AI Prompt · Copilot Tool
Pre-Mortem
Facilitator
Assume failure. Find the fatal flaws. Fix them before they find you. The most optimistic thing a smart team can do is plan for what could go wrong.
What This Does
This sequence turns your initiative into a stress test before it becomes a post-mortem. The AI acts as a ruthless Red Team — composed of distinct expert personas who argue with each other — to surface the specific, uncomfortable reasons your plan will fail. Then it forces you to defend your mitigation strategies against second-order effects. Finally, it gives you the exact psychologically safe questions to surface the risks your team is too polite to voice.
Choose Your Mode
Standard Pre-Mortem
Red Team Debate (Advanced)
How to Use This Sequence
Run these three stages in order. Replace every [BRACKETED PLACEHOLDER] with your specific context before pasting into Copilot. Each stage builds on the last — do not skip ahead. The sequence takes 20–40 minutes and is most powerful when run before a plan is finalised, not after it has organisational momentum.
1
The Red Team Setup
Assume Failure · Surface Fatal Flaws
Set the scene. Assume the project has already failed — not underperformed, but failed spectacularly. Your job is to explain why. This framing shift from "what could go wrong" to "why did this fail" is not rhetorical. It activates a fundamentally different cognitive process.
Why this works
Most planning meetings produce the risks everyone already knows about — the ones safe to name in front of the team. This prompt is engineered to surface the risks people privately suspect but won't raise: the ones that implicitly criticise the plan, the leader who championed it, or the consensus behind it.
Copy & Paste Into Copilot
Act as a highly critical, deeply experienced 'Red Team' whose sole job is to find the fatal flaws in my strategy. I am going to pitch you my team's new initiative.
SCENARIO: It is currently 6 months from now, and this initiative has failed spectacularly. Not a minor setback — a complete, embarrassing failure.
THE INITIATIVE: [Paste your project description here — include the goal, the team involved, the key assumptions, and the timeline]
Your analysis must go beyond generic risks. Give me the top 3 highly specific, uncomfortable reasons this failed, focusing on:
• INTERNAL DYNAMICS: The team tensions, unspoken hierarchies, and interpersonal dynamics that quietly derailed execution
• ASSUMPTION EXPLOSIONS: The beliefs we treated as facts that turned out to be dangerously wrong
• RESOURCE REALITY: The hidden dependencies, capacity gaps, and competing priorities that nobody wanted to name out loud
• SECOND-ORDER EFFECTS: The knock-on consequences of our own actions that we didn't anticipate — the problems our solutions created
For each failure reason: name it precisely, explain the mechanism of failure, and identify the earliest warning sign we would have seen (but probably ignored) in the first 30 days.
Do not give me solutions yet. Just hold up the mirror.
Follow-Up Prompts — Go Deeper
1
"Of the three failure reasons you identified, which one is most likely to be defended — not because it's wrong, but because it's tied to someone's identity or past decisions? Who in this team is most likely to resist hearing it?"
2
"Now run a partial failure scenario: the project succeeded, but at twice the cost and half the speed. What happened? What does this tell us about where the real fragility is?"
3
"What is the single assumption in this initiative that, if wrong, would make the entire plan collapse? Frame it as a falsifiable hypothesis: 'We believe [X] is true. We will know we are wrong if [Y] happens within [Z] weeks.'"
2
The Defence Stress-Test
Challenge Your Mitigation · Expose Second-Order Failures
Now defend yourself. Paste your mitigation strategy for the fatal flaws identified in Stage 1. The AI will not be polite. It will tell you why your fix is naive, what secondary problems your solution creates, and where you are trading one failure mode for another.
The trap to avoid
Most teams accept the first plausible mitigation and move on. This is the planning fallacy in action. The most dangerous risks are not the ones you haven't thought of — they are the ones you think you've already solved.
Copy & Paste Into Copilot
Based on the fatal flaws you just identified, I am going to defend against them. Push back hard.
MY MITIGATION STRATEGY: [Paste your defence plan here — be specific about what you intend to do differently]
For each mitigation I've proposed, tell me:
1. WHY IT'S NAIVE: What am I missing? What am I assuming about my own capability to execute this fix?
2. THE SECOND-ORDER PROBLEM: What new problem does my solution create? What does fixing this break elsewhere in the system?
