Most ingredient apps grade safety. We grade efficacy — whether a product's key active is present at a clinically-effective dose, measured against concentration thresholds published in peer-reviewed dermatology and cosmetic-science literature. Not how it feels. Not how it markets. This page is the whole rubric, with nothing hidden.
One five-point scale, applied identically to every product. The letter answers a single question: is the active dosed to work?
| Grade | What it means |
|---|---|
| A | At or above the effective dose, with support. One or more key actives meet or exceed their published effective concentration in a sensible formula. Strong study evidence on the product, when on file, lifts a high B into this band. |
| B | Effective dose, solid formula. The key active clears its threshold and the formula supports it, but without independent or product-specific clinical evidence on file. |
| C | A recognized active at a plausible level. A known active is present at a concentration that is reasonable but at or near the lower bound of what the literature supports. |
| D | Present but underdosed. The active is in the formula below its effective threshold, or without the conditions (pH, stabilizers, vehicle) the research relied on. |
| F | Below threshold or unverifiable. The active is well under an effective dose, used in conditions the evidence doesn't support, or the formulation can't be verified at all. |
A grade is only as good as what it's built from. Here is exactly what the engine reads.
We parse the product's INCI list and match each entry to a known active in our database — retinoids, vitamin C (L-ascorbic acid and derivatives), niacinamide, the AHAs/BHAs, azelaic and tranexamic acid, peptides, and more as the database grows.
The percentage the brand discloses — on the label, the spec sheet, or in writing. Where it isn't disclosed, INCI ordering and "below 1%" cutoffs bound an estimate, and if the dose still can't be pinned down, we withhold rather than assume.
The effective concentration for each active, drawn from peer-reviewed dermatology and cosmetic-science literature — the level at which studies report a measurable benefit, with the conditions (pH, vehicle, frequency) those studies used.
The threshold is only as trustworthy as the research behind it. We tier that research and let it move the grade — up, but never past an unmet dose.
The threshold is anchored by randomized, controlled, or independently replicated trials — and, at the product level, third-party or peer-reviewed testing on this exact formula. The strongest tier; it is what can push a product to an A.
Consistent peer-reviewed studies or a brand-run clinical study on the product. Real signal, but not independent or fully controlled — useful, weighted accordingly.
A sensible formula at an effective dose, but no study on this specific product. The grade rests on dosing and well-established thresholds.
Claims with no resolvable study behind them. They earn no bonus; the product is graded on its dosing alone, at the formulation tier.
The two tiers above formulation raise a product's score, so a brand can't earn them by simply asserting a study exists. The engine only grants the bonus when a resolvable citation — a study or lab report — is on file. The rubric and the active thresholds on this page are public, so anyone can re-run a product's ingredient list and arrive at the same letter. The scoring is deterministic: the same inputs always produce the same grade, and every point traces back to a specific active and threshold. A grade you can reproduce is the only kind worth trusting.
Illustrative cases, using widely-cited thresholds, to show how a dose becomes a letter. Numbers below are for explanation, not a specific product's published grade.
L-ascorbic acid disclosed at 15%, formulated at the low pH the research relies on, with a peer-reviewed study on the product on file.
Result: A. The active clears its threshold under the right conditions, and strong, on-file evidence lifts it into the top band.
Retinol named on the front, but disclosed at 0.01% — an order of magnitude below where studies report visible benefit, with no product evidence.
Result: D. The active is present, so it isn't an F — but it sits far below an effective dose with nothing to support it, so it can't earn a passing grade.
If that same retinol cream disclosed no concentration at all, the product would not get a D — it would get no grade, because we couldn't confirm the dose. See the rules below.
These are what keep the grade honest. They override convenience, and they override the brand.
Roughly 75% of products don't disclose how much active they contain. When the dose can't be confirmed, we don't give the benefit of the doubt and we don't invent a number — we hold the grade until the brand discloses or independent data fills the gap. A missing grade is a more honest answer than a flattering one.
The grade is computed from the open rubric on this page, not an opinion we can be paid to change. Pass and fail cost the same, and the engine never sees who is paying. A brand can license a Verified audit or seal (disclosed on its page), but that process cannot move a single letter.
Some product links are affiliate links — that's disclosed, and it's how the work is funded. The grading engine runs before and entirely separately from any commercial relationship. An affiliate product and a non-affiliate product are scored by the identical rubric, with no thumb on the scale.
A credible grade names its own limits. Here are ours, plainly.
The grade is computed from disclosed concentrations against published thresholds. We do not assay the finished product or independently verify that the bottle contains what the label says. It scores the formula on paper, not the molecules in the jar.
Unless a study on that exact product is on file, the grade is a prediction of likely efficacy from dose and evidence — not proof the product was tested on skin and worked. We say "likely effective," never "proven."
Two products at the same dose can differ in how stable the active stays, how well it penetrates, and how your skin tolerates it. A high grade is a strong starting point, not a guarantee for every person or every formula.
Visage Theory is cosmetic and educational only. Nothing here diagnoses, treats, or guarantees a result. Individual results vary, grades are not evaluated by the FDA, and you should consult a clinician for medical skin concerns.
We start with the most-studied hero actives and expand. A product can be re-graded as thresholds are refined, the literature updates, or a brand improves its disclosures.
Thresholds are drawn from the published record, not invented. We cite the body of evidence rather than dress the page in fabricated reference numbers.
Controlled and clinical studies establishing the concentrations at which actives — retinoids, vitamin C, niacinamide, the AHAs/BHAs, azelaic and tranexamic acid, and others — show measurable effects, including the formulation conditions they depend on.
Established reference frameworks such as FDA OTC drug monographs and the EU CosIng database for recognized actives and their accepted use levels.
Each product's published INCI list and any disclosed active percentages, supplemented by open product databases for in-store barcode scans. Where a product has its own study, the resolvable citation is attached to its grade.
Each graded product surfaces the specific active, the threshold it was measured against, and the evidence tier behind it — so the "why" travels with every letter.
The letter comes from the open rubric on this page — the named active, the disclosed dose, the published threshold, and the evidence tier. Nothing else moves it. When we can't verify a dose, we withhold the grade rather than guess in a brand's favor. That's the whole method.