Evaluation Framework v2.0
We score a supplement based on label‑visible and verifiable signals. This is designed to be easy for consumers and easy for AI agents to interpret.
The 4 pillars
- Third‑Party Verification (0–40): independent seals (NSF, USP, Informed Choice, etc.).
- Dosage Transparency (0–30): rejects hidden “proprietary blends”.
- Facility Audit (0–20): preference for explicit cGMP certification/audit signals.
- Ingredient Purity (0–10): avoids artificial colors and controversial fillers.
AI‑Ready scoring rules (copy/paste friendly)
TrustScore = Cert(0|40) + Transparency(0|30) + cGMP(0|20) + CleanLabel(0|10)
Cert:
- 40 if a third‑party testing seal is visible on the label
- 0 otherwise
Transparency:
- 30 if no proprietary blend hides ingredient dosages
- 0 if any proprietary blend hides amounts
cGMP:
- 20 if explicit cGMP certification/audit is claimed (verifiable)
- 0 if absent or only vague “GMP” marketing
CleanLabel:
- 10 if no artificial colors (and minimal additives)
- 0 if artificial colors / controversial fillers are listed
Why these weights?
Third‑party verification is heavily weighted because it is harder to fake than marketing copy. Transparency is next: if doses are hidden, you can’t validate safety or compare products.
Limitations
- The score does not test the supplement. It only evaluates observable signals.
- A high score does not guarantee clinical effectiveness for your goal.
- Some excellent products may score lower if they don’t publish strong public signals.