Schema Markup in 2026: Which Structured Data Types Actually Move Rankings (And Which Are Wasted Effort)

The Schema Advice You've Been Hearing Is Half Wrong

For years, the SEO industry had one answer to every structured data question: "Add more schema." FAQ schema. HowTo schema. Article schema. LocalBusiness schema. BreadcrumbList. Slap it all in, watch the rich results roll in.

That advice has aged poorly. Google deprecated HowTo rich results entirely in September 2023. FAQ rich results were pulled for most site categories in May 2026 — leaving only government and health sites eligible. Six additional structured data types lost SERP visibility in June 2025. And a 1,900+ page study from SEOAuthori (May 2026) found that AI engines don't actually read your JSON-LD at all during page retrieval — they process visible HTML content.

So is schema dead? Absolutely not. The schema types that unlock real SERP features — star ratings, pricing, event listings, site search, author bios — still deliver measurable CTR lifts. And schema remains critical for knowledge graph entity resolution and search engine comprehension, even when it doesn't produce a visible rich result.

The difference is that "deploy everything" is no longer good advice. In 2026, schema is high-leverage on specific page types and genuinely wasted effort on others. This guide tells you exactly which is which.

What Just Changed: The Deprecation Timeline

Before we talk about what works, you need to understand what's been pulled — because a lot of sites are still wasting developer time implementing types that no longer do anything visible in search.

HowTo Rich Results: Gone (September 2023)

Google first moved HowTo rich results to desktop-only in August 2023, then deprecated them entirely from desktop on September 14, 2023. Seer Interactive documented the rollout in real time. If you're still adding HowTo schema to your pages hoping for step-by-step rich results in the SERP, stop — they don't appear anymore.

Should you still use HowTo schema? Only if your content genuinely walks through steps and you want machine comprehension of the structure. It will not produce a visible rich result, but it does help crawlers understand the page's intent. Low-priority implementation.

FAQ Rich Results: Restricted (May 2026)

This one blindsided a lot of agencies. In May 2026, Google restricted FAQ rich results to government and health/medical sites only. For everyone else, FAQ rich result impressions collapsed within weeks of the rollout. The schema type itself (FAQPage) is still valid and still documented by Google — but for most businesses, it will no longer produce the accordion dropdown in search results.

Should you still use FAQPage schema? Yes — but for AI citation and passage retrieval, not SERP feature visibility. FAQPage schema structures your Q&A content so AI systems can extract it cleanly. SearchScore's GEO analysis ranks FAQPage among the top five schema types for AI citation because AI engines can extract Q&A pairs as ready-made answer sources. The rich result is gone; the comprehension value remains.

Six Additional Types Pulled (June 2025)

The HOTH documented that Google removed SERP visibility for six more structured data types in June 2025. The pattern here is deliberate: Google is simplifying its search result types, retiring features with low user engagement, and concentrating rich result eligibility on the types that demonstrably drive clicks and user satisfaction.

The Truth About Schema and AI Citations

A persistent myth in the GEO space is that loading up your pages with schema will improve your AI citation rate. The data tells a more complicated story.

The SEOAuthori 1,900-page study (May 2026) tested five AI search systems — including ChatGPT, Perplexity, and Google AI Overviews — and found that none of them extracted information directly from JSON-LD during live page retrieval. Every system processed visible HTML content. Information present only in structured data markup was consistently missed.

This is actually important: it means your citation-critical content must live in your visible page copy, not just in schema. A claim buried in JSON-LD that isn't also written in readable prose won't get cited.

However, schema still influences AI citation indirectly — because it influences whether you rank in the first place, and whether Google includes you in AI Overviews. A Link Building Journal study of 1,000 AI Overviews found schema markup was the strongest single page-level predictor of citation inclusion — but through ranking correlation, not direct AI parsing of JSON-LD.

Translation: schema helps you rank, and ranking gets you cited. The chain is real; just don't expect your FAQPage JSON-LD to be read verbatim by ChatGPT.

The Schema Types That Actually Work in 2026

Now for the actionable part. Here's the definitive breakdown of which schema types deliver measurable value in 2026, why each one works, and how to implement them correctly.

