AEO Schema Generator
Generate AEO-ready JSON-LD schema in seconds — with @graph architecture, entity signals, and speakable markup. Paste directly into WordPress to get cited by ChatGPT, Perplexity, and Google AI Overviews.
about to understand topical authority.@graph
Recommended
Fill in the form and click Generate Schema to produce AEO-ready JSON-LD markup.
Schema types supported
Every schema type is built to AEO Insider’s entity-first content strategy, with fields that directly influence how AI models extract, attribute, and cite your content.
Article / BlogPosting
Includes about, mentions, speakable, @id, keywords, wordCount, articleSection, and inLanguage — the full AEO citation stack.
FAQPage
Q&A pairs structured for People Also Ask, Google AI Overviews, and conversational AI answer extraction. Includes @id for entity linking.
HowTo
Step-by-step procedural schema with time, cost, and step-level instructions. High-value for AI Overview surfaces on process queries.
BreadcrumbList
Generates correct positional schema with @id references, helping AI systems understand site hierarchy and navigation context.
Author + Organization
Person + Organization pair with sameAs social profiles, job title, and logo — strengthens E-E-A-T signals for LLM source credibility.
WebSite New
Signals your site as a named entity to LLMs. Includes optional SearchAction for Sitelinks Search Box and inLanguage declaration.
Service New
Essential for agency service pages. Surfaces your audits, retainers, and productized offers in AI-generated results for commercial intent queries.
VideoObject New
Enables AI Overviews to cite your video content for how-to and explainer queries. Pairs with HowTo schema for maximum surface area.
AEO Schema Generator: The Complete Guide to JSON-LD for AI Search
An AEO schema generator produces JSON-LD structured data — the machine-readable markup that tells AI engines like ChatGPT, Perplexity, and Gemini how to read, attribute, and cite your content. This tool generates eight schema types using @graph architecture and entity signals, covering every surface where AI engines make citation decisions.
What Is @graph Architecture and Why It Matters
Most schema generators produce separate, disconnected JSON-LD blocks — one for Article, one for BreadcrumbList, one for FAQPage. This works, but it forces AI engines to process each block independently with no shared context. @graph architecture solves this by wrapping all schema types in a single structured data block with a unified @context.
The practical effect: an AI engine reading a page with @graph schema understands in one read that the Article is written by this Person, who works for this Organization, which publishes at this WebSite, and the Article answers these FAQ Questions. Every relationship is declared. There is no ambiguity about attribution, authorship, or entity identity.
The @graph toggle in the output panel is on by default because disconnected schema blocks are consistently outperformed in AI citation tests. When you copy the output from this tool, you are copying a single, unified block — not 3–4 separate ones to paste individually.
Entity Signals: What the About and Mentions Fields Actually Do
The about and mentions fields in the Article tab are the most important fields in this tool for LLM citation — and the most commonly skipped. Here is what each does:
about— declares what concepts, tools, or entities this article is primarily about. AI engines use this to understand topical authority. If your article is about Answer Engine Optimization, that entity should be explicitly declared here. An LLM reading this field is reading your editorial statement of topic ownership.mentions— declares what other entities appear in the content: tools you reference, platforms you discuss, organizations you cite. This is how AI engines build knowledge graph connections between your content and established entities in their training data.speakable— marks CSS selectors pointing to your most authoritative paragraphs. Voice assistants and AI summarisation systems use this to identify which content to read aloud or include in AI Overviews.
Filling in these three fields correctly — with specific, named entities rather than generic labels — is what separates schema that gets cited from schema that gets ignored. A generic about: "SEO" is weak. A specific about: "Answer Engine Optimization" with @type: Thing is an explicit knowledge graph signal.
Which Schema Type to Use for Each Situation
With eight schema types available, the right choice depends on what the page is and what queries you want to appear in. Use this as your decision guide:
| Schema Type | Use When | Primary AI Surface |
|---|---|---|
| Article / BlogPosting | Any editorial content: blog posts, guides, reports | AI citations, Google AI Overviews, ChatGPT sources |
| FAQPage | Pages with explicit Q&A content or FAQ sections | People Also Ask, Perplexity answer extraction, voice assistants |
| HowTo | Step-by-step guides, tutorials, processes | Voice search, Google AI Overviews for procedural queries |
| BreadcrumbList | Every page — helps AI engines understand site structure | Site navigation context, entity hierarchy |
| Author + Organization | Any bylined content — strengthens E-E-A-T and attribution | Source credibility signals for LLM citation selection |
| WebSite | Homepage or primary brand page — declare site as a named entity | Brand entity recognition across all AI platforms |
| Service | Service or product offering pages | Commercial intent queries in ChatGPT, Perplexity, Gemini |
| VideoObject | Pages with embedded video content | AI Overviews for how-to and explainer queries |
For maximum coverage on a single important page, combine Article + FAQPage + BreadcrumbList in a single @graph block. Toggle the @graph switch in the output panel — it is on by default — and generate each type in sequence, then copy the unified block.
