About this site
Pore Evidence exists because the comedogenic ratings circulating online are copied from a small number of decades-old studies, mostly on rabbit ears, and then printed as though they were product-level facts. This site shows the underlying evidence instead, including where there is none.
Who maintains it
Software engineer and health-research hobbyist. Not a clinician. Read the editorial policy and limitations.
I am a software engineer and a health-research hobbyist, not a dermatologist or a cosmetic chemist. That shapes what this site is willing to claim. The work here is locating primary sources, reading them, recording what they actually tested, and presenting that honestly — not diagnosing skin conditions and not turning a literature lookup into a medical claim.
Because the author is not a clinician, the site is designed so that you never have to take my word for anything: every result names its sources and the kind of study each one was, so you can go read them and disagree.
What it does
Paste or scan an ingredient list and the checker matches each name against a curated evidence dataset — currently 69 records drawn from 9 sources, last reviewed 2026-08-06. Instead of a score, each name gets a result class and an evidence tier: what was found, and what kind of study found it. Names with no verifiable assessment are reported as insufficient rather than quietly rated.
Everything runs in your browser. There is no account, and the list you check is never sent anywhere — see the privacy page for the specifics.
Why there are fewer pages than records
The dataset holds 69 records, but only 17 ingredients have a written page. That gap is deliberate. A page is written when there is something specific and defensible to say — a contested classification, a widely repeated claim that traces back to a single weak study, a case where human testing disagreed with the animal assay. Where the honest summary would be “this appears in the corpus and nothing interesting is known,” no page is published.
Generating a page for every record would be easy and would make the site larger, more thoroughly indexed, and worse. The methodology page sets out the inclusion rules and how results are classified.
How it is funded
The site carries advertising, which pays for hosting and for the time spent reading papers. It has no sponsors, no affiliate links, no paid placements and nothing to sell you. No advertiser sees a page before it is published, and no commercial relationship has any bearing on what the evidence pages say. If a popular ingredient’s evidence is weak, the page says so.
Corrections
The corpus is small and this field’s literature is genuinely messy, so errors are likely. Anything factual here should be checkable against a cited source; if a citation does not support what the page claims, or a source has been superseded, that is a defect worth fixing. Corrections are made in the dataset and its version rather than by quietly editing wording, so the change is visible.