How to Fact-Check a Skincare Ingredient Claim Yourself

01

Reduce the claim to one sentence first

Skincare marketing is vague on purpose. Radiance, renewal, visibly transformed. None of that can be checked. Before you can fact-check anything, reduce the pitch to one concrete sentence. This ingredient, at this kind of strength, is supposed to do this specific thing.

If a product page cannot be pinned to a sentence like that, you have already learned something. There is no claim there, only a mood.

When there is a real sentence, you can check it in about ten minutes using tools that are free to everyone.

02

A checkable claim comes with a citation

A checkable claim comes with a citation. A named study in a named journal, ideally with an identifier you can look up. On research indexes that identifier is a short number called a PMID, and anything carrying one can be found by anyone in seconds.

"Clinically proven", "dermatologist approved" and a row of asterisks are not citations. They may be true. Without a pointer to a study, treat them as decoration.

No citation does not automatically mean a claim is false. Plenty of honest cosmetic claims are modest enough not to need one. It means the seller still owes you the evidence.

03

PubMed abstracts are free to read

PubMed is the public index of peer-reviewed biomedical research, run by the US National Library of Medicine, and searching it costs nothing^[1]. Paste in the PMID, or the paper's title, or just the ingredient name and the concern you care about.

You will usually not get the full paper free, and you usually do not need it. The abstract shown on every PubMed record is nearly always enough to answer the question that matters at this stage. Is this paper about the claim it is being used to support?

The answer is no more often than you would expect. A citation can be real and still beside the point. Search a popular ingredient and you will turn up papers about swallowing it as a supplement rather than putting it on skin, studies in cell cultures rather than people, and research on medical conditions borrowed to sell a cosmetic. Niacinamide is a common example. Much of its research is on oral use in clinical settings, and that says nothing about what a five percent serum does on a face.

So read the abstract with a short checklist. Was it applied to skin, the way you would use it? Was it tested in humans? Is the thing being measured the thing the product promises, rather than a neighbouring condition that sounds similar? And does the strength used in the study match anything the product discloses? If the concentration is undisclosed or plainly lower, the citation is describing a different product.

04

Some study types count for more than others

Real, relevant papers still do not all carry the same weight. Roughly, from strongest down: systematic reviews and meta-analyses that pool many trials, then randomised controlled trials, then small uncontrolled studies, then lab and animal work, which generates ideas rather than proof.

PubMed labels most records with their publication type, so you are not left guessing. A Cochrane systematic review of emollients and moisturisers sits at the top of that ladder^[2]. A two-week study on eleven volunteers with no control group sits near the bottom, however exciting its conclusion.

One paper is a data point, not a verdict. When a claim matters to your spending, look for a second, independent study saying the same thing. Notice how often there is not one.

Sometimes the search comes back nearly empty for an ingredient you see praised everywhere. That does not prove the ingredient does nothing. Ceramides are well liked and have thinner direct trial evidence for some claimed benefits than their reputation suggests. It does mean that the claim is unproven at the confidence the marketing implies, and you can price that in. Pay workhorse prices for workhorse evidence, and treat thinly evidenced promises as experiments rather than commitments.

Learning to say "the evidence is thin" is the most useful habit you can bring to buying skincare.

05

The standard we hold ourselves to

We run the same check on every ingredient note on this site. A person reads the research, matches it to the sentence it supports, and links the reference to its public record. Where the evidence is thin, the page says so.

You never have to take our word for it, because the citations are one click away. If you find one that does not hold up the sentence above it, tell us and we will rewrite the claim or withdraw it.

Questions, answered plainly

Before you add another step.