Minimum evidence standard
A supplier may be listed for evaluation without being scored. Publication of a scored ranking requires a recorded review date, company-identity check, capture of product and policy pages, explicit separation of supplier claims from verified facts, category rationales and source links for material conclusions.
Unknown data is shown as “Not yet verified”. It is not converted to zero by the interface and is not described as a failure. A category score may be reduced only where the published rubric makes the evidentiary burden part of that category; confidence is reduced whenever material evidence is missing.
Category rules
1. Testing and COA transparency
Up to full credit for current, product-relevant, batch-specific reports with an identifiable laboratory, methods, dates and an independent verification route. Partial credit reflects incomplete relevance or verification. Generic, expired or unverifiable documents do not receive full credit.
2. Company identity, compliance and policy transparency
Assesses the operator’s identity, address, contact routes, research-use positioning, legal and commercial policies, and clarity about responsibilities. It does not infer legal compliance from a polished policy page.
3. Product traceability and information quality
Assesses lot display, correspondence between stock and documents, storage information, measured claims and product-page consistency. Missing details stay unknown.
4. EU fulfilment and delivery transparency
Assesses dispatch origin, destinations, tracking, delivery estimates, carrier detail, restrictions, returns and customs language. A European domain or currency is not evidence of EU dispatch.
5. Customer support and issue resolution
Assesses published channels, test enquiries, response evidence and clarity of issue handling. Testimonials controlled by the seller do not substitute for support testing.
6. Website security, payment and privacy
Assesses HTTPS, checkout hand-off, payment disclosure, privacy information, data minimisation and obvious security hygiene visible to an ordinary visitor. It is not a penetration test.
7. Independent reputation evidence
Assesses clearly sourced external reputation signals with observation dates and counts. Third-party scores are never merged with the editorial score.
8. Price and fee transparency
Assesses visible pricing, taxes, shipping fees, currency and conditions. Cheapness is not rewarded; clarity is.
Partial credit and material penalties
Reviewers record a reason beside every category score. Partial credit is used where relevant evidence exists but is incomplete, not independently confirmable or not current. Material penalties may apply for substantiated deceptive claims, falsified or misrepresented documents, undisclosed conflicts, unsafe security handling, or contradictory company information. A penalty requires a source, date, editor review and an opportunity for correction or response.
Selection and verification
Sites may be selected for European relevance, availability to legitimate research purchasers, search interest, evidence availability and market presence. Selection is not endorsement. Supplier statements are labelled as claims until corroborated. If anonymous purchasing or laboratory testing is introduced, the protocol, sample chain, funding, laboratory and limitations will be published before results influence scores.
Cadence, conflicts and appeals
Target cadence is quarterly, with event-driven rechecks after material changes. Affiliate status cannot affect criteria or placement. Paid ranking and guaranteed placement are prohibited. Suppliers may submit a correction, right-of-reply or appeal with supporting sources; editors publish material corrections and record score changes.
Change log
v1.0 — 3 September 2026: Initial eight-category model, unknown-data handling, confidence tiers, penalties, commercial-independence rules and appeals process published. Provisional requested positions are stored outside the scoring engine and cannot qualify for an evidence-backed rank.