Testing & Evidence Policy

Evidence-first publishing

Each shortlist includes an evidence summary and route factors that explain why picks are placed in Budget, All-Rounder, Premium, and specialist or alternative roles.

Most UK Shortlists pages are based on structured evidence review: public product information, category criteria, merchant or manufacturer listings, source packs where available, external review patterns where relevant, and editorial judgement. Where relevant, we update routes when important product details, terms, or pricing context changes.

Evidence levels we use

• Structured evidence review: organised comparison using public information, category criteria, source packs where available, and editorial judgement.
• Spec-verified comparison: key claims checked against manufacturer or merchant-listed specifications.
• Evidence-supported comparison: source-backed review supported by independent review patterns, buyer feedback themes, named expert input where documented, or multiple corroborating sources.
• First-party use checks: not a default public state; used only where UK Shortlists has direct product experience and says so on the page.
• Long-term tested: not a default public state; used only where there is documented extended use over time.

For the full reader-facing label guide, see Review Basis.

Manufacturer samples and first-party checks

UK Shortlists may accept gifted, discounted, or loaned products when they can improve household-use evidence for a relevant route. A sample does not guarantee inclusion, positive coverage, an award, a ranking position, affiliate placement, or a publication date.

If a sample materially affects the evidence basis, the page should disclose whether the product was gifted, discounted, loaned, returned, or used only for brand-provided data. Hands-on wording is used only where there is a test log covering receipt, setup, real use, limitations, and reviewer identity.

How automation and human review are combined

• Automation helps with route validation, rendered-output QA, and repeatable evidence hygiene checks.
• A named editor compiles recommendation logic and evidence framing before publication.
• A named reviewer checks trust copy, disclosure alignment, and policy consistency before review timestamps are updated.
• Automation may flag weak, stale, or inconsistent pages; it is not used as a substitute for editorial accountability.

How expert input is used

Expert input is used as a bounded evidence lane. It can inform criteria, caveats, exclusions, and claim safety, but it is not treated as a blanket claim that a product was tested, professionally endorsed, medically suitable, or safe for every reader.

Where a page relies on named external expert review, professional course material, or specialist source evidence, the page or source pack should identify the scope. See the Expert Review Framework.

What we do not claim

• We do not claim every route is based on direct hands-on testing.
• We do not claim lab testing unless a page explicitly says that a real testing protocol was used.
• We do not claim continuous live monitoring of every merchant listing.
• We do not treat affiliate payout levels as a ranking rule.
• We do not fill missing product data with guesses when a value cannot be verified safely.

When a page does not explicitly state direct testing, readers should assume the route is based on structured evidence review and editorial evaluation, not hands-on or lab testing.