The mentions and citation Properties

Beyond sameAs, structured data offers two underused properties that create explicit, machine-readable relationship signals: mentions and citation. These properties tell search engines exactly how your content connects to other entities and sources, removing the need for NLP-based inference.

When your blog post about entity SEO uses mentions to tag specific concepts and authorities, you are telling Google exactly how your content connects to the broader knowledge graph. This is the difference between implied and declared relationships.

The mentions Property

mentions is a Schema.org property that declares "this creative work mentions this entity." It can be used on Article, BlogPosting, WebPage, and other CreativeWork types. The value is a reference to another entity.

In JSON-LD, it looks like this:

{
  "@type": "Article",
  "name": "Entity SEO: A Complete Guide",
  "author": {"@type": "Person", "name": "Jane Smith"},
  "mentions": [
    {"@type": "Thing", "name": "Knowledge Graph", "sameAs": "https://www.wikidata.org/wiki/Q18167797"},
    {"@type": "Organization", "name": "Google", "sameAs": "https://www.wikidata.org/wiki/Q95"},
    {"@type": "Thing", "name": "Structured Data", "sameAs": "https://www.wikidata.org/wiki/Q26813700"}
  ]
}

Each mentions entry creates an explicit edge between your content and a recognized entity. When you link to a Wikidata item via sameAs within the mentions entry, you are connecting directly to the knowledge graph.

The citation Property

citation is different from mentions. While mentions says "this content talks about this entity," citation says "this content cites this source." It declares a formal reference relationship, similar to academic citations.

Why These Properties Matter for Recognition

When search engines crawl your content, their NLP models extract entity mentions from the text. This works but is imperfect. The NLP might miss an entity, misidentify one, or fail to distinguish between an entity being mentioned and an entity being the primary topic.

Structured data properties remove this ambiguity. When you explicitly declare that your article mentions "Knowledge Graph" and "Google" and "Structured Data," the system does not need to guess. The relationship is declared in a machine-readable format.

This diagram shows the full signal chain. Your article declares its topic (about), its mentioned entities (mentions), its cited sources (citation), and its author (author). The author entity declares expertise in the same topic (knowsAbout). The system sees a coherent, self-reinforcing graph of relationships.

Implementation Strategy

You do not need to add mentions to every page on day one. Prioritize your most important content pages: pillar content, high-traffic articles, and pages that target your core topic associations.

For each page:

1. Identify every significant entity mentioned in the content (people, organizations, concepts, tools)
2. Find Wikidata URLs for each entity where available (not all entities have Wikidata items)
3. Add mentions entries to the Article or WebPage schema
4. Add citation entries for any research, studies, or authoritative sources you reference
5. Ensure about declares the primary topic
6. Validate with Google's Rich Results Test

Connecting to Wikidata

The most powerful version of mentions connects your content directly to Wikidata entities. When you reference "Knowledge Graph" and include sameAs: "https://www.wikidata.org/wiki/Q18167797", you are creating a direct link between your content and a node in the same knowledge graph that Google's systems consume.

Not every entity you mention will have a Wikidata item. For those that do not, use the entity's official website URL or a widely recognized canonical URL. For those that do, always prefer the Wikidata reference.

Further Reading

• Schema.org: mentions Property (Schema.org)
• Schema.org: citation Property (Schema.org)
• Article Structured Data (Google for Developers)
• Wikidata Main Page (Wikidata)