A post-mortem happens after something fails. You look back, figure out what went wrong, and document the lessons.
A pre-mortem happens before something starts. You try to kill the idea. Find every reason it could fail. Test whether it survives.
Most people skip this step. They get excited about an idea, start building, and only discover the flaws three months in when they have already invested time and money. The pre-mortem catches those flaws on day one, when changing course costs nothing.
Why Your Gut Is Not Enough
Here is the problem with gut instinct: it is biased. When you are excited about an idea, your brain actively filters out evidence that it might not work. You notice the signals that say "go" and ignore the signals that say "stop."
AI does not have this bias. It does not care whether your idea succeeds or fails. It will evaluate it based on patterns from thousands of similar situations. It will tell you what your gut won't.
How to Run a Pre-Mortem with AI
Take your idea. Describe it clearly. Then ask AI to do three things:
Step 1: Kill it.
Here is my idea: [describe your product, project, or initiative]. I want you to try to kill this idea. Tell me every reason it could fail. Be brutal. No politeness.
AI will generate a list of failure modes. Some will be obvious. Some will catch you off guard. Pay attention to the ones that surprise you. Those are the blind spots.
Step 2: Evaluate what survives.
Of all the failure modes you identified, which are actually likely and which are unlikely? Which ones can I mitigate and which are deal-breakers?
This separates real risks from theoretical risks. Not every failure mode is equally probable. AI is good at calibrating likelihood when you push it to evaluate rather than just list.
Step 3: Pivot or proceed.
Based on the failure modes that survived, can you pivot this idea into something stronger? Or should I walk away?
Sometimes the fix is a small adjustment. Change the target market. Adjust the pricing model. Narrow the scope. AI will suggest these pivots based on what it knows about similar products and markets.
Sometimes the answer is to walk away. That is valuable too. Walking away from a bad idea on day one saves you months of wasted effort.
A Real Example
An attendee at a recent workshop had an idea for a full-stack application. They had spent time thinking through the features, the tech stack, the user flow. Everything looked good on paper.
Then they ran a pre-mortem. AI flagged three things they had not considered:
- The target market was already served by three well-funded competitors with similar features
- The pricing model assumed a subscription volume that required a user acquisition cost they could not afford
- The technical architecture they chose would not scale past 500 users without a complete rewrite
None of these were fatal on their own. But together, they painted a picture of an idea that needed significant rethinking. The attendee pivoted the idea based on the feedback. Narrower market, different pricing, simpler architecture. The core concept survived. The execution plan changed completely.
The Counterintuitive Part
The hardest part of a pre-mortem is emotional. You have to sit down with an idea you are excited about and intentionally look for reasons it will fail. That feels wrong. It feels pessimistic.
It is not pessimistic. It is rigorous.
Every successful product, project, and business went through this process. The founders who skipped it are the ones who launched, failed, and said "we never saw it coming." They didn't see it coming because they never looked.
AI makes this process fast. Ten minutes of structured questioning can surface what three months of building would have cost you. Use it.
When to Run a Pre-Mortem
- Before starting any new project or product
- Before making a significant career change
- Before launching a new service or offering
- Before committing more than one week of time to something new
- Before investing money in tools, ads, or infrastructure
Five minutes of stress-testing beats five months of regret.