A public product rule — not an internal memo
Entertainment on top, evidence beneath. You should be able to enjoy a show — and at any moment see what the host knows, where a statement came from, when it was last reviewed, and which statements are not adequately supported. What we promise is a verifiably documented process, not infallible truth: sources change, studies conflict, editors correct mistakes — visibly.
Identity, INCI and claims are confirmed — and at least one independently funded, independently authored study of the specific product exists. Without that, no product reaches this tier, however good it looks.
Additionally, the advertised claim set survives verification intact. A single removed headline claim is enough to keep a product below this line.
Publishable — with a narrowed claim set and mandatory contraindication language. The honest verdict for products that plausibly work but are sold with claims their evidence can’t carry. See it applied: Effaclar Duo+ M →
A safety problem, a material misrepresentation of identity, or absent ingredient disclosure. Rejections are published with reasoning and sources — and never hidden from search.
A claim is a single verifiable statement with scope (variant, market, formulation, point in time), typed sources and an owner. A charismatic line cannot appear as a product fact unless it maps to an approved claim.
Exact names per market, EAN/GTIN, PZN, legal entity, country of origin, full INCI — cross-verified across at least three sources.
Every marketing claim quoted verbatim, categorised, and matched against manufacturer and independent substantiation.
Each key ingredient examined separately: mechanism, trial quality, dose, and whether the product plausibly delivers a studied dose.
Ratings across platforms treated as experience patterns, not evidence — including contradictions we refuse to smooth over.
Who the evidence actually covers — and mandatory contraindication language for everyone it doesn’t.
Dated, sourced prices across merchants, honest alternatives — including cheaper options from the same corporate group, disclosed.
Ownership structures, funded studies, affiliate landscape, and our own relationships — stated for the record, on the page.
Three lists: removed entirely, downgraded to exact wording, approved for scripts and Q&A. Four-eyes review for anything safety-adjacent.
Twelve-month expiry plus hard triggers: INCI changes, regulatory action, recalls, new independent trials, price moves beyond ±15 %.
Evidence Quality · Product Performance · User Fit · Value · Trust — each scored 1–5 with written justification and sources. A low Evidence score next to a high User-Fit score is the informative result, not a rounding problem. Collapsing them into one star rating would destroy exactly the information you came for.
ALINA carries the visible label “AI avatar by entysia” on every deliverable. Licensed digital twins appear only under fully approved rights records. No open creator onboarding exists.
Affiliate relationships are marked before or at the click, stored with a disclosure version — and can never override claim status, ranking or a chat answer. A merchant without a confirmed AI-content policy is not activatable.
Formulation changes create a new product version — never a silent overwrite. Corrections are versioned: what changed, why, by whom, from when. In severe cases a dossier is withdrawn, not “repaired”.