Quality
and the Pre-Mortem: When Your Organization Imagines Failure Before It
Happens — and the Disaster You Predicted Became the Disaster You
Prevented
The Launch That Was Too
Perfect
In 2019, a mid-sized automotive supplier in Slovakia completed what
everyone called a flawless production launch. The new brake caliper line
had passed every gate review. FMEA was comprehensive. Control plans were
airtight. PPAP documentation was pristine. The quality team signed off.
The plant manager celebrated. The customer approved.
Fourteen months later, a recall.
The root cause was a thermal expansion coefficient mismatch between
the caliper housing and a secondary seal — a condition that only
manifested under a specific combination of sustained highway speed,
ambient temperature above 38°C, and brake wear past 70%. Nobody had
tested that combination. Nobody had even thought to test that
combination.
During the 8D investigation, a junior engineer mentioned something
chilling: “I thought about that interaction during the design phase, but
everyone seemed so confident that I didn’t want to slow things
down.”
The organization had done everything right — except imagine what
could go wrong before it did.
This is where the pre-mortem comes in. And it might be the most
powerful quality tool you’ve never used.
What Is a Pre-Mortem?
A pre-mortem is deceptively simple. Before a project launches, a
product ships, or a process goes live, you gather your team and ask one
question:
“Imagine it is one year from today. This project has been a
catastrophic failure. What went wrong?”
That’s it. That’s the exercise.
But the psychology behind it is profound — and its implications for
quality management are enormous.
The technique was formalized by psychologist Gary Klein in 2007,
building on decades of research into prospective hindsight — the curious
finding that people generate 30% more reasons for failure when they
imagine an event has already occurred than when they try to predict it
might occur.
Let that sink in. Simply changing the framing from “what
might go wrong?” to “what did go wrong?” unlocks a
fundamentally different kind of thinking.
Why Traditional
Risk Assessment Falls Short
Most quality professionals are familiar with the standard risk
toolkit. FMEA assigns severity, occurrence, and detection ratings. Risk
matrices plot likelihood against impact. HAZOP studies examine
deviations from design intent.
These tools are essential. They are also systematically limited by
three psychological traps.
Trap 1: Optimism Bias
When you ask people to predict what might happen, the human
brain automatically discounts negative outcomes. We are wired for
optimism. Research by Tali Sharot at University College London shows
that when people are presented with evidence that contradicts their
optimistic expectations, their brain’s reward system simply fails to
process the contradiction. We literally cannot update our beliefs about
bad outcomes as efficiently as we update our beliefs about good
ones.
In practice, this means your FMEA team will systematically
underestimate occurrence ratings. Not because they are incompetent, but
because they are human.
Trap 2: Social Conformity Pressure
In a group setting, the first person to voice a concern sets the
tone. If the senior engineer says “I think we’re in good shape,” the
junior engineer who noticed the thermal expansion issue will keep quiet.
Solomon Asch’s conformity experiments demonstrated this in 1951 — people
will deny the evidence of their own eyes rather than disagree with a
group.
The junior engineer in our opening story didn’t speak up because the
social cost of being the pessimist in a room full of confidence was too
high.
Trap 3: The Availability Heuristic
When generating risks, people think of what they can easily recall —
recent failures, dramatic incidents, things that happened to them
personally. Systemic risks that are novel, slow-building, or emerge from
the interaction of multiple variables are far less likely to be imagined
because they have no vivid precedent in memory.
The thermal expansion interaction in our example was precisely this
type of risk — low salience, high complexity, no memorable
precedent.
How the Pre-Mortem Breaks
Through
The pre-mortem disrupts all three traps simultaneously.
Against optimism bias: By asking people to assume
failure has already occurred, you remove the prediction problem. You are
not asking “will this fail?” You are asking “how did this fail?” The
brain treats this as a diagnostic exercise rather than a predictive one,
and diagnostic reasoning is far less susceptible to optimistic
distortion.
Against social conformity: The pre-mortem format
typically begins with silent, individual brainstorming. Every team
member writes down their own list of failure causes independently,
before any group discussion. The junior engineer writes down the thermal
expansion issue without needing to contradict the senior engineer. Only
after all ideas are collected does the group discuss them together. By
then, the idea exists on paper — it cannot be dismissed as
negativity.
Against the availability heuristic: The framing “it
has already failed” prompts a different kind of search strategy in the
brain. Instead of looking for examples of similar failures (which
requires recall), people reason forward from the current state to
construct plausible failure pathways (which requires imagination). This
shift from recall to construction dramatically expands the range of
risks identified.
