GEO checklist for an industrial supplier
20 concrete checks so a manufacturer or distributor gets cited by ChatGPT, Claude, Gemini and Perplexity.
This checklist condenses what — in our audits — separates a supplier page cited by AIs from one that is ignored. Ordered from highest to lowest impact.
AI crawler access
✓ llms.txt at the site root listing GPTBot, ClaudeBot, PerplexityBot, Google-Extended.
✓ robots.txt does not block these user-agents.
✓ XML sitemap declared, current, with lastmod set.
Structured data
✓ JSON-LD Organization with legalName, NAICS, address, sameAs (LinkedIn, registry).
✓ JSON-LD Service or Product per main offering, with areaServed.
✓ JSON-LD FAQPage for recurring buyer questions.
✓ JSON-LD BreadcrumbList on every internal page.
Factual content
✓ Quantified capabilities (tonnage, MOQ, tolerances, lead times) in the HTML, not just inside a PDF image.
✓ Certifications listed with number and body (ISO 9001:2015, IATF 16949, AS9100D, NADCAP…).
✓ Explicit served geography (provinces, states).
✓ Dated and sourced numbers ("2025", "source: Statistics Canada").
Editorial architecture
✓ One pillar page per region × sector pair.
✓ Glossary of trade and compliance terms.
✓ FAQ visible in the HTML (not click-revealed only).
✓ Multilingual pages (FR / EN) with hreflang.
External consistency
✓ Name, address, business number, certifications identical across website, LinkedIn and registries.
✓ Listed in at least 2 recognized vertical directories.
✓ At least one mention in a verifiable third-party source (press, university, government agency).
Key takeaways
- → Crawler access + JSON-LD = entry ticket. Without them, the rest does not matter.
- → Dated numbers and cross-source consistency tip comparative queries.
FAQ
How long before citations appear?
RAG engines (Perplexity, ChatGPT Search) can cite a page within days. "Memory" models (answers without live search) integrate new data at every retraining, typically 3-6 months.