3. THE REAL TEST: What would I need to see — in the first 30 days — to know whether this mitigation is actually working? Give me a specific, observable signal, not a vague metric.
4. THE HONEST VERDICT: On a scale of 1–5, how confident are you that this mitigation is sufficient? What would make it a 5?
Do not soften the feedback. The goal is not to make me feel better about the plan. The goal is to make the plan better.
Follow-Up Prompts — Go Deeper
1
"What is the one mitigation on my list that I am most emotionally attached to — the one I would defend even if the evidence said I was wrong? Why is that one dangerous?"
2
"If I had to choose between fixing the most likely failure and fixing the most catastrophic failure, which would you recommend and why? What am I trading off?"
3
"What is one risk on the original list that I am consciously choosing to accept — not because it's unlikely, but because I don't have the power to fix it? What does that mean for how I set expectations with stakeholders?"
3
The Candour Extraction
Psychological Safety · Team Conversation Design
The Red Team has shown you the blind spots. Now you need to bring this to your actual human team — the people who have been privately worried about exactly these things but haven't said so. This stage gives you the exact questions to create the psychological safety for them to tell you the ugly truth.
The facilitation principle
The questions must be assumptive (assume there IS a problem, not "is there a problem?"), specific enough that people can't dodge them, and framed so that speaking up feels like a contribution to success rather than a criticism of the plan.
Copy & Paste Into Copilot
The Red Team analysis is complete. Now I need to bring this to my actual human team tomorrow morning.
THE CONTEXT: [Briefly describe your team — size, seniority mix, and whether people typically speak up or stay quiet in planning meetings]
THE TOP RISKS IDENTIFIED: [Paste the 2–3 most important failure modes from Stage 1]
Give me 4 highly specific, psychologically safe questions I can ask my team to surface these exact risks — questions they will actually answer honestly.
Each question must be:
• ASSUMPTIVE: Assume the problem exists. Don't ask "is there a problem?" Ask "where is the problem?"
• SPECIFIC: Narrow enough that people can't give a vague, safe answer
• SAFE: Framed so that answering honestly feels like helping the team, not attacking the plan
• ACTIONABLE: The answer should lead directly to something we can do differently
Also give me: the one question I should NOT ask — the version of this question that sounds right but will actually shut people down. Explain why.
Finally: what is the most important thing I should do in the first 5 minutes of that meeting to signal that I genuinely want to hear bad news?
Follow-Up Prompts — Go Deeper
1
"Write the exact opening statement I should use to frame this conversation — 3 sentences maximum. It should signal psychological safety without sounding like a corporate script."
2
"Who in my team is most likely to have the most important insight — and least likely to share it unprompted? How do I specifically draw them out without putting them on the spot?"
3
"After the team meeting, what is the single most important thing I should do in the next 24 hours to show that I actually heard what they said — and that it changed something?"
Adversarial Collaboration — Advanced Mode
Instead of asking for a generic list of risks, this mode forces distinct expert personas to argue with each other about why your plan will fail. Structured disagreement prevents superficial answers and uncovers the hidden dependencies that cause catastrophic failure. This technique is used by U.S. military planning units, leading investment banks, and elite product teams before major strategic decisions.
Persona A
The Cynical Operator
20 years of execution. Has seen every "transformational initiative" die in the implementation phase. Focuses on: capacity, competing priorities, and the gap between what leadership says and what the organisation actually does.
Persona B
The Behavioural Psychologist
Expert in cognitive bias, groupthink, and organisational psychology. Focuses on: the assumptions people are too optimistic to question, the social dynamics that silence dissent, and the identity traps that make leaders defend bad plans.
Persona C
The Ruthless Competitor
Thinks like your most dangerous rival. Focuses on: the market moves you haven't anticipated, the stakeholder relationships you've taken for granted, and the windows of opportunity your delay is handing to others.
1
The Adversarial Debate
Three Personas · Structured Disagreement
Force three distinct expert personas to argue about why your plan will fail — and to disagree with each other. The disagreement is the point. It surfaces the hidden tensions between different types of failure risk that a single perspective would miss.
