Our methodology
Clear comparisons begin with knowing exactly what is being compared.
1. Record the source
We document manufacturer label facts, their source URL, and the date consulted. A source check is not independent laboratory verification or clinical review.
2. Preserve exact ingredient names
Selected label ingredients are compared by exact name, ignoring letter case. We do not merge synonyms, infer concentrations from ingredient order, or treat incomplete lists as complete formulas.
3. Normalize the price
Cost per mL equals recorded USD price divided by volume in mL. One US fluid ounce equals 29.5735295625 mL. Tax, shipping, promotions and personal discounts are excluded. Observations are dated snapshots, not live offers. Price observations older than 30 days need rechecking before a purchase decision.
4. Make uncertainty visible
Missing prices, undisclosed concentrations, incomplete labels and absent head-to-head outcomes are explicitly labeled. Lower cost does not establish greater effectiveness. Editorial comparisons remain draft until a human editor has reviewed the specific analysis.
5. Publish deliberately
Source-backed factual label profiles may be published. Scientific or medical claims require qualified human review and primary research. Drafts, interactive comparisons and internal search are excluded from indexing. No automatic product ratings are used.
6. Match each claim to the right evidence
Our numbered-reference pilot covers two label guides and the NIOD/The Ordinary copper comparison. References identify label facts, supplier information and human ingredient research separately. Editorial interpretation and illustrative arithmetic are labeled. A study on an ingredient is not automatically a study on a retail product. A reference number is not a quality score.
7. Evaluate relevance and study limits
For academic sources, record authors, year, journal, DOI, study design, population, duration, reported outcome, relevance to the exact formula, funding and declared conflicts where available. State whether we inspected only an abstract or the full paper. Do not infer missing concentrations, safety rates or results. Check publisher corrections/retractions; disclose the scope of that check. Peer review alone does not establish sufficient evidence, independence or reproducibility. This pilot is not a systematic review or a clinical recommendation, and qualified scientific review is pending.
8. Keep commercial interests and corrections visible
Manufacturer and supplier pages have a commercial interest. Funding and conflict statements from papers are author declarations, not independently audited facts. No affiliate partnerships are active; future compensation must be disclosed and must not alter conclusions. Reference-check dates, price observations and actual human review dates are separate. No author, credentials or reviewer approval is invented. To suggest a correction, use Contact and identify the page, disputed statement, market/variant and primary source; its availability is shown there.