Structured Data for SEO

TL;DR — Structured data is extra code you add to a page that labels what your content means — “this is the price,” “this is the author,” “this is the cook time” — so search engines don’t have to guess. It uses a shared vocabulary called schema.org, usually written in a format called JSON-LD. It can make your page eligible for fancier search listings (star ratings, product cards), but it does not make your page rank higher.

What structured data is

When you look at a web page, you can tell a phone number from a price from an author’s name just by looking. A search engine can’t — it mostly sees text. Structured data is extra code you add to a page that spells it out for machines: “this part is the product name,” “this is the rating,” “this is the publish date.”

It uses a shared dictionary called schema.org — a vocabulary that Google, Microsoft (Bing), Yahoo, and Yandex agreed on together — so publishers can use the same labels across consumers. Evidence for this claim Schema.org provides a shared vocabulary for structured data, while JSON-LD is one supported encoding format. Scope: Schema.org vocabulary and Google structured-data guidance; this does not imply every type is eligible for a Google search feature. Confidence: high · Verified: 2026-07-14 Schema.org Most of the time it’s written in a format called JSON-LD: a small block of code that sits in the page’s HTML and describes the content without changing how the page looks to a visitor.

You’ll hear “structured data,” “schema markup,” and “schema.org” used almost interchangeably. Strictly: schema.org is the vocabulary, and structured data is the broad idea of machine-readable labels. In everyday SEO talk they mean the same thing.

Why it matters

Structured data does two useful jobs:

• Rich results. When Google can read your labels clearly, your search listing can become eligible for an upgrade — star ratings under a review, a price and “in stock” on a product, a recipe’s cook time, breadcrumbs. These are called rich results, and they can make people more likely to click your listing. Evidence for this claim Valid structured data can make a page eligible for Google search features, but Google does not guarantee that a rich result will appear. Scope: Google Search structured-data guidance; eligibility is distinct from display and ranking. Confidence: high · Verified: 2026-07-14 Google: Structured data introduction
• Helping engines understand who you are. Markup like Organization and Person — especially a property called sameAs that links your brand or author to your Wikipedia or LinkedIn page — helps Google tell your “Apple” from the fruit, and your author from someone else with the same name.

The thing most people get wrong

Structured data does not directly improve your rankings. This is the single biggest myth. Adding schema doesn’t make a page rank higher for a keyword — Google’s John Mueller has said this many times. The benefit is indirect: a rich result can earn you more clicks, and clearer entity signals can help Google understand you. But there’s no “add schema, climb the rankings” lever.

Two more honest cautions:

• Rich results aren’t guaranteed. Even with perfect markup, Google decides whether to show the enhancement. You can do everything right and still get a plain blue link.
• Only mark up what’s actually on the page. Don’t add schema for content that isn’t visible to visitors — that breaks Google’s rules and can get you a penalty.

Want the deep version — the three formats, the full list of rich result types, which ones got removed, what explicit clues do and do not guarantee, and where to go next? Switch to the Advanced tab.

TL;DR — Structured data applies the schema.org vocabulary to your page in one of three formats — JSON-LD (Google’s recommendation), Microdata, or RDFa. It gives search systems explicit clues and can make an eligible page available for supported search features. Neither appearance nor CTR is guaranteed, and it does not promise Knowledge Graph inclusion, AI citations, or disambiguation. It is not a direct ranking factor — Mueller, repeatedly. The supported rich-result list keeps shrinking: FAQ and HowTo were gutted in 2023, seven more types removed June 2025, and the FAQ rich result was fully removed by May 2026. Only mark up visible content, prefer JSON-LD in the raw HTML for the widest parser coverage, and keep every property, including sameAs, accurate and supported by the page.

What structured data is — vocabulary vs. format

Two things get conflated constantly, so let’s separate them up front:

• schema.org is the vocabulary — the shared dictionary of types (Product, Article, Recipe, Organization) and properties (price, author, cookTime, sameAs). It’s a collaborative project of Google, Microsoft, Yahoo, and Yandex, and it’s enormous — over 800 types. Google only supports a subset for rich results.
• JSON-LD, Microdata, and RDFa are the formats — the three ways to actually write that vocabulary into a page. They’re interchangeable carriers of the same meaning.

Google’s own definition: “Structured data is a standardized format for providing information about a page and classifying the page content.”