Gary Klein’s research showed that teams using the pre-mortem
technique identified an average of 30% more potential failure modes than
teams using traditional risk assessment methods. In quality management,
where a single unidentified risk can trigger a recall costing millions,
that 30% is not a marginal improvement — it is the difference between a
controlled process and a catastrophic one.
Running a
Quality Pre-Mortem: A Practical Guide
When to Run One
The pre-mortem is not a replacement for FMEA or your existing risk
tools. It is a complement. The ideal timing is after your traditional
risk assessment is complete but before launch. You have your FMEA. You
have your control plan. Now ask: what did we miss?
Critical moments for a pre-mortem:
- New product launches — especially those involving
new materials, processes, or suppliers - Process changes — any significant modification to a
validated process - Supplier qualifications — before onboarding a
critical new supplier - Organizational changes — restructuring that affects
quality responsibilities - Software releases — particularly those controlling
production or inspection equipment - Regulatory submissions — before submitting to
auditors or certification bodies
Who Should Participate
The power of the pre-mortem increases with cognitive diversity.
Include:
- Quality engineers (obviously)
- Production operators who will actually run the process
- Maintenance technicians who understand equipment behavior
- Design engineers who know the product’s weak points
- Supply chain managers who understand supplier risks
- Customer-facing people who hear what customers actually complain
about - At least one person who was not involved in the project — the
outsider’s perspective is disproportionately valuable
Aim for 6-12 participants. Fewer than 6 and you lack diversity. More
than 12 and the group dynamics become unwieldy.
The Step-by-Step Process
Step 1: Frame the Failure (5 minutes)
The facilitator sets the scene: “It is [specific date, 6-12 months in
the future]. The [specific project/process/product] has failed
catastrophically. Not a minor issue — a total, embarrassing, expensive
failure. Customer complaints have escalated. We are facing a recall. Our
reputation is damaged. What happened?”
Be specific about the scenario. “The new coating line has failed”
works. “A quality problem has occurred” does not.
Step 2: Silent Individual Brainstorming (10
minutes)
Every participant writes down every reason they can think of for the
failure. No discussion. No groupthink. No social pressure. Just
individual, silent thinking and writing.
Ask for a minimum of three causes from each person. The first two
will be obvious; the third and fourth are where the insight lives.
Step 3: Round-Robin Collection (15-20 minutes)
Go around the room. Each person reads one cause from their list.
Continue rounds until all causes are on the table. No criticism, no
debate, no “that can’t happen here.” The facilitator records every cause
on a shared document or board.
Step 4: Cluster and Prioritize (15 minutes)
Group similar causes into themes. Then have the team vote on which
failure modes are most concerning — either most likely, most impactful,
or most likely to be overlooked by existing controls.
Step 5: Action Planning (20 minutes)
For the top 3-5 failure modes, ask: “What would we need to do
differently right now to prevent this from happening?” Assign specific
actions, owners, and deadlines.
The entire exercise takes 60-90 minutes. The return on that time
investment is incalculable.
The Quality Culture Shift
Beyond the immediate risk identification, the pre-mortem does
something more profound: it changes organizational culture around
failure.
In most organizations, the people who raise concerns are labeled
pessimists, complainers, or — the worst label in corporate culture —
“not team players.” The pre-mortem inverts this dynamic. Pessimism
becomes not just acceptable but required. The person who identifies the
most potential failure modes is not being negative — they are being
thorough.
I have watched this shift happen in real time. In organizations that
adopt the pre-mortem as a standard practice, the culture around risk
undergoes a quiet revolution. People stop being afraid to raise
concerns. Junior engineers start speaking up. Operators start sharing
the things they have been quietly worrying about but never formalized
into a complaint.
This is not coincidental. Psychological safety — the belief that you
can speak up without punishment — is the single strongest predictor of
team effectiveness, according to Google’s Project Aristotle research.
The pre-mortem creates a structured, legitimate space for
concern-raising. It makes speaking up not just safe but expected.
Resistance You Will
Encounter
The pre-mortem sounds simple, but implementing it as a consistent
practice requires overcoming predictable resistance.
“We already do FMEA. This is redundant.”
It is not. FMEA asks “what could go wrong with each component and
process step?” The pre-mortem asks “how could this entire system fail in
ways we haven’t considered?” These are fundamentally different
questions. FMEA is analytical and decompositional. The pre-mortem is
integrative and imaginative. You need both.
