# Schema Markup for AI: How Structured Data Gets Your Business Cited by ChatGPT and Gemini
When ChatGPT recommends a business, it doesn't guess. It pulls from indexed web data — and the businesses that give AI platforms the cleanest, most structured information win the citation. That structured information is schema markup.
Schema markup has been a Google SEO tool for years. Most businesses either ignore it or let their developer add a basic Organization tag and forget about it. But in 2026, schema markup serves a second audience: large language models. ChatGPT, Gemini, Perplexity, and Claude don't read your website the way a human does. They parse structured data. And if your website doesn't have it, you're making it harder for AI platforms to understand — and recommend — your business.
Here's how schema markup for AI works, what types you need, and how to implement it today.
How AI Platforms Use Your Structured Data
Large language models are trained on massive datasets of web content. When that content includes schema markup — specifically JSON-LD structured data — the model gets machine-readable facts about your business instead of having to interpret free-form text.
Think about it from the AI's perspective. If someone asks "Who's the best emergency plumber in San Diego?", the model needs to find businesses that match three criteria: the service (emergency plumbing), the location (San Diego), and some measure of quality. Your website might mention all of this across five different pages in paragraph form. Schema markup packages it into a single, unambiguous block of structured data.
Here's a simplified example of what a LocalBusiness schema tells an AI model:
{
"@context": "https://schema.org",
"@type": "Plumber",
"name": "Pacific Plumbing San Diego",
"address": {
"@type": "PostalAddress",
"streetAddress": "1234 Market St",
"addressLocality": "San Diego",
"addressRegion": "CA",
"postalCode": "92101"
},
"telephone": "(619) 555-0123",
"url": "https://pacificplumbing-sd.com",
"priceRange": "$$",
"aggregateRating": {
"@type": "AggregateRating",
"ratingValue": "4.8",
"reviewCount": "247"
},
"openingHoursSpecification": {
"@type": "OpeningHoursSpecification",
"dayOfWeek": ["Monday","Tuesday","Wednesday","Thursday","Friday","Saturday","Sunday"],
"opens": "00:00",
"closes": "23:59"
},
"areaServed": ["San Diego", "La Jolla", "Pacific Beach", "Hillcrest"]
}That block tells an AI model everything it needs: what the business does, where it is, its rating, its hours, and what areas it covers. No ambiguity. No interpretation required.
Compare that to a website that mentions "We serve San Diego" in a footer, has phone numbers scattered across different pages, and never explicitly states its service type in a machine-readable format. The AI model can figure it out — but it's less confident, and less likely to cite that business over one with clean structured data.
The Schema Types That Matter for AI Citations
Not all schema types are equal for AI visibility. Here are the ones that directly influence whether AI platforms cite your business.
LocalBusiness (or industry-specific subtypes)
This is the foundation. Every local business website needs a LocalBusiness schema — or better, a specific subtype like Dentist, Plumber, Restaurant, LegalService, or RealEstateAgent. The subtype gives AI models a precise category match rather than a generic "this is a business."
Must include: name, address, telephone, url, areaServed, openingHours, priceRange, aggregateRating (if you have Google reviews).
Service Schema
If you offer multiple services, each one should have its own Service schema. A dentist doesn't just do "dentistry." They do cleanings, crowns, implants, emergency dental work, and cosmetic procedures. Each of these is a distinct service someone might ask an AI about.
{
"@context": "https://schema.org",
"@type": "Service",
"name": "Emergency Dental Care",
"provider": {
"@type": "Dentist",
"name": "Hillcrest Family Dental"
},
"areaServed": "San Diego",
"description": "Same-day emergency dental appointments for toothaches, broken teeth, and dental trauma."
}When someone asks ChatGPT "Where can I get emergency dental care in Hillcrest?", the model is looking for exactly this kind of structured match: service type + location + provider.
FAQPage Schema
FAQ schema does two things. For Google, it can trigger rich results — the expandable Q&A boxes in search results. For AI models, it provides direct question-answer pairs that are easy to parse and cite.
If your FAQ section answers "How much does teeth whitening cost in San Diego?", and that's wrapped in FAQPage schema, an AI model can directly reference your answer when someone asks the same question. Without the schema, the model has to extract the answer from your paragraph text — and might choose a competitor's cleaner data instead.
Article Schema
For blog posts and content pages, Article schema tells AI models this is editorial content with a publication date, author, and topic. This helps AI models assess recency and authority when deciding what to cite.
BreadcrumbList Schema
BreadcrumbList schema maps your site structure for both search engines and AI crawlers. It shows the hierarchy: Home > Services > Emergency Plumbing. This helps AI models understand the relationship between your pages and find the most relevant one for a given query.
The AI Citation Gap: Schema vs. No Schema
ClawSignal tracks AI citations across 9 platforms — ChatGPT, Gemini, Claude, Perplexity, Grok, Microsoft Copilot, Meta AI, Apple Intelligence, and SearchGPT. When we analyzed which San Diego businesses get cited most frequently, a clear pattern emerged.
