How to read skincare evidence

Follow the claim to the experiment

Research-driven is useful only when the research is clear. A long bibliography can establish that an ingredient attracts scientific interest without showing that a particular serum produces a particular result. Start with the claim on the page, then ask what experiment would be needed to support it.

For example, a claim about a visible change in human skin calls for evidence closer to real use than a biochemical measurement in cultured cells. The 1988 GHK-Cu fibroblast study is valuable ingredient research; it is not a clinical trial of the serum sold on this website. ^[1]

Three levels to keep separate

Ingredient research examines a substance under defined conditions. Formulation research examines that substance within a particular combination of ingredients. Finished-product testing concerns the actual preparation offered to customers. The same ingredient name can appear at all three levels without making the evidence equivalent.

This distinction applies to safety as well as performance. FDA describes several kinds of information manufacturers may use to substantiate cosmetic safety, including ingredient data and additional testing when needed. Safety evidence and a study of visible benefits answer different questions; one should not be presented as the other. ^[2]

Ask what happened to whom

When human testing is cited, look for the number and characteristics of participants, what they used, the comparison group, the duration, and the outcomes. Also look for who left the study, adverse effects, and who funded it. The CONSORT 2025 reporting guidance emphasizes transparent accounts of randomized trials so readers can assess what was done. ^[3]

Randomization, a suitable control, and blinding can make an experiment more informative. Even then, check the size and uncertainty of the reported effect. A statistically significant difference need not be large enough to notice, and a positive average does not promise the same experience for every participant. ^[4]

Read the result, not just the headline

A reported increase in a cell marker, a measured change in skin appearance, and a participant saying that a product feels pleasant are different outcomes. Each may have a place, but the wording should preserve the distinction. An attractive before-and-after photograph is most informative when the capture conditions and assessment method are explained.

The comparison also matters. Robinson and colleagues’ pal-KTTKS study used the same moisturizer with and without the peptide, making the comparison more specific than simply comparing a new routine with no routine at all. That study concerns that preparation—not all peptides and not Lumiere Routine’s serum. ^[5]

Understand the language around cosmetics

FDA explains that intended use helps determine whether a product is a cosmetic, a drug, or both. Claims about treating disease or changing body structure and function differ from ordinary cosmetic appearance claims. The term cosmeceutical is not a separate legal category. A scientific-sounding label is therefore not a substitute for the evidence behind a claim. ^[6]

Leave room for an honest unknown

Ask a brand which exact formula was studied and whether you can read the methods and results. If those details are unavailable, treat the claim as unverified rather than filling in the gaps yourself. That does not prove the product has no value; it defines what you can reasonably conclude.

Keep a simple question in mind as you read: does this evidence test the thing the claim asks me to believe? A good explanation makes the connection visible and leaves an uncertainty open when it cannot. That habit is useful across skincare, whether you are exploring copper peptides, comparing moisturizers, or deciding that your current routine already serves you well.