Quality
and the Diderot Effect: When Your Organization’s One Improvement
Triggers a Cascade of Costly Upgrades Nobody Needed — and the Upgrade
You Never Required Became the Waste You Couldn’t Stop
How a single “improvement” can set off an chain reaction of
unnecessary changes, and what quality leaders can do to break the cycle
before it consumes their budget, their focus, and their
credibility.
The Scarf That Ruined
Everything
In 1765, the French philosopher Denis Diderot received a gift — a
beautiful scarlet dressing gown. It was elegant, luxurious, and far
finer than anything else he owned. Delighted, he put it on and
immediately noticed that his old wooden desk looked shabby beside it. So
he replaced the desk. Then the desk made the curtains look drab, so he
replaced those too. The curtains made the tapestries look faded, the
tapestries made the chairs look worn, and before long, Diderot had spent
himself into near poverty replacing everything in his home to match the
grandeur of one gift.
He wrote about it in an essay called “Regrets on Parting with My
Old Dressing Gown,” and the pattern has borne his name ever since:
the Diderot Effect — the tendency for one new
acquisition to trigger a cascade of replacements, upgrades, and
expansions that were never planned, never needed, and never
justified.
Organizations do this constantly. Not with dressing gowns, but with
quality systems, inspection equipment, software platforms, and
improvement initiatives. One upgrade — justified, reasonable, maybe even
necessary — sets off a chain reaction of additional upgrades that
consume budgets, distract teams, and create complexity far beyond what
the original problem required.
And the most dangerous part? Everyone along the chain genuinely
believes they are making things better.
How
the Diderot Effect Manifests in Quality Organizations
The pattern is remarkably consistent across industries. Here is how
it typically unfolds:
Stage 1: The Justified Upgrade. An organization
replaces an aging coordinate measuring machine with a modern vision
system. The old machine was inaccurate, slow, and could not handle the
tolerances the new product line demanded. The purchase is well-justified
with data, ROI calculations, and a solid business case. So far, so
good.
Stage 2: The First Cascade. The new vision system
produces more detailed reports than the old CMM ever did. The quality
engineering team realizes that the existing SPC software cannot handle
the volume of data, so they request an upgrade to a more capable
platform. This seems reasonable — better data deserves better tools.
Stage 3: The Acceleration. The new SPC platform has
features the old one did not — automated email alerts, mobile
dashboards, integration APIs. Someone in management sees the dashboards
and decides that all production lines should be connected to this
system. Now the organization is running cable, installing sensors on
machines that were previously unmonitored, and training operators on
data entry they never had to do before.
Stage 4: The Entanglement. The new sensors reveal
variation patterns that were previously invisible. This is actually
valuable information, but the organization responds not by analyzing and
acting on the patterns but by purchasing additional software to automate
the analysis. This software requires a dedicated server, which requires
IT support, which requires security protocols, which require audits.
Stage 5: The Regret. Eighteen months and $2.3
million later, the organization has a sophisticated, interconnected
quality data ecosystem that generates beautiful dashboards, sends
automated alerts to twenty-seven people, and produces reports nobody
reads. The original problem — an inaccurate CMM — was solved in month
one. Everything that followed was the Diderot Effect in action.
I have watched this happen in automotive, aerospace, medical device,
and electronics manufacturing. The details change. The pattern does
not.
Why Quality
Organizations Are Especially Vulnerable
Several factors make quality departments particularly susceptible to
the Diderot Effect:
The Perfectionist Culture. Quality professionals
are, by nature and training, dissatisfied with “good enough.” This is
what makes them effective at finding defects and driving improvements.
But it also means that once they see a gap between current capability
and potential capability, they feel compelled to close it — even when
the gap is theoretical, not practical.
The Technology Vendor Ecosystem. Quality technology
vendors understand the cascade effect intimately. They price entry-level
systems affordably and load higher-tier packages with features that seem
indispensable once you are already using the platform. The vision system
vendor recommends the SPC integration module. The SPC vendor recommends
the analytics add-on. The analytics vendor recommends the predictive
maintenance extension. Each step feels incremental. The total is
anything but.
Audit and Compliance Pressure. When an auditor flags
a gap, organizations often overcorrect. A single finding about
inadequate traceability leads to a full digital transformation of the
document control system. A single nonconformance about insufficient gage
calibration leads to a completely new metrology lab. The fear of the
next audit becomes the justification for the next purchase.
