1 · Primary evidence
Original research, analytical reports and documents tied to a specific claim or batch are preferred when available.
A transparent framework for research, supplier comparisons and editorial decisions. We would rather publish a lower-confidence conclusion than manufacture certainty.
Original research, analytical reports and documents tied to a specific claim or batch are preferred when available.
Documents that can be checked through an identified laboratory or other issuing-provider verification surface receive stronger documentary weight than an unverified copy.
Systematic reviews, authoritative reviews and technical guidance help establish context and identify areas of agreement or uncertainty.
Supplier pages, policies and certificates can document what a company says or publishes, but are not automatically independent verification.
Indexes and databases can be useful for locating or cross-checking documents, but their commercial relationships, extraction methods and source provenance must be considered.
We can inspect the relevant source or document and establish what it supports.
The document or result has an identifiable issuing-provider verification route. This supports the existence and attribution of the record, not product safety or chain of custody.
The information comes directly from the supplier but has not been independently established beyond the supplier's own publication.
An external database or index preserves or summarizes the evidence. We disclose material commercial relationships and do not treat aggregation alone as laboratory verification.
We cannot currently establish the claim from sufficient evidence. This is not treated as positive evidence.
We will compare what can actually be verified, not who has the best marketing.
Lot-specific COAs, named analytical methods, identity evidence, purity reporting and the ability to connect a document to a product lot.
Testing-lab identification, accessible documentation, business information, policies and clear dates for information that can change.
We record how much of the framework has actually been checked. More published evidence can make a supplier easier to evaluate without proving better underlying product quality.
Price and service can be compared separately from analytical quality. Shipping, support and usability should never be presented as proof of scientific quality.
A score is a summary of the evidence we have collected against defined criteria. It is not a guarantee of product quality, safety, efficacy or suitability for human use.
Where evidence coverage is incomplete, we will say so. A supplier with more published documentation may be easier to evaluate than a supplier with less documentation; that difference should not automatically be described as a difference in underlying product quality.
A COA can support a claim about what a tested sample reportedly contained. It does not automatically establish the identity of every unit sold, current inventory status, chain of custody, manufacturing consistency or suitability for human use.
Affiliate relationships may help fund the site, but they should never determine an editorial conclusion or numerical score. Affiliate status will be disclosed on relevant pages. If we cannot verify a claim, we will label it as unverified rather than filling the gap with marketing copy.
When an external evidence database has a commercial relationship with a supplier, that relationship will be disclosed and the database will be treated as a secondary source rather than automatically as independent evidence.
Supplier information, pricing, catalogues, testing documents and affiliate links can change. Pages should carry a review date when appropriate, and material corrections should be documented rather than silently overwritten.