skininsider

What the skincare industry knows and does not tell you

Dispatch 6 August 2026Published independentlyNo affiliate links here
Claims

Professional cosmetic claims briefing

For brand, formulation, regulatory and marketing teams in the UK: what a cosmetic claim needs behind it before it is published, in the order the questions actually arise.

SectionClaims
Reading11 min
Reviewed20 October 2026
EditorSkin Insider

A UK cosmetic claim needs documentary evidence capable of supporting it as a reader would understand it, held before the claim is published rather than assembled after it is challenged. This briefing sets out the four questions that decide whether a claim can go out, what kind of study answers each of them, where the wording crosses into medicinal territory, and who inside a business is accountable for the file behind the sentence. It is editorial. Nothing here is legal, regulatory or medical advice, and no brand, manufacturer or agency pays to appear in it.

Who this briefing is for

It is written for the people who have to say yes or no to a sentence: the brand owner who wants the claim, the formulator who knows what the formula does, the regulatory lead who holds the product information file, and the marketing team that has to write the headline. Those four read the same claim differently, and most disputes inside a business are not about the science but about which question is being answered.

It assumes a product sold in the United Kingdom and marketing seen by a UK audience. It does not cover medicines, medical devices or food supplements, and it does not decide whether any particular claim is compliant. That decision belongs to the business and, where it matters, to its advisers.

The four questions a claim has to survive

Take the claim exactly as a reader will meet it, headline and qualifier together, and answer these in order. A claim that fails one of them does not usually need better evidence; it needs different wording.

QuestionWhat settles itWhere it is written down
What is the claim, as a reader would understand it?The headline and every qualifier near it, read together. A footnote does not undo a headline, and an image can make a claim on its own.The approved artwork, the page as published, and a dated capture of both
Is it a cosmetic claim at all?Whether the stated purpose stays with cleaning, perfuming, changing appearance, protecting, keeping in good condition or correcting body odours, or moves to treating a conditionThe product information file, and the presentation as a whole
What kind of evidence would support it?Whether the claim is about a measured physical change, about perception, or about composition. Each needs a different study, and they are not interchangeable.The substantiation dossier, with the protocol, not only the result
Is that evidence held now?Whether the documents exist, cover the product as sold and the population the claim addresses, and can be produced on requestThe dossier, dated before the claim is published

Decision rule: publish only when all four answers are written down somewhere a colleague could find them without asking you. If one of them lives only in a conversation, the claim is not ready.

What counts as evidence, and what does not

The common failure is a real study answering a question the claim does not ask. A questionnaire can establish that most of a panel agreed with a statement. It cannot establish that a measurement changed. If the claim says skin is firmer, the reader hears a physical change, and a perception study does not reach it. If the claim says most users said their skin felt firmer, the perception study is exactly the right evidence, and the wording has to carry the panel and the question asked.

Three things that look like evidence and are not: a reference to testing with no protocol attached, an ingredient study run at a concentration or in a vehicle unlike the finished product, and a supplier claim adopted without the data behind it. A supplier dossier can support a claim about the raw material; it rarely supports a claim about the product on the shelf. Our article on how a claim is substantiated before it is printed sets out what a usable dossier contains, and what clinically proven is allowed to mean covers the phrase that attracts the most challenges.

Keep the protocol with the result. A number with no method behind it is not portable: the next person who needs it will not know what it measured, and neither will anyone reviewing the claim.

The wording that moves a cosmetic towards a medicine

The boundary is decided by presentation and by stated function, not by the ingredient list on its own. A product described as treating, curing or preventing a condition is being presented as a medicine, whatever its formula. Naming a condition is where most wording drifts: a claim about the appearance of blemishes sits differently from a claim about treating acne, and a claim about the comfort of dry skin sits differently from a claim about eczema.

The practical test for a marketing team is to read the claim aloud and ask what a reader would expect to happen to a diagnosed condition. If the answer is that it would be treated or prevented, the claim needs rewriting rather than more evidence. The Medicines and Healthcare products Regulatory Agency publishes its guidance on how a product is judged to be a medicinal product, and our article on the line between a cosmetic and a medicine explains why claims are worded the way they are.

Who signs off what

Accountability is easy to lose between functions, so write it down once. The responsible person named for the product carries the regulatory obligations attached to it, including the product information file and the safety report; our articles on who the responsible person is and what a cosmetic product safety report is and who signs it set out what that covers. The marketer who publishes the claim is the one who has to be able to produce the evidence for it. Those are different jobs and they are often different people.

Two habits prevent most of the trouble. Date everything, so the file shows what was held when the claim went out. And keep one owner per claim, named, so that a challenge does not arrive at a team rather than a person.

Running the review in practice

For a live claim, work from the record rather than from memory. Capture the claim and its setting, mark the type of proposition, record the stated evidence and the qualifiers, then mark what is missing without guessing at it. That is the sequence in our free cosmetic claims evidence checklist, which is the working document for one claim at a time; this briefing is the reasoning behind it.

For a claim still being drafted, run the four questions above before the artwork is approved, not after. Rewording a claim at draft stage costs a sentence. Rewording it after launch costs the artwork, the retailer listings and the pages that quote it.

Limits and appropriate use

This briefing describes how claims are assessed in the United Kingdom. It does not decide whether a particular claim is compliant, does not substitute for the product information file, and does not constitute legal, regulatory or medical advice. Rules change, and the primary sources below are the ones to read rather than taking our word for any of it.

We do not review claims for businesses, and we do not take payment from any brand, manufacturer, retailer or agency. If you want a review, that is work for your own regulatory adviser.

Questions people actually ask

Is holding the evidence after publication enough?

No. The point of the file is that it exists before the claim is published, so dating it matters as much as its contents. A dossier assembled in response to a challenge answers a different question from the one the claim raised.

Can a supplier dossier support a product claim?

Sometimes, and only for what it actually tested. A study on a raw material at a given concentration in a given vehicle supports a claim about that material under those conditions, not a claim about the finished product unless the product matches them.

Do percentages make a claim stronger?

Only if the method is attached. A percentage from a questionnaire describes agreement, not measurement, and the wording has to say which it is, how many people were asked and what question they answered.

Primary sources

Cited because they are public, stable and checkable. Read them rather than taking our word for any of this.

This is journalism about an industry. It is not medical, legal or regulatory advice. For a skin condition, see a clinician.

This article contains no commercial links. No affiliate links, no sponsored placements and no paid mentions appear anywhere on this site. No brand, manufacturer, retailer or agency is named in our editorial, and no company that pays us can appear in it. Published independently.

The dispatch

One email a fortnight. A new translation, a mechanism explained, and what changed in the rules. No product recommendations, because we do not make any. No affiliate links, because we do not take commission. Unsubscribe in one click.

We use your address to send the dispatch and nothing else. See privacy. Sponsorship rates and limits are published on the newsletter page.