Structured Data for AI Content Discovery: How to Make Any Content Type Visible to LLMs

> TL;DR
> - Every content type has a Schema.org schema — most sites use fewer than 3 of the 700+ available types
> - AI agents discover content by type: Event, Course, Recipe, SoftwareApplication, Service, Article all get indexed and surfaced differently
> - The right schema type for your content determines which AI queries you're eligible for
> - Audit your content schema now →

Updated: April 21, 2026

---

The Schema Vocabulary Is Much Larger Than You Think

Most people know about Product, Article, and LocalBusiness schema. But Schema.org defines over 700 types — and each one maps to a different class of AI agent query.

When an AI agent receives a query about events, it searches its Event index. When it's looking for courses, it queries Course records. When it needs a recipe, it retrieves Recipe schema. Each type has different fields, different ranking signals, and different query patterns it's eligible for.

Using the wrong type — or no type at all — puts you in the wrong index or no index. You could have the world's best content and be completely invisible to the agents querying for it.

This guide maps the most important content types to their schema, the key fields for AI discovery, and the query patterns they unlock.

---

The Content Type × Schema Map

Events → `Event`

If you run conferences, workshops, webinars, or meetups, Event schema is how AI assistants surface you when someone asks "what AI conferences are happening in April?"

{
  "@context": "https://schema.org",
  "@type": "Event",
  "name": "Agent-to-Agent Commerce Summit 2026",
  "description": "Two-day conference on AI agent infrastructure, MCP protocol adoption, and the future of autonomous commerce. Speakers from Anthropic, OpenAI, Google, and leading AI startups.",
  "startDate": "2026-09-15T09:00:00-05:00",
  "endDate": "2026-09-16T18:00:00-05:00",
  "eventStatus": "https://schema.org/EventScheduled",
  "eventAttendanceMode": "https://schema.org/MixedEventAttendanceMode",
  "location": {
    "@type": "Place",
    "name": "Javits Center",
    "address": {
      "@type": "PostalAddress",
      "streetAddress": "429 11th Ave",
      "addressLocality": "New York",
      "addressRegion": "NY",
      "postalCode": "10001"
    }
  },
  "offers": {
    "@type": "Offer",
    "price": "799",
    "priceCurrency": "USD",
    "availability": "https://schema.org/InStock",
    "validFrom": "2026-01-01",
    "url": "https://agentsummit.com/tickets"
  },
  "organizer": {
    "@type": "Organization",
    "name": "WebMCP Guide",
    "url": "https://webmcpguide.com"
  },
  "performer": [
    { "@type": "Person", "name": "Dario Amodei" },
    { "@type": "Person", "name": "Sam Altman" }
  ],
  "audience": {
    "@type": "Audience",
    "audienceType": "AI developers, enterprise architects, e-commerce operators, SaaS founders"
  }
}

Key fields for AI agent event discovery:

startDate — Critical. Agents filter on date range.

eventAttendanceMode — Distinguishes in-person, online, and hybrid

offers — Price and ticket availability

audience — Who the event is for

---

Courses → `Course` + `CourseInstance`

E-learning, workshops, bootcamps. When someone asks "best AI development course under $200," Course schema is how you're discovered.

{
  "@context": "https://schema.org",
  "@type": "Course",
  "name": "MCP Server Development: Zero to Production",
  "description": "Build production-ready Model Context Protocol servers in TypeScript. Covers tools, resources, prompts, authentication, deployment, and monitoring. For developers with Node.js experience.",
  "provider": {
    "@type": "Organization",
    "name": "WebMCP Guide",
    "url": "https://webmcpguide.com"
  },
  "hasCourseInstance": {
    "@type": "CourseInstance",
    "courseMode": "online",
    "duration": "PT8H",
    "startDate": "2026-05-01",
    "offers": {
      "@type": "Offer",
      "price": "197",
      "priceCurrency": "USD"
    }
  },
  "audience": {
    "@type": "Audience",
    "audienceType": "JavaScript developers, Node.js developers, AI engineers"
  },
  "teaches": ["MCP server development", "TypeScript", "AI agent integration", "Production deployment"],
  "coursePrerequisites": "Basic JavaScript/Node.js experience"
}

---

Software Tools → `SoftwareApplication`

Covered in depth in SoftwareApplication Schema: Complete Guide. The key fields: applicationCategory, featureList, offers with priceSpecification.unitText, and additionalProperty for free trial and integrations.

