new director's absence from the floor communicated — whether he intended it or not — that the floor wasn't where the important work happened. The important work happened in the boardroom, in the data, in the monthly review.
By week eight, the CI practitioner came to me frustrated. She was struggling to get people released for kaizen events. Supervisors were pushing back. The line between "improvement work" and "real work" — which had been almost invisible under the previous director — was hardening again.
By week twelve, the programme was unrecognisable. Project completion rates had dropped from ninety percent to under forty percent. Two projects had stalled completely. The daily stand-ups were still happening in name, but they'd become status updates instead of problem-solving sessions. The energy was gone.
Twelve weeks. Same site. Same people. Same methodology. Same tools. Same governance structure. The only thing that changed was the Gemba cadence of one person.
I've told this story to leadership teams many times. The reaction is always the same.
First, scepticism. They don't believe one person's floor presence could have that much impact. They think I'm exaggerating for effect or confusing correlation with causation.
Then I show them the data. I show them the project completion rates before and after. I show them the board update frequency. I show them the team engagement scores, which we tracked quarterly. The decline tracks the Gemba absence almost perfectly.
Then comes discomfort. Because every leader in the room starts calculating their own number. When was the last time they walked the floor? Not toured. Not visited for an announcement. Actually walked the process, stopped at the boards, talked to the teams, and followed up on what they heard.
Most of them can't remember.
Here's why Gemba walks matter so much, and why no data system can replace them.
First, they create a feedback loop that no report can replicate. When you're standing next to an operator and she tells you that the changeover takes forty minutes because three of the parts she needs are stored in a different building, you don't just learn about the changeover problem. You learn about the storage problem, the layout problem, the procurement problem, and probably two other things she mentions in passing. Reports filter. Gemba walks reveal.
Second, they communicate priority more powerfully than any email, town hall, or strategy presentation. When a VP of operations walks past the CI board without stopping, that's a signal. When a VP of operations stops, reads, asks a question, and comes back the next week to check — that's a signal too. And everyone within line of sight receives it.
Third, they hold leaders accountable to reality. It's very easy to sit in a conference room, look at green-status dashboards, and conclude that everything is fine. It's much harder to maintain that illusion when you're standing on the floor watching a process that clearly isn't working as the dashboard suggests.
I've watched leaders walk a process and say, within five minutes, "wait — this isn't what the data shows." Exactly. That's the point. The gap between the data and the reality is where the value lives, and you can't find it from your office.
Now, I'm not naive. I understand the pressures on senior leaders. Their calendars are crushing. They have board meetings, investor calls, regulatory requirements, strategic decisions that genuinely require their attention. I'm not suggesting they spend all day on the floor.
But I am suggesting this: if you can't find forty-five minutes a week to walk the process, you are telling your organisation something about your priorities that no amount of emails will undo.
Forty-five minutes. Once a week. Three questions at each stop: What's working? What's blocking you? What support do you need?
That's the minimum cadence that sustains a CI programme. I've tested this across enough organisations to be confident in the number. Below forty-five minutes a week, programmes erode. Above it, they build.
At Shell, the strongest sites had leadership Gemba cadences of two to three times per week. The weakest had cadences measured in months. The correlation between Gemba frequency and programme health was the most reliable predictor I found in thirteen years.
There's one more thing I want to say about this, and it's the thing that separates a genuine Gemba walk from a management tour.
A tour is about the leader. The leader walks around, sees things, receives information, and leaves. The information flows one way: from the floor to the leader.
A Gemba walk is about the team. The leader walks around, observes, asks questions, and then acts on what they learn. The information flows both ways. The leader takes something away. But more importantly, the team receives something: evidence that their work matters, that their voice is heard, that the leader is engaged enough to come back and follow up.
The follow-up is everything. A Gemba walk without follow-up is worse than no walk at all, because it tells people that leadership listens but doesn't act. And that's a more damaging signal than simple absence.
So here's my challenge.
Whatever your role — CEO, VP, director, plant manager — calculate your number. How many days since your last Gemba walk? Not a tour. Not an inspection. A genuine walk where you observed the process, asked the people doing the work what they needed, and followed up afterwards.
If the number is under seven, you're probably in good shape. Keep going.
If the number is over thirty, your CI programme is likely in trouble, regardless of what the project tracker says. The governance may still be there. The structure may still be there. But the life is leaving.
The fix is not complicated. It's not expensive. It doesn't require a new framework, a new training programme, or a new consultant. It requires you to put forty-five minutes on your calendar, walk the process, and follow up.
That's it. That's the intervention. And it predicts programme survival more accurately than anything else I've ever measured.
Are you willing to start counting?
If you want to build the discipline and capability to sustain CI programmes at scale — including the leadership rhythms that make Gemba walks effective, not just performative — the CI Practitioner Mastery Program covers this in depth across sixteen modules. Find it at stormholt.org/products/stormholt-self-paced-continuous-improvement-foundation.
New to Stormholt? Start here — it locates your stage in minutes.
Building capability? Stage III — Skill — the tools, the courses, the simulations.