Structured data and schema markup: what still earns rich results in 2026

Structured data uses the schema.org vocabulary to describe a page's content to search engines, which makes the page eligible for a rich result — an enhanced listing like a review star, product price, or video thumbnail. Eligibility is never a guarantee, and in 2026 several once-popular types no longer produce any rich result at all.

What structured data does — and does not — do

Structured data does not directly raise rankings. It makes a page eligible for enhanced SERP features when Google judges them appropriate. Google recommends JSON-LD over Microdata or RDFa because it is the least error-prone format; JavaScript-injected JSON-LD works as long as it renders into the DOM. The old data-vocabulary.org format is dead and produces no feature.

The general guidelines are enforced. Violating them can trigger a manual action that strips your rich-result eligibility. In short:

• Mark up content that is visible to users — never invisible, fake, or misleading data.
• Use the most specific type that fits (Restaurant, not the generic LocalBusiness).
• Prefer fewer complete, accurate properties over many half-filled ones.
• Keep the markup on the page it describes, and make sure that page is crawlable and indexable — do not block it via robots.txt, noindex, or a login.

Which types still earn a rich result in 2026

These types currently produce a visual rich result or a recognized enhancement: Article / NewsArticle / BlogPosting (Top Stories eligibility), Breadcrumb, Product (product snippet and merchant listings), Review snippet / AggregateRating, VideoObject, Recipe, Event, JobPosting, and LocalBusiness, along with Course, Discussion forum, Q&A, Software app, Speakable, Movie, Profile page, and image metadata. Organization and WebSite markup do not draw a visual result but feed Google's entity and knowledge understanding.

Deprecated and retired types — stop relying on these

Some types that dominated SEO advice for years no longer do anything in Search. Investing in them for SERP gains is wasted effort.

Type	Status	Date
FAQPage	Restricted to government/health sites, then removed entirely	Restricted Aug 2023; stopped showing 7 May 2026 (report and Rich Results Test support gone by mid-to-late 2026)
HowTo	Removed from Search entirely	14 September 2023
Practice Problem (Quiz)	Support removed	January 2026
Sitelinks Searchbox (WebSite + SearchAction)	Retired; markup is inert	November 2024
data-vocabulary.org	No feature at all	—

Google's own position: leftover or unsupported markup causes no errors and no ranking harm, so there is no urgency to rip it out — it may still feed AI entity understanding. But stop investing in these types expecting SERP features. If you have been maintaining hand-written FAQ or HowTo schema for rich snippets, that budget is now better spent elsewhere.

How to validate schema markup

Google retired its long-standing Structured Data Testing Tool, which is why so many people search for a replacement. Use these instead:

• Rich Results Test — checks Google eligibility and previews the result. This is the authoritative check for "will this earn a rich result?"
• Schema.org validator — validates the generic vocabulary regardless of what Google supports.
• URL Inspection in Search Console — confirms what Google actually found and rendered on the live URL.
• Rich result status reports in Search Console — monitor after deploying, and allow several days for recrawl.

A common trap: markup that validates in a generic checker but earns nothing, because the type is one of the deprecated ones above. Validating syntax is not the same as confirming eligibility.

A minimal, correct Article example

{
  "@context": "https://schema.org",
  "@type": "BlogPosting",
  "headline": "Structured data and schema markup in 2026",
  "author": { "@type": "Person", "name": "Jane Doe" },
  "datePublished": "2026-07-12T09:00:00-04:00",
  "image": ["https://example.com/cover-1200.jpg"]
}

Notes that trip people up: headline should stay within about 110 characters, images should be at least 1200px wide (ideally in several aspect ratios), dates use ISO-8601 with a timezone, and author.name is a plain name — no titles or honorifics.

Structured data and AI search

Valid schema does not demonstrably increase AI citation rates. Controlled experiments (Ahrefs, 2026: 1,885 pages against 4,000 controls) found no statistically meaningful lift in AI Overviews, ChatGPT, or AI Mode citations from adding structured data. The correlations between schema-marked pages and citations reflect overall site quality, not schema causation. Google's AI optimization guide (updated 10 July 2026) is explicit: "there are no additional requirements" and "no special schema.org markup for AI." Schema earns rich results in standard Search — that is its job. Do not build AI-specific schema expecting citation lift. For the full picture, see the GEO/AEO guide.

How Crawlinx helps

Crawlinx checks structured data site-wide on a single crawl, which catches the two failures Google-hosted single-URL testers miss: systemic gaps and errors repeated across templates.

• schema.invalid — flags malformed JSON-LD and missing required properties before they cost you eligibility.
• schema.none — surfaces page types (articles, products, videos) that have no structured data where a supported type would apply.
• video.contenturl_self_host — checks that VideoObject markup points at a real, crawlable video URL.
• mobile.structured_data_missing — catches markup present on desktop but absent on the mobile version, which is the version Google indexes.
• images.missing_alt — supports the visible, accurate content that structured data is supposed to describe.

Takeaway

Keep valid JSON-LD on the types that still earn results — Article, Product, Review, Video, Recipe, Event, JobPosting, LocalBusiness — and stop building FAQ, HowTo, and Quiz markup for SERP features, since Google removed all three. Validate with the Rich Results Test, and audit the whole site for systemic errors rather than checking one URL at a time.

Related reading: Core Web Vitals and GEO/AEO for AI search.