AI-Ready Schema Generator

AI-Ready Schema Generator

Generate copy-paste JSON-LD (FAQ, Organization, Local Business, How-To, Article) tuned for AI answer extraction. Free, no API key.

Free · no API key · paste the output into your page <head>

The Schema Markup Generator that tells you which types are already dead

Pick a page type and get back valid, copyable JSON-LD in seconds — Organization, LocalBusiness, Article, Product, and more. We mark which fields Google actually requires versus the ones that are only recommended, and we won't hand you FAQPage or HowTo markup without telling you both are retired from Google Search.

Valid JSON-LD in secondsRequired vs recommended, markedFlags retired schema typesFree, no signup

Built and reviewed by Sayed Hasan, founder of SEOs Hut · Updated 28 September 2026 · Free, no signup

Structured data / schema for AI citations: real lever or overrated?

Overrated as a direct lever, based on the evidence so far. Google states structured data doesn't affect web ranking, and its AI optimisation guide files schema under what you don't need to do for generative search. The one documented exception is Bing, which says structured data 'may support clearer grounding' for Copilot — hedged, not guaranteed, and the only platform to have published anything on this at all.

Read this before you add anything

What Google actually says about schema and ranking

“Structured data won't make your site rank better. It's used for displaying the search features listed in developers.google.com/search/docs/… Use it if your pages map to & are appropriate for any of those features.”

<b>John Mueller, Google, 13 April 2025</b> &mdash; Search Engine Roundtable

Danny Sullivan said the same thing back in January 2020: "structured data like this has no impact on ranking in web search." Google's own structured data policies page still puts it in writing today: using structured data enables a feature to be present, it does not guarantee that it will be present, and a manual action against your structured data affects rich result eligibility, not your ranking. Three sources, six years apart, saying the same thing is not a coincidence — it's the one point in this whole topic Google has never wavered on.

The table nobody else publishes

Ten schema types, and what's actually still alive

Google currently documents 25 structured data features in its gallery. FAQ, HowTo, and the sitelinks search box are not among them, no matter what your last tutorial told you. Seven of ten competitor generators we checked still produce FAQPage on request, and four still produce HowTo — neither will earn you anything in Google Search today.

Schema typeGoogle rich resultRequired propertiesBing groundingOther AI consumer (published)
OrganizationLive in galleryNone required30 recommended incl. sameAs, logo, legalName"May support clearer grounding"None published
LocalBusinessLive in galleryaddress, name"May support clearer grounding"None published
ArticleLive in galleryNone requiredrecommended: author, headline, image, dateModified"May support clearer grounding"None published
Product (snippet)Live in galleryname + one of review / aggregateRating / offers"May support clearer grounding"None published
BreadcrumbListLive in gallerydesktop onlyitemListElement (item, name, position)"May support clearer grounding"None published
EventLive in gallerylocation, location.address, name, startDate"May support clearer grounding"None published
Review snippetLive in galleryauthor, itemReviewed, itemReviewed.name, reviewRating.ratingValue"May support clearer grounding"None published
FAQPageDeadRestricted Aug 2023, removed 7 May 2026, docs deleted 15 Jun 2026N/A — feature retiredNot addressedNone published
HowToDead since 13 Sep 2023docs removed same monthN/A — feature retiredNot addressedNone published
WebSite + SearchAction (sitelinks box)Removed Nov 2024nositelinkssearchbox tag archivedN/A — feature retiredNot addressedNone published

Scroll the table sideways to see every column.

"Other AI consumer" means a platform that has published, in writing, that it uses structured data for anything. Bing has (Webmaster Guidelines §14, hedged to "may"). OpenAI, Anthropic, Perplexity and schema.org itself have published nothing on the subject either way.

Copy this, not a screenshot of it

One @graph, four types, wired together

This is what combining blocks actually looks like — shared @id values instead of four disconnected script tags.