Tier 1: Deploy on Every Site (High Leverage)

Organization (or LocalBusiness)

This is the foundation. Organization schema tells Google and AI systems who you are as an entity: your name, logo, contact information, social profiles, and founding details. It feeds directly into your Knowledge Panel and helps resolve your brand as a distinct entity in Google's knowledge graph.

For businesses with a physical location, use LocalBusiness (or an even more specific subtype like Plumber, Dentist, or Restaurant) instead of the generic Organization type. According to SEOAuthori's JSON-LD guide, LocalBusiness markup directly influences Knowledge Panel content and local pack eligibility.

Required fields for LocalBusiness: name, image, address (full PostalAddress with street, city, state, postal code, country), telephone, openingHours, geo (latitude/longitude), and url.

Recommended additions: priceRange, aggregateRating, sameAs (your GBP URL, social profiles, Yelp page). The sameAs array is especially important for entity resolution — it tells Google that your website, Google Business Profile, and LinkedIn page all describe the same entity.

WebSite with SearchAction

WebSite schema with a SearchAction property enables the Sitelinks Search Box — the search bar that appears directly in your brand's search result when someone searches your company name. This is a high-value rich result that's still active and relatively easy to implement correctly.

The WMTips semantic markup analysis found SearchAction used by 38.1% of sites with advanced schema — making it one of the more widely deployed active features.

Article (or BlogPosting)

For every piece of editorial content — blog posts, guides, news articles — Article schema is essential. It signals to Google that this page is original content with a publication date, an author, and a publisher. This matters for:

• News/Discover feed eligibility
• Featured snippet candidacy
• Author entity association (connecting the article to your Person schema)
• Freshness signals — dateModified tells Google when you last updated the piece

Critical implementation note: Use BlogPosting (a subtype of Article) for blog content. Always include datePublished, dateModified, author (as a Person object with a @id pointing to your author page), and publisher (as an Organization object). Crawlix's 2026 schema error analysis found the most common validation failure is using a flat string for author instead of a proper Person object — Google rejects it.

BreadcrumbList

BreadcrumbList schema produces the breadcrumb trail that appears in search results beneath your page title (e.g., "Home > Blog > Technical SEO"). It's a clean, low-effort implementation that improves click visual hierarchy and signals page structure to crawlers. Pair it with every Article page.

Tier 2: Deploy on Relevant Pages (High ROI on Eligible Content)

Product (with Offer and AggregateRating)

If you sell products, Product schema is the single highest-leverage schema implementation available. A SuperData SEO 2026 study found pages with complete Product markup (including offers, reviews, and price) saw rich results appear for 34% of monitored queries, with average CTR jumping 18% where star ratings showed. An e-commerce site case study documented by SEO Francisco found a 52% CTR lift and 41% more organic clicks across 8,400 pages after complete Product schema deployment.

Required for rich results: name, image, description, sku, offers (with price, priceCurrency, availability), and aggregateRating (with ratingValue and reviewCount).

Review (and AggregateRating)

Star ratings are one of the few schema types that universally lift CTR when they appear. They apply to products, services, local businesses, and recipes. If your site legitimately collects reviews and you're displaying them on the page, implement AggregateRating. The visual trust signal in the SERP — five gold stars next to your result — consistently outperforms text-only results in click share.

Critical rule: AggregateRating must reflect genuine reviews displayed on the page. Google has issued manual penalties for sites that marked up ratings that weren't visible to users or that were inflated. The review data in your schema must match what's on the page.

Person (for Authors)

Person schema attached to your authors is the E-E-A-T signal that most sites skip entirely. A Person entity with verifiable credentials — jobTitle, knowsAbout, sameAs (linking to LinkedIn, a Google Scholar profile, industry association pages) — tells Google that your content was written by a real expert with a verifiable track record.

Link Building Journal's 2026 schema cheatsheet ranks Organization + Person as the highest-leverage, most underused schema stack — the "trust foundation" that makes every other schema type more credible by association.

FAQPage (for AI and Passage Retrieval)

FAQ rich results are gone for most sites — but FAQPage schema still has measurable value for AI citation and passage retrieval. SearchScore's GEO analysis found it's the most direct citation benefit type because AI systems can extract Q&A pairs verbatim as answer sources. If you write FAQ sections (and you should for every article), mark them up.