The three formats — and why JSON-LD wins

All three work. Google recommends JSON-LD as “the easiest solution for website owners to implement and maintain at scale (in other words, less prone to user errors).” Evidence for this claim Google Search supports JSON-LD, Microdata, and RDFa for structured data and generally recommends JSON-LD. Scope: Google Search documentation; individual features can impose additional requirements. Confidence: high · Verified: 2026-07-14 Google: Structured data introduction Because it’s a self-contained block, a typo doesn’t break your page layout, and you can add, update, or remove it without touching visible content.

One implementation caveat that bites people: JSON-LD can be injected with JavaScript (e.g., via Google Tag Manager), and Google documents support for JavaScript-generated structured data. Other crawler providers generally do not publish a complete JavaScript-rendering contract. If broad machine readability matters, render JSON-LD server-side so it is present in the raw HTML response — a coverage choice, not proof that every named AI crawler is unable to execute JavaScript. I go deeper on that in Schema Markup for AI.

What structured data can support

Supported use 1: search features. Structured data can make a page eligible for an enhanced SERP listing. Google’s case studies report outcomes for named implementations: Rotten Tomatoes saw a 25% higher CTR, Nestlé 82% higher CTR on rich-result pages, Food Network a 35% increase in visits. But note the word eligible — Google is explicit: “Google does not guarantee that your structured data will show up in search results, even if your page is marked up correctly.” Those case studies do not guarantee the same appearance or CTR change for another site.

Supported use 2: explicit clues. Beyond appearance, Google says it uses structured data to understand — “to gather information about the web and the world in general, such as information about the people, books, or companies that are included in the markup.” That statement does not promise a Knowledge Panel, Knowledge Graph inclusion, AI Overview association, brand-recognition change, or citation. sameAs identifies another URL about the same entity; use it accurately, but do not call it a guaranteed disambiguation lever.

The myth to kill: structured data is not a ranking factor

This is the accuracy spine of the whole topic. Structured data is not a direct ranking signal for organic results. John Mueller has confirmed this many times over the years. A page does not rank higher because it has schema. The SEO value is eligibility for supported search features and clearer machine-readable clues. Appearance, CTR impact, entity treatment, and ranking changes all require separate observation.

If a tool or agency sells you “add schema to rank higher,” they’ve got the mechanism wrong. Rich results may change presentation; measure their actual effect.

Rich result types Google supports — and the ones it removed

Google maintains a Search Gallery of supported types, but the list shrinks over time, and most articles online are out of date. As of the gallery’s last update (2026-06-15), the supported set is 25 types: Article, Breadcrumb, Carousel, Course info, Dataset, Discussion forum, Education Q&A, Employer aggregate rating, Event, Image metadata, Job posting, Local business, Math solver, Movie, Organization, Product, Profile page, Q&A, Recipe, Review snippet, Software app, Speakable, Subscription/paywalled content, Vacation rental, and Video.

What got removed — this is where most guides are wrong:

• HowTo rich results were removed from desktop in 2023 and limited thereafter.
• FAQ rich results were restricted in 2023 to government and health sites, then fully removed by May 2026. The FAQPage type is still valid markup — it just no longer produces a SERP enhancement for the vast majority of sites.
• Six more types were removed in June 2025: Claim Review, Estimated Salary, Learning Video, Special Announcement, Vehicle Listing, and the old single-course “Course Info” format. Google announced a seventh — Book Actions — in the same batch, but it was excluded from the actual September 2025 removal and Google later pulled the deprecation banner from that doc; it remains a narrow, registration-gated feature (see Book schema for the full story).

The takeaway: don’t build a strategy around a rich result that no longer exists, and don’t panic-strip valid markup just because the visual enhancement is gone.

How to implement structured data

A reliable order of operations:

1. Pick the right type for what’s genuinely on the page (Product for a product page, Article for a blog post, LocalBusiness for a location page).
2. Write JSON-LD — by hand, with a generator (Google’s Structured Data Markup Helper, Dentsu’s generator), or via a CMS plugin (Yoast, Rank Math).
3. Fill required properties first, then recommended ones. Google: “It is more important to supply fewer but complete and accurate recommended properties rather than trying to provide every possible recommended property.” Accuracy beats breadth.
4. Validate with the Rich Results Test and the schema.org validator.
5. Prefer raw-HTML delivery for parsers that do not document a rendering stage.
6. Monitor the rich result reports in Google Search Console after deployment.

