The Paradox of Approval
The Production Part Approval Process was designed to answer a simple
question: Can this supplier consistently produce parts that meet all
engineering requirements at the required production rate?
That question demands evidence. Real evidence — produced under actual
production conditions, using the real processes, real tools, real
operators, and real materials that will be used for ongoing production.
The answer matters because the cost of getting it wrong is enormous. A
non-conforming part in an automotive assembly line doesn’t just mean a
defective product. It means stopped production, warranty claims, safety
incidents, and potentially catastrophic failures in the field.
PPAP was the automotive industry’s answer to this risk. Standardize
the evidence. Make it auditable. Ensure that every supplier, everywhere,
proves their process before they ship production volumes. The Automotive
Industry Action Group (AIAG) formalized it into 18 elements — a
comprehensive dossier that covers everything from design records to
material certificates to capability studies.
But here’s what happened. Over decades, the evidence became
the product. The dossier became the deliverable. And the
approval — the actual confidence that a supplier can consistently
produce conforming parts — became secondary to whether the paperwork was
complete.
What PPAP Was Actually Built
For
To understand how far the practice has drifted, we need to understand
the original intent.
PPAP emerged from the automotive industry’s hard lessons in the 1970s
and 1980s. Japanese manufacturers were producing vehicles with
dramatically fewer defects than their Western counterparts, and the gap
was traced partly to how suppliers were qualified. American and European
OEMs were approving suppliers based on prototypes — hand-built parts
that bore no resemblance to what the production process would actually
produce. When production began, quality predictably collapsed.
The solution was elegantly pragmatic. Require suppliers to produce a
significant production run — typically 300 consecutive parts — using the
actual production process, and then submit comprehensive evidence that
those parts met every requirement. Not prototypes. Not samples from a
pilot line. Real parts from the real process, at the real production
rate.
The 18 elements of PPAP weren’t arbitrary bureaucracy. Each one
answered a specific question:
- Design Records: Do you have the correct, current
engineering specifications? - Engineering Change Documents: Have you documented
any authorized deviations? - Customer Engineering Approval: Has the customer
signed off on the design? - Design FMEA: Have you identified how the design
could fail? - Process Flow Diagrams: Do you understand every step
of your manufacturing process? - Process FMEA: Have you identified how each process
step could fail? - Control Plan: Do you have documented methods to
control each identified risk? - Measurement System Analysis: Can you trust your
measurement data? - Dimensional Results: Do the actual parts match the
drawing? - Material Performance Test Results: Do the materials
meet specifications? - Initial Process Studies: Is your process capable?
(Ppk ≥ 1.67 for initial studies) - Qualified Laboratory Documentation: Are your test
labs accredited? - Appearance Approval Report: Do appearance items
meet visual standards? - Sample Production Parts: Have you provided physical
samples? - Master Sample: Is there a retained reference
standard? - Checking Aids: Do you have the right inspection
tools? - Customer-Specific Requirements: Have you met all
unique customer demands? - Part Submission Warrant: The formal cover sheet
signing off on everything.
Each element connects to the others. The Process FMEA feeds the
Control Plan. The Control Plan defines what gets measured. The
Measurement System Analysis validates that the measurements are
trustworthy. The Initial Process Studies use that validated measurement
system to prove capability. It’s a system — interconnected,
interdependent, and designed to create a complete picture of process
readiness.
The Documentation Drift
So how did this rigorous, well-intentioned system become what it is
in many organizations today?
The answer is the same answer that explains the degradation of every
quality system we’ve examined in this series: the form replaced
the substance, and nobody noticed because the form kept getting more
elaborate.
Here’s how it happens, step by step, in organization after
organization:
Phase 1: Compliance-Driven Preparation. A supplier
receives a PPAP requirement from a customer. The immediate goal becomes
producing a PPAP package that will be accepted — not proving the process
works. The distinction is subtle but enormous. When the goal is
acceptance, the questions change. “What does the reviewer want to see?”
replaces “What do we need to prove?” The focus shifts from technical
substance to documentary completeness.
Phase 2: Template Copying. Someone in the quality
department creates a PPAP template. It has all 18 elements. It has the
right headers, the right formatting, the right reference numbers. Then,
for every new part or every new customer, the template gets copied and
populated. The Process FMEA from a similar part gets adapted. The
Control Plan gets tweaked. The wording stays 80% the same because 80%
was accepted last time. Each iteration drifts slightly further from the
specific reality of the current process.
Phase 3: Data Filling. The capability studies need
data. So someone collects 300 parts, measures them, and runs the
numbers. If Ppk comes back below 1.67, there’s enormous pressure to find
a reason to exclude data points, adjust the specification limits, or
redefine the study boundaries — because the goal is approval, not
learning. The measurement system analysis needs Gage R&R results. If
the results are poor, the response is often to find a more favorable
calculation method rather than to genuinely improve the measurement
system.
