How an analysis is done

Every analysis on this site follows the same procedure. It is published so that you can tell where a conclusion came from, and so that two analyses are comparable.

1. Capture the ingredient list

The INCI list is transcribed from the manufacturer’s own product page for the market the product is sold in. The URL and the date of access are recorded and published with the analysis.

Retailer listings are not used as a primary source. They are frequently stale, and they are not the manufacturer’s statement.

2. Record concentrations only where stated

A concentration is recorded when the manufacturer publishes it — on the label, on the product page, or in technical documentation.

Where no concentration is published, the entry reads not published. It is not estimated from position in the list, not inferred from a competitor product, and not carried across from an earlier formulation.

Order in an INCI list constrains concentration loosely above 1% and not at all below it. Treating list position as a concentration estimate is the most common error in ingredient commentary, and it is avoided here deliberately.

3. Assign a function to every ingredient

Each ingredient is classified by what it is doing in this formula — active, solvent, humectant, emulsifier, preservative, chelator, pH adjuster, thickener.

The same molecule can play different roles in different products, so the classification is per-formula rather than looked up once.

4. Assess each claim separately

Marketing claims are quoted as the brand words them, then assessed one at a time against the ingredient list and the published evidence, using four gradings:

• Well supported — repeatable published evidence at a comparable concentration
• Plausible — a sound mechanism, with evidence that is thinner or at a different dose
• Weakly supported — evidence exists but is small, contested, or supplier-funded
• Not verifiable from the ingredient list — the buyer has no way to check it

The fourth grading is a statement about checkability, not about honesty.

5. State what could not be established

Every analysis ends with what the method could not reach. Typically: finished product pH, actual concentration of undisclosed actives, stability over shelf life, supplier grade of a branded raw material, and anything requiring laboratory measurement.

What this method deliberately does not attempt

Laboratory analysis. Nothing here is chromatographed, assayed or measured. Every conclusion derives from published composition and published evidence.

Clinical outcomes. No trials are run, no subjects recruited, no before-and-after photography commissioned. An analysis tells you what a formula plausibly does, not what it did for anyone.

Ranking. Products are not scored against each other. Scoring formulations with different purposes on one scale produces a number that looks precise and means very little.