And the policies that matter: only mark up visible content, keep it accurate (no fake reviews/ratings), don’t block your structured-data pages from Googlebot, and apply the same markup across canonical and duplicate versions of a page.

Structured data and AI search

It’s tempting to assume schema is the key to getting cited in AI answers. The best evidence says it isn’t a direct lever — an Ahrefs study of 1,885 pages found adding JSON-LD barely moved AI citations on any platform. Google states there’s “no special schema.org structured data that you need to add” for AI features. Accurate schema supplies explicit machine-readable assertions, but provider use, entity resolution, and citation outcomes are not guaranteed and need separate evidence. Full treatment in Schema Markup for AI.

Does Bing use structured data?

Yes. Bing reads schema.org markup in all three formats, uses it for its own rich results (recipes, events, products, FAQs, how-tos, videos), and feeds it into its Satori knowledge base for entity disambiguation. Bing Webmaster Tools includes a Markup Validator, and Microsoft’s Fabrice Canel has said schema helps Bing’s LLMs understand content — the one AI-platform rep to confirm that on the record.

Where to go next

This hub is the map. The deep dives in this sub-cluster:

• Schema markup — the schema.org vocabulary itself: the common types every SEO should know (Organization, Product, Article, BreadcrumbList), how @id and @graph connect them, and how to choose the right type for a page.
• JSON-LD — the recommended format up close: syntax, where to place the <script> block, server-side vs. client-side rendering, and how to write and validate it.
• Rich results — the SERP enhancements themselves: which types still exist, eligibility, the removals/deprecations, and how to track them in Search Console.
• ImageObject schema — the media type used standalone and nested inside Product/Article/Recipe.
• Carousel rich results and ItemList schema — the mechanism behind Google’s carousel feature and the generic list type underneath it (BreadcrumbList included).
• MerchantReturnPolicy and OfferShippingDetails — the Offer-level properties Merchant Center listings increasingly require.
• Speakable schema — the narrow, still-beta markup for voice-assistant readback.
• Nesting schema with @id and @graph — the deeper how-to on linking entities instead of re-declaring them.
• Common structured data errors — what the Rich Results Test and Search Console actually flag, and how to fix it.
• Schema & rich result deprecations — the running, dated timeline of what’s been removed or narrowed.

This sub-cluster lives in the broader on-page SEO cluster — alongside meta tags, header tags, image SEO, and the other on-page signals. Closely related entity work lives in Entity SEO and Schema Markup for AI.

AI summary

A condensed take on the Advanced version:

• Vocabulary vs. format. schema.org is the shared vocabulary (800+ types); JSON-LD, Microdata, and RDFa are the three formats that carry it. Google recommends JSON-LD as least error-prone — all three are supported.
• Documented value: eligibility for supported search features and explicit clues about page meaning. Display, CTR lift, Knowledge Graph inclusion, AI citations, and ranking changes are not guaranteed. sameAs must be accurate but is not a promised lever.
• Not a ranking factor. Structured data does not directly improve rankings — Mueller, repeatedly. Feature appearance and other outcomes require separate measurement.
• The supported list shrinks. HowTo removed from desktop (2023); FAQ restricted in 2023 and fully removed by May 2026 (the FAQPage type stays valid); six types removed June 2025 (Claim Review, Estimated Salary, Learning Video, Special Announcement, Vehicle Listing, old Course); Book Actions was announced in the same batch but was NOT actually removed.
• Implementation: pick the right type, fill required properties first, validate (Rich Results Test + schema.org validator), prefer raw-HTML delivery for broad parser coverage, monitor in Search Console, and only mark up visible content.
• AI search: schema is entity infrastructure, not a citation lever — Google says there’s “no special schema” for AI features; an Ahrefs 1,885-page study found no citation lift.
• Bing reads schema for its own rich results and Satori knowledge base; Fabrice Canel confirmed it helps Bing’s LLMs.
• Where to go next: schema-markup (the vocabulary), json-ld (the format), and rich-results (the SERP enhancements).

Official documentation

Primary-source documentation from the search engines and schema.org.

Google

• Intro to How Structured Data Markup Works — the definition, the three formats, why JSON-LD, and the rich-result business case.
• Structured Data Search Gallery — the current list of supported rich result types (the canonical source as the list changes).
• Structured Data General Guidelines / Policies — visibility, accuracy, and spam rules; “no guarantee” of rich results.
• AI Features and Your Website — the “no special schema.org structured data that you need to add” statement.

