Cosmetic Efficacy Testing: Matching Claims to Studies and Reports

Cosmetic efficacy testing is easiest to buy when the claim comes first. A brand that asks a laboratory for "an efficacy test" usually receives a study that measures something real but does not quite support the sentence on the pack. A brand that arrives with the exact wording, the product version and the market can commission a study whose report actually closes the question.

This guide is about that commissioning decision: turning a proposed claim into an endpoint, briefing the study and reading what the report does and does not show. It does not compare laboratories or accreditation; our cosmetic testing laboratory guide covers that. Safety evidence is a separate file, explained in our cosmetic safety testing guide, and country-by-country claim rules are compared in our claims compliance guide for the US, EU and China.

Write the claim before you choose the study

Start with the full sentence you intend to use, not the benefit category. "Hydrating" and "visibly more hydrated skin after one application" lead to different studies. So do "reduces the look of fine lines" and "reduces wrinkle depth by a measured amount". The second version in each pair names something that can be measured, a time point and an implied comparison.

Then list everything the claim implies. A duration, a percentage, a user group ("for sensitive skin"), a comparison with an untreated area or another product, and any reference to the type of test ("clinically tested", "dermatologist tested") each add a condition the evidence must meet.

The EU claims technical document, a non-binding best-practice tool agreed by the Commission's working group on cosmetic products in 2017, illustrates the point with duration: a "48-hour hydration" claim should not be made if the evidence only supports a shorter period. It also says the level of evidence should be consistent with the type of claim being made. EU technical document on cosmetic claims (2017).

In the United States, the FTC expects advertisers to have a reasonable basis for objective claims before they are disseminated. When an advertisement refers to a level of support, such as "tests prove" or "studies show", the FTC expects at least that level of substantiation. FTC policy statement on advertising substantiation. Other markets set their own expectations, so have the target-market reviewer check the final wording.

Turn each claim into an endpoint

An endpoint is the specific result the study will record. Agreeing it with the laboratory before the study starts is the single decision that most affects whether the report will be useful.

On a phone, swipe the table sideways to see every column.

The table is a starting point for the conversation with a laboratory, not a protocol. The laboratory should select and justify the method, the number of participants, the time points and the statistics for the actual product and claim. One sample size or study length does not fit every claim.

Instrumental, expert-graded and consumer evidence answer different questions

The EU technical document groups claim support into experimental studies, consumer perception tests and published information, used alone or in combination. Experimental studies include in silico, in vitro and ex vivo work, instrumental and biochemical methods, studies on volunteers, investigator evaluations and sensory evaluations.

Each route has a different job. An instrument records a defined physical parameter. A trained assessor grades what is visible under controlled conditions. A consumer perception test records what users noticed and felt. None is universally stronger; each is stronger for the claim it was designed to answer.

That is why mixing routes needs care. A consumer questionnaire can support "felt smoother", but the technical document states that consumer self-perception studies are not appropriate support for a "dermatologically tested" claim. Equally, an instrumental change may be statistically clear yet too small for users to notice. If the pack copy combines a measured claim and a perception claim, plan evidence for both.

For consumer perception tests, the wording of the questionnaire is critical. The technical document asks for clear questions, a balanced answer scale that does not steer the participant, and a claim that follows directly from the answers to the relevant question. Ask to see the questionnaire before the study begins, not after.

Ingredient data is not a finished-product result

Supplier data packs are useful for choosing candidates. They usually describe an ingredient at a stated level, in a model formula or in vitro. Your product has a different base, level, pack and routine.

The EU technical document says ingredient claims should not imply that the finished product has the same properties when it does not, and that a claim extrapolating ingredient properties to the finished product needs adequate and verifiable evidence, such as showing the ingredient is present at an effective concentration. It also notes that results from in vitro or in silico studies should not be presented as if they were in vivo results.

In practice, decide at the brief stage which claims rest on the ingredient and which on the finished product. "Contains" claims need the ingredient to be deliberately present. Benefit claims for the product usually need evidence on the product, or a justified argument that the ingredient evidence applies to it.

Brief the study around the product you will sell

A useful commissioning brief lets the laboratory design a study whose report links to the product on the market. Include:

• The exact claim wording for each market, and which words carry the measurable promise
• The formula version and batch to be tested, and confirmation that it matches the intended commercial formula
• The pack, if application or dosing depends on it
• Directions for use that match the label: amount, frequency, rinse-off or leave-on, contact time
• The target users, including any group named in the claim, with inclusion and exclusion criteria to be set by the laboratory
• The comparison you need: baseline, untreated area, placebo base or a named reference product
• The time points the claim implies, including any duration
• How results will be reported, and who owns the report and raw data

Studies on volunteers should follow ethical principles, and the technical document says products tested on people should already have been assessed as safe. Build the safety step into the timeline rather than discovering it when the laboratory asks.