JSON-LD
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Organization",
      "@id": "https://example.com/#organization",
      "name": "Acme Roofing Ltd",
      "url": "https://example.com/",
      "logo": "https://example.com/logo.png",
      "sameAs": [
        "https://www.facebook.com/acmeroofing",
        "https://www.linkedin.com/company/acmeroofing"
      ]
    },
    {
      "@type": "LocalBusiness",
      "@id": "https://example.com/#localbusiness",
      "name": "Acme Roofing Ltd",
      "image": "https://example.com/storefront.jpg",
      "address": {
        "@type": "PostalAddress",
        "streetAddress": "14 Elm Street",
        "addressLocality": "Leeds",
        "postalCode": "LS1 4AB",
        "addressCountry": "GB"
      },
      "telephone": "+44-113-000-0000",
      "priceRange": "$$",
      "parentOrganization": { "@id": "https://example.com/#organization" }
    },
    {
      "@type": "BreadcrumbList",
      "@id": "https://example.com/roofing-repairs#breadcrumb",
      "itemListElement": [
        { "@type": "ListItem", "position": 1, "name": "Home", "item": "https://example.com/" },
        { "@type": "ListItem", "position": 2, "name": "Services", "item": "https://example.com/services" },
        { "@type": "ListItem", "position": 3, "name": "Roofing Repairs", "item": "https://example.com/roofing-repairs" }
      ]
    },
    {
      "@type": "WebPage",
      "@id": "https://example.com/roofing-repairs#webpage",
      "url": "https://example.com/roofing-repairs",
      "name": "Roofing Repairs | Acme Roofing Ltd",
      "about": { "@id": "https://example.com/#localbusiness" },
      "breadcrumb": { "@id": "https://example.com/roofing-repairs#breadcrumb" }
    }
  ]
}

Note what's absent: no FAQPage, no HowTo. Both would validate as syntax and earn nothing as a rich result.

Decide before you generate

What kind of page is this, actually?

The generator will build any of these. Whether it's worth building is a separate question, answered here.

1

What is this page fundamentally about?

Your homepage or an about page describing the business itself
Organization (once per site)Emit one Organization block, ideally inside the same @graph as your other types. No properties are required, but Google names thirty recommended ones — sameAs, logo, legalName, and, for disambiguation specifically, iso6523Code or naics rather than a standalone duns or leiCode.
A physical shop, service-area business, or single location
LocalBusiness, in addition to OrganizationThis one does have two required properties — address and name. Everything else, including phone and hours, is recommended rather than required, and it should match your Google Business Profile exactly, not approximately.
A blog post, guide, or news piece
ArticleNo required properties here either, but a rich result realistically needs author, headline, image and dateModified to be worth adding. Make sure the title tag and meta description say the same thing the markup does — check that with the Meta Tag Analyzer.
A service page describing what you offer — consulting, SEO, plumbing call-outs
DON'T mark it up as ProductProduct schema is for physical or purchasable goods carrying a name plus a review, rating, or offer — not a description of a service. Misusing it won't trigger a manual action, but it won't earn the result either, since it isn't the thing the type describes.
2

Are you about to add FAQPage or HowTo because a template still offers them?

You want the FAQ rich result — the visual accordion in search
DON'T — it's goneRestricted to government and health sites in August 2023, fully removed from Search on 7 May 2026, and Google deleted the documentation entirely on 15 June 2026. The markup won't cause harm, but it will not earn a rich result.
You want HowTo's step-by-step rich result
DON'T — dead since September 2023Same story, over two years earlier. If your CMS still generates it by default, that's legacy code running, not a live feature.
You're not sure whether your CMS plugin already emits Organization
Check before you duplicateTwo competing Organization blocks on one page is a common, entirely avoidable mess. Validate what's actually live on the page with the Schema Markup Analyzer before adding a second one.
Numbers worth knowing

What Google actually documents, in three figures

25
structured data features Google currently documents in its gallery — FAQ, HowTo and the sitelinks search box aren't among them
Google Search Central
0
required properties for Organization schema — all thirty recommended fields are optional
Google Developers
3
independent methods (a matched study, a cross-sectional study, and a direct test) that each found no detectable independent lift in AI citations from adding schema
see sources
The one real exception

Bing published something. Nobody else has.