---

Recipes → `Recipe`

If you run a food blog or any culinary content, Recipe schema unlocks AI assistant cooking queries — "give me a 30-minute pasta recipe" — with direct citation of your content.

{
  "@context": "https://schema.org",
  "@type": "Recipe",
  "name": "30-Minute One-Pan Pasta Primavera",
  "cookTime": "PT20M",
  "prepTime": "PT10M",
  "totalTime": "PT30M",
  "recipeYield": "4 servings",
  "recipeCategory": "Dinner",
  "recipeCuisine": "Italian",
  "recipeIngredient": ["400g pasta", "2 zucchini", "1 cup cherry tomatoes", "3 cloves garlic"],
  "recipeInstructions": [
    { "@type": "HowToStep", "text": "Boil pasta in salted water until al dente." },
    { "@type": "HowToStep", "text": "Sauté garlic in olive oil, add zucchini and tomatoes." }
  ],
  "nutrition": {
    "@type": "NutritionInformation",
    "calories": "420 calories",
    "proteinContent": "14g"
  },
  "aggregateRating": {
    "@type": "AggregateRating",
    "ratingValue": "4.8",
    "reviewCount": "312"
  }
}

---

Job Listings → `JobPosting`

Recruitment sites, company careers pages, and job boards. AI assistants and job-search agents look for JobPosting schema to answer queries like "remote Python developer jobs at AI companies."

{
  "@context": "https://schema.org",
  "@type": "JobPosting",
  "title": "Senior MCP Server Engineer",
  "description": "Build and maintain MCP servers that expose AI agent capabilities to enterprise clients.",
  "hiringOrganization": {
    "@type": "Organization",
    "name": "WebMCP Guide",
    "sameAs": "https://webmcpguide.com"
  },
  "jobLocation": {
    "@type": "Place",
    "address": {
      "@type": "PostalAddress",
      "addressLocality": "New York",
      "addressRegion": "NY"
    }
  },
  "jobLocationType": "TELECOMMUTE",
  "employmentType": "FULL_TIME",
  "baseSalary": {
    "@type": "MonetaryAmount",
    "currency": "USD",
    "value": {
      "@type": "QuantitativeValue",
      "minValue": 150000,
      "maxValue": 200000,
      "unitText": "YEAR"
    }
  },
  "datePosted": "2026-04-21",
  "validThrough": "2026-06-30",
  "skills": "TypeScript, Node.js, MCP protocol, REST APIs, AI agent development"
}

---

Products → `Product`

Covered in depth in JSON-LD Product Schema: Complete Guide. Essential fields: offers with return policy and shipping, aggregateRating, brand as entity, additionalProperty for AI enrichment.

---

Local Services → `LocalBusiness` + `Service`

Covered in LocalBusiness Schema for AI Assistants and Service Schema for AI Booking Agents.

---

The Schema Type Selection Framework

Choosing the wrong schema type is a silent failure — your content looks fine to humans but ends up in the wrong AI index.

Decision tree:

Is it something you can buy or order? → Product or Offer
Is it a software or app? → SoftwareApplication
Is it a place-based service? → LocalBusiness + Service
Is it a scheduled event? → Event
Is it a structured learning experience? → Course
Is it authored content (article, guide, tutorial)? → Article or TechArticle
Is it a step-by-step process? → HowTo
Is it a collection of questions? → FAQPage
Is it audio content? → PodcastEpisode or AudioObject
Is it video content? → VideoObject
Is it a job listing? → JobPosting
Is it a food recipe? → Recipe

Multiple types apply? Use an array: "@type": ["LocalBusiness", "Attorney"]

---

The Headless Schema Engine Approach

For teams managing diverse content types across a headless CMS or multi-site setup, the traditional approach — plugin per schema type, manual configuration per page — doesn't scale.

The headless approach:
1. Schema injection API — Call the API with a URL, get back the correct schema type and fields automatically detected
2. Type detection — The engine reads og:type, URL patterns, page metadata, and content structure to determine the correct schema type
3. Field generation — Fields are populated from your existing page metadata — no duplicate data entry
4. Injection — Schema is injected client-side or returned as a string for server-side embedding

SchemaInject is built on this architecture. One integration covers every content type across your entire domain.

→ Run a free audit on any URL to see detected schema type →
→ Schema injection API for developers →

---

Related articles:

SoftwareApplication Schema: Complete Guide

Article Schema for LLM Citation Optimization

Why Perplexity Cites Some Sites and Ignores Others

LocalBusiness Schema for AI Assistants

Free AI Visibility Audit →