Case Study 36.1: The Pre-Mortem as Luck Architecture Review

How Gary Klein's Technique Reveals Structural Weaknesses Before They Become Real Failures

Overview

Research context: Organizational psychology, decision-making, and risk analysis Key figure: Gary Klein, cognitive psychologist and decision-making researcher Core application: Using prospective hindsight to identify luck architecture vulnerabilities before they produce bad outcomes Textbook connections: Chapter 36 (luck audit framework), Chapter 17 (resilience and bad luck), Chapter 35 (from noticing to acting)

Background: What Is a Pre-Mortem?

In a conventional post-mortem, analysts look backward at a project that has already failed. They examine what went wrong, why decisions were made as they were, and what could have been done differently. Post-mortems are valuable — but they are expensive. The failure has already happened. The damage has already been done.

Gary Klein, a cognitive psychologist who spent decades studying how experienced professionals make decisions under uncertainty, developed an alternative: the pre-mortem.

The technique, formalized in a 2007 Harvard Business Review article and elaborated in his book Seeing What Others Don't (2013), works as follows:

1. Imagine the future failure. Before a plan is implemented, the team conducts a structured exercise in which they assume, as a premise, that the plan has been implemented and has failed. Not "might fail" — has failed. Spectacularly.

2. Write down why it failed. Each participant independently writes a brief explanation for the failure. What specifically went wrong? Why? What did we miss? The exercise is conducted individually first, before group discussion, to prevent anchoring to others' explanations.

3. Compile and analyze. The explanations are collected and examined. Patterns are identified. Structural vulnerabilities that appear in multiple participants' explanations receive particular attention.

4. Redesign accordingly. The plan is revised to address the most credible structural weaknesses identified in the exercise.

The key cognitive insight behind the pre-mortem is prospective hindsight — a concept Klein drew from research by Deborah Mitchell, J. Edward Russo, and Nancy Pennington (1989). Their research found that when people are asked to imagine that an event has already happened (rather than asking them to imagine that it might happen), they generate significantly more reasons and explanations for it. The past-tense framing — "it failed" rather than "it might fail" — activates a different, more generative cognitive mode.

Specifically, thinking about failure in the past tense bypasses several cognitive biases that normally inhibit realistic risk assessment:

• Planning fallacy: We tend to underestimate the time, cost, and complications of projects. The pre-mortem's definitive failure assumption overrides this bias.
• Overconfidence: We tend to be more confident in our plans than is warranted. The pre-mortem creates psychological permission to be skeptical.
• Motivated reasoning: We tend to explain away threats to plans we're invested in. The pre-mortem makes threat-identification the goal rather than the obstacle.

The Pre-Mortem Applied to Luck Architecture

The luck audit (Chapter 36) identifies the current state of your luck-generating systems. The pre-mortem is a complementary tool that asks a different question: What could cause your luck architecture to fail?

This is different from simply finding your lowest-scoring domain. The pre-mortem asks you to imagine not just "what's weak" but "what could go disastrously wrong if I continue on my current path."

The luck pre-mortem protocol:

1. Assume it is two years from now. Your luck architecture has failed. You have had significantly worse luck outcomes than you expected — fewer opportunities, fewer meaningful connections, fewer fortunate breaks. You are stuck, stagnant, or materially worse off than you are today.

2. Write (in past tense, as if the failure has already happened): Why did it fail? What went wrong? What structural weakness proved to be the critical failure point?

3. For each identified failure cause, ask: Is this risk already present in my current luck architecture? What is the probability it could produce this outcome if I do nothing?

4. Redesign: What specific change to my current luck architecture most directly addresses the most credible failure scenario?

Case Illustration 1: The Content Creator's Pre-Mortem

Consider a college student who has been building a social media audience for three years. Her platform following has grown steadily. She runs the full luck audit and scores reasonably well — 125 out of 175. Her weakest domain is Environmental Design (she works largely in isolation), but overall she feels her luck architecture is sound.

She runs the luck pre-mortem.

The failure scenario she imagines: It is two years from now. My account has stagnated. I'm still at roughly the same follower count. Brand partnerships have dried up. I'm spending more time creating content than ever but getting diminishing returns. I feel trapped.

Why did it fail?

Writing from imagined hindsight, she generates these explanations:

• The platform algorithm changed and I didn't see it coming because I had no network outside the creator community who would have known about it earlier.
• I kept making the same type of content because it was "working" and never experimented enough to build future relevance.
• I got so focused on content production that I stopped attending events and conferences where I would have met collaborators who could have expanded my reach.
• I burned out. I was making content seven days a week for two years and my creativity collapsed.

The structural risk this reveals:

The pre-mortem has surfaced something the standard audit didn't fully capture: the compounding nature of her most significant risk. Her network weakness, her environmental isolation, her algorithm-dependence, and her sustainability risk aren't just separate weak domains — they reinforce each other. If any one of them produces a significant failure event (platform algorithm shift, burnout, stagnation), the others provide no buffer.

This is different from knowing her Environmental Design score is a 3 out of 5. It's understanding that her environmental isolation is the structural precondition for several of her worst possible outcomes.

The redesign:

She makes a change she wouldn't have made based on the audit alone: she commits to attending one in-person industry event per month, not as a "nice to have" but as structural insurance against the failure scenario she'd just imagined in detail. The pre-mortem made the environmental fix feel urgent in a way the audit score didn't.

Case Illustration 2: The Startup Founder's Pre-Mortem

A 19-year-old startup founder has been running a mobile app for two years. He has some traction — a few hundred paying users. He runs the luck audit and identifies his weakest domain as Risk Portfolio: everything he has is concentrated in the app.

