Quality
APQP: When Your Organization Stops Launching Products by Hope and Starts
Planning Excellence Before the First Part Is Ever Made
The Launch That Wasn’t
In the spring of 2019, a mid-sized automotive supplier in central
Europe won a contract they’d been chasing for three years. A major OEM
had selected them to produce a critical brake component for a new
vehicle platform. The celebration in the boardroom lasted an entire
afternoon. The managing editor called it a turning point. The quality
manager smiled politely and said nothing.
He said nothing because he’d seen this movie before.
Eighteen months later, that same supplier lost the contract. Not
because the part was bad — they never actually got to full-rate
production. They lost it because they couldn’t consistently produce
parts that met specification during the launch phase. The OEM’s
production line stalled twice waiting for shipments. A thousand parts
were quarantined after a dimension drifted during a tool change nobody
had anticipated. The supplier’s own incoming inspection caught defects
their process should never have produced in the first place.
The post-mortem was brutal. The root cause wasn’t a single failure.
It was dozens of small decisions, made in isolation, by people who never
talked to each other. The design team had specified tolerances the
production equipment couldn’t hold consistently. The tooling engineer
had ordered fixtures based on prototype volumes, not production volumes.
The quality engineer had written control plans using assumptions from a
similar part that turned out to be nothing like this one. And nobody had
validated the measurement system before using it to accept or reject
parts that were already behind schedule.
Every single one of those problems was predictable. Every single one
was preventable. And every single one could have been caught by a
structured approach to product quality planning — an approach that the
automotive industry formalized decades ago and called APQP.
What APQP Actually Is
Advanced Product Quality Planning and Control Plan is, at its core, a
structured framework for ensuring that products are designed and
manufactured to meet customer requirements. It was developed by the
American automotive industry — specifically by Chrysler, Ford, and
General Motors — and first published as a reference manual in 1994 under
the umbrella of what was then AIAG, the Automotive Industry Action
Group.
But reducing APQP to its definition misses the point entirely. APQP
is not a checklist. It is not a set of forms to fill out and file away.
It is a disciplined way of thinking about product launches that starts
long before production begins and doesn’t end until the customer
confirms, with data, that the product meets their expectations
consistently.
The framework is organized into five phases, and understanding what
each phase demands — and what happens when you skip it — is the
difference between a launch that builds momentum and a launch that burns
through your credibility.
Phase 1:
Planning — The Conversation Before the Work
The first phase of APQP is where the most damage is usually done,
because it’s the phase most organizations rush through. Planning is not
about creating a Gantt chart. It’s about establishing shared
understanding.
During this phase, the team must capture the voice of the customer in
concrete, measurable terms. Not “the customer wants a good part.” Not
“they need it to be reliable.” Those are aspirations, not requirements.
APQP demands that you translate aspirations into specifications: this
dimension must be 12.7 millimeters plus or minus 0.05. This surface must
have a roughness of no more than 1.6 micrometers. This component must
survive 500,000 load cycles without fatigue failure.
The planning phase also establishes the project timeline, identifies
the team members and their responsibilities, and — critically — defines
the criteria for success before anyone starts spending money.
Organizations that skip this phase or treat it as a formality share a
common symptom: they discover requirements during production that should
have been understood during planning. Every requirement discovered late
costs ten to a hundred times more to address than one identified early.
This is not a theoretical relationship. It is one of the most
consistently validated findings in product development research.
Phase 2: Product Design
and Development
Once the requirements are clear, the second phase focuses on
translating them into a product design that can actually be
manufactured. This is where engineering robustness is built in — or
where it’s accidentally designed out.
Design FMEA lives in this phase. So does design verification and
validation. The team evaluates whether the proposed design can meet the
customer’s requirements under real-world conditions, and they do it
before committing to tooling.
One of the most powerful tools in this phase is Design for
Manufacturing and Assembly, or DFMA. It asks a simple question that most
design engineers don’t naturally ask: can this part actually be made
consistently on the equipment we have, by the operators we employ, at
the volumes the customer needs?
The answer, surprisingly often, is no. Not because the design is bad,
but because the design was optimized for function without considering
production reality. A tolerance that looks perfectly reasonable on a CAD
screen may be impossible to hold on a Tuesday morning when the humidity
changes and the machine’s thermal compensation hasn’t caught up. A
feature that assembles beautifully in a prototype shop may require a
custom fixture and three extra minutes per part in a high-volume
production environment.
APQP forces these conversations to happen during design, when changes
are still inexpensive, rather than during production, when they’re
not.
Phase 3: Process Design
and Development
This is where the rubber meets the road — or, more accurately, where
the tooling meets the material. Phase three translates the product
design into a manufacturing process, and it is where many organizations
discover that their optimism about Phase 2 was, shall we say,
insufficiently grounded.
Process FMEA belongs here. So do process flow diagrams, control
plans, and work instructions. The team must define every step of the
manufacturing process, identify what could go wrong at each step,
determine how those failures would be detected, and establish what
controls will prevent them from reaching the customer.
This phase also includes Measurement Systems Analysis. Before you can
control a process, you must be able to measure it. And before you can
trust your measurements, you must prove that your measurement system is
capable of distinguishing between good parts and bad parts with
acceptable repeatability and reproducibility.
