1000 verified brands committed to quality, value, and transparency.



























































| Signal | What it indicates | Limitations | Weight it when |
|---|---|---|---|
| Average composite | Overall catalogue consistency | Hidden by small catalogues | The brand has several scored products |
| Pillar spread | Where the brand reliably wins or loses | Category-dependent | You care about one pillar in particular |
| Company certifications | Governance and process maturity | Not product-specific | Buying across the catalogue |
| Recall history | How failures are handled | Absence may mean low scrutiny | The product category carries safety risk |
| Controversy record | Gap between marketing and practice | Coverage varies by brand size | The brand markets heavily on ethics |
Every published, scored product from the brand is pooled, so one strong item cannot mask a weak range.
Only company-level certifications resolvable to a registry appear at brand level; product-specific marks stay on product pages.
Recalls and controversies require a primary source such as a regulator filing or brand statement, and keep their original dates.
Resolved issues remain on the page rather than being deleted, so a brand's trajectory stays readable over time.
An average from two products is a data point, not a trend. Check how many scored items sit behind the number.
No recall history can mean strong quality control or simply less regulatory attention in that category.
Most large manufacturers eventually issue a recall. How quickly and openly they closed the loop is the more informative signal.
Two brands with the same average can be opposites: one wins on durability, the other on price and support.
It is the mean composite score of every published product for that brand, so it moves when products are added, rescored or delisted.
No. Resolved entries are marked resolved and dated but stay on the record, because the pattern over time is the useful information.
They can submit corrections with a verifiable public source. Verified corrections are published within seven days; nothing can be removed by request alone.