Reverse a decision that turned out to be wrong
Sticking with a decision to avoid looking wrong usually costs more than admitting it and changing course.
- Outcome
- An honest confirmation that the decision is actually wrong rather than just difficult, a reversal planned to minimize disruption to whoever built on the original decision, and a way of communicating it that preserves trust instead of looking erratic.
- Time
- A few hours to plan, though the reversal itself may take longer to execute
This workflow is part of PromptKit Pro
A handful of prompts and workflows stay free for everyone — Pro unlocks the full library, including this one, plus the guided runner.
Reversing a decision that turned out to be wrong is often avoided longer than it should be, because reversing looks like admitting failure, while sticking with a wrong decision just to appear consistent quietly compounds the cost. This playbook first confirms honestly that the decision is actually wrong rather than just difficult to execute, then plans the reversal to minimize disruption for anyone who built on it, and communicates it in a way that reads as good judgment rather than erratic reversal.
When not to use this
If the decision is just proving difficult to execute but is still the right call, reversing would be the actual mistake here — this is for when the decision itself, not just its execution, is genuinely wrong.
Did this work?
The reversal happens with disruption minimized for anyone who built on the original decision, and is communicated in a way that preserves trust rather than looking erratic or evasive.
Tested on claude-opus-5. Evidence status is draft; it moves to battle-tested only on recorded runs, never by hand.