Organizations that skip MSA during process design share another
common symptom: they spend weeks or months arguing about whether parts
are in or out of specification, only to discover that their measurement
system couldn’t tell the difference reliably. The parts might have been
fine. The measurement system was the problem. But by the time they
figure that out, the production schedule has been destroyed, the
customer’s confidence has eroded, and someone has already been blamed
for a problem that was never actually a problem.
Phase 4: Product and
Process Validation
Phase four is where the planning, the design, and the process come
together for a trial run. This is the PPAP phase in automotive —
Production Part Approval Process — where the organization must
demonstrate, with evidence, that its process can produce parts that
consistently meet all customer requirements at production volume.
Significant production runs. Initial process studies. Measurement
system verification. Appearance approval if applicable. Every element of
the PPAP submission serves a specific purpose: to prove that the process
works, not in theory, not in a prototype environment, but in the actual
production environment with actual production personnel and actual
production variability.
This phase is where organizations that treated APQP as paperwork
discover the gap between their documentation and their reality. The
control plan says the operator checks dimension A every two hours, but
nobody trained the operator on how to measure it. The process flow says
the part moves from station three to station four in a container, but
nobody specified the container, and now the parts are getting scratched
in transit. The FMEA identified a potential failure mode with a severity
rating of 8, but the recommended action was “operator awareness,” which
is not a control — it’s a hope.
The validation phase exists to catch exactly these disconnects. It is
the last line of defense between your organization and the customer. And
it is the phase most likely to be compressed when the project is behind
schedule, which is precisely when it is most needed.
Phase 5:
Feedback, Assessment, and Corrective Action
The fifth phase is the one that separates organizations that improve
from organizations that merely repeat. After launch, the team evaluates
what went well, what didn’t, and what should be done differently next
time.
This is not a feel-good exercise. It’s a data-driven review of actual
performance against planned performance. Were the process capabilities
what the team predicted? Did the control plan catch the variation modes
the FMEA identified? Were there any surprises — any failure modes that
nobody anticipated?
Lessons learned that are not documented are lessons that will be
learned again, by a different team, on a different project, at a
different time, with the same expensive consequences. APQP Phase 5
exists to break that cycle.
The Hidden Cost of Skipping
Steps
Here is the pattern I’ve observed across dozens of organizations, in
automotive and beyond: everyone says they follow APQP. Almost no one
actually does.
Not because they’re dishonest. Because following APQP properly
requires discipline that conflicts with the pressure to move fast. When
the customer wants parts in twelve weeks and your process development
timeline says sixteen, someone will suggest compressing Phase 3. When
the design team is behind on their deliverables, someone will suggest
running the prototype tooling while the design FMEA is still in draft.
When the validation run produces parts that are marginally out of
specification, someone will argue that they’re close enough and that
production will sort it out.
Every one of those decisions is understandable in the moment. Every
one of them increases risk. And the cumulative effect of those
decisions, across an entire launch, is the difference between a supplier
that earns the customer’s trust and a supplier that earns the customer’s
audit.
The organizations that do APQP well share a common trait: they treat
the framework not as a burden but as a communication tool. APQP’s
greatest value is not in the forms it produces but in the conversations
it forces. When the design engineer, the process engineer, the quality
engineer, and the production supervisor sit in the same room and walk
through a process FMEA together, they discover assumptions they didn’t
know they were making. They find gaps that no individual would have
spotted alone. They build shared understanding that no amount of email
communication can replicate.
APQP Beyond Automotive
Although APQP was born in the automotive industry, its principles are
universal. Any organization that designs and manufactures products —
medical devices, aerospace components, consumer electronics, industrial
equipment — benefits from the structured approach that APQP
provides.
The medical device industry has its own version, embedded in design
controls and process validation requirements under regulations like FDA
21 CFR 820 and ISO 13485. The aerospace industry uses a similar approach
through AS9100 and various first-article inspection requirements. The
specific terminology differs. The underlying logic is the same: plan
thoroughly, validate rigorously, and never assume that good intentions
are a substitute for good evidence.
The Quality Architect’s
Perspective
In twenty-five years of working with organizations across automotive,
aerospace, and pharmaceutical manufacturing, I’ve seen APQP save
launches that were headed for disaster. I’ve also seen APQP fail — not
because the framework was inadequate, but because the organization
treated it as a compliance exercise rather than a thinking exercise.
The difference is always culture. In organizations where APQP is a
living, breathing process — where teams challenge each other’s
assumptions, where the FMEA is updated when new information emerges,
where the control plan evolves as the process matures — launches succeed
consistently. In organizations where APQP is a binder on a shelf,
launches succeed or fail based on luck.
Luck is not a strategy. APQP is.
Peter Stasko is a Quality Architect with 25+ years
of experience transforming organizations across automotive, aerospace,
and pharmaceutical industries. He specializes in building quality
systems that don’t just comply with standards — they create competitive
advantage. His approach combines deep technical expertise in tools like
APQP, FMEA, and SPC with an understanding of the human and
organizational dynamics that determine whether quality systems actually
work or merely exist on paper.