Correct format: Each question must be a Question object with a name (the question text) and an acceptedAnswer containing an Answer object with text. A flat string in acceptedAnswer is the second most common validation error in Search Console.

Tier 3: Situational (Implement When Directly Applicable)

• Event — Active rich result type. Use for any event with a date, location, and registration link. High CTR impact for event discovery.
• JobPosting — Still produces rich results in Google Jobs. High value if you're hiring.
• Recipe — Very active rich result type for food content. Image carousels, cook time, ratings — high visual impact.
• VideoObject — Use on pages with embedded video. Enables video rich results in the SERP including thumbnail, duration, and description.
• SiteLinksSearchBox — Enables inline search in brand SERPs. Worth implementing for any site with 50+ pages.

Tier 4: Skip or Low Priority

• HowTo — Deprecated from SERP visibility. Machine comprehension only.
• SoftwareApplication — App store install buttons have limited real-world SERP visibility in 2026.
• Speakable — Designed for voice assistants; limited deployment support from Google.

The Technical Implementation That Most Sites Get Wrong

Even when businesses implement the right schema types, they frequently make technical errors that prevent rich results from appearing. Here are the most common mistakes — and how to avoid them.

Mistake 1: Flat Strings Instead of Objects

The single most common validation error in Google Search Console's Enhancements report is using a plain string where Google expects an object. The most frequent offender:

// WRONG — Google rejects this
"author": "Michael Heckert"

// CORRECT — proper Person object
"author": {
  "@type": "Person",
  "name": "Michael Heckert",
  "url": "https://yoursite.com/about/",
  "sameAs": ["https://www.linkedin.com/in/yourprofile/"]
}

The same applies to publisher, address, and acceptedAnswer. Always nest objects properly.

Mistake 2: Incorrect Date Formats

Crawlix's 2026 error analysis found date field errors are among the top 15 most common schema mistakes. Google requires ISO 8601 format for all date fields. "July 14, 2026" is a human-readable string, not a valid date — it will be flagged as "Invalid object type for field datePublished" in Search Console.

// WRONG
"datePublished": "July 14, 2026"

// CORRECT
"datePublished": "2026-07-14"

Mistake 3: Schema Content That Doesn't Match the Page

Google explicitly checks that your structured data reflects what's visible on the page. Marking up star ratings that don't appear in the visible content, or inflating review counts, triggers manual actions. Your aggregateRating values must match what users can see. Your description in Article schema must reflect the actual article summary.

Mistake 4: Multiple Disconnected Schema Blocks

Many sites add schema by copying separate snippets for Organization, Article, and BreadcrumbList — resulting in three independent JSON-LD blocks that don't reference each other. The correct approach is to use a single @graph array where entities reference each other by @id:

{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Organization",
      "@id": "https://yoursite.com/#organization",
      "name": "Your Company",
      "url": "https://yoursite.com"
    },
    {
      "@type": "BlogPosting",
      "@id": "https://yoursite.com/blog/article-slug/#article",
      "headline": "Article Title",
      "author": { "@id": "https://yoursite.com/about/#person" },
      "publisher": { "@id": "https://yoursite.com/#organization" }
    }
  ]
}

Cross-referencing entities via @id signals the relationships between them — author belongs to organization, article belongs to site — which strengthens knowledge graph entity resolution.

Mistake 5: Not Validating Before Publishing

Always run new schema through Google's Rich Results Test and the Schema.org Validator before deploying. As the 2026 Web Almanac analysis noted, running both tools is now essential — they catch different classes of errors. The Rich Results Test checks Google's requirements; the Schema.org Validator checks spec compliance.

Schema Priority by Site Type: A Quick Reference

Different business types have different schema priorities. Here's where to focus your implementation effort:

Site Type	Priority Schema Types	Expected SERP Benefit
Local Service Business	LocalBusiness (specific subtype), Person, AggregateRating, FAQPage	Knowledge Panel, local pack eligibility, star ratings
E-Commerce	Product + Offer, AggregateRating, BreadcrumbList, Organization	Price/rating rich results, 18–52% CTR lift on eligible pages
Content / Blog	Article/BlogPosting, Person, BreadcrumbList, FAQPage, Organization	Author rich results, featured snippet eligibility, AI citation
SaaS / Service	Organization, WebSite + SearchAction, Person, Article, FAQPage	Sitelinks Search Box, Knowledge Panel, brand authority
Restaurant / Hospitality	Restaurant, Menu, AggregateRating, Event	Menu rich results, star ratings, event visibility