Phase 4: Submission and Forgetting. The package is
submitted. The customer reviews it — or more likely, checks that all 18
elements are present and the key metrics meet thresholds. Approval is
granted. The PPAP binder goes on a shelf. The production process begins.
And the organization moves on, confident that they have “done PPAP” —
never revisiting the assumptions, the risks, or the evidence until
something goes wrong or a new PPAP is required.
The Capability Illusion
The Initial Process Study is perhaps the most consequential element
of PPAP — and the most commonly manipulated.
The requirement is straightforward: produce 300 parts under
production conditions, measure the critical characteristics, and
calculate Ppk ( Preliminary Process Capability). A Ppk of 1.67 or higher
indicates that the process is capable — it has enough margin between the
process spread and the specification limits to produce consistently
conforming parts.
But here’s what actually happens in practice:
Sample selection bias. The 300 parts are supposed to
be representative of normal production. But suppliers know which
machine, which operator, which batch of material produces the best
results. The “production run” for PPAP is often the best-case scenario,
not the typical scenario. The process that gets approved is the process
under optimal conditions — conditions that may never be replicated in
day-to-day production.
Measurement manipulation. When results are
borderline, there’s a strong incentive to find reasons to exclude
outliers. “That part was damaged in handling.” “The operator was still
learning.” “That measurement was taken before the process stabilized.”
Each exclusion is individually defensible. Collectively, they transform
a marginal process into one that appears capable.
Specification gaming. When a process genuinely can’t
meet the specification, the first response should be to improve the
process. The actual first response, in far too many organizations, is to
question the specification. “Is this tolerance really necessary?” “Can
we open this up?” Sometimes tolerance reviews are legitimate — but when
they’re driven by PPAP pressure rather than engineering analysis,
they’re a retreat from quality, not an advance toward it.
The snapshot problem. Even an honest, well-conducted
initial process study is a snapshot. It captures the process at one
moment, under one set of conditions, with one set of operators and one
batch of material. Real production varies. Operators change. Material
lots differ. Tools wear. Environmental conditions shift. A process that
showed Ppk of 2.0 in the PPAP study might degrade to 1.2 within months —
and nobody notices because the PPAP said it was capable.
The FMEA Disconnect
The Design FMEA and Process FMEA are supposed to be living documents
— the accumulated knowledge of how a product or process could fail,
updated as new failure modes are discovered. In PPAP practice, they’re
often historical fiction.
Here’s the pattern. A supplier has a PFMEA template for “machined
components” or “injection molded parts.” When a new part comes along,
the template is copied, the part number is changed, and the failure
modes listed are the generic failure modes from the template — not the
specific failure modes identified through analysis of this particular
process.
The result is that the PFMEA doesn’t actually capture the risks of
the real process. It captures the risks that someone, at some point in
the past, identified for a category of process. The Control Plan, which
is supposed to be derived from the PFMEA, therefore controls for generic
risks while potentially missing the specific risks that matter most for
this particular part.
And when a real failure occurs in production — one that wasn’t
identified in the PFMEA — what happens? The PFMEA should be updated. The
Control Plan should be revised. The PPAP should be re-examined. In
practice, the failure gets fixed, the corrective action is documented
(probably in 8D format), and the PFMEA stays exactly as it was. The
institutional knowledge gained from the failure doesn’t make it back
into the risk analysis system.
The Control Plan Theater
The Control Plan is where PPAP meets reality — or should. It defines,
for every characteristic of every part, how it will be controlled: the
measurement method, the frequency, the sample size, the reaction plan
when something goes out of control.
But in practice, Control Plans often become lists of characteristics
with generic control methods. “Visual inspection, every piece, reject if
non-conforming.” That’s not a control method — that’s a wish. It doesn’t
define what visual characteristics to inspect, under what lighting, at
what magnification, by whom, with what training, using what reference
standard. It doesn’t define how to ensure the inspection is actually
performed. It doesn’t define what happens when an inspector finds a
borderline case.
The reaction plans are the weakest point. “If out of specification,
quarantine and notify supervisor.” This is a reaction plan that assumes
the only scenario is a clear, unambiguous failure. But real processes
drift. They gradually move toward the specification limit. They show
increased variation before they produce actual nonconformities. A
meaningful reaction plan should define what to do when the process shows
early warning signs — not just what to do when it has already
failed.
The Real Cost of PPAP
Theater
The cost of treating PPAP as a documentation exercise rather than a
genuine process validation extends far beyond the time and money spent
preparing the package.
False confidence. When a PPAP is approved, the
organization believes the process is validated. Resources and attention
shift elsewhere. But if the PPAP was a documentary exercise rather than
a genuine validation, the process may be marginal or incapable. The
false confidence means problems aren’t detected until they’re severe —
and by then, the consequences are far greater.
