A percentage is a measurement. It is not a potency score.
Beauty likes numbers because they look objective. Ten percent looks more serious than five. Twenty looks very committed. None of that tells you whether the material is the same, the vehicle is comparable, the formula is stable or the extra concentration buys anything except more irritation.
Percentage of what?
“10% niacinamide” and “10% brightening complex” are not equivalent disclosures. One may describe the named ingredient itself. The other may describe a proprietary blend whose individual components remain undisclosed.
The same problem appears with botanical extracts, peptide complexes, retinoid systems and encapsulated materials. Before comparing two numbers, make sure the denominator did not change.
Different molecules do not share one strength ladder
0.1% tretinoin, 0.1% retinaldehyde and 0.1% retinol are different molecules with different evidence, delivery and regulatory contexts. Ten percent L-ascorbic acid is not automatically comparable with 10% of a vitamin C derivative.
The number becomes useful only after the material itself is comparable.
Even the same molecule can punish number-chasing
A classic 48-week double-blind trial compared 0.1% tretinoin, 0.025% tretinoin and vehicle in 99 people with photoaged skin. Both tretinoin strengths improved photoaging compared with vehicle. The study did not find a significant clinical-efficacy difference between the two tretinoin strengths, while irritation was significantly greater with 0.1%.
That is not a universal rule that lower is always equal. It is a very good reason to stop assuming higher automatically means better.
Pain is not a concentration assay.
The vehicle changes the practical dose
A cream, gel, serum, suspension and anhydrous formula can handle the same active differently. Solubility, pH, stability, release, spreadability, evaporation and skin tolerance can all change what the label percentage means in use.
A 20% formula is not necessarily the 10% formula with twice as much useful material poured in.
Exposure is concentration multiplied by behavior
Skin experiences amount per application, frequency, surface area, contact time and adherence. A stronger leave-on product used once a week because it is miserable may deliver less useful exposure than a lower concentration used consistently.
This is why tolerability belongs inside the efficacy discussion instead of being treated like an unfortunate personality issue.
The ingredient list cannot secretly fill in the missing numbers
FDA generally requires cosmetic ingredients above 1% to appear in descending order of predominance. Ingredients at 1% or less may be listed without regard to predominance, and color additives have additional flexibility.
Ingredient order gives broad formula clues. It does not let you calculate that the third-from-last ingredient is 0.6% because someone on social media drew an imaginary line through the list.
Sunscreen is the useful counterexample
U.S. OTC sunscreen labels identify active sunscreen ingredients and their strengths. Even there, the active percentages do not mathematically produce the SPF. Finished-product performance comes from standardized testing.
That is the larger lesson: a percentage can describe part of the formulation. The outcome still needs outcome evidence.
A faster way to compare two products
| Question | What matters |
|---|---|
| Same material? | Pure active, derivative, extract, blend or proprietary complex. |
| Same vehicle? | Cream, gel, serum, suspension, anhydrous base or delivery system. |
| Same use? | Amount, frequency, leave-on vs rinse-off and duration. |
| Same evidence? | Did the study test this concentration and a comparable formulation? |
| Same burden? | Irritation, reduced frequency, extra recovery products or abandonment. |
Let the number identify itself
Numbers are excellent when they identify the right variable. They are excellent marketing when they merely look precise.
Make the percentage identify its material. Make the formula explain the delivery. Make the study match the concentration. Then see whether people could tolerate the schedule long enough for the result to matter. The largest number on the bottle does not win by default.
Checked against FDA guidance and peer-reviewed dermatology literature. Reviewed August 2026.
Related: Ingredient-List Literacy · How to Read a Beauty Claim Before You Trust It.
Source file
What this article is standing on
- FDA: Summary of Cosmetics Labeling RequirementsIngredients run in order of predominance, anything at one percent or less may be listed in any order, and concentration is never declared. Which leaves a percentage on the front of a bottle unverifiable from the label the law requires.
- FTC: Health Products Compliance GuidanceSubstantiation scales with the specificity of the claim. A numeric claim needs evidence about that number, not about the ingredient in general.
- FTC: Policy Statement Regarding Advertising SubstantiationThe evidence has to be in the advertiser's hands before the claim runs.
- FDA: "Cosmeceutical" Label on CosmeticsNo regulator reviews a cosmetic's percentage claim before sale. That makes the number a marketing decision as much as a formulation one.