Businesses with comprehensive schema markup — LocalBusiness + Service + FAQ across their key pages — were cited by AI platforms 4x more often than businesses with no schema or only basic Organization markup. The correlation held across every industry we tracked.
This makes intuitive sense. AI models prioritize information they can parse with high confidence. Schema markup is the highest-confidence data format on the web. It's literally designed to be machine-readable.
The businesses with no schema aren't invisible to AI. But they're at a disadvantage. When an AI model has to choose between a business with clean structured data and one without, the structured data wins.
How to Implement Schema Markup for AI Visibility
Here's a practical implementation plan, starting from the highest-impact actions.
Step 1: Audit your current schema
Use Google's Rich Results Test (search.google.com/test/rich-results) to check what schema your website currently has. Most small business websites have either nothing or a basic Organization schema that their website builder added automatically.
ClawSignal's audit tool checks schema markup as part of its site analysis — including whether your schema types match what AI platforms need, not just what Google requires.
Step 2: Add LocalBusiness schema to your homepage
This is the single highest-impact change. Use a specific subtype if one exists for your industry. Include every field: name, address, phone, URL, hours, area served, aggregate rating, and price range. Place it in a script tag in your page's head.
Step 3: Add Service schema to each service page
Every distinct service you offer should have its own Service schema on its corresponding page. Don't dump all services into one schema block. One service per page, one schema per service.
Step 4: Add FAQPage schema to pages with FAQ sections
If you have FAQ sections on your service pages or blog posts — and you should — wrap them in FAQPage schema. Each question-answer pair becomes a structured data point that AI models can directly reference.
Step 5: Add Article schema to blog posts
Every blog post should have Article schema with headline, datePublished, dateModified, author, and publisher fields. This helps AI models assess whether your content is current and authoritative.
Step 6: Add BreadcrumbList schema to all pages
Map your site hierarchy with BreadcrumbList on every page. This is low effort and high reward — it helps both Google and AI models understand your site structure.
Step 7: Validate and monitor
After implementing, validate every page with Google's Rich Results Test. Then monitor your AI citations to see if visibility improves. ClawSignal's ai-scan feature tracks your mentions across AI platforms over time, so you can correlate schema implementation with citation growth.
Common Schema Mistakes That Hurt AI Visibility
Using Organization instead of LocalBusiness. Organization schema doesn't include location data, hours, or service area. For a local business, this is like introducing yourself without saying where you are.
Missing areaServed. If your schema doesn't specify what geographic areas you cover, AI models have to infer it from your content. Be explicit.
Stale aggregateRating data. If your schema says you have 50 reviews but Google shows 200, the discrepancy confuses crawlers. Keep your rating data current or pull it dynamically.
Duplicate schema across pages. Every page should have its own schema relevant to that page's content. Don't copy-paste the same LocalBusiness block on every page without adding page-specific Service or FAQ schema.
Not including FAQ schema on blog posts. Blog posts with FAQ sections are perfect candidates for FAQPage schema. Most businesses add FAQs to their posts but forget the schema. That's leaving AI citation potential on the table.
Schema Markup Is Your AI Foundation
Traditional SEO is about convincing Google's algorithm you're relevant. AI visibility is about giving language models the structured data they need to confidently cite you. Schema markup is the bridge between the two.
The businesses that implement comprehensive schema markup today are building the foundation for AI visibility that will pay off for years. AI search is growing. The platforms are multiplying. The one constant is that structured data makes it easier for every platform — current and future — to understand and recommend your business.
If you want to see where your schema stands and how AI platforms currently perceive your business, get your free schema audit at clawsignal.co. We'll show you exactly what's there, what's missing, and what to fix first.
FAQ
What is schema markup for AI? Schema markup is structured data code (JSON-LD format) added to your website that helps both search engines and AI platforms understand your business details — services, location, hours, reviews, and more — in a machine-readable format. AI platforms use this structured data when deciding which businesses to cite in their responses.
Does schema markup help you show up in ChatGPT? Yes. ChatGPT draws from indexed web data, and schema markup provides the cleanest, most structured version of your business information. Businesses with comprehensive schema markup are cited more frequently by ChatGPT and other AI platforms than those without it.
What schema types does a local business need? At minimum: LocalBusiness (or an industry-specific subtype), Service for each service you offer, FAQPage for any FAQ sections, Article for blog posts, and BreadcrumbList for site navigation. Each type gives AI models a different dimension of understanding about your business.
How do I track if AI platforms are citing my business? AI citation monitoring tools like ClawSignal scan multiple AI platforms — ChatGPT, Gemini, Claude, Perplexity, and others — to track when and how your business is mentioned. This lets you measure the impact of SEO and schema improvements on your AI visibility over time.