Use Article, Organization and breadcrumb structured data without conflicting claims

Understand vocabulary, implementation and search eligibility

Google’s introduction to structured data describes machine-readable information about a page. Schema.org supplies a vocabulary; JSON-LD, Microdata and RDFa are ways to express it. Google supports those formats and generally recommends JSON-LD where the site’s setup allows it.

A valid vocabulary statement and eligibility for a particular Google feature are different checks. Schema.org describes more types and properties than Google uses for its documented search features. Use the relevant Google feature guide for Google-specific behaviour, then the vocabulary reference to understand the entities and relationships.

Structured data also needs an owner. In WordPress, a theme, SEO plugin, specialist extension or custom code may already generate it. Inspect a representative public page before adding anything. The plugin dependency audit helps identify which component owns a capability and what would be affected by changing it.

Match Article fields to real editorial information

Google’s Article documentation covers Article, NewsArticle and BlogPosting. Its current guide lists recommended properties rather than a universal set of required Article fields. Choose the type that truthfully describes the content; calling every tutorial a news report is not an improvement.

Check the headline against the displayed title. Use genuine author attribution and a profile URL that identifies that author when available. An organisation can be an author when that accurately represents the authorship. Do not add a fictional editor, professional title or credential to fill an empty field.

Record the actual first publication and meaningful modification dates. Do not refresh the public modification date merely because a routine build ran. Use a representative article image when supplied, rather than substituting the publisher’s logo for every guide. A missing genuine fact is a reason to improve the editorial record, not fabricate one.

A practical acceptance sheet lists each generated field beside its visible counterpart and responsible editor. If the page says one author while the markup names another, resolve the attribution before treating a validation result as success.

Describe the publisher without inventing a business identity

Google’s Organization guide recommends placing organisation information on the homepage or a single descriptive page such as About. It does not require the complete organisation description on every page. Use the most specific subtype that fits the actual organisation.

Keep the name, public URL and relevant logo consistent with the publisher readers can identify. Link only identity profiles that belong to that organisation. A social profile with a similar name is insufficient evidence for a sameAs relationship.

Do not copy registration numbers, addresses, awards, employee counts or service areas from a template. If the website’s public owner identity is unresolved, record that as an editorial decision to settle before release. A tidy machine-readable graph cannot establish ownership or credentials.

Where multiple nodes refer to the same publisher, a stable identifier can connect them. Google’s general guidelines explain using @id to connect related items. Decide a consistent identifier once and check that separate components have not created contradictory publishers under different names.

Make breadcrumbs describe a useful path

The BreadcrumbList vocabulary uses an ordered list of items. The Google breadcrumb guide recommends a typical user path rather than mechanically reproducing every URL segment.

For Google’s documented breadcrumb implementation, use at least two ListItems with the relevant names, ordered positions and destinations. Its last breadcrumb item may omit item; Google then uses the containing page’s URL. Therefore, a missing last-item URL is not automatically the error reported elsewhere in a trail.

Check the visitor-facing path too. A category label should lead to the intended category, and the current article label should identify this article. If an article belongs to several categories, choose a meaningful editorial path or deliberately supported multiple paths. Two different legitimate trails are different from accidental conflicting output.

Do not add nonexistent categories solely to produce longer breadcrumbs. Renaming a section also means checking its navigation label, breadcrumb output and destination together.

{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Organization",
      "@id": "https://example.com/#publisher",
      "name": "Fictional Example Journal",
      "url": "https://example.com/"
    },
    {
      "@type": "Article",
      "@id": "https://example.com/guides/sample/#article",
      "headline": "Fictional sample guide",
      "publisher": { "@id": "https://example.com/#publisher" }
    },
    {
      "@type": "BreadcrumbList",
      "itemListElement": [
        { "@type": "ListItem", "position": 1,
          "name": "Guides", "item": "https://example.com/guides/" },
        { "@type": "ListItem", "position": 2,
          "name": "Fictional sample guide",
          "item": "https://example.com/guides/sample/" }
      ]
    }
  ]
}

Validate syntax, truth and actual search state separately

Google’s testing-tool guidance distinguishes the Rich Results Test, which checks supported Google features, from the Schema Markup Validator, which checks schema.org markup without Google-specific validation. Use the tool that answers the question and read the exact entity and property behind an error.

Google’s general policies require relevant, accurate information representing visible content. They also explain that a correct test result does not guarantee search appearance. Never add made-up reviews, ratings or affiliations to seek stars.

Keep a private trial private. Do not remove login protection or noindex from staging to satisfy a public-URL test; review code privately through an approved process. When an authorised public release occurs, check the served output and Google’s view of representative URLs. Our launch checklist covers the wider release, and Search Console basics explains subsequent observation. Record validation, discovery and displayed search features as different outcomes.