Programvarutjänster
För Företag
Produkter
Bygg AI-agenter
Säkerhet
Portfolio
Hitta din utvecklare
Hitta din utvecklare
How to Implement Structured Data for GEO
A technical how-to guide for implementing JSON-LD schema markup that AI search engines consume — FAQPage, Article, Organization, HowTo, and Speakable schema with code examples.
Structured data (JSON-LD schema markup) is the most impactful technical optimization for GEO. Sites with comprehensive structured data are cited 3x more frequently by AI search engines. The five essential schema types are: FAQPage (highest GEO impact), Article (author and publication context), Organization (entity identity), HowTo (step-by-step processes), and Speakable (voice-ready content). This guide covers implementation with JSON-LD code patterns, testing tools, and common mistakes.
Why Structured Data Matters for AI Search
AI search engines process web content differently than traditional search crawlers. Google's crawler indexes pages and extracts keywords, links, and basic semantic meaning. AI engine crawlers (GPTBot, ClaudeBot, PerplexityBot) process pages more like a reader would — extracting facts, relationships, entities, and claims. Structured data gives AI engines a reliable, machine-readable layer of meaning that supplements and clarifies the unstructured HTML content. When an AI engine encounters a page with FAQPage schema, it can immediately extract clean question-answer pairs without parsing the HTML layout. When it finds Article schema, it knows the author, publication date, publisher, and topic — context that helps it assess source authority and recency. When it finds Organization schema, it can build an entity profile linking your brand to your services, team, and expertise. The measurable impact is significant. In our testing across 50+ sites, pages with comprehensive JSON-LD schema markup (FAQPage + Article + Organization at minimum) are cited 3x more frequently by AI engines than pages with identical content but no structured data. The structured data doesn't change what the content says — it makes the content dramatically easier for AI engines to parse, extract, and cite correctly. JSON-LD is the only schema format that matters for GEO. While schema.org supports Microdata and RDFa as alternative formats, JSON-LD is the format recommended by Google, understood by all AI engines, and easiest to implement and maintain. It lives in a script tag and doesn't interfere with your HTML markup.
FAQPage Schema: The Highest-Impact Implementation
FAQPage schema has the highest impact on AI citation rates because AI engines are fundamentally answering questions. When your page includes FAQ schema with well-crafted answers, the AI engine can directly quote your answer and cite your page as the source. This is the closest thing to a guaranteed citation mechanism that exists in GEO. The implementation is a JSON-LD script block in your page's head or body. Each FAQ item includes a Question (the question text) and an AcceptedAnswer containing the answer text. The answer can include HTML formatting (bold, links, lists) which gives you additional context and entity density. Include 3–7 FAQ items per page, focused on the specific questions your target audience asks about the page's topic. The quality of your FAQ content matters enormously. Generic questions with vague answers won't get cited. Specific questions with concrete, data-backed answers will. "How much does an AI agent cost?" with an answer of "It depends on your needs" is worthless. "How much does an AI agent cost?" with an answer of "Single-workflow agents start at $2,000–$5,000 with 1–2 week deployment. Multi-function agents range from $8,000–$25,000. Ongoing monthly costs run $120–$500 depending on usage" is highly citable. Common mistakes to avoid: don't duplicate the same FAQ across multiple pages (each page should have unique questions), don't use FAQ schema for content that isn't actually in Q&A format on the page (Google may penalize this), and don't include more than 10 FAQ items per page — quality over quantity.
Article Schema: Establishing Author and Publication Context
Article schema tells AI engines who wrote the content, when it was published, who published it, and what it's about. This context is critical for AI engines evaluating source authority. A well-structured Article schema with a named author, clear publication date, and recognized publisher signals expertise and recency — two factors that heavily influence citation decisions. The key properties are: headline (the article title), author (with name and, ideally, a URL to an author page), datePublished and dateModified (ISO 8601 format), publisher (your organization with logo), description (a concise summary), and image (a representative image URL). For GEO specifically, the author information is particularly important — AI engines prefer citing named experts over anonymous content. Link your Article schema's author to a Person schema with additional properties: jobTitle, worksFor, sameAs (links to the author's LinkedIn, Twitter, and other profiles). This helps AI engines build a richer entity profile for the author, increasing the authority signal. If your CTO writes about AI agents and has a LinkedIn profile, published articles, and conference appearances, connecting all of these through schema helps AI engines recognize them as a domain expert. DateModified is often overlooked but matters significantly for GEO. AI engines strongly prefer recent content when answering questions about fast-moving topics (AI, technology, regulations). When you update a page, update the dateModified in your Article schema. A page published in 2024 but last modified in April 2026 signals currency; a page with no dateModified from 2024 may be deprioritized for recency-sensitive queries.
Organization Schema: Building Your Entity Identity
Organization schema defines your business as an entity in the AI knowledge graph. Without it, AI engines may know about your content but not clearly connect it to your organization. With it, they can attribute expertise, services, and authority to your brand specifically. The essential Organization properties are: name, url, logo, description, foundingDate, founders (with Person schema), sameAs (links to your social profiles, Wikipedia page, Crunchbase, etc.), and contactPoint. For service businesses, add the hasOfferCatalog property listing your services, each with a name, description, and URL. This explicitly tells AI engines what you do and creates direct connections between your brand and your service categories. The sameAs property deserves special attention. It lists all the places on the web where your organization has an official presence — LinkedIn company page, Twitter/X profile, GitHub organization, Crunchbase, Wikipedia article (if one exists), YouTube channel, and industry directory listings. AI engines use these cross-references to validate your entity identity and build a more complete profile. The more verified connections you establish, the stronger your entity authority. Organization schema should appear on every page of your site (typically in the site-wide layout), not just the homepage. This consistent entity signal across all your content reinforces the connection between your organization and every topic you publish about. For multi-location businesses, use LocalBusiness schema (a subtype of Organization) with location-specific details on location pages.
