The Promise That Sounded
Perfect
Kaizen is the Toyota Production System’s heartbeat. The word itself —
kai (change) plus zen (good) — translates simply:
change for the better. Not revolutionary change. Not disruptive change.
Small, incremental, continuous change. The idea is democratic: every
worker, every day, looks at their workspace and asks, “What is one thing
I could improve?” Over weeks, months, years, those small improvements
compound into transformational results.
It is beautiful in theory. It is also one of the most corrupted
concepts in modern manufacturing.
Walk into any factory that claims to practice kaizen and you will
find one of two failures. Either kaizen is a word used in meetings but
never practiced on the floor — a slogan on a banner above a production
line where nothing has changed in three years — or kaizen has
metastasized into a relentless churn of changes that never settle, never
stabilize, and never deliver the compounded benefits they promised. The
first failure is obvious. The second is the one nobody talks about.
This article is about the second one.
What Kaizen
Actually Means (Before We Destroy It)
To understand how kaizen breaks, you first need to understand what it
was supposed to be.
At Toyota, kaizen was never just “make changes.” It was a structured
discipline built on the foundation of standard work. You could not
improve what you had not standardized. The sequence was: establish a
standard, follow the standard, find a better way, make that the new
standard, repeat. The standard was the anchor. The improvement was the
wave. Without the anchor, you are not improving — you are drifting.
Taiichi Ohno, the architect of the Toyota Production System, was
emphatic about this. He did not want creative people inventing new
methods every day. He wanted people who would follow the standard
exactly, observe the gaps between standard and reality, propose a
specific improvement, test it, and — only if it worked — make it the new
standard. The creativity was in the observation and the experiment, not
in the constant reinvention.
This is the part that most manufacturers miss.
Failure
Mode One: The Kaizen Event Industrial Complex
The most common corruption of kaizen in Western manufacturing is the
“kaizen event” — also called a kaizen blitz, a rapid improvement event,
or a continuous improvement workshop. The format is familiar: pull five
to ten people off their normal jobs for three to five days, give them a
facilitator, point them at a problem, and expect dramatic results by
Friday afternoon.
Some of these events produce genuine improvements. Many produce
theater. Here is why.
The deadline distorts the problem. When you have
five days to show results, you do not tackle the hardest problem. You
tackle the problem you can solve in five days. This means the structural
issues — the ones causing 80% of the waste — sit untouched while teams
rearrange workbenches and add shadow boards. The visible energy of
improvement is high. The actual impact is low. But leadership sees
people busy and declares victory.
The team has no skin in the game. A kaizen event
team is often assembled from people who do not own the process being
improved. They are smart, they mean well, but they will walk away on
Friday and never touch this process again. The people who do
own the process — the operators who run it every day — are frequently
excluded from the event or included as a single voice drowned out by
engineers and managers. So you get solutions designed for operators
without operators.
The sustainment plan is a lie. Every kaizen event
ends with a report-out. The report-out includes a sustainment plan. The
sustainment plan is a slide that says “We will audit monthly and hold
the gains.” Nobody audits monthly. Nobody holds the gains. Within six
weeks, the shadow boards are empty, the new standard work is ignored,
and the process has reverted to exactly what it was before — except now
there is residual cynicism about the next kaizen event, because everyone
remembers the last one changed nothing that lasted.
When you run kaizen events every month and none of the changes stick,
you are not practicing continuous improvement. You are practicing
continuous disruption. You are pulling people off their jobs, generating
paperwork, creating stress, and delivering nothing durable. The cost is
enormous — not just in labor hours, but in trust.
Failure
Mode Two: Change Fatigue and the Loss of Standard Work
The deeper damage from corrupted kaizen is change fatigue. And it is
the most dangerous failure mode in manufacturing quality, because it is
invisible until it is catastrophic.
Here is how it happens. A factory decides to “embed kaizen into the
culture.” Every team is expected to implement at least one improvement
per month. Some set a quota: five improvements per person per year. The
metrics look great. The improvement boards are covered with sticky
notes. Leadership celebrates the count.