Copy & Paste Into Copilot
I want you to run an Adversarial Collaboration pre-mortem on my initiative. You will play three distinct expert personas simultaneously, and they must argue with each other.
THE INITIATIVE: [Paste your project description here — include the goal, the team involved, the key assumptions, and the timeline]
THE SCENARIO: It is 6 months from now. The initiative has failed completely.
The three personas are:
PERSONA A — THE CYNICAL OPERATOR: 20 years of execution experience. Has watched every "transformational initiative" die in implementation. Focuses on capacity gaps, competing priorities, and the chasm between what leadership says and what the organisation actually does.
PERSONA B — THE BEHAVIOURAL PSYCHOLOGIST: Expert in cognitive bias, groupthink, and organisational dynamics. Focuses on the assumptions people are too optimistic to question, the social dynamics that silence dissent, and the identity traps that make leaders defend bad plans even when the evidence is against them.
PERSONA C — THE RUTHLESS COMPETITOR: Thinks like our most dangerous rival. Focuses on the market moves we haven't anticipated, the stakeholder relationships we've taken for granted, and the windows of opportunity our delay is handing to others.
Structure the output as a debate:
1. Each persona gives their primary reason for failure (2–3 sentences, sharp and specific)
2. Each persona challenges one of the other two personas' reasoning — where do they disagree?
3. From the debate, identify the ONE failure mode that all three personas agree is the most dangerous — the point of convergence
Be specific to my initiative. No generic risks. Push hard.
Follow-Up Prompts — Go Deeper
1
"The Cynical Operator and the Behavioural Psychologist disagree on [the point of disagreement from the debate]. Who is right — and what evidence would settle the argument?"
2
"Add a fourth persona: The Loyal Insider — someone who genuinely believes in this initiative but has been quietly worried about one specific thing for weeks. What are they worried about that they haven't said out loud?"
2
The Convergence Defence
Address the Point of Consensus · Stress-Test the Fix
All three personas agreed on one failure mode. That is your highest-priority risk. Now defend against it — and let the personas critique your defence from their distinct perspectives.
Copy & Paste Into Copilot
All three personas agreed that [the convergence point from Stage 1] is the most dangerous failure mode.
MY MITIGATION STRATEGY: [Paste your defence here]
Now have each persona critique my mitigation from their specific perspective:
• THE CYNICAL OPERATOR: Why will this fail in execution? What am I underestimating about the organisation's actual capacity to do this?
• THE BEHAVIOURAL PSYCHOLOGIST: What cognitive bias is driving my confidence in this mitigation? Am I solving the problem I have, or the problem I wish I had?
• THE RUTHLESS COMPETITOR: How does my mitigation strategy create a new vulnerability that a competitor or external actor could exploit?
End with a synthesis: given all three critiques, what is the minimum viable version of my mitigation that addresses the most important objections from each persona?
3
The Candour Extraction
Psychological Safety · Team Conversation Design
The adversarial debate has surfaced the real risks. Now design the conversation that brings this intelligence to your team — without triggering defensiveness or shutting down the people who most need to speak.
Copy & Paste Into Copilot
The adversarial pre-mortem is complete. The three personas have identified the following as the most critical risks:
TOP RISKS: [Paste the 2–3 most important failure modes from the debate]
MY TEAM CONTEXT: [Briefly describe your team — size, seniority mix, and the typical culture around raising concerns]
Design a 15-minute team conversation that surfaces these risks from the people who already know about them but haven't said so.
Give me:
1. THE OPENING FRAME: A 2-sentence statement that signals I want to hear bad news — not as a threat, but as an act of care for the team's success
2. THREE QUESTIONS: Specific, assumptive, psychologically safe questions that directly target each of the top risks
3. THE SILENCE PROTOCOL: What to do if no one speaks up in the first 60 seconds — the exact move that breaks the silence without creating pressure
4. THE CLOSING COMMITMENT: The one thing I must say at the end of the meeting to demonstrate that I will actually act on what I heard
Write in plain language. No corporate script. These are real people in a real meeting.
