# Pivot or persist The go/no-go after scoring and validation. **Which rule decides:** the /15 score in [SKILL.md](../SKILL.md) (≥ 8 go, 6–7 conditional, < 6 no-go). The only exception is the no-go list below: any one of those stops the idea whatever it scored. The five go checks are not a second vote; use them as evidence when you score. ## Go checks: evidence for the score 1. Demand score ≥ 2 on the [customer-signal rubric](customer-signal-rubric.md) 2. Feasibility is "straightforward" or "hard but solvable"; no open research problem 3. A testable MVP in two weeks or less 4. The founder has an unfair advantage: domain, existing users, unique data or relationships 5. Crypto is necessary: the product is worse or impossible without the chain ## No-go: any one is enough 1. Demand score is 0 2. A well-funded team already shipped a good version (you are late and worse) 3. The core technical problem is unsolved (that is research, not a product) 4. Crypto is ornamental: the chain adds friction, not value ## Every no-go gets a pivot | Pivot | Keep | Change | |---|---|---| | Adjacent | the problem | the user segment | | Wedge | the users | a smaller first problem | | Stack | the concept | the layer (infra, app or protocol) | | Chain | the idea | the ecosystem, where timing or liquidity is better | ## State the confidence | Confidence | Meaning | |---|---| | 0.8–1.0 | Signals point clearly one way | | 0.5–0.7 | Mixed; one more week of validation is worth it | | 0.0–0.4 | Not enough data to decide; research before deciding |