“We don’t have time for another meeting.”
You have time for a 60-minute pre-mortem before launch. You do not
have time for a six-month recall investigation after launch. The math is
not complicated.
“This will create a culture of negativity.”
Research shows the opposite. Teams that conduct pre-mortems report
higher confidence in their projects after the exercise, not lower. Why?
Because the risks that were previously unnamed and unmanaged are now
visible and addressed. Fear of the unknown is always worse than fear of
the known.
“Leadership won’t take it seriously.”
Start small. Run one pre-mortem on a moderate-risk project. Document
what it catches that FMEA missed. Present the results. The evidence will
make the case more effectively than any argument.
Case Study: The
Pre-Mortem That Saved a Launch
A pharmaceutical company was preparing to validate a new sterile
filling line. The traditional risk assessment — FMEA, hazard analysis,
process validation protocols — had been completed. Everything looked
solid.
The quality director, who had recently trained on the pre-mortem
technique, insisted on running one before the validation campaign.
During the silent brainstorming, a maintenance technician wrote: “The
new laminar flow hood vibrates slightly when the HVAC system cycles to
heating mode at 6:00 AM. It is barely perceptible, but I noticed it
during installation testing. If that vibration disturbs the airflow
pattern during a fill, we could have a contamination event that our
environmental monitoring might not catch in time.”
This risk was not in the FMEA. It was not in the hazard analysis. It
was not in the validation protocol. It existed in the gap between what
the engineering team had designed and what the physical system actually
did when interacting with the building’s HVAC.
The investigation that followed confirmed the vibration was real. The
HVAC cycling was adjusted, the hood was stabilized, and validation
proceeded successfully.
Without the pre-mortem, that vibration would have been discovered
during a media fill failure — or worse, during a sterility failure in
the field. The cost difference between “adjust the HVAC timing” and
“recall three batches of sterile product” is measured in millions of
dollars and potentially in patient safety.
The maintenance technician later said: “I noticed it weeks ago but
didn’t think it was important enough to mention. The pre-mortem made me
realize that anything I’ve noticed is worth mentioning.”
That sentence is the entire philosophy of the pre-mortem, distilled
to its essence.
Integration With
Existing Quality Systems
The pre-mortem does not exist in isolation. It integrates with your
quality management system at multiple points:
-
APQP: Insert a pre-mortem between the Product
Design Validation and Process Design Validation phases. This catches
product-process interaction risks that separate FMEAs miss. -
Change Management: Make the pre-mortem a
mandatory step in any change control process that affects validated
processes or critical quality attributes. -
Management Review: Include pre-mortem findings
in management review inputs. The risks identified are strategic
intelligence, not just project-level data. -
CAPA: When closing a CAPA, run a pre-mortem on
the corrective action itself. The question becomes: “Imagine the
corrective action failed. Why?” This prevents corrective actions that
create new problems. -
Audit Preparation: Before your certification or
customer audit, run a pre-mortem. “Imagine the auditor found a major
nonconformity. What did they find?” This is more effective than any
checklist.
The Deeper Lesson
The pre-mortem works because it acknowledges a fundamental truth
about quality management: our cognitive architecture is not
designed to anticipate failure naturally. We are built to
expect things to work. Our brains are prediction engines optimized for
success, not for failure detection.
This is not a flaw. It is a feature of human cognition that served us
well for hundreds of thousands of years. But in the context of quality
management — where a single overlooked failure mode can cascade into a
recall, a lawsuit, or a tragedy — our natural optimism is a systematic
vulnerability.
The pre-mortem is a deliberate countermeasure. It does not ask people
to stop being optimistic. It asks them to channel their optimism into a
different question: “If we want this to succeed, what failure modes do
we need to eliminate first?”
The organizations that embrace this practice discover something
unexpected. The pre-mortem does not make people more pessimistic. It
makes them more confident — because confidence grounded in honest risk
assessment is always stronger than confidence grounded in hopeful
ignorance.
The Slovakian supplier from our opening story? They now run
pre-mortems before every new product launch. The junior engineer who
kept quiet about the thermal expansion issue? He now facilitates
them.
The recall cost them €2.3 million. The pre-mortem costs them 90
minutes per launch.
Some investments do not require a business case.
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 work in the real world — not just on paper — and believes that the
most powerful quality tools are often the simplest ones.