Follow-Up Prompts — Go Deeper
1
"Write the exact email I send to the team 24 hours before this meeting — 4 sentences — that primes them to think about risks without triggering defensiveness."
2
"After the meeting, I need to update the project plan to reflect what I learned. What are the three most important changes I should make — and how do I communicate them to stakeholders without undermining confidence in the initiative?"
"The pre-mortem is not a pessimism exercise. It is the most optimistic thing a smart team can do. It assumes your team is intelligent enough to see the risks — and professional enough to say them out loud. Your job is simply to create the conditions where that is safe."
Use this prompt sequence with Microsoft Copilot. Replace all bracketed placeholders with your specific context. Bring the insights back to your team. The risks you surface in this exercise are not reasons to abandon the plan — they are the work that makes the plan worth executing.
AI Prompt · Copilot Tool
Solution
Ideation Engine
Go wide before you go deep. Generate ideas from multiple angles, stress-test them against reality, and shape the best ones into something worth building.
What This Does
This sequence takes the problem surfaced in your pre-mortem and turns it into a structured ideation sprint. Stage 1 goes wide — generating ideas across multiple innovation lenses including the 10 Types of Innovation, forced associations from unrelated industries, and SCAMPER. Stage 2 goes unconventional — using opposite thinking to find the non-obvious ideas hiding in plain sight. Stage 3 evaluates everything on the How-Now-Wow-Ciao matrix and shapes your best ideas into more powerful versions of themselves.
The How-Now-Wow-Ciao Matrix
WOW
High Impact · High Feasibility
The goldmine. These ideas are both transformative and executable within your current capabilities. Protect them, resource them, and move fast.
→ Build now. Protect the idea.
HOW
High Impact · Low Feasibility
Transformative but currently out of reach. These need shaping — find the version of this idea that is achievable with a different approach, partner, or phased timeline.
→ Shape it. Find the feasible version.
NOW
Low Impact · High Feasibility
Quick wins and incremental improvements. Useful for building momentum and demonstrating progress, but don't mistake ease of execution for strategic importance.
→ Execute quickly. Don't over-invest.
CIAO
LOW IMPACT · LOW FEASIBILITY
Neither transformative nor achievable. Say ciao and move on. The discipline to kill weak ideas is what keeps the strong ones alive.
→ CIAO. FREE UP THE COGNITIVE SPACE.
How to Use This Sequence
Run these four stages in order. Start with the problem statement from your pre-mortem — the specific challenge you are trying to solve. Replace every [BRACKETED PLACEHOLDER] with your specific context before pasting into Copilot. The goal of Stages 1 and 2 is volume and diversity — do not evaluate yet. Save evaluation for Stage 3.
1
Go Wide — Volume & Diversity
10 Types · Forced Association · SCAMPER
The first rule of ideation: quantity before quality. The goal here is to generate ideas from multiple, structurally different lenses — not to find the one right answer, but to populate the full landscape of possibility. Most teams stop after the first plausible idea. This prompt forces you past that.
Profit Model
Network
Structure
Process
Product Performance
Product System
Service
Channel
Brand
Customer Engagement
Why forced association works
When you borrow the structure of a solution from an unrelated domain — Formula 1 pit stops, Michelin Star kitchen mise en place, ICU triage protocols — you bypass the cognitive ruts that come from thinking inside your own industry. The best ideas often come from asking: "How would a completely different world solve this exact problem?"
Copy & Paste Into Copilot
I need to generate a wide range of solutions to a specific problem. Do not evaluate ideas yet — the goal is volume and diversity. I will evaluate in a later step.
THE PROBLEM STATEMENT: [Paste your "How Might We" challenge here — the specific problem from your pre-mortem that you are trying to solve]
THE CONTEXT: [Briefly describe your team, your organisation, and any key constraints — budget, timeline, stakeholders, or existing assets]
Generate ideas using ALL of the following lenses. Give at least 2 distinct ideas per lens:
LENS 1 — 10 TYPES OF INNOVATION (Doblin Framework):
For each of the following types, generate one idea that applies that innovation type to my problem:
• Profit Model: How might we change the economics or value exchange?
