3.3 Schema.org and structured data

Structured data — through schema.org — explicitly tells an AI system what your content means. Not just what it says, but what it is: a question, a product, an organisation, a procedure, a service.

Why schema.org matters for GEO

AI systems have to interpret what they read. Structured data removes ambiguity. A page with FAQPage markup is explicitly a question-and-answer structure. A page with Recipe markup is a recipe with ingredients and preparation time. A page with Service markup describes a service and its characteristics. That explicit classification makes citation more reliable and faster — and above all: less dependent on the model’s NLP capabilities.

The universal foundation

Four schema types are relevant to almost every organisation, regardless of sector. Start here.

FAQPage

For pages with question-and-answer structures. One of the most powerful types for AI visibility because the question-answer pairs are explicitly marked as direct answer sources. Almost every organisation has substantive questions that customers ask — from a law firm to a retailer to a software company.

HowTo

For step-by-step procedures. Whether it concerns “How do I apply for a building permit?”, “How do I install a water pipe?”, “How do I register my business?” or “How do I book an appointment?” — structured as HowTo schema, it increases the chance that AI systems will cite your explanation when users ask process-related questions.

Organization and Person

For the basic definition of your organisation and of individual authors or experts. Directly linked to E-E-A-T: who writes about what, with what expertise. The sameAs property linking to LinkedIn, Wikidata or official registers is critical here — that is how AI systems connect your entity to external knowledge.

Article and NewsArticle

For editorial content. Makes author, publication date and topic explicit. The standard for blog and knowledge articles. NewsArticle for news-style content with an editorial context.

Sector-specific schemas

Beyond these, dozens of sector-specific schema types exist. Which one you use depends entirely on what you do. A few examples from different domains, not intended as exhaustive.

For product sales and e-commerce

Product, Offer, AggregateRating and Review are the foundation. For specific product categories, refinements exist such as Vehicle, FoodEstablishment or SoftwareApplication.

For financial services

FinancialProduct, BankAccount, LoanOrCredit, MortgageLoan, InvestmentOrDeposit, InsuranceAgency. Marks specific product properties such as interest rates, terms and conditions explicitly.

For healthcare

MedicalCondition, MedicalProcedure, Drug, Physician, Hospital. Particularly relevant for content about specific conditions, treatments or medical practices. AI systems apply stricter reliability criteria here — schema markup helps signal your expertise explicitly.

For legal and professional services

LegalService, Attorney, AccountingService, ProfessionalService. For law firms, accountancy practices and consultancy. Combine with Service for specific offerings.

For local and physical services

LocalBusiness with its many subtypes: Restaurant, Store, AutoRepair, Dentist, RealEstateAgent, TravelAgency. Essential for organisations with a physical location or local customer base.

For education and knowledge institutions

Course, EducationalOrganization, LearningResource. Relevant for educational programmes, courses, libraries and knowledge platforms — including publications like Groundbase itself.

The full schema.org vocabulary contains hundreds of types. The trick is not to apply everything, but to choose the right types for what your content actually is. One well-implemented, sector-appropriate schema is worth more than five generic ones.

Implementation priorities

Not every page needs every schema. Start with your most valuable pages: home page, service or product pages, FAQ pages, about-us and author pages. Then expand based on content type. Always validate with Google’s Rich Results Test before going live — an error in JSON-LD means the schema simply isn’t read, with no visible error message.

Maintenance

Structured data is not a one-off task. When products, prices, opening hours or terms change, the markup must follow. Inconsistency between visible content and markup undermines your reliability — for AI as well as for Google. A product that still shows €99 in markup but €129 on the page is a credibility issue.

Related in the hub

→ Want to understand how brand voice and authority signals work in GEO? Read 3.4.