He runs the luck pre-mortem.

The failure scenario: It is two years from now. The startup has failed. Not a pivot, not a slowdown — it has shut down. I have nothing to show for two years of work except some code nobody uses.

Why did it fail?

• A well-funded competitor entered the market six months after my launch and outspent me on marketing. I couldn't compete.
• I didn't go to college. When the startup failed, I had no degree, no professional network, and no credentials to fall back on.
• I ran out of runway before I figured out what the right product-market fit was. I didn't have enough time or capital.
• I was so focused on building that I never built the investor relationships that could have given me more runway.

The structural risk this reveals:

The pre-mortem reveals something the risk portfolio audit pointed toward but didn't make visceral: concentration risk combined with no floor. If the startup fails — which is the most probable outcome for any given startup, statistically — he has no fallback position. No degree, no professional network in the investor world, minimal savings. The downside risk isn't "startup fails, I try something else." It's "startup fails, I have no foundation from which to try something else."

The pre-mortem also surfaces a second finding: the investor relationship deficit isn't just about capital. It's about the resilience signal that investor relationships provide — they're a form of social proof, a network extension, and an information pipeline that he currently lacks entirely.

The redesign:

He makes two decisions he wouldn't have made based on the audit's Risk Portfolio score alone: he applies to college as a parallel track (not abandoning the startup, but building a fallback floor), and he commits to spending 10% of his time specifically on investor and mentor relationship-building, even though there's no immediate payoff. The pre-mortem made the floor-building feel non-optional.

Case Illustration 3: The Academic's Publication Pre-Mortem

A behavioral economist is writing a book manuscript on institutional luck. She has been developing it for two years. She runs the luck audit and identifies her lowest domain as Recovery and Resilience (one previous rejection that became a full stop).

She runs the luck pre-mortem.

The failure scenario: It is two years from now. The book never got published. The research sits in my hard drive. Nobody read it.

Why did it fail?

• I never found an agent because I never submitted the proposal. I kept "not quite being ready."
• I sent it to one agent, got a rejection, and stopped. Same pattern as the journal article.
• The research on institutional luck became less novel because two other researchers published on similar topics while I was still revising.
• I let a fear of the trade publishing world — which is completely different from academic publishing — paralyze me from even starting.

The structural risk:

The pre-mortem reveals something specific: the failure isn't about the manuscript's quality. It's about a systematic pattern — rejection leading to withdrawal — that has already cost her one submission cycle on the journal paper and could cost her the book if the same pattern plays out again.

The pre-mortem makes the failure type visible: it's not "the book isn't good enough." It's "the behavior pattern that once stopped me from resubmitting the article will stop me from submitting the manuscript." Named clearly, it can be countered specifically.

The redesign:

She adds a structural protection to her submission process: she commits to sending the manuscript to five agents before allowing herself to process any individual rejection as meaningful data. Single-rejection withdrawal is explicitly defined as the prohibited behavior. The pre-mortem gave her the courage to name the pattern and build a specific countermeasure.

The Pre-Mortem Versus the Standard Audit: Complementary Tools

The luck audit and the luck pre-mortem answer different questions, and both are necessary.

The ideal practice combines both: run the luck audit to identify your current state and improvement priorities, then run the luck pre-mortem to identify the structural vulnerabilities that could turn your weakest domains into real failures.

Research Grounding

Klein's original 2007 HBR article on pre-mortems, titled "Performing a Project Premortem," documented the technique in organizational contexts. Subsequent research has supported the core mechanism.

Deborah Mitchell and colleagues' research on "prospective hindsight" (published in Organizational Behavior and Human Decision Processes, 1989) found that imagining future events as already past increased the accuracy of causal explanation generation. In their experiment, participants who were told to assume an outcome had occurred generated 30% more causes for it than those who were simply asked to predict whether it might occur.

A 2012 meta-analysis by Hirt and Markman on mental simulation strategies found that "prospective hindsight" (imagining the event as having occurred) was significantly more effective at improving planning accuracy than standard prediction methods or brainstorming, particularly for complex, multi-causal scenarios — exactly the kind of scenario that luck architecture failures represent.

The application of pre-mortem thinking to personal life planning is less formally researched, but the mechanism is the same: the definitive past-tense assumption of failure bypasses the motivated reasoning that would otherwise prevent honest identification of structural weaknesses.

Key Takeaway

The luck audit tells you where you are. The luck pre-mortem tells you where you're most likely to fail if you don't change.

Used together, they form a complete forward-looking diagnostic: not just "what is my luck architecture's current state?" but "what happens to my luck architecture if I stay on this path?" The pre-mortem adds a temporal dimension that the static audit alone cannot provide.

Gary Klein's insight — that humans think more clearly about future failures when they imagine them as past events — turns out to be exactly what we need for honest luck architecture review. The imagined failure isn't pessimism. It's structural integrity testing. And the structures that pass it are the ones worth building.

Discussion Questions

1. Why does the past-tense framing of the pre-mortem generate more and better failure explanations than future-tense prediction? What cognitive mechanisms are at work?

2. In the three case illustrations, the pre-mortem surfaced a different (or more specific) insight than the standard audit in each case. What does this suggest about the limits of the standard audit as a standalone tool?

3. The technique assumes that imagining failure doesn't demoralize people — that it instead generates useful protective action. Is this assumption always valid? Under what circumstances might the pre-mortem backfire?

4. How would you adapt the luck pre-mortem for a team or organization, rather than an individual? What modifications would be necessary?