# Shelfgrade

> Self-serve AI shopping visibility audits. Shelfgrade checks a store's product
> feed (OpenAI-native and Google-compatible parser modes), Product/Offer
> JSON-LD, and crawler access (OAI-SearchBot, PerplexityBot, Googlebot) against
> the vendors' published specifications, then ships corrected files: a fixed
> feed, a patched robots.txt, paste-ready JSON-LD, and a white-label report.
> Weekly citation monitoring measures whether stores are cited on ChatGPT,
> Perplexity and Gemini via the vendors' official APIs.

Key facts our audits are built on (each verified against primary docs and
re-checked weekly):
- OAI-SearchBot gates ChatGPT search/shopping answers; GPTBot is training-only.
- Google-Extended does not affect Google Search, Shopping, or AI Overviews.
- gtin is optional in OpenAI's native feed schema; gtin-or-mpn is required in
  Google-compatible mode unless identifier_exists is "no".
- OpenAI publishes no JSON-LD requirement; markup matters via Google's
  merchant-listing rules and Perplexity's crawl-fed index.

## Pages
- [Home + free scanner](https://shelfgrade.ai/): instant readiness scan by domain
- [Full audit](https://shelfgrade.ai/audit): feed upload, findings, corrected artefacts
- [The AEO Quickstart](https://shelfgrade.ai/guide): the full framework — how AI answers are assembled, which crawlers gate visibility, markup/feed requirements per surface, measurement
- [Research](https://shelfgrade.ai/research): primary-source explainers and the weekly Spec Watch (vendor spec-change log with verbatim before/after quotes)
- [Sample report](https://shelfgrade.ai/sample-report.html): what the deliverable looks like