What Local Business Schema Types AI Crawlers Actually Read
Schema markup is one of the most misunderstood levers in AI search visibility. Most businesses have schema markup on their site that AI systems partially ignore. The difference between schema that influences citations and schema that just decorates HTML comes down to which types, which properties, and which structure you use.
Not sure what schema your site actually has or whether AI can read it? The free Blind Spot Report includes a schema audit.
- Why Schema Markup Matters for AI Citations
- The 6 Schema Types That Actually Move the Needle
- FAQPage Schema: The Highest-Impact Type for AI
- LocalBusiness and Specific Business Type Schema
- Service Schema: Matching Specific Queries to Specific Pages
- Schema Types AI Mostly Ignores
- Common Schema Mistakes That Hurt AI Visibility
- Frequently Asked Questions
Why Schema Markup Matters for AI Citations
When ChatGPT or Perplexity crawls a business website to decide whether to cite it, the system has two options: interpret the unstructured prose on the page and infer what the business does, or read structured data the site owner explicitly provided. Structured data is always faster and more reliable. Schema markup is how you hand AI the answer directly instead of making it guess.
The reason schema markup is underutilized in AI search strategy is the same reason it was underutilized in traditional SEO for years: it requires technical implementation and has no visible payoff on the site itself. Business owners and marketers do not see schema markup when they visit their own site. AI systems see it at every crawl.
AI retrieval systems use schema markup for three functions: entity recognition (confirming what business this is, where it is, and what it does), query matching (determining which user questions this page is relevant to), and trust verification (confirming consistency between schema data and other signals like Google Business Profile and directory listings). Each function is distinct and requires different schema types and properties.
Schema markup is not a magic switch. It works by giving AI systems more accurate, more parseable information about your business. The quality of your schema data is what determines the citation impact. A LocalBusiness schema with all required properties correctly filled in does far more than a generic WebPage schema that simply repeats the page title.
For more on how AI systems decide which businesses to cite in the first place, see our guide on what an AI citation actually is and how it works.
Want to know which schema types are missing from your site? Call (213) 444-2229 for a quick schema diagnostic.
The 6 Schema Types That Actually Move the Needle
Schema.org defines over 800 types and thousands of properties. For local service businesses, six schema types are responsible for the vast majority of AI citation impact. The rest are relevant for other contexts (e-commerce, academic, entertainment) but do little for a local plumber, lawyer, or dentist trying to appear in AI recommendations.
| Schema Type | Primary AI Function | Citation Impact | Implementation Priority |
|---|---|---|---|
| FAQPage | Feeds Q&A data directly to AI retrieval | Very High | Priority 1 on all service pages |
| LocalBusiness | Entity identity and NAP verification | Very High | Priority 1 on homepage and contact page |
| Service | Service-to-query matching | High | Priority 2 on each dedicated service page |
| Review / AggregateRating | Social proof signal and trust ranking | High | Priority 2 on homepage and key pages |
| BreadcrumbList | Site structure for AI navigation | Moderate | Priority 3 on all pages |
| Article / BlogPosting | Content authority for informational queries | Moderate | Priority 3 on blog and resource pages |
The majority of local business schema is implemented by website builders and plugins that generate minimal, template-based markup. This produces a LocalBusiness schema with only name, address, and phone, none of the properties that give AI systems meaningful differentiation signals. Schema that tells AI only what is already obvious from the page title adds little citation value over having no schema at all.
Curious what schema your competitors are running? The Blind Spot Report includes a competitive schema comparison.
FAQPage Schema: The Highest-Impact Type for AI
FAQPage schema has consistently shown the strongest correlation with AI citation rates across service business categories. The reason is structural: AI language models are trained to produce answers to questions. FAQPage schema provides pre-structured question-and-answer pairs that retrieval systems can directly use, without the processing overhead of extracting that structure from unformatted prose.
When a user asks ChatGPT "how much does emergency plumbing cost in Phoenix," the model's retrieval system searches for pages that directly address that question. A plumbing company with FAQPage schema containing "What does emergency plumbing cost in Phoenix?" with a clear answer is a near-perfect retrieval match. A plumbing company whose website mentions emergency pricing in a paragraph buried under marketing copy requires the AI to do more work and is more likely to be skipped.
| Property | Why It Matters for AI | Common Mistake |
|---|---|---|
| Question name | Matched directly against user query language | Questions phrased as business talking to self, not as user searches |
| AcceptedAnswer text | Directly extracted as citation content | One-sentence answers with no detail or local specifics |
| Question count (5+) | More questions cover more query variants | 2-3 generic questions that don't match real search behavior |
| Answer specificity | Local details (city, price range, timeline) increase relevance | Generic national answers that apply to any business in any city |
| Valid JSON-LD format | Invalid syntax causes crawlers to skip the block entirely | Unclosed brackets, unescaped characters, nested @context errors |
The questions in FAQPage schema should be written as natural language queries, matching how a customer would actually phrase a question to an AI. "Do you offer same-day appointments?" is a business FAQ. "Can I get a same-day plumbing appointment in Scottsdale?" is an AI query. The second format drives citations. The first format does not.