• Network: How might we leverage partnerships or ecosystems we're not currently using?
• Structure: How might we reorganise our team, assets, or capabilities differently?
• Process: How might we change how the work gets done — not what we do, but how?
• Product Performance: How might we change the core offering itself?
• Service: How might we change the support, information, or experience around the offering?
• Channel: How might we change how we reach or deliver to the people who need this?
• Brand: How might we change the story, identity, or perception?
• Customer Engagement: How might we change the relationship or emotional connection?
LENS 2 — FORCED ASSOCIATION (Borrow from Other Worlds):
How would each of the following solve my problem? Borrow their approach and translate it:
• A Formula 1 pit crew (precision, speed, choreography under pressure)
• A Michelin Star restaurant kitchen (mise en place, quality obsession, invisible systems)
• An ICU triage team (rapid prioritisation, life-or-death clarity, no wasted motion)
• A jazz improvisation ensemble (structured freedom, listening, real-time adaptation)
• A Navy SEAL pre-mission brief (contingency thinking, shared mental model, no surprises)
LENS 3 — SCAMPER:
Apply each SCAMPER verb to my problem:
• Substitute: What could we replace or swap?
• Combine: What could we merge or integrate?
• Adapt: What could we borrow and modify from elsewhere?
• Modify / Magnify: What could we amplify, exaggerate, or change the scale of?
• Put to other uses: What existing asset or capability could we repurpose?
• Eliminate: What could we remove entirely to make the solution simpler?
• Reverse / Rearrange: What if we flipped the sequence, the roles, or the logic?
LENS 4 — EXTREME USER LENS:
How would you solve this problem for:
• The most time-pressured, impatient user imaginable?
• The most skeptical, resistant stakeholder in the room?
• Someone who has never encountered this problem before and has no preconceptions?
Present the ideas in a clear list. No evaluation yet. Just generate.
Follow-Up Prompts — Push Further
1
"Of the 10 Types of Innovation, which type is most underrepresented in how my industry typically solves this kind of problem? Generate 3 more ideas specifically from that type."
2
"Pick the most unexpected forced association from Lens 2 and go deeper. Give me 5 specific, concrete ideas that borrow directly from that world's approach."
3
"What would a 10-year-old suggest? What would a 90-year-old who has seen everything suggest? What would a competitor who has nothing to lose suggest? Generate one idea from each perspective."
2
Go Unconventional — Opposite Thinking
Reverse Brainstorming · Inversion · Worst Possible Idea
The most powerful ideas are often hiding inside their own opposites. Opposite thinking forces you to generate the worst possible solutions — then inverts them to find the non-obvious gems that conventional brainstorming never reaches. This is not a party trick. It is a structured technique for breaking the cognitive lock-in that comes from thinking inside your own category.
The inversion principle
Charlie Munger famously said: "Invert, always invert." The ideas that would horrify a senior BlackRock leader are often pointing directly at the most powerful innovations — because they are the ones that challenge the deepest assumptions about how things must be done. The inversion step converts the horror into the insight.
Copy & Paste Into Copilot
I want to use opposite thinking to find the non-obvious ideas that conventional brainstorming misses.
THE PROBLEM STATEMENT: [Same problem as Stage 1]
STEP 1 — WORST POSSIBLE IDEAS:
Generate the 8 worst possible "solutions" to this problem — ideas that would be catastrophically bad, embarrassing, or completely counterproductive. Make them specific and plausible enough to be genuinely terrible, not just absurd. These are the ideas that would make a senior leader wince.
STEP 2 — THE INVERSION:
For each of the 8 worst ideas, invert it. What is the opposite of this terrible approach? What does the inversion reveal about what we should actually do? Some inversions will be obvious — discard those. Look for the ones where the inversion is surprising, counterintuitive, or challenges a core assumption we've been making.
STEP 3 — ASSUMPTION REVERSAL:
List the 5 most fundamental assumptions embedded in how we currently approach this problem — the things we treat as fixed constraints that we've never questioned. For each assumption, generate one idea that emerges if we assume the opposite is true.
