Article Schema

Element Code: RI-004

What Article schema actually is

Article schema is structured data from schema.org that labels a page as editorial content and describes its key facts in machine-readable form: the headline, the author, the publish date, the last modified date, the lead image. You add it as a JSON-LD block in the page, and crawlers read it alongside the visible content.

Three types matter in practice. Article is the generic parent. NewsArticle is for time-sensitive reporting. BlogPosting is for, well, blog posts. Google's article structured data documentation treats all three as valid for article features, so the choice is about honest labeling, not gaming a richer result. My rule: news site gets NewsArticle, blog gets BlogPosting, everything editorial that fits neither gets plain Article.

Why it matters for SEO

Two reasons, one flashy and one quiet.

The flashy one: eligibility for enhanced article presentation. Google uses article markup to help display headline and image treatment in surfaces like Top Stories on mobile and in Google News. Markup alone does not put you there, and non-AMP pages have been eligible for Top Stories since 2021, but clean NewsArticle markup with a proper large image is table stakes if you compete for news visibility at all.

The quiet one: disambiguation. Dates and authorship on the open web are messy. A page can show three different dates in the byline, the footer, and the URL. The datePublished and dateModified properties give Google an explicit, timezone-stamped answer, which feeds the date shown next to your snippet. Same for authorship: a proper author object with a name and a URL pointing to an author page helps Google connect the article to a real person. On sites I have worked on, fixing contradictory date signals through schema was often what stopped Google from displaying a stale or plain wrong date in the SERP.

One thing it will not do: Article schema is not a ranking switch. Do not expect positions to move because you added it. Expect cleaner SERP presentation, correct dates, and eligibility for article features you were locked out of.

Which type should this page use?

The properties Google actually reads

Google lists these as recommended for article rich results. Skipping them will not invalidate the markup, but each one you fill in makes the result more useful.

A working example

This is the shape I ship on client blogs, as JSON-LD in the head or body (both work; Google reads either):

{
  "@context": "https://schema.org",
  "@type": "BlogPosting",
  "headline": "How We Cut Crawl Waste by Fixing Faceted URLs",
  "image": [
    "https://example.com/img/crawl-16x9.jpg",
    "https://example.com/img/crawl-4x3.jpg",
    "https://example.com/img/crawl-1x1.jpg"
  ],
  "datePublished": "2026-05-14T09:30:00+02:00",
  "dateModified": "2026-06-02T11:00:00+02:00",
  "author": {
    "@type": "Person",
    "name": "Jane Doerr",
    "url": "https://example.com/authors/jane-doerr/"
  },
  "publisher": {
    "@type": "Organization",
    "name": "Example Publishing",
    "logo": { "@type": "ImageObject", "url": "https://example.com/logo.png" }
  },
  "mainEntityOfPage": "https://example.com/blog/faceted-urls/"
}

How to detect problems

1. Rich Results Test on a handful of representative URLs. It tells you whether the page is eligible for article features and flags malformed properties.
2. Schema.org validator for pure syntax checking when you care about more than Google.
3. Screaming Frog with structured data extraction and validation enabled. This is how you find template-level problems at scale: every post missing dateModified, every author resolving to "admin", every image below 1200px.
4. Search Console shows parsing errors under the enhancements and structured data reporting once Google has crawled the markup.

The failure modes I see most: dates without timezones, a headline that differs from the on-page H1 because a plugin pulled the SEO title instead, author marked up as a string rather than a Person object, and Article schema slapped on commercial category pages where it flatly does not belong.

What good looks like

Every editorial template emits one Article-type block, generated from CMS fields so it never drifts from the visible content. Dates are ISO 8601 with timezones and match the byline. Authors are Person objects linking to live author pages. The lead image exists in three crops at 1200px or wider. Screaming Frog reports zero validation errors across the archive, and the dates Google shows in the SERP match what you published. Set it up once at the template level and it maintains itself.

FAQ