Skip to content
MouthMetric

Editorial policy

How MouthMetric decides what to publish, how we handle evidence and errors, and where commercial relationships stop.

Independence

  • Scores are not for sale. Commission terms never enter dimension inputs.
  • If two products score within noise, commission cannot decide a “top pick.”
  • The site must remain useful if every affiliate program ends.

Accuracy

  • Prices are “as observed” on a dated check — not eternal guarantees.
  • Product names, ingredients, and policies are verified before merge.
  • When we cannot verify something, we label the gap instead of papering over it.

Evidence handling

  • Independent evidence is preferred over vendor marketing materials.
  • Claims use the five-label taxonomy: vendor claim · independent evidence · editorial interpretation · user report · verified fact.
  • Claim Evidence (strong / moderate / limited / vendor-only) applies to individual claims — distinct from Score Confidence.
  • Red-flag health terms trigger verification; unverifiable claims are quoted only as vendor claims with explicit non-endorsement.

User reports

Aggregated external reviews inform the User Feedback dimension with volume and pattern rules (defined for the scoring engine). We do not present cherry-picked testimonials as clinical proof.

Affiliate relationships

Disclosures appear site-wide, on money pages above the first affiliate link, and on link markup (rel="sponsored nofollow"). Details:affiliate disclosure.

AI-assisted research and drafting

We do not claim automated processes that do not exist yet. Human final approval is required before anything merges to main (see our editorial gates documentation).

Human editorial judgment

  • No fabricated experts, dentists, lab tests, or “we spent N hours testing” claims unless performed.
  • Bylines use MouthMetric Editorial Team until a real publisher identity is published (open decision Q18).
  • Corrections are logged publicly — see corrections.

What we will not publish

  • Medical diagnoses or treatment instructions.
  • Scores described as medical effectiveness or guaranteed outcomes.
  • Sponsored copy disguised as independent evaluation.
  • Invented statistics, study results, or credentials.