Quality
and Conway’s Law: When Your Organization’s Structure Becomes Your
Quality System’s Architecture — and the Silos You Built Between
Departments Became the Gaps Your Defects Flow Through

Melvin Conway wrote a single sentence in 1967 that would eventually
reshape how the technology industry thinks about design. His observation
was devastatingly simple: “Any organization that designs a system
will produce a design whose structure is a copy of the organization’s
communication structure.” He was talking about software. But he
might as well have been talking about your quality management
system.

Because here is the uncomfortable truth that most quality
professionals learn too late: your QMS doesn’t look the way it does
because of ISO 9001. It doesn’t look the way it does because of your
customers’ requirements. It doesn’t even look the way it does because of
your quality policy. It looks the way it does because of your org
chart.

And that is a problem.

The Architecture of
Accidents

Consider a typical manufacturing organization. Engineering designs
the product. Purchasing sources the components. Production builds the
assembly. Quality inspects the result. Shipping sends it to the
customer. Each department has its own manager, its own metrics, its own
meetings, its own language. And each department has a boundary — an
invisible wall where its responsibility ends and another department’s
begins.

Now look at your quality management system. I’ll wager it has the
same structure. There’s a design control process owned by Engineering. A
supplier quality process owned by Purchasing. A production quality
process owned by Manufacturing. A final inspection process owned by
Quality. A customer complaint process owned by — well, that depends on
who gets the email first.

Your QMS mirrors your org chart because the people who wrote it could
only coordinate within the communication channels they had. The
engineers wrote the design procedures because they talk to each other.
The purchasing team wrote the supplier procedures because they talk to
their suppliers. The quality team wrote the inspection procedures
because they talk to their inspectors. Nobody wrote the procedure for
what happens between these domains because nobody talks to
anyone outside their own department long enough to realize those gaps
exist.

Conway’s Law predicts this precisely. The structure of your quality
system is not the structure of your process. It is the structure of your
communication. And wherever communication breaks down, your quality
system has a hole that defects pass through like water through a cracked
dam.

The Handoff Problem

Let me tell you about a company I worked with — a mid-size automotive
supplier producing precision-machined housings for transmission systems.
They had a robust APQP process. Their PFMEA was 400 lines long. Their
control plan covered every dimension, every tolerance, every feature.
They had passed every customer audit for five consecutive years.

And yet they were shipping parts with a persistent burr defect on a
critical sealing surface — a defect that caused oil leaks in the field,
warranty claims in the millions, and a quality alert from their customer
that threatened to cost them the program entirely.

The root cause investigation was revealing. Engineering had designed
the feature with a sharp edge transition because their CAD software made
it easy and their design review focused on dimensions, not edge
conditions. Manufacturing had programmed the toolpath to minimize cycle
time, and the sharp transition created a natural burr that the tool left
behind. Quality inspected the part against the drawing — and the drawing
didn’t specify burr limits. Shipping packed the parts in bulk, and the
rubbing during transit actually made the burrs worse.

Every department had done its job correctly within the boundaries
of its own procedures. Every department’s process was internally
consistent, properly documented, and regularly audited. The defect
wasn’t a failure of any single department. It was a failure of the space
between departments — the handoff points where information should have
flowed but didn’t because the organizational structure didn’t require it
to.

The engineering team didn’t know that their edge transition created
burrs because they never talked to the machinists. The machinists didn’t
report the burrs because the burrs weren’t on the inspection checklist.
The inspectors didn’t flag the burrs because the drawing didn’t call
them out. And the customer suffered the consequences of a defect that
existed entirely in the gaps between organizational silos.

This is Conway’s Law in action. Your quality system is designed to
match your communication structure. If your communication structure has
gaps, your quality system has gaps. And defects don’t respect
departmental boundaries. They flow through the gaps like electricity
through the path of least resistance.

The Mirror in Your
Procedures

Here is an exercise that will change how you see your quality system.
Take your QMS documentation and map it to your organizational chart. Not
the other way around — don’t map your org chart to your procedures.
Start with the procedures and ask: who wrote this? Who owns this? Who
reviews this? Who is this procedure designed to serve?

What you will find, more often than not, is that each procedure
serves the department that wrote it. The incoming inspection procedure
is written from Quality’s perspective — it tells inspectors what to
check, not what Purchasing should have verified before the material
arrived. The corrective action procedure is written from the quality
manager’s perspective — it describes the CAPA form and the closure
criteria, not the cross-functional investigation that should have
happened on the shop floor. The design control procedure is written from
Engineering’s perspective — it describes the phase-gate reviews and the
approval signatures, not the manufacturing feasibility questions that
Production needed answered before the design was frozen.

Each procedure is a reflection of the department that created it. And
the interfaces between procedures — the critical moments where
information must pass from one domain to another — are either poorly
defined, inconsistently executed, or completely absent.

This is not an accident. It is Conway’s Law at work. People write
procedures for the people they communicate with. They don’t write
procedures for the people they don’t communicate with — because they
don’t know what those people need.

The Cross-Functional Myth

I can already hear the objection: “But we have cross-functional
teams. We have production meetings. We have management reviews. We
communicate across departments all the time.”

No. You don’t.

What you have are periodic events where representatives from
different departments sit in the same room and take turns presenting
information to each other. That is not cross-functional communication.
That is parallel monologue.

