Editorial methodology

How our local shortlists work

ThreeNearMe narrows a category to three editorial selections. The shortlist is intended as a practical starting point, not a guarantee or substitute for your own checks.

Research and selection

ThreeNearMe deliberately limits each shortlist to three researched choices so readers do not have to work through an endless directory. Editors use publicly available information, service coverage, business websites, independent sources, and category-specific criteria. Desk research has limits: details can change and we do not claim personal experience unless explicitly stated and documented.

No paid ranking

Businesses cannot buy inclusion or position in the Top Three. A page may carry one clearly labelled advertisement, but sponsor records are stored, queried, and displayed separately from recommendation selections.

Human judgement

Category-specific criteria and internal comparison scores help editors check consistency and identify evidence gaps. Scores support editorial comparison; they do not replace judgement. A human reviewer approves every final shortlist, and AI assistance cannot select or publish the Top Three.

Evidence and reasons

Each selected entity needs retained source evidence and a clear public reason. We record what was checked and when, assess source quality, and retain contradictory information for review. Private sources, weights, scores, internal notes, and personal assignment details are not published.

Review cycle

Published pages record selection and review dates. We periodically revisit availability and supporting information, while acknowledging that changes may occur between reviews.

Corrections

Readers and businesses may privately submit corrections or suggestions. A submission enters an editorial queue and never changes published content automatically.

Contact the editorial team

Submissions are private and reviewed by the editorial team. They never change published content automatically.

Include the page, claim, or business details we should check.