Benchmarking Against Peers. Industry conferences and
trade shows are accelerants for the Diderot Effect. A quality manager
sees what a competitor or peer company has implemented and returns home
convinced that their organization is falling behind. The fact that the
peer company has three times the production volume, ten times the
quality staff, and a completely different product architecture does not
enter the analysis. Envy drives acquisition.
The Sunk Cost Rationalization. Once the cascade
begins, each new purchase is justified by the previous ones. “We have
already invested in the vision system and the SPC platform — it would be
wasteful not to connect them.” “We have the data infrastructure in place
— we might as well add the analytics module.” The sunk cost fallacy and
the Diderot Effect reinforce each other in a feedback loop that is very
difficult to interrupt.
The Cost of the Cascade
The financial cost of the Diderot Effect is significant, but it is
not the most damaging consequence. The true costs are:
Attention Fragmentation. Every new system requires
training, maintenance, troubleshooting, and attention. Quality teams
that were focused on solving production problems are now managing
software licenses, attending vendor webinars, and configuring
dashboards. The core mission — preventing defects, reducing variation,
improving processes — gets crowded out by system administration.
Change Fatigue. Operators, technicians, and
engineers who are subjected to a continuous stream of “improvements”
develop a specific kind of exhaustion. They stop engaging with new
initiatives because they have learned that today’s transformation will
be replaced by tomorrow’s. This is how organizations breed cynicism, and
cynicism is the enemy of genuine quality improvement.
Complexity Risk. Every additional layer in the
quality system introduces new failure modes. The integrated
vision-SPC-analytics-dashboard ecosystem has dozens of integration
points, each one a potential source of error. The organization that once
had a simple, understandable measurement process now has a system that
nobody fully understands and nobody can troubleshoot without calling
three different vendors.
Credibility Erosion. When quality leaders champion
each successive upgrade with the same enthusiasm and the same promises
of transformation, the rest of the organization stops believing them.
The production manager who has sat through six “this will change
everything” presentations in two years has stopped listening. The plant
manager who approved the last three quality technology purchases and saw
no measurable improvement in defect rates has stopped trusting the
quality department’s judgment.
Recognizing the
Diderot Effect in Real Time
The most insidious aspect of the Diderot Effect is that each
individual decision seems reasonable in isolation. The cascade is only
visible from a distance. Here are the warning signs:
The “While We’re At It” Justification. If the
primary argument for a purchase begins with “While we are upgrading X,
we might as well…” — you are in the cascade.
The Escalating Business Case. If the ROI calculation
for each successive purchase is weaker than the one before it, but the
purchases keep getting approved because they are “consistent with the
direction we are heading” — you are in the cascade.
The Unfinished Previous Initiative. If your
organization has not fully implemented the last quality improvement
before proposing the next one — you are in the cascade. Full
implementation means the system is operational, people are trained, the
data is being used to make decisions, and measurable results have been
documented. If you cannot check all four boxes, you are not ready for
the next thing.
The Feature-Chasing Pattern. If your team spends
more time discussing what a new system could do than what your
operation actually needs — you are in the cascade.
The Vendor-Driven Roadmap. If your quality
technology roadmap looks suspiciously like your vendor’s product catalog
— you are in the cascade.
Breaking the Chain:
Practical Strategies
If you recognize the Diderot Effect in your organization, here are
concrete steps to interrupt it:
1. Institute the “Solve
the Problem” Rule
Before approving any quality technology purchase, require a clear,
one-sentence statement of the specific problem it solves. Not the
opportunity it creates. Not the capability it enables. The
problem it solves. If the problem cannot be stated in
one sentence, the purchase is a want, not a need.
2. Apply the Three-Month Test
Ask: “If we implement this and then change nothing else for three
months — no additional purchases, no system expansions, no cascade —
will this purchase still deliver its promised value?” If the answer is
no, the purchase is dependent on the cascade, and the cascade is the
real cost.
3. Mandate a “Last Thing
First” Review
Before proposing any new quality initiative, require the team to
demonstrate the measurable results of the previous initiative. This
forces the organization to close the loop on each improvement before
opening a new one. It also provides natural resistance to cascade
thinking, because the team must show that the last purchase actually
delivered value before being allowed to propose the next one.