True cross-functional communication means that the person designing
the part understands the constraint of the machine that will produce it.
It means the person inspecting the product understands the customer’s
application well enough to know which deviations matter and which don’t.
It means the person buying the material understands the process well
enough to know that a cheaper supplier with a slightly different alloy
will cause a downstream quality problem that costs ten times the
savings.

This kind of deep, embedded, continuous cross-functional
understanding cannot be achieved through weekly meetings. It requires
something far more structural: shared accountability, shared metrics,
shared workspaces, and — most importantly — shared problems.

Inverse Conway’s Maneuver
for Quality

The technology industry has developed a response to Conway’s Law that
quality professionals should steal immediately. It’s called the Inverse
Conway Maneuver, and the idea is brilliantly simple: instead of letting
your organizational structure dictate your system architecture, you
deliberately restructure your organization to produce the system
architecture you want.

Applied to quality, this means starting with the question: What
kind of quality system do we need? And then restructuring the
organization so that its communication channels naturally produce that
system.

If you want a quality system that catches defects at the source, you
need an organization where the people who create the product and the
people who verify the product are in constant, natural communication —
not separated by departmental walls and formal inspection request forms.
If you want a quality system that prevents problems rather than
detecting them, you need an organization where engineering and
manufacturing share not just meetings but objectives, not just project
timelines but physical workspace.

Some practical applications of the Inverse Conway Maneuver for
quality:

Create quality pods, not quality departments.
Instead of a centralized quality function that serves as an external
auditor to production, embed quality engineers directly into production
cells. Give them the same goals as the production team. Make them
responsible for the cell’s output — not just its inspection results, but
its delivery, its cost, and its customer satisfaction. When quality
people and production people share the same daily experience, the
procedures they write together will naturally address the interfaces
between their domains.

Make handoff points explicit and owned. In software,
this is called defining APIs — the contracts between different
components of a system. In quality, it means defining exactly what
information, in what format, with what level of confidence, must pass
between each functional boundary. And it means assigning ownership of
those handoffs to a named individual — not a department, a person.

Measure the gaps, not just the functions. Most
quality metrics measure departmental performance: first-pass yield,
scrap rate, on-time delivery, customer complaints. These are important,
but they don’t measure the quality of the interfaces between
departments. Add metrics that specifically target the handoff points:
how many defects originated at a functional boundary? How many
corrective actions trace back to a communication failure between
departments? How much time is spent on rework caused by incomplete
information transfer?

Rotate people across boundaries. Nothing breaks down
organizational silos faster than people who have worked on both sides of
them. When an engineer has spent six months on the production floor,
they design differently. When an inspector has spent time in the
customer’s plant, they evaluate defects differently. When a purchasing
agent has run the machine that processes the material they source, they
select suppliers differently. These rotations create personal
communication channels that persist long after the rotation ends — and
those channels become the pathways through which the quality system
naturally integrates.

The Silo Multiplier Effect

There is a mathematical dimension to Conway’s Law that most
organizations underestimate. The number of communication channels in an
organization grows factorially with the number of people — but the
number of effective communication channels grows much more
slowly, because most channels are blocked by organizational
boundaries.

What this means in practice is that adding departments doesn’t add
communication capacity — it subtracts it. Every new departmental
boundary creates a new interface that must be managed, a new handoff
point where information can be lost, a new place where defects can hide.
An organization with four departments has six pairwise interfaces to
manage. An organization with eight departments has twenty-eight. The
complexity doesn’t grow linearly. It grows combinatorially.

And here is the cruelest part: each interface is a potential quality
gap, but the interfaces are exactly where the fewest resources are
invested. Departments have budgets, headcount, and management attention.
The spaces between departments have — usually — nothing. Maybe a shared
spreadsheet. Maybe a standing meeting that everyone dreads. Maybe a
procedure that hasn’t been updated since the last audit.

Your quality system is only as strong as its weakest interface. And
Conway’s Law guarantees that the interfaces are where your organization
is weakest.

The Customer Sees One System

Here is the ultimate irony of Conway’s Law in quality: your customer
doesn’t experience your departments. They don’t experience your
engineering process, your manufacturing process, and your inspection
process as separate events. They experience one product. One
interaction. One quality outcome.

When a defect reaches the customer, the customer doesn’t care whether
it originated in design, manufacturing, or inspection. They don’t care
about your org chart. They care that the product doesn’t work. And every
defect that crosses a departmental boundary — every gap that exists
because your communication structure doesn’t match your process
structure — is a defect that your organization structured itself to
miss.

The customer experiences an integrated system. Your job is to build
one. And you cannot build an integrated system with a disintegrated
organization.

Breaking the Mirror

Conway’s Law is not a law of physics. It’s a pattern — a strong
tendency, but one that can be overcome with deliberate effort. The
organizations that produce quality systems transcending their
organizational structures are the ones that recognized the mirror effect
and chose to break it.

They break it by making communication explicit where it was implicit.
By assigning ownership where there was ambiguity. By measuring what
happens between departments, not just within them. By rotating people,
sharing objectives, and creating shared problems that no single
department can solve alone.

They break it by understanding that a quality management system is
not a document. It is a communication system. And the most important
communication in your organization doesn’t happen within departments. It
happens between them.

Your org chart is not your quality system. But if you’re not careful,
it will become one.

Peter Stasko is a Quality Architect with 25+ years
of experience transforming organizations across automotive, aerospace,
and pharmaceutical industries. He specializes in bridging the gaps
between departments, functions, and processes — because he’s learned
that the most expensive quality failures don’t happen within a single
function. They happen in the spaces between them.