HowTo Schema: Capturing Process-Based Queries
HowTo schema structures step-by-step processes in a format that AI engines can directly consume and cite. When someone asks an AI engine "how to implement AI agents for my business," the engine looks for HowTo schema that answers the question with clear, sequential steps. Pages with HowTo markup are significantly more likely to be cited for process-oriented queries. Each HowTo includes a name (the process being described), a description (brief overview), totalTime (ISO 8601 duration format), estimatedCost (optional but valuable for AI citations), and an ordered list of steps. Each step includes a name (step title), text (step description), and optionally an image and URL. The implementation strategy is to identify every process you describe on your site — onboarding workflows, setup guides, evaluation frameworks, implementation steps — and structure them with HowTo schema. A page about "How to Build an AI Agent" that includes HowTo schema with 5 clear steps (define scope, choose framework, build integrations, test, deploy) is vastly more citable than the same content without schema. Include practical details in your HowTo steps that AI engines can extract as specific claims: estimated time per step, tools or technologies needed, cost estimates, and expected outcomes. These specific details make your HowTo more valuable to cite than a competitor's generic process description. AI engines favor actionable specificity over abstract guidance.
Speakable Schema: Optimizing for Voice and AI Responses
Speakable schema is the least commonly implemented but increasingly relevant schema type for GEO. It explicitly marks sections of your content as suitable for text-to-speech delivery — the exact format AI engines use when generating spoken responses. As voice-based AI interactions grow (smart speakers, AI phone agents, in-car assistants), Speakable-marked content has a direct advantage. The Speakable property can be added to your Article or WebPage schema and uses CSS selectors or XPath to identify which content blocks are suitable for spoken delivery. Mark content that is concise (1–3 sentences), self-contained (makes sense without surrounding context), and information-dense (contains a specific fact, statistic, or recommendation). Avoid marking content that relies on visual formatting (tables, bulleted lists, charts) or requires context from surrounding paragraphs. A practical implementation approach: identify the 2–3 most important claims or facts on each page and wrap them in elements with a consistent CSS class (e.g., .speakable). Then reference that class in your Speakable schema. This also serves as a useful content quality exercise — if you can't identify 2–3 concise, standalone facts on a page, the page may lack the entity density needed for GEO. Speakable is currently supported by Google and is being adopted by AI engine providers. While its direct impact on GEO is still emerging, implementing it now positions your content for a channel that's clearly growing and costs minimal additional effort once you have your other schema in place.
Testing, Validation, and Monitoring
Validation is critical — invalid schema markup is worse than no schema because it may confuse AI engines and won't generate rich results in Google. Use three testing tools in sequence. Google's Rich Results Test (search.google.com/test/rich-results) validates that your schema is syntactically correct and eligible for Google rich results. It shows you exactly which schema types are detected and highlights errors or warnings. Run every page template through this tool before deploying. Schema.org's Validator (validator.schema.org) checks your markup against the full schema.org specification, catching issues that Google's tool may not flag. It's stricter about property types and relationships, which matters for AI engines that process schema more thoroughly than Google's rich results system. Manual AI engine testing is the final validation step. After deploying schema markup, ask ChatGPT, Perplexity, and Claude questions that your content answers. Check whether your structured data appears to influence the response — are your specific statistics cited? Is your brand mentioned? Is your FAQ answer quoted? Track these results over 4–6 weeks to measure the impact of your schema implementation. Ongoing monitoring should check for schema validation errors after CMS updates (which can break JSON-LD injection), ensure dateModified values are updated when content changes, and verify that new pages receive the appropriate schema templates. Add schema validation to your CI/CD pipeline or use monitoring tools like SchemaApp or Merkle's Schema Markup Generator to maintain quality over time.
Need help with this?
Our team has built 200+ projects across AI agents, SaaS, and enterprise platforms.
Frequently Asked Questions
JSON-LD (JavaScript Object Notation for Linked Data) is a method of encoding structured data in a script tag within your HTML. It's preferred because it doesn't interfere with your HTML markup, is recommended by Google, and is the format most reliably consumed by AI search engines.
Include 3–7 FAQ items per page, each with a unique question directly related to the page's topic. Avoid duplicating FAQ items across pages. Quality and specificity matter more than quantity — each answer should contain concrete facts, statistics, or actionable information.
Yes. FAQPage, HowTo, and Article schema generate rich results in Google SERPs (FAQ dropdowns, how-to carousels, knowledge panels). These rich results increase click-through rates by 20–30%. Structured data is one of the highest-overlap optimizations between GEO and SEO.
For a site with established page templates (most CMS-based sites), implementing schema across all templates takes 2–4 hours per template type. A typical site with 5 page templates (homepage, service pages, blog posts, answer pages, landing pages) takes 10–20 hours total.
Partially. CMS plugins (Yoast for WordPress, next-seo for Next.js) can generate Article and Organization schema automatically. FAQPage and HowTo schema typically require manual content creation for each page. The template structure can be automated, but the content within it should be hand-crafted.
The top mistakes are: using FAQ schema for content that isn't actually on the page (violates Google guidelines), missing dateModified on Article schema, not linking author to a Person schema with credentials, duplicating identical FAQ items across pages, and not validating markup after CMS updates break the injection.
The direct impact is still emerging, but Speakable schema explicitly marks content as suitable for AI-spoken responses — the format all voice AI and AI search engines use. It's a low-effort, high-optionality implementation that positions your content for a growing channel.
Ready to build?
Talk to our team about your project.