schemas

Structured Data & Schema Markup: 2026 Guide

How to implement JSON-LD schema markup in 2026: rich results, AI citations, deprecated types, validation tools, and CTR impact data.

Only 12.4% of all registered domains use Schema.org markup — yet pages with structured data earn 35% higher click-through rates and are cited up to 3x more often in AI-generated answers than unstructured pages (Digital Applied, ALM Corp). That gap is your opportunity.

The March 2026 Google core update fundamentally changed why schema matters. Rich results are still real, but the bigger prize is now AI trust infrastructure — getting cited inside AI Overviews, Google's AI Mode, and third-party LLMs.

Quick answer:

JSON-LD is the only format Google recommends. Implement Organization, Article, Product, and BreadcrumbList schema first. FAQ and HowTo schema no longer generate rich results for most sites after August 2023 / March 2026. Validate with Google's Rich Results Test and validator.schema.org. The 2026 priority is accurate, entity-linked schema — not schema volume.

Why Schema Matters More in 2026

Structured data has moved from "earn a star rating in SERPs" to "get cited by AI systems." Here is what the data shows:

65% of pages cited by Google AI Mode include structured data (SE Ranking via NAV43)
71% of pages cited by ChatGPT include structured data (SE Ranking via NAV43)
Pages with schema see a 3.2x citation lift in Google's AI Mode (Digital Applied)
97% of AI Overview citations come from pages in the top 10 organic results (seoClarity)
Google's Knowledge Graph holds 500 billion facts about 5 billion entities — schema is how you get into it (Tonic Worldwide)

Rich result CTR gains are also still measurable:

Product schema: +40–60% CTR
Article / BlogPosting schema: +15–25% CTR
BreadcrumbList: +5–10% CTR
Rich results overall capture 58% of clicks vs 41% for plain blue links (Digital Applied)

JSON-LD vs. Microdata vs. RDFa

Google explicitly recommends JSON-LD over Microdata and RDFa (Google Search Central). JSON-LD adoption on web pages grew from 34% in 2022 to 41% in 2024 and continues rising; Microdata has not grown in efficacy over the same period (Moz).

Format	Google recommendation	Separation from HTML	Maintainability
JSON-LD	Preferred	Clean — separate `<script>` tag	High — edit without touching markup
Microdata	Supported	Embedded in HTML attributes	Low — tightly coupled to layout
RDFa	Supported	Embedded in HTML attributes	Low — verbose, error-prone

Use JSON-LD. Every tool, CMS plugin (Yoast, RankMath, Schema App), and code generator outputs it. Microdata and RDFa are legacy formats — stop using them for new implementations.

Basic JSON-LD structure

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Acme Corp",
  "url": "https://acme.example.com",
  "logo": "https://acme.example.com/logo.png",
  "sameAs": [
    "https://www.linkedin.com/company/acme",
    "https://www.wikidata.org/wiki/Q12345"
  ],
  "contactPoint": {
    "@type": "ContactPoint",
    "telephone": "+1-555-123-4567",
    "contactType": "Customer Service"
  }
}

Place JSON-LD inside a <script type="application/ld+json"> tag in the document <head>. Body placement works too but head is preferred.

The @id property: entity graph linking

The most underused JSON-LD technique is @id. It assigns a stable canonical URI to each entity so you can cross-reference it from every page on your site.

Define your Organization once on the homepage with "@id": "https://example.com/#organization"
Every Article page references it via "publisher": { "@id": "https://example.com/#organization" }
Without @id, Google may treat the same organization on different pages as separate entities, fragmenting authority signals

{
  "@type": "Article",
  "headline": "How to Implement Schema Markup",
  "author": {
    "@id": "https://example.com/#person-jane-doe"
  },
  "publisher": {
    "@id": "https://example.com/#organization"
  }
}

Schema Types: Priority Tiers for 2026

Tier 1 — Foundation (implement on every site)

Schema type	Where	Key properties	Impact
Organization	Homepage	`name`, `url`, `logo`, `sameAs`, `knowsAbout`	Knowledge Panel, entity recognition
Person	Author pages	`name`, `jobTitle`, `sameAs`, `knowsAbout`, `worksFor`	E-E-A-T signals, AI citation
WebSite	Homepage	`name`, `url`, `potentialAction` (Sitelinks search box)	Sitelinks search box
BreadcrumbList	All interior pages	`itemListElement` with `ListItem` items	+5–10% CTR in SERPs

Organization sameAs tip: Link to Wikidata first — it is the primary input to Google's Knowledge Graph. Also link to LinkedIn, Wikipedia (if applicable), and official social profiles (Tonic Worldwide). Sites with comprehensive Organization schema are 3.7x more likely to earn Knowledge Panels (ALM Corp).

