Discussion
  • AI Search
  • Entity SEO

Are about and mentions properties on articles worth adding? Any evidence they influence topic understanding in Google or AI answers?

Published 1 reply

Discussion starter: an entity SEO question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.

Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.

1 reply

  • Niraj Raut
    Niraj Raut
    Original poster

    They are optional and low priority. The about and mentions properties let an article’s structured data state which entities it is primarily about and which it mentions, often by linking to identifiers such as Wikidata items. Google does not document a search feature for them, and I am not aware of public evidence that adding them changes rankings, topic understanding in Google or AI citations in a measurable way. Google’s May 2026 guide to optimising for its generative AI features says special schema is not needed for those features. If your CMS already tags articles with well-defined entities, generating these properties automatically costs little and is harmless when accurate. Adding them manually, or treating them as a lever, is not a good use of time.

    What the properties do

    Property Meaning Example
    about The main subject or subjects of the article An article about a drug links to that drug’s entity
    mentions Entities referred to but not central The same article mentions a condition and a regulator

    Why the evidence is thin

    Google can already identify entities in text using its own systems, and its understanding of what a page is about comes mainly from the content, headings, links and context. Declaring entities in markup duplicates information Google can usually infer. Because no search feature uses these properties, there is nothing visible to measure, and public tests that isolate their effect have not been published. AI systems that retrieve and summarise text mostly work from the visible content, and many extraction pipelines discard structured data.

    When they may be worth adding

    • Your CMS already has an entity tagging system linked to stable identifiers.
    • Your content covers ambiguous entities, where clarifying which one you mean has some value for machine readers.
    • You publish structured data for other consumers, such as partners or internal knowledge systems, that use these properties.
    • The markup can be generated and validated automatically.

    When to skip them

    • You would need to add them by hand.
    • Your tagging is inconsistent, so the markup would be inaccurate.
    • Higher-value structured data, such as Product, Organization or Article basics, is incomplete or has errors.

    Hypothetical example:

    A health publisher’s CMS tags articles with conditions and treatments linked to Wikidata identifiers. A developer adds about and mentions to Article markup in a day, generated from the tags. Six months later, the publisher compares articles with and without the properties, as some older articles had never been tagged. Rankings, impressions and AI citations show no clear difference. The markup stays, because it is accurate and costs nothing to maintain, but the team stops describing it as an SEO initiative.

    Watch for this:

    Overloading mentions with every entity a tool detects, or declaring entities the article does not really cover, makes the markup inaccurate. Structured data should describe the page truthfully; inaccurate markup is worse than none.

    If you do add them, get the identifiers right

    The value of about and mentions, where there is any, comes from linking entities to unambiguous identifiers. Use stable identifiers, such as Wikidata item URLs or links to authoritative pages describing the entity, and map each CMS tag to one identifier through a maintained table. Review the mapping when tags are added, since a single wrong mapping can mislabel hundreds of articles. Validate the output with a structured data validator, and keep the number of declared entities small and accurate: two or three main subjects in about, and only genuinely relevant ones in mentions.

    A quick decision rule

    If it takes more than a day of development to generate accurate about and mentions markup from existing tags, it is probably not worth doing until more valuable work is finished. If it takes less, and the tags are good, add it, validate it, and move on.

    What helps topic understanding instead

    Clear titles and headings that name the topic, precise use of entity names in text, links to authoritative pages about the entities, a consistent internal linking structure around topics, and author expertise shown on the page all help readers and machines understand what an article covers. These are visible, useful to people, and more likely to influence how content is understood than declarations in markup.

    Add about and mentions only if accurate entity tagging already exists and the markup can be generated automatically, do not expect measurable effects, and invest first in clear content, internal linking and the structured data types Google actually uses.

    Need help with this on your own site?

    Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.

    See SEO services

Add a reply

Replies are open to approved members. Log in or apply to join.