Not every social link on your site is yours
sameAs is how you tell a search engine that this website, that
LinkedIn page and that Instagram account are one organisation. It is an identity claim, not a
list of links you happen to have — and the difference matters, because the wrong entry is worse
than a missing one.
The advice that would have caused harm
A tool crawls a newsroom, collects every social link it finds, and reports that those
profiles are missing from the organisation's sameAs. Reasonable-sounding, and
completely wrong: on that site the social links were reporters' own accounts and links to cited
sources. The suggested markup — ready to paste — would have declared a named journalist's
personal profile, and somebody else's video, as the publisher's own identity.
That is not a tidy-up gone slightly astray. A missing sameAs costs you a little
entity evidence. A wrong one asserts that a company and a private individual are the same thing,
in machine-readable form, on the company's own site.
The two rules that keep the claim honest
A content permalink is not a profile. A comment thread, a short video, a
status post — these live on a social host and identify no account. They are things somebody
published, not somebody's identity. Only a profile URL can be a sameAs.
A profile you own appears across the site, because it sits in the template. Your accounts are in the footer, so they turn up on every page crawled. A reporter's account appears on that reporter's articles. A cited source appears once. The ones that belong to the organisation are the ones that repeat.
Why position does not work and repetition does
The obvious filter is structural: ignore links inside the header and footer, keep the rest —
or the reverse. It fails on real sites. That newsroom nests its author cards inside a
<footer> element that sits within each <article>, so the
byline links are inside a boilerplate tag and survive any filter keyed on position.
Repetition separates them and position does not. It is worth knowing as a general habit: when you need to tell a site's own furniture from its content, count how often something appears rather than reasoning about where it sits. Markup is a weaker signal of intent than frequency, because the markup is whatever the theme author chose and the frequency is what the site actually does.
Checking yours
Look at the sameAs block in your organisation markup and ask of each entry:
is this a profile, and is it ours? Then look at what you are missing — the accounts in
your own footer that never made it into the markup are the ones worth adding, and they are
usually the whole answer.
What an audit cannot settle
Whether an account belongs to you. A crawler sees a URL on a page; it does not know your org chart, your agency relationships, or which of two similarly-named accounts is the real one. A tool can tell you which links repeat across the site, and that is a strong hint — it is not ownership, and anything that presents it as ownership is inviting you to publish a claim about somebody else.
This check was wrong in exactly that direction before the two rules above were added. A page citing a check is worth more for saying so than for implying the check has always been right.
Common questions
What is sameAs for?
It tells search engines that a website and a set of social profiles are one organisation. It is an identity claim rather than a list of links you happen to have on the page.
Is a missing sameAs worse than a wrong one?
No — the other way round. A missing entry costs a little entity evidence. A wrong one asserts in machine-readable form that your company and somebody else, possibly a private individual, are the same thing.
Can I use a link to a post or a video?
No. A comment thread, a short video or a status post lives on a social host and identifies no account. It is something somebody published, not somebody's identity. Only profile URLs belong in sameAs.
How do I tell my own accounts from other people's?
Count how often each appears. Your accounts sit in the template, so they turn up across the crawl; an author's account appears on that author's articles and a cited source appears once.
Why not just ignore links in the header and footer?
Because position is unreliable. Author cards are sometimes nested in a footer element inside each article, so byline links sit inside a boilerplate tag and survive that filter. Repetition separates them; position does not.