How to Add the Generated Schema to WordPress
The output from this tool is a complete, paste-ready JSON-LD block. There are three reliable ways to add it to a WordPress page:
- RankMath (recommended): Open the page in the WordPress editor. Go to the RankMath sidebar → Schema tab → Add Schema → Custom Schema. Paste the JSON-LD output directly — no script tags needed. RankMath injects it into
<head>automatically and keeps it clean. - WPCode plugin: Install the free WPCode plugin (formerly Insert Headers and Footers). Create a new snippet, set type to JavaScript, paste the JSON-LD wrapped in
<script type="application/ld+json">tags, set location to Header, and scope it to the specific page URL using Smart Conditional Logic. This bypasses WordPress content filters entirely — the most reliable method if your security setup strips inline scripts. - Custom HTML block in Gutenberg: Add a Custom HTML block to the page, paste the JSON-LD wrapped in script tags. Works for admin users with unfiltered HTML permissions. Note that some security plugins strip inline script tags on save — use WPCode if this happens.
After publishing, validate at the Google Rich Results Test and Schema.org Validator. Then use the AEO Readiness Checker to confirm the page is correctly structured for AI engine citation across all signal layers, not just schema.
Common JSON-LD Mistakes That Reduce AI Citation Rates
- Skipping the
@idfield. Adding your page URL as@idlinks your schema entities to a specific URL — a critical attribution signal. Without it, AI engines cannot definitively attribute your answers to your domain. - Generic entity names in
aboutandmentions. “SEO” is not a useful entity signal. “Answer Engine Optimization”, “Google AI Overviews”, and “Perplexity” are named entities with knowledge graph presence. Be specific. - Using multiple separate schema blocks instead of
@graph. Separate blocks are processed independently with no shared context.@graphlets AI engines understand all relationships between entities on a single pass. - Omitting the Author schema on editorial content. Author and Organization schema are the primary E-E-A-T signals AI engines use when choosing which sources to cite. A page with no author attribution is harder to cite than one with a declared person and organisation.
- Setting
datePublishedbut notdateModified. AI engines weight freshness. A page published two years ago with a recentdateModifiedsignals active maintenance. WithoutdateModified, the publication date is the only freshness signal available. - Speakable selectors pointing to navigation, headers, or footers. Only target your core content sections — intro paragraphs, key answer blocks, conclusion summaries. Speakable signals on non-content elements reduce the quality of voice and AI extraction.
How This Tool Differs From Standard Schema Generators
Most JSON-LD generators produce basic schema with the minimum required fields. This tool is built specifically for AEO — Answer Engine Optimization — and adds fields that standard generators omit entirely: about and mentions entity arrays for LLM topical signalling, speakable CSS selector specification for voice and AI extraction, inLanguage declaration for multilingual entity disambiguation, and sameAs social profile arrays on Author schema for cross-platform entity resolution.
The @graph output combines all generated types into a single block. The Compact / Pretty toggle in the output panel controls formatting — use Compact for production (smaller page weight), Pretty for readability during validation. The Copy All button wraps the output in <script type="application/ld+json"> tags ready for direct paste. For FAQPage and Speakable schema specifically, see the Schema Markup for AI Engines tool — it covers those types with additional preset templates and an AI Citation Readiness Score.
about, mentions, and speakable that directly influence citation rates in AI-generated answers.about property in Article schema — a direct declaration of what topics, tools, or concepts the content is primarily about. Large language models use this field to understand topical authority and decide whether your content is a relevant source for a given query. The Mentioned Entities field populates the mentions property — a list of other named entities that appear in your content, such as tools, platforms, and organisations. Both fields create explicit knowledge graph connections between your content and established entities.h1, .intro-paragraph, #key-answer, .article-summary. These selectors tell voice assistants and AI engines which content to extract for spoken responses and AI Overviews. Only target content sections, not navigation, headers, or footers. Separate multiple selectors with commas.about and mentions entities improves source selection for topical queries. The effect is most pronounced on pages where the schema entities match the query intent precisely — which is why filling in the About and Mentioned Entities fields correctly is more important than the schema type itself.- Schema.org — Article specification. Official schema.org documentation for Article, BlogPosting, and TechArticle types including
about,mentions, andspeakableproperties. - Google Developers — Article structured data. Google’s implementation guide for Article schema including required and recommended fields for rich result and AI Overview eligibility.
- Google Developers — Speakable structured data. Official documentation on Speakable schema implementation, CSS selector requirements, and voice assistant integration.
Want This Implemented — Properly?
AEO Insider offers hands-on AEO audits and implementation. We fix your schema, restructure your content architecture, and wire up entity signals for AI citation.
- Full-site AEO + schema audit
- Entity architecture and @graph implementation
- Monthly AI visibility monitoring