Quality
and the Premortem: When Your Organization Stops Asking Why Things Failed
and Starts Imagining How They Could Fail — Before They Do
The investigation was flawless. The team had reconstructed the
timeline with forensic precision. The root cause analysis diagram
stretched across two whiteboards, every branch meticulously labeled,
every contributing factor catalogued. The corrective action plan was
detailed, approved, and tracked. By every measure, it was a textbook
response to a quality escape that had cost the organization $2.3 million
and a customer’s trust.
The quality manager closed the investigation file, leaned back in her
chair, and stared at the wall.
“We did everything right,” she said to no one in particular. “Except
see it coming.”
She was right. The failure mode that had escaped detection was not
exotic. It was not unprecedented. Three similar incidents had occurred
across the industry in the previous eighteen months. Two had been
published in trade journals. One had been discussed at a conference her
own team had attended. The information was there. The warning signs were
visible. But nobody had imagined this specific failure happening to them
— and because nobody had imagined it, nobody had prepared for it.
This is the quality profession’s most uncomfortable truth: we
are spectacular at learning from the past and nearly useless at
imagining the future.
Our entire methodology is retrospective. FMEA asks what
could go wrong, but teams fill it out by remembering what
has gone wrong. 8D investigations dissect failures after they
happen. CAPA systems document corrective actions for problems that
already occurred. We are archaeologists of failure, digging through the
rubble of disasters past, building monuments to lessons learned — and
then walking blindfolded into the next one.
There is a better way. It comes from an unlikely source: cognitive
psychology. And it has the power to transform how your organization
anticipates, prevents, and prepares for quality failures that haven’t
happened yet.
It’s called the premortem.
The Psychology of Hindsight
Before we understand the premortem, we have to understand why we need
it.
In 1975, a psychologist named Baruch Fischhoff published a paper that
would change how we understand human judgment. He called it “hindsight
bias” — the tendency for people to believe, after an event has occurred,
that they “knew it all along.” Fischhoff’s experiments were elegant in
their simplicity. He would give participants a description of a
historical event without the outcome, ask them to predict what happened,
and then reveal the actual outcome. When he later asked participants to
recall their original predictions, they consistently remembered being
more accurate than they actually were.
The implications for quality are staggering.
Every postmortem investigation is contaminated by hindsight bias.
When you know how the story ends, the path to that ending seems
inevitable. The clues seem obvious. The failures seem preventable. And
the people who missed them seem negligent. This is why postmortems so
often conclude with some variant of “we should have known” — because in
hindsight, of course you should have. Every piece of evidence now points
in one direction. But in the moment, the evidence pointed in seventeen
directions simultaneously, and the team was looking at three of
them.
Hindsight bias doesn’t just distort our understanding of past
failures. It creates a false confidence about future ones. When every
past failure looks predictable in retrospect, we begin to believe that
future failures will be equally predictable. We build our quality
systems around the assumption that we will recognize the warning signs —
because we always recognize them in hindsight. We create checklists
based on problems we’ve already seen. We build inspection criteria
around defects we’ve already caught. We train our teams on failures that
have already happened.
And then a new failure walks through the door — one that looks
nothing like anything in our experience — and we are shocked. Every
time.
The Premortem: A
Different Kind of Thinking
The premortem was formalized by psychologist Gary Klein in 2007,
though the underlying concept has roots in Stoic philosophy and military
planning. The idea is deceptively simple:
Before a project or process launches, gather the team and ask
them to imagine that the project has already failed — catastrophically.
Then ask each person to write down, independently, what caused the
failure.
That’s it. That’s the technique. But the psychology behind why it
works is profound.
The premortem exploits two cognitive mechanisms that standard
planning completely ignores:
First, it legitimizes dissent. In most planning
sessions, optimism is the default. Anyone who raises concerns is labeled
negative, risk-averse, or “not a team player.” The social cost of
pessimism is high. But in a premortem, pessimism is the assignment.
You’re not being negative — you’re following instructions. This single
reframing unlocks a flood of concerns that would never surface in a
standard risk assessment. People who would never speak up in a meeting
will write detailed, thoughtful failure scenarios when asked to imagine
what went wrong.
