The Supfact Score is a 0–100 measure of verification quality, calculated from structured, reviewed evidence under a published methodology — the same one for every product. The weights below are provisional pending independent scientific review, which will be completed before any product is certified.

Facility certifications such as cGMP, audit standing, and quality systems — judged at the facility that actually makes the product. An expired certification scores as expired, not as partial credit for having once held it.
Heavy metals, microbial contamination, and adulterant screening appropriate to the product's ingredients — with current results, traceable to identified production lots.
The heaviest category. Whether tested amounts match the Supplement Facts panel and label claims, ingredient by ingredient, for the exact label version sold. A materially inaccurate claim is a critical failure.
The qualifications of the labs behind the results — ISO/IEC 17025 accreditation for the methods used — plus complete original reports, with authenticity confirmed against the issuer.
Lot traceability, manufacturing origin disclosure, and how openly the brand shares its evidence — including whether buyers can see batch test results.
Evidence is comprehensive, current, and strong across every category.
Strong evidence across the board, with minor gaps that are shown openly.
The evidence bar is met. Weaker areas are visible on the product's page.
The product does not qualify for verification at this time.
Missing, inconclusive, and failed are three different things. They're recorded differently, shown differently, and never blurred into one another.
The evidence meets the criterion in full. Full points.
The evidence is real but incomplete — an expired-within-months certificate, a panel that covers most but not all of what's required. Partial points, and the gap is stated.
The evidence was checked and did not meet the criterion — an assay short of the label claim, a contaminant over its limit. Zero points.
No evidence was provided or found. Missing evidence earns zero — an absence never counts in a product's favor.
Evidence exists but conflicts, or can't be confirmed — two reports that disagree, a panel that doesn't match the product's risk profile. Zero points until it's resolved.
The criterion genuinely doesn't apply to this product. Only allowed with a written justification, and the category is rescaled openly on the product's page.
A confirmed critical safety failure, a materially inaccurate label claim, or fraudulent evidence blocks or revokes verification — regardless of the total score. A product can earn 91 points and still not be verified. Mandatory evidence is also mandatory: without reviewed safety testing and label-accuracy evidence, no product can be verified at any score.
Label accuracy must rest on at least one batch tested within the last 12 months — older evidence turns inconclusive rather than quietly carrying a score. Scores expire when their key evidence does, and expired verifications say so, live.
The methodology version, the input facts, the review date, and the expiry date are preserved with every score. Testing one batch never implies every batch was tested — each page states which lots its evidence actually covers.