“Structured data may support clearer grounding but does not guarantee visibility or grounding traffic. Markup must accurately reflect visible content.”

<b>Bing Webmaster Guidelines, §14</b> &mdash; Bing Webmaster Guidelines

Read the hedge carefully: "may support," not "is used for" or "determines." This is still the only sentence in this entire research area written by a platform describing its own product, rather than inferred by an outside test. Its scope is also explicit — Bing search experiences, Copilot, and grounding API results — not the wider AI ecosystem. That second line, "markup must accurately reflect visible content," is worth reading as a warning as much as a feature: mismatched or inflated schema is the failure mode Bing is naming, not a reason to add more of it.

Seen in the wild

Four things every other schema page still gets wrong

None of these is exotic. All four showed up repeatedly in the pages currently ranking for this term.

Still generating FAQPage because a tutorial says to

What happens: Most generators we checked still ship it by default. The rich result died in stages: restricted August 2023, fully gone 7 May 2026, documentation deleted 15 June 2026.

Do this instead: Drop it from new pages. If it's already live somewhere, it's inert, not harmful — stop treating it as an active task.

Linking Google's retired Structured Data Testing Tool

What happens: A majority of the pages we reviewed still send readers to a tool Google has already retired.

Do this instead: Use the Rich Results Test and validator.schema.org — the two tools Google actually maintains today.

Quoting "completeness is one of the most important ranking signals"

What happens: This line is real, but it's from Google's Enriched search results documentation, about position inside a carousel — not about organic web ranking. It's the single most misquoted line in Google's structured data docs.

Do this instead: If you want a ranking claim, cite Mueller or Danny Sullivan instead: both stated plainly that structured data like this has no impact on ranking in web search.

Claiming Bing "confirmed its LLMs use schema"

What happens: There's no written primary source that says this. What Bing has actually published is narrower — grounding "may" benefit, not a confirmed usage claim — and it doesn't extend to OpenAI, Anthropic, or Perplexity, none of which have published anything on the subject.

Do this instead: Cite the Webmaster Guidelines line directly, and be honest that Bing is the only platform that has published anything on this at all.

The honest read

Does schema get you cited by AI engines? The evidence says no, so far

Three independent attempts to measure this converge on the same answer, using three different methods.

Ahrefs ran the cleanest version: a matched difference-in-differences study across 1,885 pages that added JSON-LD, against roughly 4,000 matched pages that didn't. Google AI Overviews citations moved -4.6% — a real, negative, statistically significant effect. AI Mode moved +2.4% and ChatGPT +2.2%, both statistically indistinguishable from zero. Their own caveat matters: every sampled page already had over 100 AI Overview citations before the test, so this measures incremental lift on pages already being cited, not whether schema helps a page get noticed in the first place.

A separate cross-sectional study of 1,006 pages and 730 citations found schema presence had no significant effect (p=.296); the dominant predictor of getting cited was plain Google organic rank. It isn't peer-reviewed and the author runs a GEO agency, so treat it as a data point, not a verdict.

The most direct test came from searchVIU: they put a price in JSON-LD only, with nothing matching in the visible page text, then asked five engines — ChatGPT, Claude, Perplexity, Gemini, and Google AI Mode — to find it. Zero did. All five found the same price when it also appeared in visible Microdata or RDFa. That's the most literal evidence available: these engines are reading rendered text, not parsing your markup.

None of this means schema is worthless — it means the value sits elsewhere. It still creates eligibility for rich results, and once you've asserted an identity with an Organization block, the Entity & Knowledge Panel Checker tells you whether Google actually resolved it. If the page itself needs to be more quotable in the first place, that's a separate job, covered by the AEO Readiness Checker.

The practical takeaway for a small business is unglamorous: generate the markup because it costs minutes and keeps you eligible for rich results, not because a blog post promised it would get you quoted by an AI engine. If your organic ranking is the thing actually driving citations, as Fischman's data suggests, then the schema on the page is doing far less work than the content and the ranking underneath it.