Second, it leverages prospective hindsight. Research
by Deborah Mitchell, Jay Russo, and Nancy Pennington in 1989
demonstrated that imagining why a future event did happen
generates 30% more correct reasons than imagining why it might
happen. The shift from “what could go wrong” (a hypothetical) to “what
went wrong” (a narrative) engages different cognitive pathways. Our
brains are wired for stories, not probabilities. When you ask someone to
construct a narrative about failure, they draw on a richer set of
experiences, intuitions, and pattern recognitions than they would from a
standard risk brainstorming session.
The difference between a traditional FMEA and a premortem is the
difference between reading a list of symptoms and imagining yourself
actually being sick. One is abstract. The other is visceral.
How the Quality
Premortem Works in Practice
Let me walk you through how I’ve implemented quality premortems
across organizations, from a 200-person automotive supplier to a
4,000-person pharmaceutical manufacturer.
Step 1: Frame the exercise. Before launching a new
process, product line, quality system, or significant change, schedule a
90-minute session with the cross-functional team. Frame it clearly:
“This project has failed. It’s two years from now, and everything we
hoped for has collapsed. What happened?”
Step 2: Independent generation. Give every
participant ten minutes to write their own failure scenarios. No
discussion. No groupthink. No anchoring on the first idea someone blurts
out. Each person works silently, constructing their own narrative of
catastrophe. I provide a simple template:
- What was the specific quality failure that occurred?
- What was the chain of events that led to it?
- What early warning signs were present but ignored?
- Why didn’t our existing quality system catch it?
- What did we assume that turned out to be wrong?
Step 3: Collective intelligence. Each person reads
their scenario to the group. A facilitator captures every failure mode
on a board. This is where the magic happens. One person’s scenario
triggers an insight in another person. The operator mentions a machine
behavior that the engineer didn’t know about. The quality engineer
identifies a gap in the control plan that the process engineer assumed
was covered. The supply chain specialist sees a vulnerability that
nobody in manufacturing considered. The collective intelligence that
emerges from this exercise routinely surpasses what any individual — or
any traditional risk assessment — could produce.
Step 4: Probability and impact assessment. Once all
failure scenarios are captured, the team assesses each one. Which are
most likely? Which would be most damaging? Which are we least prepared
for? This creates a prioritized list of threats that is fundamentally
different from what a standard FMEA produces — because it includes
threats that the team’s institutional knowledge and experience
recognize, even if they don’t appear in any database of known failure
modes.
Step 5: Preemptive action. The final step is the
most important: what do we do about it? For each high-priority failure
scenario, the team designs specific preventive or detective controls.
These aren’t generic “monitor and review” actions. They are concrete,
measurable, and assigned to owners with deadlines. The premortem doesn’t
just identify risks — it transforms them into actions taken before the
risk materializes.
The FMEA Blind Spot
At this point, you might be thinking: “This sounds like FMEA. Don’t
we already do this?”
No. And understanding why is critical.
FMEA is a powerful tool, but it has a systematic blind spot: it is
anchored to known failure modes. When a team sits down to fill out an
FMEA, they start with what they know. The failure mode library. The
historical defect data. The industry-standard risk categories. They
brainstorm within the boundaries of their experience and their reference
material.
The premortem has no such boundaries. It explicitly asks people to
imagine the unexpected. It invites the failure you haven’t seen before.
It surfaces the assumption nobody questioned. It finds the blind spot
that your FMEA can’t see — because the blind spot is your
FMEA’s framework.
Consider a real example. An automotive supplier was launching a new
production line for electric vehicle battery housings. Their FMEA was
comprehensive — 847 failure modes identified, risk priority numbers
calculated, control plans designed for each. The quality team was
confident. They had followed the methodology to the letter.
In the premortem, a maintenance technician raised his hand. “The
coolant system on the new CNC machines has a known issue with
particulate contamination after about 90 days of operation,” he said. “I
saw it at my last company. The filter design is inadequate for the chip
load. The contamination gradually shifts the tool offset, and the
dimensional drift is so slow that your SPC chart won’t catch it until
you’re already out of spec on a critical tolerance.”
This failure mode was not in the FMEA. It was not in any database the
engineering team had access to. It would not have been caught by the
control plan. And it would have resulted in a latent defect — housings
that passed dimensional inspection but failed in the field under thermal
cycling — that could have triggered a massive recall.
The technician knew about it because he had lived through it. But
nobody had asked him the right question. The FMEA asked “what could go
wrong?” and he didn’t think to mention it in that context — it felt too
speculative, too specific to his personal experience. The premortem
asked “what went wrong?” and suddenly it was the first thing that came
to mind.