Independent clinical testing providers commonly offer this kind of commissioned service from protocol design to reporting. Intertek, for example, describes protocol development, ethics submissions, trial management, data management, statistical analysis and report writing for cosmetics and personal care products. Intertek cosmetics clinical testing. That is an example of the commissioning model, not a recommendation or a partnership.

Cosmetic efficacy study brief template

Use the template below when you ask laboratories to quote, and send the same completed version to each of them. It keeps the brand's decisions apart from the study design, so proposals can be compared on method rather than on how each laboratory interpreted an incomplete request.

The middle column holds decisions only the brand can make: what will be claimed, on which product, for whom and in which market. A laboratory should not have to guess them. The third column is where a qualified laboratory proposes the method and explains why it fits. Leave it blank when you send the brief. A reply that fills it with a standard package, without saying how that package answers your claim, has not yet answered the question.

On a phone, swipe the table sideways to see every column.

Three habits make the template more useful. Complete it before the formula is frozen, so a claim that cannot be supported can still change the brief instead of the label. Ask every laboratory to answer against the same rows, which turns quotations into a like-for-like comparison. Then keep the approved version with the final report, because it records what the study was designed to show.

The template deliberately sets no sample size, study length or method. Those depend on the claim, the endpoint and the variability the laboratory expects, and a good proposal explains them. If a laboratory cannot say how its design answers one of the rows, ask before accepting the quotation.

Read the report against the claim

When the report arrives, check it against the claim sentence, line by line. The technical document lists what a report should include: clear identification of the product so it can be linked to the product on the market, the study objective, the test schedule and protocol, results and their interpretation, statistics, and the signature of the person in charge.

On a phone, swipe the table sideways to see every column.

The technical document also warns that data processing and presentation should not overstep the limits of the test or overstate the effect. If a chart in the report looks more dramatic than the numbers beneath it, use the numbers.

Where the report and the planned wording do not match, change the wording rather than stretching the report. A narrower accurate claim is easier to defend than a broad claim with partial support.

Keep the evidence valid after changes

Evidence belongs to a specific product. The EU technical document says the responsible person should make sure the evidential support still applies when the formulation changes. The same logic applies to changes in directions, pack or dispenser if they affect how much product reaches the skin or hair.

Agree in advance what happens after a change. A fragrance swap, a new grade of a key ingredient or a different pump may need a documented justification, a bridging check or a repeat study, depending on the claim. Record the decision with the change, so the claim file and the product stay in step.

Treat the claim file as part of the launch documents. Keep the claim wording, the study reports, the questionnaire, the formula and pack versions tested, and the change log together, with an owner.

Worked example: one product, two claims

The following is an illustrative scenario, not a client project or test result.

A brand plans a leave-on facial lotion with two claims: "boosts skin hydration for 24 hours" and "skin feels softer from the first use". The brief separates them. The first is a measured claim with a duration, so it needs an instrumental study on the tested lotion with measurements that cover the full period. The second is a perception claim, so a consumer test with a clear question about softness after first use fits it.

The instrumental report shows a clear increase at the early time points, but not at the final one. Instead of keeping "24 hours", the brand changes the wording to match the last time point the evidence supports. The consumer test supports the softness claim as worded. Both reports are filed with the formula version tested, and the brand agrees that a planned fragrance change will be reviewed before the claims are reused.

How we fit into the testing plan

We are a manufacturer, not a testing laboratory. On ODM projects, claim substantiation, in vitro work, clinical studies and accredited third-party testing can be scoped through suitable laboratories when the formula, claims or destination market require them. For CDMO development, we can define a substantiation plan once the formula, claims and markets are confirmed, and coordinate the agreed testing through qualified laboratories. Methods, samples, timelines and report ownership are confirmed per project.

What we can do on our side is supply the tested product as a defined formula version and batch, keep the formula and pack records the report needs to link to, and flag formula or packaging changes that may affect existing claims.

Frequently Asked Questions

What is cosmetic efficacy testing?

It is testing that shows whether a cosmetic product delivers a specific benefit claimed for it, under defined conditions. The right method depends on the claim: instrumental measurement, expert grading, consumer perception or a combination.

Can one efficacy test support all of our claims?

Rarely. Each claim implies its own endpoint, users, time points and comparison. One study can sometimes cover several claims if it is designed for them from the start, so list all claims before commissioning.

Can we use our ingredient supplier's studies as product claims?

Ingredient data can support choosing an ingredient and "contains" wording. A benefit claim for the finished product usually needs evidence that applies to the finished product.

Does efficacy testing replace the safety assessment?

No. Efficacy and safety answer different questions. Products tested on volunteers should already have been assessed as safe, and the safety assessment remains a separate requirement.

Plan the evidence before the samples are final

Send us the product, the target markets and the exact claims you are considering, and say which ones are essential. We can review which claims are realistic for the formula, what evidence route each implies, and where testing would be scoped through a suitable laboratory, before the formula and pack are frozen. When you are ready, send your claims plan for review.