4. Create a “Good Enough”
Threshold
Not every capability gap needs to be closed. Define explicit
performance thresholds for quality systems and processes, and declare
anything above those thresholds “good enough” — not as a permanent
judgment, but as a deliberate boundary that prevents cascade-driven
upgrades. Revisit the thresholds annually, not quarterly.
5. Separate Improvement
from Expansion
Make a clear organizational distinction between improving existing
systems (better calibration, faster cycle times, more consistent
execution) and expanding into new systems (additional sensors, new
software modules, integration projects). Apply different approval
criteria to each. Improvement should be encouraged. Expansion should be
scrutinized.
6. Calculate Total Cascade
Cost
When a quality technology purchase is proposed, require the business
case to include not just the cost of the purchase itself but the
estimated cost of all likely downstream purchases it will trigger. If
buying the SPC platform will inevitably lead to the analytics module,
the predictive maintenance extension, and the mobile dashboard — include
all of those costs in the initial business case. If the total cascade
cost is not justified, neither is the initial purchase.
7. Implement a “Cooling Off”
Period
For any quality technology purchase above a defined threshold,
require a mandatory waiting period between the proposal and the
approval. Thirty days is usually sufficient. The purpose is not to delay
genuine needs but to interrupt the emotional momentum that drives
cascade purchasing. If the purchase still looks as compelling after
thirty days of reflection as it did in the vendor demo, it is probably
justified.
The
Counter-Argument: When the Cascade Is Justified
To be fair, not every cascade is irrational. Sometimes an
organization’s quality infrastructure genuinely is outdated, and a
comprehensive upgrade — even one that touches multiple systems — is the
right thing to do. The key distinction is between planned
transformation and reactive cascade.
A planned transformation is characterized by: – A clear end state
defined before the first purchase – A total budget established before
the first purchase – A realistic timeline with defined milestones –
Leadership commitment to completing the transformation before starting
anything new – A genuine business case for the entire transformation,
not just the first step
A reactive cascade is characterized by: – No clear end state (“we
will see what we need after this first step”) – No total budget (“we
will request additional funding as needs emerge”) – No defined timeline
(“this is an ongoing journey”) – Leadership approval for each individual
step without visibility into the total – Business cases that become
progressively weaker as the cascade continues
If your organization is in the middle of a planned transformation,
stay the course. If you are in a reactive cascade, stop and
reassess.
A Personal Observation
In twenty-five years of quality work across dozens of organizations,
I have seen the Diderot Effect play out more times than I can count. And
I have been complicit in it myself. Early in my career, I championed
technology purchases and system expansions that I believed were
genuinely necessary, only to realize years later that I had been caught
in the cascade. The dressing gown was beautiful. The desk did look
shabby. The curtains did need replacing. Each individual decision was
defensible. The total was indefensible.
The organizations I have seen achieve the most sustained quality
improvements are not the ones with the most sophisticated technology
stacks. They are the ones that mastered a small number of tools, applied
them consistently, and resisted the temptation to upgrade before they
had extracted full value from what they already had.
They understood something that Diderot did not: the best way to enjoy
a beautiful dressing gown is to put it on, sit down at your old wooden
desk, and get to work.
The Bottom Line
The Diderot Effect teaches us that individual rational decisions can
combine into collectively irrational outcomes. In quality management,
this means that a series of seemingly reasonable upgrades — each one
justified, each one modest, each one an incremental improvement — can
cascade into a transformation the organization never planned, cannot
afford, and does not need.
The antidote is not to stop improving. It is to improve deliberately
— with clear problems defined, full costs calculated, previous
improvements validated, and the discipline to say “good enough” when
good enough is exactly what it sounds like.
Your old dressing gown kept you warm. Think carefully before you
trade it in.
About the Author
Peter Stasko is a Quality Architect with over 25 years of experience
in manufacturing quality across automotive, aerospace, electronics, and
medical device industries. He specializes in helping organizations build
quality systems that are effective without being excessive — systems
that solve real problems without creating new ones. His work focuses on
the intersection of human psychology and operational excellence, drawing
on decades of field experience to help leaders see the patterns that
drive both quality success and quality failure.
