How grading works

The scale

Every product starts at 100 points. Points come off for each problem found on the label, once per type of problem rather than once per ingredient. The final score maps to a letter grade on the same scale in every review:

Each review's own page spells out what each grade typically means for that product category.

Evidence versus preference

Every deduction is tagged by why it exists: evidence means published research of harm or irritation; preference means a clean-label choice without strong evidence either way; bonus points come from a real third-party seal. A product's "Evidence only" grade drops every preference-based deduction, so you can see how much of a grade rests on evidence and how much rests on preference — one tap away on every product and review page.

Capping

A score can never show above 100 or below 0 on the page, even when the raw total (adding every deduction and bonus) would go higher or lower. When that happens, the breakdown still shows the raw total next to the capped score, so the numbers always add up.

How a top pick is chosen

A review's top pick is the product with the best grade that also passes the review's main gate (its central red-flag or safety check). Ties go to the product that's easiest to find, then to the higher raw score, then to the lower price per unit, and finally to name order.

A note on stores

Some listed places aren't stores you can pick, such as Brand site, Drugstores, Everywhere, Import sites, Salons. They describe where a product is sold rather than somewhere to search, so they're never offered as a choice in the store picker.