Failure Mode Analysis
Reasoning
Level 2 · Advanced reasoningSystematically identify how a plan could fail — on your own work, or applied to analyze a competitor's.
Failure Mode Analysis works backward from failure: given a plan, it identifies the most likely ways it could go wrong, ranked by severity, with a mitigation or early warning sign for each. It has two variants that share the same underlying logic — reflective (stress-testing your own plan before committing to it) and applied externally (analyzing a competitor's strategy for its weak points) — shown here as two examples of the same technique, not two separate ones. Distinct from Adversarial Review: this technique enumerates risks to a plan *before* a draft even exists, while Adversarial Review critiques a draft that's already been written.
When to use it
Use the reflective variant before committing to a decision or plan; use the external variant when analyzing a competitor's strategy or a rival plan.
When to avoid it
Avoid it for low-stakes decisions where a wrong outcome costs little to correct.
Advantages
- Surfaces risks before resources are committed, not after something has already failed
- Ranking by severity and likelihood focuses attention on what actually matters
- The external variant turns the same mechanism into a lightweight competitive-analysis tool
Disadvantages
- Can produce a generic list of risks if the plan itself is too vague
- Doesn't guarantee it caught the actual failure mode that will occur
Examples
Reflective — Input: launching a subscription tier at $15/month with a 14-day free trial, targeted at existing free users. Identify the most likely ways this could fail, ranked by severity and likelihood. For each, suggest one mitigation or early warning sign.
Applied externally — Input: a competitor just announced a similar subscription tier at $12/month with an unlimited free trial. Identify the most likely ways this competitor's strategy could fail, ranked by severity and likelihood. For each, suggest what evidence would confirm it's happening.
Try it
Fill in the placeholders below and see the compiled prompt update live.
What you want to achieve.
The material to work from.
- 25
- Words
- 168
- Characters
- 0/2
- Placeholders filled
- 1 min
- Reading time
- Completeness0
- Clarity0
- Context0
- Specificity0
- Constraints0
- Format0
- Output0
- Constraints note0
- Success criteria0
- Scope focus100
- Verification step0
- Negative example0
- Confidence0
- Placeholder syntax100
- Markdown structure0
- Fill in the remaining placeholders.
- Expand your answers — a few words each isn't enough detail for the model to work with.
- Add background context so the model understands the situation.