STEP 4 — THE HIDDEN GEMS:
From all of the above, identify the 3 most genuinely surprising ideas — the ones that feel uncomfortable, counterintuitive, or slightly dangerous. These are the ones worth keeping. Explain why each one is worth taking seriously despite the discomfort.
Follow-Up Prompts — Push Further
1
"Take the most uncomfortable idea from Step 4. What would it look like if we ran a small, low-risk experiment to test whether the discomfort is justified — or just habit?"
2
"What assumption from Step 3 is the most load-bearing — the one that, if wrong, would change the most about how we approach this problem? What evidence would tell us whether it's actually true?"
3
Evaluate — How-Now-Wow-Ciao
Impact vs. Feasibility · Prioritisation · Honest Assessment
Now evaluate. Take the best ideas from Stages 1 and 2 and place them on the matrix. The goal is not to validate your favourites — it is to be brutally honest about where each idea actually sits. The most common mistake is placing everything in WOW because you are emotionally attached to the ideas you generated.
WOW
High Impact + High Feasibility → Build now. What is the single most important next step?
HOW
High Impact + Low Feasibility → Shape it. What is the feasible version of this idea?
NOW
Low Impact + High Feasibility → Execute quickly. Don't over-invest in it.
CIAO
Low Impact + Low Feasibility → Ciao. Say goodbye without guilt.
Copy & Paste Into Copilot
I have generated a set of ideas and now need to evaluate them honestly using the How-Now-Wow-Ciao matrix.
MY IDEAS: [Paste the ideas from Stages 1 and 2 — or your own list]
THE EVALUATION CRITERIA:
• IMPACT: How significantly would this change the outcome if it worked? (Not how much we like it — how much it would actually move the needle)
• FEASIBILITY: How realistic is execution given our current team, budget, timeline, and organisational constraints?
For each idea, assign it to one of the four quadrants and explain the reasoning in 1–2 sentences:
• WOW: High Impact, High Feasibility — the priority ideas
• HOW: High Impact, Low Feasibility — the ambitious ideas that need shaping
• NOW: Low Impact, High Feasibility — the quick wins
• CIAO: Low Impact, Low Feasibility — the ideas to say goodbye to
Be honest. Do not place ideas in WOW out of enthusiasm. If an idea is genuinely a HOW or a CIAO, say so and explain why.
After the evaluation:
1. For each WOW idea: What is the single most important next step to begin execution this week?
2. For each HOW idea: What is the specific constraint making this infeasible — and what would need to change (partner, technology, timeline, budget) to move it into WOW?
3. For the NOW ideas: What is the risk of over-investing in these at the expense of the WOW ideas?
Follow-Up Prompts — Push Further
1
"I think I am being too optimistic about [specific idea]. Challenge my assessment. What am I missing about the feasibility? What would a cynical operator say about our ability to actually execute this?"
2
"Of the HOW ideas, which one is closest to the WOW quadrant — the one where feasibility is the only barrier? What is the minimum viable version of this idea that we could test in 30 days without the full resource commitment?"
3
"We have [X] WOW ideas. If we could only pursue one, which would you recommend — and why? What does choosing one mean we are explicitly deciding not to do?"
Follow-Up Prompts — Push Further
1
"Write the one-page brief for this idea — the document I would send to a senior stakeholder to get a 30-minute meeting. Include: the problem, the idea, why now, the first step, and what I need from them."
2
"What are the three most important questions a skeptical CFO would ask about this idea — and what are the honest answers? Don't give me the polished version. Give me the real answers."
3
"If I had to run a 2-week experiment to test the core assumption of this idea — with no budget and no formal approval — what would I do? What would I be looking for? What would tell me I was right or wrong?"
"The goal of ideation is not to find the one right answer. It is to populate the full landscape of possibility — and then be ruthlessly honest about which part of that landscape is worth building in. Volume first. Evaluation second. Shaping third. In that order, always."
Use this prompt sequence with Microsoft Copilot. Start with the problem statement from your pre-mortem. Replace all bracketed placeholders with your specific context. The How-Now-Wow-Ciao matrix is not a voting exercise — it is a discipline. The ideas that survive this sequence are the ones worth taking to leadership.