sample edition · harvested from the live gbg engine · 2026-07-08
brands, graded like public companies.
grade scale (shown in full — no hidden curve)
A ≥ +3.0 · B ≥ +2.0 · C ≥ +1.0 · D ≥ 0.0 · F < 0
portfolio composite → gray chip 1–10
method
composite = unweighted mean of each product's 14-metric engine vector, no skin profile applied ·
env-flag and caution rates read verbatim from engine product flags ·
greenwash gap is a sample-based heuristic, defined where shown
the index — every brand in the sample, ranked
sample report — 28 analyses · 1–3 analyzed products per brand · click any row to open its report · click a column to re-rank
the brand report
what a brand — or a journalist — would pay to see. every number verbatim from the engine.
where this goes
at engine scale, this becomes the industry's report card.
this sample graded a shelf's worth of brands from 28 real analyses. the engine already scores every ingredient it meets against a curated evidence corpus — so each new product analysis lands in a brand's portfolio automatically, and the report card writes itself. the ranking is the pr engine; the detailed report — per-product vectors, flags, reformulation guidance — is what brands pay for.
the precedent: yuka turned label-scanning into 70–85 million users — with no brand-level report card.1
the shelf it grades: india beauty & personal care ≈ usd 31–33b (2025)2; global skincare usd 155.8b (2025).3
1 apple app store + press, 2025–26 · 2 imarc & statista, 2025 · 3 grand view research, 2025 — full citations in the README. every other number on this page: live gbg engine.