Missed learning. A genuine PPAP process generates
enormous learning about the process: where it’s robust, where it’s
fragile, what factors influence quality, what failure modes are
possible. When PPAP is treated as paperwork, this learning is never
captured. The same problems are rediscovered, the same failures recur,
and the same solutions are reinvented — because the institutional
knowledge that PPAP should have created was never actually built.
Supplier development opportunity lost. PPAP was
meant to be a dialogue between customer and supplier — a shared
understanding of process capability and risk. When it becomes a one-way
document submission, that dialogue never happens. The customer doesn’t
gain insight into the supplier’s process. The supplier doesn’t receive
feedback that could help them improve. The relationship remains
transactional when it should be collaborative.
Competitive disadvantage. Organizations that
genuinely use PPAP as a tool for process understanding develop a deep,
cumulative knowledge of their manufacturing processes. They know which
processes are robust and which are fragile. They know where to invest in
improvement. They can quote new business with realistic estimates of
capability and cost. Organizations that treat PPAP as paperwork have
none of this knowledge — and they compete blindly against competitors
who do.
What Genuine PPAP Looks Like
An organization that uses PPAP as it was intended looks fundamentally
different:
The FMEA is specific, not generic. It lists failure
modes that were identified through actual analysis of the specific
process — process flow diagrams that reflect the actual steps, with the
actual equipment, operated by the actual personnel. Each failure mode
has a real cause analysis, a real severity rating based on actual
consequences, and a real detection control that’s actually
implemented.
The capability study is honest. If the process can’t
achieve Ppk 1.67, that information flows immediately to engineering and
operations. The response is process improvement — not data manipulation.
If the specification is genuinely too tight, that conversation happens
openly with the customer, with engineering data to support it, rather
than through quiet tolerance relaxation.
The Control Plan is a living document. When a new
failure mode is discovered in production, the Control Plan is updated
within days, not years. When a process change is made, the Control Plan
reflects it immediately. When audit findings reveal gaps, the Control
Plan is the first thing updated — not the last.
The PPAP package tells a true story. A reviewer
reading the PPAP should be able to understand the process, its risks,
its capabilities, and its controls. They should be able to trace from
the design requirements through the process flow, the FMEA, the Control
Plan, and the capability data, and see a coherent, connected narrative —
not a collection of independent documents that happen to share a part
number.
The Path Back to Substance
Recovering the original intent of PPAP doesn’t require abandoning the
framework. It requires reorienting around the questions the framework
was designed to answer:
-
Can this process consistently produce conforming
parts? Answer this with real data from real production — not
optimized samples. If the answer is “not yet,” that’s valuable
information. Act on it. -
Do we understand how this process could fail?
Answer this through genuine analysis — not template copying. Walk the
process floor. Talk to operators. Review historical defect data from
similar processes. Identify failure modes that are specific to this
process, this equipment, this material. -
Do we have controls that would detect or prevent those
failures? Answer this by evaluating the actual controls in
place — not the controls listed on paper. Go to the gemba. Watch the
inspection being performed. Ask the operator what they do when they find
a problem. Compare their answer to the reaction plan in the Control
Plan. -
Are we learning and improving? Answer this by
examining what happens after PPAP approval. Are capability studies
repeated periodically? Is the PFMEA updated with new failure modes? Does
the Control Plan evolve as the process matures?
The Deeper Lesson
PPAP at its best is a structured learning process — a way of building
deep, systematic knowledge about a manufacturing process before
committing to production volumes. The documentation exists to serve the
learning, not to replace it.
But like every quality tool we’ve examined, PPAP is vulnerable to the
same pattern: the framework gets codified, the format gets standardized,
the content gets templated, and the thinking — the actual engineering
thinking that the framework was designed to structure and capture —
quietly disappears. What remains is an elaborate documentation ritual
that produces binders, satisfies auditors, and grants approval — without
delivering the process understanding and confidence that were the entire
point.
The organizations that recognize this pattern — that separate the
documentation from the learning, and invest in the learning even when
the documentation would suffice — are the ones that build genuine
manufacturing excellence. The ones that don’t will keep producing PPAP
packages that look impeccable and processes that don’t hold up.
The question isn’t whether your PPAP was approved. The question is
whether, when production starts for real, your process will actually
perform the way your PPAP said it would. And if you’re not confident it
will — then your PPAP told you a story you chose to believe, rather than
the truth you needed to hear.
About the Author: Peter Stasko is a Quality
Architect with over 25 years of experience in manufacturing quality
management, process validation, and continuous improvement. He has
implemented and assessed PPAP processes across automotive, electronics,
and industrial manufacturing sectors throughout Europe and North
America.