Bing / Microsoft

• Marking Up Your Site with Structured Data — Bing Webmaster Help — Bing’s structured data guidance; JSON-LD recommended, must match page context.
• Bing Webmaster Tools — Markup Validator — Bing’s tool for testing schema (within Webmaster Tools).

schema.org

• schema.org — the shared vocabulary itself.
• Organization · Product · Article · sameAs — common types and a same-entity URL relationship.

Quotes from the source

On-the-record statements from Google and schema.org. Where a page exposes the text, the link is a deep link that jumps to the quoted passage.

Google — what structured data is and does

• “Structured data is a standardized format for providing information about a page and classifying the page content.” — Google Search Central. Jump to quote
• “Google uses structured data that it finds on the web to understand the content of the page, as well as to gather information about the web and the world in general, such as information about the people, books, or companies that are included in the markup.” Jump to quote
• On JSON-LD: the recommended format is “the easiest solution for website owners to implement and maintain at scale (in other words, less prone to user errors).” Jump to quote
• “It is more important to supply fewer but complete and accurate recommended properties rather than trying to provide every possible recommended property with less complete, badly-formed, or inaccurate data.” Jump to quote

Google — the “no guarantee” rule and policies

• “Google does not guarantee that your structured data will show up in search results, even if your page is marked up correctly.” Jump to quote
• “Don’t block your structured data pages to Googlebot using robots.txt, noindex, or any other access control methods.” Jump to quote

Google — no special schema for AI features

• “You don’t need to create new machine readable files, AI text files, or markup to appear in these features. There’s also no special schema.org structured data that you need to add.” — Google Search Central, AI features documentation. Source

Structured data implementation checklist

A pass to confirm your structured data is correct, eligible, and actually readable:

• Right type for the page — Product on product pages, Article on posts, LocalBusiness on location pages. Don’t add types that don’t match the page.
• Only visible content is marked up — no fields describing things not on the page (Google policy; mismatches risk a manual action).
• Format is JSON-LD — Google’s recommendation; least error-prone.
• Required properties filled first, then recommended ones — completeness and accuracy beat breadth.
• Rendered server-side — confirm it’s in the raw HTML (View Source / curl), not injected only via client-side JS / Tag Manager, so AI crawlers and non-rendering parsers can read it.
• Organization markup sitewide with a populated sameAs array (Wikidata, Wikipedia, LinkedIn) — and every sameAs value double-checked so you don’t link your entity to the wrong organization.
• No reliance on removed rich results — FAQ (fully removed by May 2026) and HowTo (off desktop since 2023) no longer produce SERP enhancements for most sites; don’t build a strategy around them.
• Validated in the Rich Results Test and the schema.org validator.
• Consistent across canonical + duplicate pages — same markup on all versions.
• Time-sensitive data kept current — outdated info (past events, stale prices) prevents rich result display.
• Monitored in Search Console — rich result / enhancement reports checked for errors and warnings after deployment.
• Expectations set — schema makes you eligible for rich results (never guaranteed) and aids entity understanding; it is not a direct ranking lever.

Structured data — cheat sheet

Vocabulary vs. format

Currently supported Google rich result types (25, per Google’s Search Gallery, last updated 2026-06-15)

Article · Breadcrumb · Carousel · Course info · Dataset · Discussion forum · Education Q&A · Employer aggregate rating · Event · Image metadata · Job posting · Local business · Math solver · Movie · Organization · Product · Profile page · Q&A · Recipe · Review snippet · Software app · Speakable · Subscription/paywalled content · Vacation rental · Video.

Removed / deprecated — don’t strategize around these

Fast facts

• Not a ranking factor — feature appearance and CTR impact must be measured.
• JSON-LD recommended; goes in a <script type="application/ld+json"> block.
• Prefer raw-HTML delivery — provider rendering support varies and is often undocumented; verify a named crawler before making a crawler-specific claim.
• sameAs identifies another URL about the same entity; avoid outcome promises.
• Rich results never guaranteed, even with perfect markup.
• Validate: Rich Results Test + schema.org validator. Monitor: Search Console enhancement reports.

Test yourself: Structured Data

Five quick questions on what structured data does — and doesn’t — do for SEO. Pick an answer for each, then check.

