Is E-E-A-T a ranking factor?

Not as a measurable score. Experience, expertise, authoritativeness and trust are the criteria human quality raters are instructed to apply, and raters do not rank pages: their judgements are used to evaluate whether the ranking systems are working. The signals that correlate with it, such as verifiable authorship, are real; a dial marked E-E-A-T is not.

Does every page need an author byline?

No, and adding one everywhere devalues it. Documentation, product pages and changelogs are institutional and read better without. A byline belongs where an individual judgement, experience or credential is what makes the page worth trusting, which is most of what people mean by content.

What makes an author page worth having?

Something a reader could check: what the person has done, where else they have published, a link to a profile that exists. An author page containing a stock photograph and two sentences of description is worse than no author page, because it makes the byline look manufactured rather than merely absent.

A byline is only a signal if it resolves to somebody

Adding a name to the top of an article takes ten seconds, which is the problem: a byline is only worth something if it leads to a person, and most of them lead to a filter page listing the same author's other posts.

The short answer

A byline that leads nowhere is a string of text. Author information is worth something when the name resolves to a bio with real credentials and a way to verify the person exists, and it matters most on subjects where being wrong has consequences: health, finance, legal and safety. On a changelog, it is optional.

ImportantAudit check · Author information

A name that leads nowhere is a string

There are three levels of attribution and only the third does anything. The first is no byline at all. The second is a name printed on the page. The third is a name that links to somewhere establishing why this person is worth reading on this subject.

Most sites stop at the second and believe they have done the work. The gap matters because a bare name is unfalsifiable: it could be a staff writer, a freelancer, a pen name, or a label somebody typed into a CMS field. Nothing about the string distinguishes those, and nothing about it explains why the article should be believed.

What closes the gap is not much. A link from the byline to a page with two paragraphs on it: what this person does, what they have done that is relevant, and one or two outside references, a professional profile, a publication record, a registration number if the field has one. Two paragraphs and two links is the difference between a name and an author.

The corroboration is the part that gets skipped and the part that carries the weight. Anyone can write a bio. A bio that links to a profile somebody else controls is a claim you have exposed to contradiction, which is why it counts for more.

Which pages need this, and which are better without it

Attribution is not a site-wide obligation, and applying it everywhere devalues it. The question is whether the page rests on somebody's judgement.

Needs it: advice, analysis and anything consequential. If a reader could act on the page and be harmed by it being wrong, the person behind it is part of the information. Health, finance, legal, safety, and any technical recommendation strong enough to follow.
Needs it: opinion and review. A verdict is only interpretable if you know whose it is. An unsigned review is a review by the company selling the thing, whatever the intention was.
Does not need it: reference and documentation. An API reference or a specification page is not an argument and does not need a byline. The organisation is the author, which is what the publisher property is for.
Does not need it: product, pricing and transactional pages. Nobody is looking for the author of a checkout page, and putting one there reads as a template applied without thought.

The signal that a site has got this wrong is a byline on every URL including the contact page. That is a CMS setting rather than an editorial decision, and it makes the bylines on the pages that needed them mean less.

The author page that makes things worse

Adding attribution badly is a real outcome, not a hypothetical one. Four patterns to avoid, all common:

The byline links to a tag archive.Clicking the name gives you a list of the same person's other articles and no information about them. This is the default behaviour of most publishing platforms and it satisfies the letter of the check while providing nothing.
The bio is a job title. Sarah is a content strategist with a passion for storytelling establishes nothing about whether she knows about carbon steel. A bio has to be relevant to the subject of the article or it is decoration.
The author is invented. A stock photograph and a plausible name attached to generated articles. This is the pattern that made attribution a scrutinised signal in the first place, and it is worth knowing that reverse image search on a headshot is a thing readers do.
One author has written four hundred pages in six weeks. The volume is itself a claim, and it is not a credible one. Where content is produced at scale, the honest attribution is the organisation, not a person who could not have written it.

The hard case: writing that genuinely has no single author.

Plenty of good content is assembled by several people, edited by a team, and updated over years. Inventing a single byline for it is a fiction, and leaving it anonymous loses the signal. The honest answer is to attribute it to the organisation and name the reviewer: who checked this, and when. A line reading "reviewed by" with a name, a role and a date does the trustworthy part of what a byline does without pretending one person wrote it.