The highest-performing schema implementations put FAQPage schema on individual service pages, not just the homepage. A plumbing company's "water heater replacement" page with FAQPage schema about water heater questions captures a completely different set of AI queries than homepage FAQPage schema about the company's general services. Each service page with its own FAQPage schema becomes an independent citation target.
LocalBusiness and Specific Business Type Schema
LocalBusiness schema is how AI systems verify that a website represents a real, specific business at a real, specific location. It is the entity identity layer for local search. But most implementations include only the three most obvious properties, name, address, and phone, and miss the properties that actually differentiate businesses for citation decisions.
- name (exact business name matching GBP)
- address with full PostalAddress object
- telephone with E.164 format
- openingHoursSpecification (structured hours, not a string)
- areaServed with ServiceArea or GeoCircle
- priceRange (guides AI on tier queries like "affordable" or "premium")
- hasMap (Google Maps URL)
- sameAs (GBP URL, Yelp URL, BBB URL, LinkedIn)
- aggregateRating with ratingValue, reviewCount
- knowsAbout (specific topics/services as text array)
- logo (AI cannot process images)
- image without alt text
- description that just repeats the business name
- openingHours as a generic text string (not structured)
- contactPoint with only generic "customer service" type
- address as a single text string (not PostalAddress object)
- founder without additional entity context
- paymentAccepted with just "cash, credit" (too generic)
The specific business type matters significantly. Schema.org has dedicated types for dozens of local service categories: Plumber, Electrician, GeneralContractor, LegalService, FinancialService, MedicalBusiness, DentalClinic, and more. Using the specific type sends a more direct signal to AI about what category of queries this business should appear in.
For context on how AI systems use these business category signals as part of broader citation decisions, see our guide on how local businesses build citation authority for AI search.
Service Schema: Matching Specific Queries to Specific Pages
Service schema is the structured data layer that connects individual service pages to the specific queries those pages should answer. While LocalBusiness schema establishes what your business is, Service schema on a dedicated service page tells AI which specific service is offered on that page, what the service area is, what it costs, and how to book it.
The practical impact is highest for businesses with multiple distinct services. A roofing company that offers roof repair, roof replacement, gutter installation, and skylights benefits from Service schema on each page because it creates distinct, specific citation targets for different query types. A user asking "roof repair [city]" triggers different AI retrieval than a user asking "new roof installation cost [city]." Service schema ensures each page is the right citation target for its specific query type.
Most Service schema implementations include only name and description. The properties that drive the highest citation impact are the ones most commonly omitted: areaServed (the geographic scope of the service), provider (linked to the LocalBusiness entity), serviceType (the specific category), and offers (including price range when known). These properties transform generic service markup into specific, query-matchable citation data.
Not sure if your service pages have effective schema? Get the free Blind Spot Report for a full schema gap breakdown.
Schema Types AI Mostly Ignores for Local Business Citations
Not all schema markup is equal. Several schema types are commonly added to local business websites through plugins, themes, or misguided SEO recommendations, and have minimal impact on AI citation behavior for the local service context.
| Schema Type | Commonly Used For | AI Citation Impact | Why It Underperforms |
|---|---|---|---|
| WebPage / WebSite | Generic page and site identification | Minimal | Adds no differentiation. Every website is a website. |
| ImageObject | Photo galleries, team photos | Minimal | AI cannot process image content. Alt text matters more. |
| SiteLinksSearchBox | Search box in SERP | None for AI | Google-specific feature. Irrelevant to AI citation systems. |
| VideoObject | YouTube embeds, testimonials | Minimal | AI cannot watch videos. Transcript content matters, not the video schema. |
| SocialMediaPosting | Social feed integrations | None | Not relevant to local business citation decisions. |
| Event | Promotions, webinars, workshops | Low unless query-specific | Rarely matches local service intent queries. |
The practical lesson: removing or ignoring low-value schema types is less important than ensuring your high-value types (FAQPage, LocalBusiness, Service) are correctly implemented and filled with substantive, specific data. Adding more schema types does not compensate for weak core schema properties.