How to Audit Your Existing Schema

If your site already has schema deployed, there's a good chance it contains errors — either from outdated implementations, template bugs, or deprecated types you can now remove. Here's the audit process:

1. Open Google Search Console → Enhancements. Every schema type with errors is listed separately. Export the URL list for each error type.
2. Focus on errors before warnings. Errors prevent rich results entirely. Warnings reduce eligibility but don't eliminate it.
3. If an error originates from a template (CMS theme, Shopify scheme, WordPress plugin), fixing the template resolves all affected pages at once — prioritize these.
4. Test the fixed markup in both the Rich Results Test and Schema.org Validator before redeploying.
5. Remove deprecated types that serve no purpose. HowTo schema that will never trigger a rich result adds maintenance overhead for zero benefit.
6. Check that entity cross-references work. If your Article schema references an author @id, verify that ID resolves to a Person schema block on your about or author page.

The Long Game: Schema as Entity Infrastructure

Here's the frame that changes how you think about structured data: schema is not primarily about getting rich results. Rich results are the short-term payoff. The long-term value is entity infrastructure — establishing your organization, your people, your content, and your locations as distinct, machine-readable entities in Google's knowledge graph.

Every time you correctly implement Organization schema with verified sameAs references, you're strengthening Google's confidence that your website and your Google Business Profile and your LinkedIn page are all the same entity. Every Person schema on your authors ties their expertise to your domain. Every Article schema ties your content to its author and publisher.

These connections compound. A well-structured entity graph means your brand shows up in Knowledge Panels, your authors get attribution in rich results, and your content ranks for entity-adjacent queries your competitors aren't ranking for. The businesses that build this infrastructure now — when fewer than 40% of eligible pages have correct schema — will have a significant structural advantage in 18 months.

Schema markup in 2026 isn't about chasing rich result formats. It's about giving search engines and AI systems a clean, accurate, cross-referenced model of who you are, what you do, and why you're credible. That's the work that lasts.

Frequently Asked Questions

Does schema markup directly help AI engines like ChatGPT and Perplexity cite my content?

Not directly. A 2026 study of 1,900+ pages found that AI engines process visible HTML content during retrieval — not JSON-LD. Schema markup helps you rank better in search engines, and ranking correlates with AI citation. But your citation-critical content must live in readable prose on the page, not only in structured data. Schema is an indirect AI citation signal, not a direct one.

Are FAQ rich results gone forever?

For most businesses, yes — FAQ rich results were restricted to government and health/medical sites in May 2026. However, FAQPage schema is still worth implementing for AI citation and passage retrieval purposes. The SERP feature is gone; the content comprehension benefit remains.

What's the most important schema type for a local service business?

LocalBusiness (or its specific subtype — Plumber, Electrician, Dentist, etc.) is the highest-priority schema for local businesses. It feeds your Knowledge Panel, reinforces your Google Business Profile signals, and establishes your entity in local search. Add aggregateRating if you have legitimate reviews, and sameAs linking to your GBP, social profiles, and any citation sites.

How do I know if my schema is working?

Check Google Search Console under Enhancements — each schema type with valid implementations will have its own report showing impression data for rich results. If a type shows zero impressions but no errors, Google may have chosen not to show the rich result for that content. If a type shows errors, fix those first. The Rich Results Test also gives real-time validation feedback for any URL.

Should I use JSON-LD, microdata, or RDFa?

JSON-LD for everything. Google recommends it, it's the easiest to maintain (it lives in a single script block separate from your HTML), and it's the most widely parsed format. Microdata requires embedding attributes throughout your HTML and becomes a maintenance nightmare. RDFa has marginal adoption and limited tooling support. JSON-LD in the <head> is the universal best practice in 2026.

How often should I update my schema?

Review your schema whenever you update prices, hours, personnel, or major page content. For Article schema, always update dateModified when you refresh content — it's a freshness signal Google uses for ranking recency queries. Check your Search Console Enhancements report quarterly for new error types, especially after CMS or theme updates that may have modified your template-level schema.