What 27 Years of CI Deployment Actually Taught Me
The uncomfortable truth about Continuous Improvement
After 27 years of deploying Continuous Improvement at GE Lighting, Royal Mail, Credit Suisse, NAGE, Shell, KCA Deutag and Johnson Controls, I have worked with every major methodology the industry has produced. Lean, Six Sigma, Kaizen, Scrum, Agile, Hoshin Kanri. I have trained Master Black Belts in three continents and installed cadence disciplines in 16 countries.
Here is the uncomfortable truth. The methodology does not matter as much as people think it does.
Lean works. Six Sigma works. Kaizen works. Scrum works. Agile works. I have seen every one of these frameworks produce real, sustainable, measurable improvement when deployed well.
I have also seen every one of them fail spectacularly when deployed badly.
The three reasons CI programmes fail
Strip away the methodology debate and CI programmes die for three reasons. The same three reasons, across every industry I have worked in — light source manufacturing, retail banking, private banking, upstream oil and gas, drilling operations, HVAC services, carbon capture, back office, HR.
Reason one: leadership behaviour
The number one predictor of CI failure is the gap between what leaders demand from teams and what leaders are willing to do themselves.
I watched a European manufacturer invest €2M in Black Belt certifications over four years. Thirty certified Black Belts. Zero measurable improvement in the operation.
The reason was not the training. The training was excellent. The reason was that while the Black Belts were learning A3 problem solving, their CEO was still holding three-hour operations reviews with a projector every Monday morning. Every Black Belt in that room understood the message: improvement is something THEY do, not something WE do.
The CI programmes I have seen succeed — at Royal Mail, at NAGE, at Shell Malaysia — all share one trait. The senior leaders changed THEIR behaviour first. They shifted their own meetings. They walked the Gemba personally. They wrote Kaizen cards. They stopped using production reviews to assign blame.
Your team will mirror what you actually do. Not what you ask them to do.
Reason two: absence of cadence
"We did Kaizen last year."
If you have said this, you did not do Kaizen. You did a workshop. Kaizen — same as Lean, same as Scrum — is a cadence, not an event. The point is not the improvement. The point is the discipline of repeatedly identifying and acting on improvements.
The cadence that dies first is always the same. Weekly improvement review. Someone has a crisis. The review gets cancelled. The team gets the message that the cadence is optional. Within three months, the cadence is dead.
At Shell Malaysia between 2016 and 2020, the leadership team never cancelled the weekly cadence. Not during COVID, not during oil price volatility, not during turnarounds. Four consecutive years of unbroken discipline. That is what earned the Malaysian Ministry of International Trade and Industry Gold Certification — the first for oil and gas in the country. That is what made Masaaki Imai award it Best CI Deployment in Malaysia at the Kaizen International Convention in Putrajaya 2018.
No better template. No cleverer framework. Just a cadence that never stopped.
Reason three: tool obsession
Every failed CI programme I have diagnosed spent too much time on the tools and not enough on the thinking.
"We need the right A3 template before we start." "We need to choose between Scrum and Kanban." "We need to standardise on a value stream mapping tool."
None of this matters in the first 90 days. What matters is that someone with authority is paying attention to the right thing every week, and that the team has permission to act on what they find.
A whiteboard and the right question outperform any enterprise software in the opening phase. The software comes later, once the discipline is established.
What the successful programmes have in common
Three patterns across every CI programme I have seen sustain past five years.
Leaders who change their own meetings first
Before asking teams to run daily stand-ups, the CEO commits to a weekly improvement review that they run personally. Not their deputy. Them. For the first 24 months minimum.
Cadence that survives crisis
The weekly improvement review that does not get cancelled when things go wrong. Especially when things go wrong. This is the test. Any CI programme that gets paused during acute pressure will die shortly afterwards.
Capability in operations, not a CI function
The CI team's job is to build capability in operations and work themselves out of a job. Most CI programmes fail because the CI team becomes a permanent parallel function. When they eventually leave or lose resources, operations did not learn how to run improvement on their own.
The methodology question
Once leaders are committed, cadence is unbroken, and operations owns capability — the methodology you choose matters less than you would think.
Use Lean when the problem is waste in flow. Use Six Sigma when the problem is variation in outputs. Use Kaizen when the problem is that small improvements are not being captured. Use Scrum when the work is project-based and cross-functional. Use Agile when priorities change faster than plans can keep up.
All of these tools should be in a senior CI practitioner's kit. The religious debates between them — Lean purists vs Six Sigma purists vs Agile purists — are a distraction from the real work.
The discipline is the differentiator. Not the framework.
If your CI programme is struggling right now
Four diagnostic questions. The honest answers will reveal what is actually broken.
Is a senior leader personally running a weekly improvement review? Not delegating — running.
Has that review been cancelled or shortened in the last 90 days for any reason?
Who owns sustaining the improvements once a project wraps up — the CI team or the operation itself?
When did your last CEO last write a Kaizen card or spend 30 minutes walking the Gemba without a meeting to rush to afterwards?
If the answers are uncomfortable, you have found the problem.
The Stormholt view
ENDURANCE OVER HEROICS
CI programmes that work are boring. Same cadence for 500 weeks. Visual management boards that stay current. Daily stand-ups that happen at 7:45am every day regardless of who is in the room. A3 reviews where the biggest problem gets named out loud. The CEO who writes a Kaizen card in year ten because they noticed it mattered in month one and never stopped noticing. This is what Stormholt exists to teach and sustain.
What to do tomorrow morning
Pick one of the following and commit to it for 12 weeks.
Start a 15-minute standing daily review on the operations floor — 7:45am, no chairs, no laptops, five-part agenda, same every day.
Block 30 minutes every Monday morning to walk the Gemba without an agenda. Do not bring a meeting to rush to afterwards.
Commit to a weekly improvement review that you personally chair. Do not cancel it. Not once.
If you do any one of these for 12 weeks, the pattern will start to appear. If you do all three, the compound effect is remarkable.
This is not about methodology. This is about discipline. Every methodology works when the discipline is there. No methodology works when it is not.
What is the CI practice you know you should restart but have not yet? Why not this week?
#ContinuousImprovement #Leadership #Lean #Kaizen #Scrum #OperationalExcellence
Stay sharp.
If you want one operator-grade Stormholt note a week, the Operator Network is open. No fluff. No selling. One email. One question. stormholt.org/op
If you want to walk one of your real problems end to end with the founder, the Architecture Session is the smallest commitment: €750, 60 minutes, 1-page A3 in 48 hours.
Endurance over heroics. — Stormholt