By SearchBrand.ai · 2025-10-15 · ⏱️ 7–10 min

Introduction

LLMs don't “read pages”; they resolve entities and relationships. If your website is a maze without signposts, you'll be left out of generative answers and lose attribution. This playbook guides you to create an internal entity graph (coherent, linked, and measurable) that Google and AI engines can understand—in line with AEO/GEO and without fluff.

For: SEO/Content teams that need to organize entities and relationships to gain citations and attribution within 90 days.

TL;DR

  • Canonicalize entities with @id/sameAs and reuse them with JSON‑LD.
  • Express relationships with hasPart/isPartOf, about/mentions, brand, author, mainEntityOfPage.
  • Control AI bots in robots.txt and monitor logs/CDN.
  • llms.txt is optional (non-standard).
  • Measure SGV, Attribution, Entity Resolution, and Edge Coverage.

Quick Answer

A good entity graph = Entity Registry + consistent JSON‑LD + semantic linking + controls for AI crawlers + AEO/GEO metrics.

Step-by-step Development

1) Entity Registry (YAML)

Entity Registry (YAML) — educational demo

💡 What it is: A base document with entities (Organization, Person, Product) and their @id/sameAs. 💡 Why it matters: Allows reusing nodes in JSON‑LD and maintaining link/attribution consistency. 📍 Where it goes: Internal repository / CMS; it's poured into markup and internal links.

How to verify?
  • Consistency of @id and sameAs
  • Reusability in JSON‑LD and semantic links

2) Minimal JSON‑LD (Organization + Person)

JSON‑LD (educational demo)

💡 What it is: Structured markup to disambiguate entities and relationships. 💡 Why it matters: Helps search engines/agents understand and cite your content. 📍 Where it goes: In the <head> as <script type="application/ld+json"> (actual implementation).

How to verify?
  • Rich Results Test with your URL
  • Consistency with visible content

3) Hubs and hierarchy

Model thematic hubs (AEO/GEO/Entity Graph) and link to guides, FAQs, and HowTo content with hasPart/isPartOf, about/mentions.

4) AI bot control (robots.txt)

robots.txt by bot (educational demo)

💡 What it is: Rules for crawlers (GPTBot, Google‑Extended, PerplexityBot, etc.). 💡 Why it matters: Controls AI discovery/use and protects sensitive areas. 📍 Where it goes: Actual file in /robots.txt (site root).

How to verify?
  • Open /robots.txt (200 OK)
  • Confirm user‑agents and cache in logs/CDN

5) llms.txt (optional)

llms.txt (curated)

💡 What it is: Human‑readable guide with canonical URLs and hubs. 💡 Why it matters: Can help agents locate/cite key pieces. 📍 Where it goes: Actual file in /llms.txt (site root).

How to verify?
  • Open /llms.txt (200)
  • Review accessible canonical URLs

6) AEO/GEO Metrics

KPIDefinitionHow to measure
SGV% of queries where you appear cited/linkedPrompt sampling; citation/link checklist
Attribution Rate% of mentions with a linkCount of responses with vs. without a link
Entity Resolution Rate% of prompts with correct identificationTest suite + verification
Edge Coverage% of expected relationships presentJSON‑LD + interlinking audit
Crawler Compliance% of AI bots that respect robotsCDN/WAF logs by user‑agent

FAQ

  • Do I need llms.txt right away? No. It's optional and emerging; prioritize schema and interlinking. Add it when you have the capacity.
  • If I block Google‑Extended, will my SEO drop? It shouldn't: it manages Gemini training/grounding, not Google Search indexing.
  • Is robots.txt enough to stop AI? It helps but isn't sufficient: combine it with WAF/CDN and continuous monitoring.
  • How do I start with a small team? Homepage with Organization/@id/sameAs; 3 marked entities; 1 thematic hub; AI robots; dashboard with SGV/Attribution.

See also

Sources 2024–2025

Updated: 2025-10-17

Haz que la IA recomiende tu marca y se la respuesta