Ratings and Data
Olive Branch Global Product Reviews publishes product information, comparisons and buying guides. Evidence, review status and testing scope vary by individual record. This page states the research standard; it does not show that every legacy product or article has completed review.
Manufacturer documentation establishes what the manufacturer states; it does not independently establish performance. A regulatory record applies to its named jurisdiction, product and date. An independent test applies to its identified sample, configuration, method and period.
Legacy content is under record-by-record review for exact model/version and market, direct claim-matched sources, verification dates, testing evidence and trust fields. Missing, conflicting, mismatched, stale or unverified material should be treated as unknown and labelled Needs editor verification.
Specifications, prices, availability, recalls, safety notices and regulatory status can change. Check the relevant current manufacturer or authority source for the named market before an important purchasing or safety decision.
Our Methodology
The claim-to-source process below is the standard for new and materially revised content created with human or AI assistance. It does not certify that all existing pages already meet the process.
1. Define scope. Identify the exact product/entity, model or edition, version/configuration, country/market or jurisdiction, audience and as-of date. Hold claims whose answer depends on missing scope.
2. Split material claims. Separate identity, specifications, price/stock, safety, compatibility, performance, user observations and editorial inference so each statement can be checked on its own.
3. Match evidence to the claim. Use manufacturer documents for manufacturer-stated facts; regulators or standards bodies for scoped legal/safety records; original transparent testing for comparative performance; dated seller pages only for seller-specific commercial terms. Search/AI summaries and uncited aggregators are discovery leads only.
4. Open and record the direct source. Capture its title, publisher, type, URL, publication/update and check dates, exact section/table, model/version, market, what supports the claim and what it cannot establish. A generic brand page cannot support an exact-model claim.
5. Test scope, freshness and conflicts. Check claim-source fit, reachability and freshness. Corroborate material safety, regulatory, comparative or disputed claims with an appropriate independent/official source. Preserve conflicting evidence with dates, model and market; withhold unresolved conclusions.
Scoring and Evidence Limits
6. Cite and review. Put each direct source link beside the claim it supports. Label evidence as primary-source fact, regulatory record, independent test, attributed user observation, editorial inference or unknown. A human editor verifies every material claim and its limits before approving new or revised copy; unsupported claims stay out.
AI tools may help find, extract, compare and flag source material, but they do not supply missing facts from model memory, invent citations or approve their own claims. Citation presence alone does not demonstrate support.
The published Editorial Policy describes a weighted scoring approach. Its application to legacy records is being reconciled with the individual criteria, evidence inputs, product/market scope, method, calculation and editorial review.
Some legacy pages may still contain scores or review labels whose complete basis is not verified. They should not be treated as validated, independent test results, verified consumer ratings or comparable scores across categories until their product-specific evidence and calculation are documented.
Any new or revised numerical assessment requires documented criteria and weights, traceable inputs for each material criterion, a reproducible calculation, exact model/market/date scope and human editorial approval. If those elements are missing or the evidence conflicts, omit or hold the score and mark the record for editor verification.
Evidence grade, review status and testing status answer different questions. Desk research is not hands-on testing; manufacturer statements are not independent evidence. Use the Product Database link below to inspect record-specific sources and limitations. This page describes the standard and does not guarantee coverage or AI-search citation.