Every objective claim on a cosmetic pack is supposed to be attached to a document. In a well run business it is. The document is rarely published, almost never requested, and its existence is the reason the industry can say, accurately, that its claims are substantiated. What that word conceals is the step where a careful scientific finding becomes a confident sentence.
The requirement, plainly stated
Two frameworks operate at once. Advertising rules require that marketers hold documentary evidence to prove claims that consumers are likely to regard as objective and that are capable of objective substantiation. Cosmetic specific rules require that claims about products meet common criteria: legal compliance, truthfulness, evidential support, fairness, and enabling informed decision making.
Neither framework requires pre approval. Nobody checks the dossier before the product ships. The check happens if a complaint is made, at which point the advertiser has to produce the evidence and it is assessed against the claim as consumers would understand it. That distinction, between holding evidence and having it examined, explains a great deal about how claims are written.
What actually goes into a claims dossier
| Element | What it can support | Where it is over extended |
|---|---|---|
| Instrumental study on the finished product | Specific measured changes over the tested period in the tested population | Extrapolating to different durations, populations or measures |
| Consumer perception study | What users reported thinking about the product | Presenting perception as measured physical change |
| Ingredient literature | That a material has shown an effect in published work at a stated concentration | Assuming the finished product delivers the same, at the same level, in the same vehicle |
| Supplier claim documentation | Whatever the supplier has evidenced about its raw material | Transferring a supplier's ingredient claim to a finished product claim |
| Tolerance or patch testing | That the product was well tolerated in the tested panel | Turning tolerance into suitability for sensitive skin, or into efficacy |
| Stability and preservation data | That the product remains as specified over its shelf life | Reading it as evidence of anything about performance |
The row that causes the most trouble is ingredient literature. A material shown to do something at a given concentration in a given vehicle in a published study is not the same as a finished product containing that material at an unstated level in a different vehicle. Bridging that gap requires either testing the finished product or wording the claim so it is about the ingredient rather than about the product.
Formulated with a proven active
That the product contains an ingredient for which some body of evidence exists. The claim is about the ingredient, not about the product, and it is worded that way deliberately.
This is a legitimate construction where the ingredient literature genuinely exists and where the claim does not imply that the finished product has been tested.
It does not mean the finished product was tested. It does not mean the ingredient is present at the concentration used in the evidence, or in a form that remains active, or at a pH where it works. It does not mean the rest of the formula does not interfere with it. It does not mean the product produces the effect described in the literature.
The finished product would have to be tested, and the claim would have to be about the product rather than the ingredient. Alternatively the brand could state the concentration and form of the active, which at least lets a reader compare the product against the published evidence themselves.
The step where the risk actually is
A study produces a narrow statement: in this population, over this period, using this method, this measure changed by this amount relative to this comparator. A pack has room for six words. Somebody has to perform the compression, and that person is usually not the person who ran the study.
The compression is where claims go wrong, and it goes wrong in recognisable ways.
- Dropping the comparator. Improved against a no treatment control becomes simply improved.
- Dropping the population. A result in a selected group becomes a result for everyone.
- Dropping the measure. A specific instrumental reading becomes a general quality like radiance or firmness.
- Dropping the duration. A twelve week result becomes an unqualified statement, or worse, is paired with imagery suggesting immediacy.
- Changing the subject. A perception result becomes a statement about skin.
None of these requires anyone to lie. Each is a small editorial decision that survives internal review because, taken alone, it is defensible. The aggregate is a sentence the study does not support.
Reviewing your own homework
Claims review is almost always internal. The people who assess whether the evidence supports the sentence are employed by the business that needs the sentence to work, and they are reviewing under the same launch deadline as everyone else.
The external check exists but is slow and rare relative to the volume of claims published. A complaint has to be made, investigated and ruled on, by which point a campaign has run. The rational commercial position is therefore to write to the edge of what is defensible, because the expected cost of being wrong is low and the benefit of being bold is immediate.
- Marketing teams. Get the strongest sentence the process will allow
- Testing houses. Sell studies scoped to the claim rather than to the question
- Competitors. Are pushed to match the boldest claim in the category or look weak
- Regulators. Carry an enforcement burden that is structurally larger than their capacity
- Consumers. Encounter the strongest permissible version of every claim, all at once
A description of how a complaint led system behaves. It is not an allegation about any particular advertiser.
Who signs it off
In a large business, claims pass through regulatory affairs, legal, and sometimes an external adviser, with scientific input from whoever commissioned the studies. In a small business it may be one person, or nobody with relevant expertise at all. The requirement is the same in both cases: hold the evidence before you publish.
This asymmetry is worth knowing because it inverts a common assumption. Large companies are frequently more conservative in their claims than small ones, not because they are more scrupulous, but because they have more to lose, more people who can say no, and a compliance function whose job is to say it.
What happens if you ask
You can ask a brand for the substantiation behind a claim. There is no obligation to provide it to a consumer, and many will decline or send marketing material instead. Some will send a study summary, and that is informative in itself.
The route with teeth is a complaint to the Advertising Standards Authority, which can require the advertiser to produce its evidence and which publishes its rulings. Reading a few published rulings in this sector is the single most useful hour anyone can spend on this subject, because the reasoning shows exactly where the line falls in practice.
The practical version
When you meet a claim, separate it into three parts: what is being claimed, about what, on what basis. Most cosmetic claims fail at the third part, not because evidence does not exist, but because it is not shown. A brand that shows its working is doing something unusual and it deserves credit for it, independently of whether the product suits you.