Tier 2 — High-impact rich results

Schema type	CTR uplift	Required properties	Notes
Product	+40–60%	`name`, `offers` (price, currency, availability), one of: `aggregateRating` or `review`	Add `gtin` for Google Shopping eligibility
Article / BlogPosting	+15–25%	None required — all recommended	Include `author`, `datePublished`, `image` (min 50K pixels)
LocalBusiness	Varies	`name`, `address`, `telephone`	Critical for map pack; link `@id` to Organization
Event	Varies	`name`, `startDate`, `location`	Eventbrite saw 100% YoY traffic increase after implementation (Google Search Central)
VideoObject	Varies	`name`, `thumbnailUrl`, `uploadDate`	Add `transcript` for AI parsing

Tier 3 — Niche and emerging

Service / ProfessionalService — for B2B firms; underutilized despite strong AI search value (NAV43)
SoftwareApplication — SaaS tools; drives ratings and pricing in rich results (schema.org)
Recipe — high rich-result display rate; requires name, image, recipeIngredient
JobPosting — drives job pack appearances; jobDuration property added in Schema.org v30.0 (2026)

What Changed in 2025–2026

FAQ and HowTo are effectively dead for most sites

This is the most important deprecation of the past two years:

August 2023: Google restricted FAQ rich results to government and health sites only (Google Search Central Blog)
September 2023: HowTo rich results removed from desktop; mobile only
March 2026: Desktop HowTo rich results removed entirely; mobile limited to primary content pages; FAQ eligibility narrowed further (Digital Applied)
47% drop in FAQ rich result impressions across tracked sites post-March 2026 (Digital Applied)

Verdict: If you are not a government or major health publisher, FAQPage and HowTo schema produce no SERP display benefit. They are not penalized, but they are wasted markup for most sites. Remove them from your template strategy.

Schema types deprecated in June 2025

Google announced on June 12, 2025 that these types no longer generate rich results (Google Search Central):

Book Actions
Course Info
Estimated Salary
Learning Video
Special Announcement
Vehicle Listing
Claim Review (for SERP display — still valid for AI citation)

These were removed from Search Console reports and the Rich Results Test starting September 2025.

The March 2026 penalty risk

The update also introduced active algorithmic demotion for misleading schema:

Review schema on editorial comparison posts without genuine user reviews → manual action risk
Schema type that doesn't match the page's primary content purpose (e.g., Article on a product landing page) → now classified as misleading markup
Inflated AggregateRating with unverifiable review counts → enhanced scrutiny

Google's policy is explicit: marking up hidden content or content that doesn't match what users see violates structured data guidelines (Google Search Central).

Schema.org vocabulary updates 2025–2026

The Schema.org vocabulary is updated regularly. Key additions:

v30.0 (March 2026): floorLevel for LocalBusiness/Residence; jobDuration for JobPosting; EU Digital Product Passport examples; new Credential and Error classes
v29.4 (December 2025): ConferenceEvent, PerformingArtsEvent, SequentialArt (replaces deprecated GraphicNovel), lifeEvent for Person
v29.2 (May 2025): companyRegistration, legalAddress, legalRepresentative for Organization; pronouns for Person
v29.0 (March 2025): wordCount for CreativeWork — now useful for AI content depth signals

Full release history at schema.org/docs/releases.html.

Structured Data for AI Overviews and LLMs

This is the most important new use case for schema in 2026.

AI systems — Google AI Mode (Gemini), ChatGPT, Perplexity — use structured data as a trust signal during answer synthesis. They read schema to verify claims, establish entity relationships, and assess source credibility. When your schema accurately represents your content, AI systems are more likely to cite you even when no traditional rich result is displayed.

Key findings from 2025–2026 research:

Pages with FAQPage schema achieve 3.2x higher citation rates in AI Overviews — even though FAQPage no longer generates SERP rich results for most sites (Demand Local)
73% higher selection rate in AI Overviews for pages with schema (Demand Local)
LLMs grounded with structured data achieve 300% higher accuracy than those using unstructured data alone (ALM Corp)
Both Google and Microsoft confirmed (Spring 2025) that structured data is critical for generative AI features because it is "efficient, precise, and easy for machines to process" (Tonic Worldwide)

Properties that help AI specifically

knowsAbout on Organization and Person — signals topical authority to AI systems
sameAs linking to Wikidata — directly feeds Google's Knowledge Graph
Speakable — flags citable passages for AI synthesis (currently English news content only)
DefinedTerm — improves AI Mode citation for definition queries
wordCount (v29.0) — provides content depth signal

For more on entity-based optimization strategy, see Entities & Semantic SEO.

