Building Your Knowledge Graph

What is a Knowledge Graph?#

A knowledge graph is a structured representation of entities (people, products, concepts) and the relationships between them. When you provide a knowledge-graph.json file, you are giving AI systems a map of your domain expertise. Instead of crawling every page and trying to infer what your site knows about, AI can read your knowledge graph and immediately understand your authority areas, the key entities you discuss, and how they relate to each other. This is particularly valuable for sites that cover complex topics where the relationships between concepts matter as much as the concepts themselves.

Creating knowledge-graph.json#

Your knowledge graph should list your key entities grouped by type, with relationships connecting them. Keep it focused on your areas of genuine expertise.

{
  "version": "2.1",
  "entities": [
    {
      "id": "ai-visibility",
      "type": "concept",
      "name": "AI Visibility",
      "description": "How well AI systems can discover, interpret, and cite a website",
      "url": "/learn/adp-overview",
      "related": ["adp", "llms-txt", "schema-org"]
    },
    {
      "id": "adp",
      "type": "protocol",
      "name": "AI Discovery Protocol",
      "description": "20-endpoint specification for AI discoverability",
      "url": "/learn/adp-overview",
      "related": ["ai-visibility", "llms-txt"]
    },
    {
      "id": "mcp",
      "type": "protocol",
      "name": "Model Context Protocol",
      "description": "Standard for AI agents to discover and use tools",
      "url": "/learn/mcp-manifest",
      "related": ["agentic-commerce", "a2a"]
    }
  ],
  "expertise_areas": [
    "AI discovery protocols",
    "Domain intelligence",
    "Citation readiness optimization"
  ]
}

Knowledge Graph Best Practices#

Quality over quantity is the key principle. A focused knowledge graph with 10-30 well-described entities is far more useful to AI systems than a sprawling graph with hundreds of superficial entries. Each entity should have a clear type, description, and URL. Relationships should be meaningful and bidirectional. Update your knowledge graph when you publish significant new content or when your expertise areas evolve.

• Focus on 10-30 key entities that represent your core expertise.
• Use consistent entity types: concept, product, protocol, person, organization.
• Write descriptions as factual one-sentence summaries.
• Link entities to their most relevant page on your site.
• Keep relationships meaningful — every connection should add information.