Validation
Kill weak ideas early. Fund only what shows demand.
- Target markets where AI changes the economics, not the interface
- Prove willingness to pay before increasing product depth
- Confirm a distribution path exists before committing to build
- Avoid markets that require education to create demand