Common Schema Mistakes That Hurt AI Visibility
Schema errors do not display as visible errors to site visitors. They exist silently in the page source, causing AI crawlers to skip structured data or, worse, misread it. The following mistakes appear on a significant portion of local business sites and directly reduce citation rates.
When the phone number, business name, or address in LocalBusiness schema differs from what appears on Google Business Profile, AI entity resolution treats these as potentially different businesses. The conflict reduces confidence in both sources. The most common version: schema uses "ABC Plumbing LLC" while GBP uses "ABC Plumbing" or "ABC Plumbing Co." Exact consistency across all sources is required for AI to confidently identify and cite your business.
Valid JSON-LD syntax with no unclosed brackets or unescaped characters. Specific business type, not generic LocalBusiness when a specific type applies. FAQPage schema on every service page with 5+ questions phrased as user queries, not business FAQs. LocalBusiness schema with openingHoursSpecification, areaServed, sameAs, and aggregateRating. Service schema on service pages with provider linked to LocalBusiness entity. NAP identical across schema, GBP, Yelp, BBB, and other directories. These are the markers of schema that moves AI citation rates.
Schema validity errors are the most common and most overlooked issue. A missing closing bracket in a JSON-LD block causes the entire block to be ignored by most parsers. Most website owners have never checked whether their schema is valid because schema errors are not visible in the browser.
For more on how the broader local business content strategy supports schema signals, see our guide on how much content local businesses need to build consistent AI citations.
Want a complete schema audit alongside your AI visibility check? Email support@theanswerengine.ai or call (213) 444-2229.
Find Out What Schema Gaps Are Costing You AI Citations
The free Blind Spot Report includes a schema audit: which types are present, which properties are missing, whether your JSON-LD is valid, and how your schema compares to businesses currently appearing in AI responses for your category and location.
Get Your Free Blind Spot ReportFrequently Asked Questions
Does schema markup actually help you get cited by ChatGPT?
Yes, but not all schema types equally. FAQPage schema has the strongest documented influence on AI citation behavior because it directly feeds structured Q&A data to AI retrieval systems. LocalBusiness schema helps with entity recognition and NAP verification. Generic schema that simply restates page title and description has minimal citation impact.
What is the most important schema type for local service businesses?
For local service businesses, FAQPage schema combined with LocalBusiness schema is the most impactful combination. FAQPage schema feeds AI systems the exact Q&A format they use when generating answers. LocalBusiness schema establishes entity identity with address, phone, service area, hours, and business category.
Does Perplexity read schema markup differently than ChatGPT?
Perplexity places particularly high value on FAQPage and HowTo schema because its retrieval system is optimized to surface pages that directly answer specific questions. ChatGPT uses schema as one of several trust signals. Both benefit from clean, valid JSON-LD schema, but Perplexity shows the strongest response to FAQPage schema on service pages.
Should I use @type: LocalBusiness or a more specific type like Plumber or Restaurant?
Use the most specific schema type available for your category. If Schema.org has a dedicated type (Plumber, Restaurant, MedicalBusiness, LegalService, FinancialService), use it. AI systems use schema type specificity to match businesses to categorical queries more directly than LocalBusiness allows.
How many FAQ items should I include in FAQPage schema for AI citations?
Include a minimum of 5 questions, and typically 8 to 12 performs better. More important than count is specificity: FAQ items should address actual queries people submit about your service category with local details like city name, cost ranges, and timing information.
Can invalid or duplicate schema markup hurt my AI search visibility?
Yes. Invalid JSON-LD causes AI crawlers to skip structured data entirely. Duplicate schema with conflicting NAP data actively confuses entity resolution, which can cause AI to undercount your presence or assign citations to the wrong entity. Schema validity tools can identify these issues.
Find Out If AI Can Read Your Schema
The free Blind Spot Report checks your schema validity, identifies missing properties, and compares your structured data to businesses currently appearing in AI answers for your category and market. Takes 60 seconds to request.
Get My Free Blind Spot ReportAEO for Real Estate Agents
The playbook we built on our own real estate site — 1.14M+ monthly impressions, cited by 4 of 4 AI platforms. Now available to one agent per market.
See the real estate playbook →Free // 30-Minute AuditRun a Free Answer Engine Audit
See how ChatGPT, Claude, Perplexity, and Google AI answer questions about your category today — and where your name should be.
Book the audit →