But each improvement changes the process. Each change requires
retraining. Each retraining requires time. And each change interacts
with every other change — often in ways nobody anticipated, because
nobody is managing the system of changes holistically. Operator A’s
improvement to the feed mechanism interacts with Operator B’s
improvement to the inspection station, and the result is a new failure
mode that neither of them caused individually.
When changes happen faster than the organization can stabilize them,
you lose standard work. And when you lose standard work, you lose
quality. This is not a hypothesis — it is a mathematical certainty.
Quality requires repeatability. Repeatability requires stability.
Stability requires a process that stays the same long enough to be
understood. If the process is different every week — because someone is
always “improving” it — you can never establish a baseline, never
identify special-cause variation, and never know whether a quality
problem is caused by the process or by the latest well-intentioned
change to the process.
I have seen factories where the improvement program was so aggressive
that defect rates increased after it was launched. Each new
improvement introduced a new variable. Each new variable created new
variation. Each new variation produced new defects. The factory was
improving itself into chaos.
Failure Mode
Three: Improvement as Performance Art
The third failure mode is the most insidious because it is the
hardest to detect from the outside. This is when kaizen becomes a
performance — something done to demonstrate engagement rather than to
achieve results.
You have seen this if you have spent any time in manufacturing. The
improvement board has a section for “before” photos and “after” photos.
The before photos show a messy workbench. The after photos show a tidy
workbench with colored tape on the floor. The improvement was: we
cleaned up. This is not improvement. This is housekeeping presented as
innovation.
But the organization rewards it, because it is measurable (we can
count the sticky notes), visible (we can photograph the changes), and
safe (nobody challenged a process owner or questioned a design
decision). Real kaizen is uncomfortable. Real kaizen asks why a process
was designed badly in the first place. Real kaizen challenges the
engineer who specified the tolerance, the supplier who provided the
material, the manager who scheduled the run. Performance-art kaizen
avoids all of that and sticks to rearranging the tool cabinet.
When the reward system celebrates quantity of improvements over
quality of improvements, you get exactly what you incentivize: a flood
of trivial changes and a drought of meaningful ones. The engineers who
could redesign the fixture to eliminate the adjustment entirely are too
busy facilitating kaizen events to do engineering. The operators who
know the process is flawed are too busy filling out improvement forms to
fix it.
The Cost of Continuous
Disruption
Let me quantify what this costs, because the cost is the part that
gets ignored.
A typical kaizen event consumes 40 person-days of effort (8 people ×
5 days). At a loaded labor rate of $60/hour, that is roughly $19,200 per
event in direct labor alone. Add facilitator costs, materials, lost
production, and the overhead of reporting and follow-up, and you are
easily at 30, 000–50,000 per event.
Now ask: how many of last year’s kaizen events produced improvements
that are still in effect today? If the answer is less than half — and in
most organizations it is — then you are spending $25,000+ per event to
generate temporary changes that revert within months. That is not
continuous improvement. That is a recurring expense disguised as an
investment.
The opportunity cost is worse. While your best people are running
kaizen events on minor problems, the major problems — the ones that
actually threaten your competitiveness — go unaddressed. The supplier
quality issue that causes 60% of your field returns does not get solved
in a five-day event. The equipment reliability problem that causes your
biggest capacity constraint does not get solved by a team that includes
no maintenance engineers and no capital budget. Real problem-solving
requires depth, time, and authority — the three things that kaizen
events are structurally designed to avoid.
How to Tell If Your Kaizen
Is Broken
Here are the diagnostic signs. If three or more are true in your
organization, your kaizen program is likely producing disruption rather
than improvement:
-
You count improvements rather than measuring their
impact. If your kaizen metric is “number of improvements
implemented,” you are optimizing for activity, not results. The
organizations that benefit from kaizen measure outcomes — defect rates,
cycle times, cost per unit — and treat improvements as means, not
ends. -
Your standard work documents change more than once a
quarter. If operators cannot remember the current standard
because it changed last week, you do not have standard work. You have a
draft. -
The same problems appear on the kaizen board year after
year. If the problem list never gets shorter — if you are
“continuously improving” the same bottleneck for the third year in a row
— your improvements are not compounding. They are replacing each
other. -
Improvements are concentrated in low-impact
areas. If your kaizen events focus on 5S, workspace
organization, and visual management while your top three quality issues
(which everyone can name) have never been the subject of a kaizen event,
you are improving the easy 20% and ignoring the important 80%. -
Operators cannot explain why the current standard is what
it is. This means the standard was imposed, not understood. And
a standard that is not understood will not be followed — which means it
cannot be improved. -
Your improvement rate goes up while your quality metrics
stay flat. This is the clearest sign. More improvements with no
quality improvement means the improvements are not improvements. They
are changes.