Schema is not a substitute for the basics

None of this matters if the page carrying the markup can't be crawled or indexed in the first place — run a SEO Audit before reaching for a schema generator. And whatever LocalBusiness fields you fill in, they need to match your Google Business Profile and every other listing exactly, not approximately: check that with the NAP Consistency Checker.

Paste it, validate it, then judge the content underneath

Generated markup is only ever a wrapper around what's already on the page. Once it validates, the higher-leverage question is whether the content inside is any good to begin with — that's what the AI Content Scorer checks.

Questions

AI-Ready Schema Generator FAQ

FAQ schema is officially removed from Google Search Console — should I remove it from my website or leave it for LLMs?
Remove it from new pages; leaving it on old ones isn't harmful, just inert. The FAQ rich result was restricted to government and health sites in August 2023, fully removed from Search on 7 May 2026, and Google deleted its documentation on 15 June 2026. No AI vendor has published that it reads FAQPage markup specifically, so "leave it for LLMs" isn't a documented reason to keep it.
Structured data / schema for AI citations: real lever or overrated?
Overrated as a direct lever. Three independent attempts to measure it — a matched study of 1,885 pages, a cross-sectional study of 1,006 pages, and a direct test putting facts only in markup — all found no detectable independent effect on AI citations. Google itself files schema under "what you don't need to do" for generative search. The one exception: Bing says it may support clearer grounding for Copilot.
How do you combine several JSON-LD markups?
Wrap them in one @graph array under a single @context, and link them with shared @id values — your Organization's @id referenced from LocalBusiness's parentOrganization, your WebPage referencing your BreadcrumbList, and so on. That's how search engines expect related entities on one page to connect, rather than as several disconnected script blocks.
Do I need JSON-LD on all pages or just the home page?
Organization typically belongs once, often on the homepage or an about page. Page-specific types belong on the pages they describe: LocalBusiness where relevant, Article on each post, BreadcrumbList on any page with a real hierarchy above it. There's no single required page — it's driven by what each page actually is.
Has schema pro become useless if you use rank math or yoast?
Not useless, but check for duplication first — both plugins already emit Organization and some Article or Product markup automatically. Adding a generator's output on top without checking creates two competing blocks on the same page, which is worth catching with a validator before you publish, not after.
Is FAQ schema still relevant?
Not for the rich result — that's fully retired as of 7 May 2026, with documentation removed 15 June 2026, after being restricted to government and health sites from August 2023. There's no published statement from any AI vendor that FAQPage markup specifically affects being cited, so there's no current reason tied to rich results or AI to keep adding it.

Primary sources used on this page

  1. Google Search Gallery — the 25 structured data features currently documented — developers.google.com
  2. Google Search Central changelog — FAQ, HowTo and sitelinks searchbox removal dates — developers.google.com
  3. Google structured data type reference (Organization, LocalBusiness, Article, Product, BreadcrumbList, Event, Review) — required vs recommended properties — developers.google.com
  4. Google structured data policies — using structured data enables eligibility, doesn't guarantee display, doesn't affect ranking — developers.google.com
  5. Search Engine Roundtable — John Mueller: structured data won't make your site rank better (13 April 2025) — www.seroundtable.com
  6. Google Search Central: AI optimisation guide — schema filed under "what you don't need to do" — developers.google.com
  7. Bing Webmaster Guidelines §14 — structured data "may support clearer grounding" — www.bing.com
  8. Ahrefs — matched study of schema and AI citations (11 May 2026) — ahrefs.com
Keep going

Validate it, then check what it's resting on

Generating the markup is the easy five minutes. These check whether it's live, whether it's correct, and whether the facts inside it agree everywhere else.

Read the guide: Local Business Schema: How to Add LocalBusiness Markup and Website Breadcrumbs: Types, Benefits and Best Practices. Want it handled for you? See our Local SEO services.

Scroll to Top