That’s the difference.
The Organizational
Resistance
If the premortem is so powerful, why isn’t it standard practice in
every quality organization?
Because it is profoundly uncomfortable.
Organizations are psychologically invested in the success of their
plans. Leaders have championed initiatives, committed resources, and
built timelines. Asking a team to imagine those plans lying in ruins
feels disloyal. It feels like planning for failure. It feels like
weakness.
I have seen senior leaders shut down premortem exercises because “we
don’t need that kind of negativity.” I have seen project managers resist
because the failure scenarios might delay their launch date. I have seen
quality professionals hesitate because identifying risks without
ready-made solutions makes them look like they don’t have the
answers.
This resistance is itself a quality risk — and it’s one the premortem
is uniquely positioned to surface.
The most effective organizations reframe the premortem not as
planning for failure but as investing in success. Every
failure scenario identified before launch is a problem that never
happens. Every preventive action taken in advance is a crisis that never
occurs. The premortem isn’t pessimism — it’s the most aggressive form of
optimism. It says: I believe this project can succeed so completely
that I’m willing to invest time preparing for every possible threat to
that success.
The organizations that embrace this mindset don’t just avoid
failures. They develop a reputation for reliability that becomes a
competitive advantage. Customers notice when your launches are clean.
Suppliers notice when your requirements are thorough. Regulators notice
when your risk assessments are thoughtful. The premortem doesn’t just
prevent problems — it builds trust.
Building a Premortem Culture
The ultimate goal is not to run occasional premortem exercises. It’s
to embed prospective hindsight into the organization’s DNA — to create a
culture where imagining failure is not a special event but a habitual
practice.
Here’s how to get there:
Start with high-stakes decisions. Don’t try to run a
premortem for every process change. Start with new product launches,
major system implementations, organizational restructuring, or any
decision with significant quality implications. Demonstrate the value
where the stakes are highest.
Include diverse voices. The power of the premortem
comes from perspectives that traditional planning misses. Include
operators, maintenance technicians, suppliers, customers, and anyone
with direct experience of the conditions the project will face. The most
valuable insights often come from the people who are never invited to
planning meetings.
Protect the psychological safety of the exercise.
Every failure scenario raised in a premortem must be treated as a gift,
not a threat. If people are punished or dismissed for raising concerns,
the exercise becomes theater. The facilitator’s most important job is to
ensure that every voice is heard and every scenario is taken
seriously.
Close the loop. The biggest failure mode of the
premortem is that failure scenarios are identified, actions are
assigned, and then nothing happens. Build accountability into the
process. Track premortem actions with the same rigor you’d track
corrective actions from a real failure. Review them at milestone
meetings. Verify that preventive controls were actually implemented
before launch.
Archive and learn. Every premortem produces a
catalog of failure scenarios specific to your organization, your
processes, and your context. This catalog is gold. It becomes a living
supplement to your FMEA library. It captures the institutional knowledge
that normally lives in people’s heads and makes it available to
everyone. Over time, your premortem archive becomes one of your most
valuable quality assets.
The Quality Professional’s
New Job
The premortem represents a fundamental shift in the role of the
quality professional. For decades, our value was defined by our ability
to investigate failures, find root causes, and implement corrective
actions. We were the organizational firefighters — respected for our
skill in battling blazes, but never quite valued as highly as the people
who seemed to prevent fires by never having them.
The premortem changes this calculus. It positions the quality
professional not as a firefighter but as a fire marshal — someone whose
job is to ensure that fires never start. It elevates quality from a
reactive function to a predictive capability. And it gives quality
professionals a tool that is uniquely suited to the most important
challenge in modern manufacturing: navigating uncertainty.
In a world of complex supply chains, rapidly evolving technology,
shrinking product development cycles, and increasing customer
expectations, the organizations that will thrive are not the ones with
the best postmortems. They are the ones with the best premortems — the
ones that can imagine failure vividly enough to prevent it,
comprehensively enough to surprise-proof their systems, and routinely
enough that anticipating the unexpected becomes ordinary.
The quality manager in our opening story — the one staring at the
wall after a flawless investigation of a failure she should have seen
coming — she started running premortems the following week. Six months
later, her team identified a potential failure mode during a premortem
that was identical to the $2.3 million escape they had just
investigated.
This time, they caught it before it happened.
The investigation they never had to run was the best investigation of
their careers.
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 respond to failure — they anticipate it.
