# Atlas methodology

Atlas labels every panel with where its data came from and whether
it's real. "live" means data actually fetched from a real data partner;
"sample" means deterministic, generated placeholder data — legible, but not a
factual claim about any brand. Atlas never fabricates a review, a
rating, a certification, or a number dressed up to look real.

## Getting listed: the vetting standard

Before a brand is listed, its core identity — domain, name, category — is
checked, not assumed: does the domain resolve to a real, non-parked site that
matches the brand it claims to be. Temporarily unreachable brands are queued
for a human look rather than silently dropped. Listings are re-checked on a
90-day cycle.

## What's real today

280 brands have passed the listing check. 274 of them
additionally carry real, purchased enrichment data — brand kit, competitor
lists, product catalogs — fetched once from a paid data partner and
cached. Every other panel is currently sample data, clearly tagged, each with
a real flag-gated adapter switched on deliberately, one source at a time.

## What each panel is built from

Kinds of sources, not vendor names — the label that matters is whether the
data is real.

| Panel | Built from | Refresh | Status |
|---|---|---|---|
| Brand kit (logo, colors, fonts, socials) | a paid brand-enrichment data partner, fetched once and cached | on demand (first page view), then cached | live |
| Hiring | the brand's own public job board (no-auth public APIs) | on demand + nightly open-roles snapshot | live |
| Leadership (public-bio only) | an open, CC0 structured-knowledge base (CEO/chairperson claim only) for brand pages, real and flag-gated; the same namespace also backs the founder/leadership panel on technology vendor profile pages (/tool/[slug]), which stays on generated sample data only — no real fetch path is wired for tool-vendor leadership yet. | on demand, refreshed occasionally | live |
| News | a free public news index; clustering and ranking are ours | on demand + daily mention-volume snapshot | live |
| Similar brands | the same paid brand-enrichment data partner as the brand kit | on demand (first page view), then cached | live |
| Tech stack + change log | open-source fingerprints against each brand's own homepage | weekly (one fetch per brand per week) | live |
| What they sell | the same paid brand-enrichment data partner as the brand kit | on demand (first page view), then cached | live |
| Ads | an ad-library data partner covering the major ad platforms | on demand + nightly ad-count snapshot | sample |
| AI mentions | an AI-answer-visibility data partner (aggregate mention counts + sampled cited questions only) | on demand | sample |
| Brand video | the brand's own verified video channel, official embeds only | on demand, 30-day refresh rule | sample |
| Company facts (size, revenue band, HQ) | a company-data data partner keyed by domain | on demand, refreshed occasionally | sample |
| Creator video | public video search ranked by raw view count, official embeds only | on demand + periodic view-count snapshot | sample |
| Employer reviews | an employer-review data partner — aggregate rating/approval numbers only, never review text | on demand | sample |
| Founding / timeline | an open, CC0 structured-knowledge base (inception date only) | on demand | sample |
| Funding | a company-signals data partner (company endpoints only) | on demand | sample |
| Homepage history | the Internet Archive's free CDX index of prior homepage snapshot timestamps — a list lookup only, never a fetch of the archived page itself | on demand | sample |
| Ingredients | a public-domain government food database (food & beverage brands only) | on demand, category-scoped | sample |
| iOS app | the public app-store lookup API | on demand | sample |
| M&A / deals | official securities filings + news clusters, provenance-labeled | on demand, 90-day standalone window | sample |
| Marketplace presence | a marketplace-listings data partner | on demand | sample |
| Patents | the national patent office's official open-data API | on demand; shown only when a brand holds any | sample |
| Podcast sponsorships | a free public podcast index's episode show notes, text-matched against the brand's own domain or a declared promo code (no audio, no transcription) | on demand | sample |
| Product Hunt | a product-launch community data partner (comment bodies never read, only counts) | on demand | sample |
| Public-company results | official securities filings (public record) | on demand | sample |
| Ratings | schema.org aggregateRating structured data the brand already publishes on its own homepage | riding the weekly homepage fetch already made for the tech-stack panel — no separate fetch | sample |
| Search traffic (estimates) | an SEO-metrics data partner — estimates, clearly labeled | on demand | sample |
| Social channels | a social-metadata data partner over public profiles (captions never read) | on demand | sample |
| Third-party ratings | a business-review data partner covering Trustpilot/Google/Yelp — aggregate ratings only, never review text | on demand | sample |
| TikTok content themes | AI topic/hook tagging over public post metadata | async after posts are fetched | sample |
| TikTok labeling pipeline (internal batch state, never rendered) | internal bookkeeping for the async labeling pipeline — no external source | follows the posts fetch; internal state only | sample |
| TikTok posts (raw) | a public-post scraping data partner (public metadata only) | on demand | sample |
| Where to buy | the brand's own where-to-buy locator + an open places dataset | monthly locator check; map joined at render | sample |
| Who else runs this (tool pages) | a tech reverse-lookup data partner, else our own edges table | on tool-page view | sample |
| Certifications | public certification and values registries (organic, B Corp and similar; bulk sync pending) | — | sample (placeholder) |
| Supply chain | US customs bill-of-lading records (placeholder — no real fetch path wired) | on demand once wired | sample (placeholder) |
| Trademarks | no official name-search API exists yet — generated sample only | — | sample (placeholder) |
| Wikipedia presence | the same open, CC0 structured-knowledge base as founding/leadership (article existence + monthly pageview volume) — real adapter intentionally stubbed pending a live rate-limit probe, so this stays generated sample data for now | on demand, plus a monthly-pageview snapshot | sample (placeholder) |

## What's never collected

No personal contact data — emails, phone numbers, relationship graphs —
about anyone, including people who work at the brands indexed here. Founder
and leadership info, where shown, is public-bio level only.

See also: [about](https://consumer-brand-index.vercel.app/about.md) · [HTML version](https://consumer-brand-index.vercel.app/methodology)