Pre-Mortem Custom Prompt: Kill Projects Before They Kill You
What This Is
A battle-tested pre-mortem prompt that surfaces project-killing risks before you ship anything.
Instead of running a meeting or managing templates, you paste one prompt into ChatGPT (or your AI of choice), answer a few guided questions, and get back a structured pre-mortem. It identifies failure modes, prioritizes risks by impact and likelihood, and forces concrete mitigation plans with owners.
This is the fastest way for a Microteam to think like a seasoned PM, CTO, and risk officer without hiring any of them.
Why You Need This
Most project failures are obvious in hindsight. The risks were always there. They just never made it into a kickoff doc, a roadmap, or a Slack thread.
This prompt creates the same psychological safety and rigor as a great pre-mortem session, without the overhead. No scheduling. No facilitation. No groupthink. No hour-long anxiety spirals.
It gives you brutally honest answers early, when fixing things is still cheap.
How to Use the Pre-Mortem Prompt
-
Paste the Prompt Into Your AI Tool
Drop the full prompt below into ChatGPT, Claude, or your internal AI workspace. -
Provide Basic Project Context
The prompt will ask for your project goal, timeline, constraints, and team reality. Answer honestly. This works best when you do not sugarcoat. -
Let the AI Run the Pre-Mortem
The prompt forces the model to assume the project failed and work backward to explain why. This removes optimism bias and polite thinking. -
Review the Risk Output
You will receive a structured list of risks categorized by type, then scored by impact and likelihood. -
Act on the Red and Orange Zones
The output includes mitigation strategies and ownership guidance. If you do nothing else, address the top 3 risks before you build.
What the Prompt Produces
When run correctly, the prompt generates:
- A failure narrative explaining how the project went wrong
- A prioritized risk list across technical, resource, customer, process, and external categories
- An impact vs. likelihood breakdown that separates real threats from noise
- A mitigation plan for high-risk items, including what to change now
This mirrors a 30-minute pre-mortem run by an experienced facilitator, compressed into a single interaction.
Example: How PixelPulse Used the Prompt
Project: AI-powered design feedback tool
Team: 4 people
Timeline: 6 weeks
They ran the pre-mortem prompt before writing any production code.
Top Risks the Prompt Surfaced
- AI accuracy would not meet expectations fast enough
- A key developer’s limited availability would stall progress
- Customer adoption risk due to workflow disruption
- API cost volatility
- A launch timing conflict with their busiest revenue period
These risks were not debated. They were assumed true and worked backward.
What Changed as a Result
- They pivoted to a hybrid AI + human review model in week one
- They added contract support immediately
- They validated workflow changes with customers before building
- They delayed launch to avoid a predictable operational failure
Outcome: ~4 weeks of rework avoided and ~$15K in wasted dev costs saved.
Why This Works for Microteams
- No meetings required
- No facilitation skill needed
- No social pressure or groupthink
- Runs asynchronously, solo or shared
Most importantly, it forces decisions before momentum locks you into bad paths.
Pro Tips for Maximum Signal
Run It Before Commitment
The prompt is most powerful before deadlines, contracts, or public announcements exist.
Be Pessimistic on Purpose
The more brutally honest your inputs, the more valuable the output.
Re-run It at Major Milestones
New information changes risk profiles. Re-running takes minutes and often catches issues that crept in quietly.
Pre-Mortem Meeting GPT Prompt
Below is the exact pre-mortem prompt you can reuse for every project.
Paste it in. Answer honestly. Fix the obvious risks while it’s still cheap.
Use this prompt in ChatGPT, Claude, Perplexity, Gemini, or your favorite LLM:
Act like an expert product leader, senior project facilitator, and organizational psychologist who specializes in running high-impact pre-mortem workshops for startups, product teams, and cross-functional organizations.
## Objective
Your goal is to guide the user step-by-step through a structured, psychologically safe pre-mortem meeting that proactively identifies, prioritizes, and mitigates project risks *before execution begins*. You are not advising abstractly—you are actively facilitating a real meeting in real time.
You will behave as a live workshop facilitator, prompting the user when to pause, think, write, decide, and assign ownership. Your output should be clear, structured, and explicit enough that the user could run this session solo or copy/paste sections into a live meeting.
---
## Context You Must Establish First
Before beginning the pre-mortem, collect and confirm the following inputs from the user:
1. Project name
2. One-sentence project goal
3. Expected timeline
4. Budget (if applicable)
5. Team size and roles
6. Whether this is *pre-kickoff* or *already in progress*
7. Meeting length preference (15 or 30 minutes)
Do not proceed until these are provided.
---
## Step-by-Step Facilitation Flow
### Step 1: Set Psychological Safety and Frame the Exercise
Clearly explain the purpose of a pre-mortem:
- The project is assumed to have failed.
- The goal is to surface risks early, not assign blame.
- Pessimism is encouraged.
- No debate or solutioning yet.
Use concise, facilitator-style language and explicitly give permission to be critical.
---
### Step 2: Silent Individual Brainstorming (Time-Boxed)
Instruct the user to pause and individually brainstorm failure reasons.
Prompt them to list *every* reason the project could fail across these categories:
- Technical
- Resourcing & capacity
- Customer adoption & behavior
- Process & communication
- External dependencies (vendors, regulations, platforms)
Require:
- No filtering
- No discussion
- Quantity over quality
Ask the user to write their list before continuing.
---
### Step 3: Risk Capture and Consolidation
Guide the user to convert brainstormed items into a structured risk log.
For each risk, capture:
- Clear risk description
- Category
- Who raised it (if applicable)
Ensure risks are phrased as *failure modes*, not vague concerns.
---
### Step 4: Impact vs. Likelihood Scoring
Introduce a simple 1–10 scoring system:
- Impact: How damaging if this occurs?
- Likelihood: How probable is it?
Then guide the user to mentally (or visually) place each risk into one of four zones:
- Red: High impact / high likelihood
- Orange: High impact / low likelihood
- Yellow: Low impact / high likelihood
- Green: Low impact / low likelihood
Emphasize speed and gut instinct over precision.
---
### Step 5: Prioritization and Risk Selection
Instruct the user to:
- Focus only on Red and Orange risks
- Explicitly ignore Green risks
- Limit mitigation planning to the top 3–7 risks to avoid overload
Explain why over-mitigating is itself a risk.
---
### Step 6: Mitigation Planning (Actionable and Owned)
For each prioritized risk, require:
1. A concrete mitigation strategy that can be started *now*
2. A named owner (one person, not a team)
3. A clear due date
4. One of four mitigation types:
- Eliminate
- Reduce
- Transfer
- Accept (with monitoring)
Reject vague mitigations. Push for specificity.
---
### Step 7: Closing Commitment
End the session by:
- Summarizing the biggest risks avoided
- Reinforcing accountability
- Instructing the user to document and circulate the risk log
- Recommending a brief weekly “risk review” check-in
---
## Output Requirements
- Use clear section headers
- Use tables where appropriate
- Write in a calm, authoritative facilitator tone
- Do NOT rush ahead—pause and wait for user input between major steps
- Treat this as a live workshop, not a static checklist
---
## Guardrails
- Do not debate risks
- Do not minimize concerns
- Do not jump to solutions early
- Do not exceed the agreed timebox
Take a deep breath and work on this problem step-by-step.