Validation Workflow

Use these four tools in order:

Schema.org Validator — check syntax, catch deprecated properties, validate against the full vocabulary
Google Rich Results Test — confirm which rich result types your page qualifies for
Google Search Console → Enhancements — monitor errors, warnings, and impressions by schema type across the full site
Manual SERP spot checks — verify rich results are actually appearing post-deployment

Common errors that kill rich results

Missing name property (most common critical error)
Invalid JSON syntax — missing commas, unclosed brackets, smart quotes instead of standard double quotes
Duplicate Organization schema from competing CMS plugins
Stale price or availability in Product schema
ratingValue formatted with a comma instead of a dot (4,4 instead of 4.4)
author.name exceeding 100 characters (blocks Review snippet eligibility)
Incomplete required properties — e.g., missing location for Event, missing offers.price for Product

Quarterly maintenance cadence

Schema drifts over time as pages change. Build these into your workflow:

Check Search Console Enhancements reports monthly
Re-validate after any major CMS update or template change
Audit Product schema prices and availability data quarterly — stale data suppresses rich results (Digital Applied)
Cross-reference Wikidata entity entries if you have a Knowledge Panel

For enterprise sites, build schema validation into your CI/CD pipeline — generate schema from your CMS data model and validate it before deployment.

Common Mistakes to Avoid

Marking hidden content — only add structured data for information visible to the user. Violates Google's guidelines and risks manual action (Google Search Central)
Self-serving reviews — if the entity being reviewed controls the reviews (e.g., a business rating its own products), the page is ineligible for star rating rich results (Google Search Central)
FAQPage on every page — provides no rich result benefit for non-government/health sites post-2023; wastes crawl budget post-2026
Multiple conflicting schema blocks — two Organization schemas on the same page (often from theme + plugin conflicts) create fragmented entity signals
Using schema without @id — the most impactful missed optimization; without it, entities across pages are treated as unrelated
Ignoring Schema.org version updates — GraphicNovel is deprecated, DeliveryTimeSettings replaced — old schema patterns can generate validation errors

Technical SEO connection

Structured data is one layer of a broader Technical SEO foundation. Schema tells Google what your content means; crawlability, Core Web Vitals, and internal linking tell Google how to find and evaluate it. Both layers need to be healthy for rich results to appear consistently.

Frequently asked questions

What is the difference between structured data, schema markup, and JSON-LD?

Structured data is the broad concept — machine-readable code that describes page content. Schema markup refers specifically to using Schema.org vocabulary to define entities and their properties. JSON-LD (JavaScript Object Notation for Linked Data) is the implementation format Google recommends for delivering schema markup — it lives in a <script> tag separate from your HTML.

Does schema markup directly improve rankings?

Not directly. Google has confirmed schema is not a direct ranking factor. However, it influences search performance indirectly: rich results improve CTR, better CTR sends positive behavioral signals, and entity-linked schema helps Google understand your content deeply enough to surface it for more relevant queries. The AI citation benefit is also real and growing.

What happened to FAQ schema — is it worth implementing?

FAQ schema no longer generates rich results for most websites. Google restricted FAQ rich results to government and health sites in August 2023 (Google Search Central) and tightened eligibility further in March 2026. For non-government sites, FAQPage markup produces no SERP display benefit. It may still support AI citation (pages with FAQPage schema show higher citation rates in AI Overviews), but that benefit is uncertain and unmeasured for most verticals.

How do I validate my schema markup?

Use two tools together: validator.schema.org for syntax and vocabulary correctness, and Google Rich Results Test for rich result eligibility. After deployment, monitor Google Search Console's Enhancements section for site-wide error counts. Do manual SERP spot checks 24–48 hours after deploying new schema types.

Which schema type has the highest ROI?

For e-commerce: Product schema — up to +60% CTR and 4.2x higher likelihood of appearing in Google Shopping results (ALM Corp). For content sites: Article/BlogPosting schema combined with Person author schema — supports AI Mode citation and Top Stories eligibility. For all sites: Organization schema with sameAs and knowsAbout — the foundation for Knowledge Panel eligibility and AI entity recognition. InSinkErator saw a 69% increase in non-branded clicks after implementing entity linking in their Product schema (Schema App).

Originally published in the EcomExperts SEO library.