The same applies to content with AI involvement. The useful disclosure is who is accountable for it being correct, which is a person either way.

A byline has to be followed to be worth anything

The check is not whether a name appears. It is whether the name resolves, and what is behind it when it does, which means a second fetch of the page the byline points at.

This one needs the full crawl

A byline only counts if it leads somewhere, so the check follows the name to see whether a page exists behind it and what that page contains.

The instant search on this site audits a single page, so rather than show you a verdict it cannot support, this guide sends you to the place the check actually runs.

Run the full audit, free

The three-second version on your own site: click your own byline and see where you land. If it is a list of articles, you have the second level of attribution and not the third.

Does this affect AI search?

Yes, and it is one of the places where the machine case is genuinely stronger than the search case. An assistant answering a question has to decide which sources to rely on, and authorship is one of the few properties of a document that speaks to reliability without requiring the content to be evaluated. A named, corroborated author on a subject they demonstrably work in is a cheap and strong signal.

There is a mechanism worth understanding behind that. Attribution that resolves to outside profiles lets a name be treated as an entity rather than as text, which means an author's work across different sites can be recognised as the same person's. That accumulation is only available to authors whose identity is stated in a way something can follow, which is the argument for the outbound links in a bio.

The honest limit: nobody outside those systems can measure this, and there is no report that will show you an author signal improving. What you can observe is the negative case, which is that anonymous content on consequential subjects is exactly the profile of the material these systems are being tuned to be careful with.

Why anonymity costs more on some subjects than others

The rating is Important and it is an average over a check whose stakes vary enormously. On a page giving dosage advice or tax guidance, anonymity is close to disqualifying and the assessment frameworks are explicit about it. On a page about which whetstone to buy, it is a missed opportunity worth an afternoon.

It is not Critical because nothing mechanical fails. Anonymous pages are crawled, indexed and ranked, and a great deal of anonymous content ranks perfectly well on subjects where nobody is asking who wrote it. There is no exclusion here, only a ceiling on how far a page can be trusted.

What lifts it clear of a refinement is that the fix is small and the pages that need it are usually the pages that matter most: the ones making the strongest claims, aimed at the highest-intent readers. Two paragraphs and two links, on the twenty pages where somebody might act on what you said, is one of the better returns available in an audit.

Following your own bylines to see where they go

The check that matters is whether the chain from article to person to corroboration is unbroken.

check-bylines.sh

URL=https://example.com/guides/carbon-steel

# where does the byline point?
target=$(curl -s "$URL" \
  | grep -o 'rel="author"[^>]*href="[^"]*"\|href="[^"]*/author/[^"]*"' \
  | grep -o 'href="[^"]*"' | head -1 | sed 's/href="//;s/"$//')

echo "byline links to: ${target:-NOTHING}"

# is there a real bio there, or just a list of posts?
body=$(curl -s "https://example.com${target#https://example.com}" \
  | sed 's/<[^>]*>/ /g' | tr -s ' \n' ' ')

printf 'words on the author page: %s\n' "$(printf '%s' "$body" | wc -w)"

# does it corroborate anything outside your own domain?
curl -s "https://example.com${target#https://example.com}" \
  | grep -o 'href="https\?://[^"]*"' \
  | grep -v 'example.com' | sort -u

Two outcomes are common and both are findings. If the first command prints nothing, your byline is plain text. If the last command prints nothing, your author page makes claims that rest entirely on your own say-so.

The structured-data half of this, getting the author into an author property as an entity rather than a string, is a separate check, and it is worth doing after this one rather than before. Markup describing a person with no page behind them is a reference to nothing.

Questions this check raises

Is E-E-A-T a ranking factor?: Not as a measurable score. Experience, expertise, authoritativeness and trust are the criteria human quality raters are instructed to apply, and raters do not rank pages: their judgements are used to evaluate whether the ranking systems are working. The signals that correlate with it, such as verifiable authorship, are real; a dial marked E-E-A-T is not.
Does every page need an author byline?: No, and adding one everywhere devalues it. Documentation, product pages and changelogs are institutional and read better without. A byline belongs where an individual judgement, experience or credential is what makes the page worth trusting, which is most of what people mean by content.
What makes an author page worth having?: Something a reader could check: what the person has done, where else they have published, a link to a profile that exists. An author page containing a stock photograph and two sentences of description is worse than no author page, because it makes the byline look manufactured rather than merely absent.