What Real Kaizen Looks Like
Real kaizen is slower than you want it to be. That is not a bug — it
is the design.
At Toyota, a worker on the line who notices a problem does not
immediately change the process. They pull the andon cord, call their
team leader, and the problem is addressed in the moment. If the problem
is recurring, it goes to a problem-solving team that investigates root
cause, proposes a countermeasure, tests it in a controlled way, and —
only after the countermeasure is proven — updates the standard work.
This can take weeks for a single improvement. It is methodical. It is
deliberate. And because of that, the improvements stick.
The rate of improvement at Toyota is not dramatic on any given day.
But over a year, over a decade, the compounding is extraordinary. That
is the power of kaizen done right: not the speed of change, but the
durability of change.
The organizations that try to accelerate kaizen — that push for more
events, more improvements, more sticky notes, more activity — almost
always end up with less actual improvement than the organizations that
slow down, standardize rigorously, solve one problem at a time, and make
each solution permanent before moving to the next.
The Recovery Plan
If your kaizen program has become a disruption engine, here is how to
reset:
Stop. Freeze all kaizen events for 60 days. Use the
time to audit which previous improvements are still in effect. Be
honest. If fewer than 50% have survived, your process for generating
improvements is broken and needs to be redesigned before you generate
more.
Re-establish standard work. Before you improve
anything, you need a stable baseline. Document the current state of your
top three production processes. Not the ideal state — the actual state,
as practiced on the floor, including all the workarounds and deviations.
This is your starting point.
Shift from event-based to problem-based. Instead of
scheduling a kaizen event every month and looking for a problem to
solve, identify your top three quality or efficiency problems and
charter a team to solve each one. Give the team as long as it takes.
Measure success by whether the problem is solved, not by whether the
team finished in five days.
Change the metric. Stop counting improvements. Start
measuring: What percentage of our standard work documents have been
stable for more than 90 days? What percentage of our top-ten quality
issues from last year have been permanently resolved? How many of our
kaizen improvements from twelve months ago are still being followed
today? These metrics tell you whether you are improving or just
churning.
Protect the standard. Make it harder to change the
standard work, not easier. Require that any proposed improvement be
tested in a controlled trial, measured against the current standard, and
reviewed by the process owner before adoption. This is not bureaucracy —
it is quality control for your improvement process itself.
The Uncomfortable Truth
Kaizen is not about change. It is about better. Those are not the
same thing, and the conflation of the two is the most expensive
misunderstanding in modern manufacturing.
A factory that changes everything and stabilizes nothing is not
improving. A factory that stabilizes everything and changes nothing is
not improving either. The discipline — the hard, unglamorous, slow
discipline of real kaizen — is knowing the difference between a change
that makes things better and a change that just makes things different,
and having the patience to pursue only the former.
Your operators already know this. They watch improvement programs
come and go. They fill out the forms, participate in the events, pose
for the photos, and then go back to doing the job the way they have
always done it — because they know that next month’s improvement will
undo this month’s improvement anyway.
The question is not whether your organization can generate more
improvements. The question is whether your organization can make one
improvement that is still working on the day you retire.
That is kaizen. Everything else is noise.
About the Author: Peter Stasko is a Quality
Architect with over 25 years of experience in manufacturing quality
management, process improvement, and production system design. He has
implemented and audited quality systems across automotive, electronics,
and industrial manufacturing sectors, and writes about the gap between
quality theory and factory floor reality at iaec.online.