Resources worth your time

My structured data writing

• Structured Data: What It Is and How to Use It — my Ahrefs guide to schema types, implementation methods, validation tools, the AI-search angle, and the sameAs risk.
• The Beginner’s Guide to Technical SEO — where structured data fits into the broader technical picture.

Official

• Google’s Intro to How Structured Data Markup Works — the canonical reference: definition, formats, why JSON-LD, business case.
• Google’s Structured Data Search Gallery — the live list of supported rich result types.
• Google’s Structured Data General Guidelines — the policies (visibility, accuracy, “no guarantee”).
• schema.org — the vocabulary itself.

From around the industry

• Marking Up Your Site with Structured Data — Bing Webmaster Help — Bing’s structured data guidance and Markup Validator.
• Microsoft Bing/Copilot use schema for its LLMs — Search Engine Land — coverage of Fabrice Canel’s confirmation that schema helps Bing’s LLMs.
• Google Drops FAQ Rich Results From Search — Search Engine Journal — coverage of the FAQ rich-result removal and what stays valid.
• How schema markup fits into AI search — without the hype — Search Engine Land — the case that schema is entity infrastructure, not a citation lever.
• Dentsu Schema Markup Generator — a free tool for generating JSON-LD for common types.

Prove the markup is valid and eligible

Two different things break here, and they need two different tests: markup can parse fine yet be ineligible for a rich result because a required property is missing, and a template change can silently break markup that was working. Valid ≠ eligible ≠ shown.

Test 1 — The JSON-LD parses with no errors

• Test to run — Paste the URL or the JSON-LD into the Schema Validator (or the schema.org validator).
• Expected result — The type is detected and there are no syntax/parse errors — valid JSON-LD, correct @type, no malformed nesting.
• Failure interpretation — Parse errors usually mean a templating bug — an unescaped quote, a null value where a string was expected, or two conflicting blocks on the page.
• Monitoring window — Immediate.
• Rollback trigger — The validator can’t detect the type at all after a template deploy — the markup isn’t rendering; roll back the template change.

Test 2 — The page is eligible for the rich result

• Test to run — Run the URL through the Rich-Result Eligibility Checker and Google’s Rich Results Test.
• Expected result — The target feature reports “eligible” with all required properties present (recommended-property warnings are fine).
• Failure interpretation — “Valid but not eligible” means a required property is missing for that feature; note that eligibility is not a guarantee Google shows the result, and some types (e.g. FAQ) have been deprecated as rich results entirely.
• Monitoring window — Immediate for eligibility; if you’re tracking whether the result actually appears, watch the GSC rich-result enhancement report and Search Appearance over 2–4 weeks.
• Rollback trigger — Valid items in the GSC enhancement report drop after a template change, or “eligible” flips to errors — a change broke a required property; revert and re-test.

Structured Data

Structured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding.

Related: Schema Markup, Rich Results, JSON-LD

Structured Data

Structured data is a standardized format for labeling the information on a web page so that machines can understand its meaning, not just read its words. It applies a shared vocabulary — primarily schema.org, a project of Google, Microsoft, Yahoo, and Yandex — to explicitly identify the entities, attributes, and relationships on a page: a product’s price, a recipe’s cook time, an event’s date, an article’s author.

The vocabulary (schema.org) and the format (how you write it) are separate things. There are three supported formats: JSON-LD (a <script> block, which Google recommends as the easiest to maintain and least error-prone), Microdata, and RDFa (both interleaved into your HTML). All three work; JSON-LD is the modern default.

Two things structured data does — and one it doesn’t:

• It powers rich results. When search engines can parse your data unambiguously, your listing becomes eligible for enhanced SERP features (star ratings, product cards, breadcrumbs). These can lift click-through rate.
• It helps entity understanding. Properties like sameAs connect your organization or author to authoritative profiles, feeding the Knowledge Graph and helping disambiguate similarly named entities.
• It is not a direct ranking factor. John Mueller has confirmed this repeatedly. Adding schema does not make a page rank higher for a query; any ranking benefit is indirect (higher CTR from rich results).

Eligibility for a rich result is never guaranteed even with correct markup, and the supported types change over time — FAQ and HowTo rich results were sharply restricted in 2023, and a batch of other types was removed in 2025–2026.

Related: Schema Markup, Rich Results, JSON-LD

Changelog