Author bios and AI citations: a real test, not a theory

A test, not a theory

Most advice on E-E-A-T and AI visibility is a description of best practice, written by people who have not measured whether the practice does anything. Seer Interactive's AI optimization team ran a real test instead, documented by Jamir Ong, an AI optimization intern at the agency. They added six specific byline elements, a headshot, a full name with a profile link, a stated title and company affiliation, a specific expertise claim, an external credential reference, a stated tenure, and Person schema markup, to 123 posts on the agency's own Insights blog, then compared the results against an untreated control group over the same measurement window.

What actually moved

Two numbers moved cleanly. Citations tracked through Bing Webmaster Tools roughly doubled on the treated pages, against a 34% gain on the untreated control, a real gap even before adjusting for anything else. Googlebot's crawl rate on the treated pages ran 11.7 percentage points higher than on the control. Google's AI Overviews did not move. Treated pages performed roughly in line with the untreated ones over the same window, with no statistically significant difference reported.

Bing's citation reporting also underpins Microsoft Copilot, which answers from Bing's own index the same way our guide to getting cited by Copilot describes. A signal that moves Bing's crawl and citation behavior plausibly moves Copilot's too, though Seer's report does not break Copilot out separately, so that link is our inference, not their finding.

The stranger result: unknown authors gained more than a famous one

Seer sorted its 95 solo-byline pages into three tiers by how recognizable the author was. Wil Reynolds, the agency's best-known name, carried 47 pages and gained 21%, the smallest increase of the three groups. Two other recognized practice leads, across 33 pages, landed in the middle. Three team members with no public profile, across just 15 pages, gained 113%, the largest jump by a wide margin.

That runs backwards from the industry's default advice, which is to build a personal brand and let E-E-A-T follow. What the test shows looks closer to the opposite mechanism. A well-known byline already carries authority signals an engine has presumably resolved long before any single page gets a bio added to it, so there is less room left to move. An anonymous team member's page had nowhere to go but up: the moment it states clearly who wrote it, what their job is, and what they claim to know, the engine has information it did not have before. On this reading, E-E-A-T is not about becoming famous. It is about closing a gap most pages leave wide open, since a page with a real byline is simply more identifiable than a page with none. That same idea, that clarity about who and what you are outweighs size, is the whole argument of our guide to entity SEO.

Why that lines up with what Google itself says about E-E-A-T

Google's own explanation backs that interpretation. When Google added the extra E for Experience to its quality rater guidelines in December 2022, it said directly that E-E-A-T itself is not a specific ranking factor. It is a framework the company's human quality raters use to evaluate search results. Google's automated systems then try to identify a mix of signals correlated with what raters call good E-E-A-T, one of which is something as ordinary as serving pages over HTTPS. There is no single E-E-A-T score sitting in the algorithm waiting to be raised. What a page can control is whether the underlying signals, a stated author, a real credential, a visible history, are present at all. An engine cannot infer identity or expertise from a page that never states either, the same structural point our guide to where AI engines actually quote from on a page makes about the words themselves.

What to actually do about it

1. Add a real byline. A full name, not "Team" or "Staff," even at a company where nobody is well known yet. Named beat anonymous by a wide margin in the one test that measured it.
2. State the claim, not the adjective. A stated title, company and one concrete area of expertise measured better than vague authority language.
3. Ship Person schema alongside the visible bio, not instead of it. Seer tested markup as one of six bundled elements, not as its own isolated variable, so this data cannot tell you how much of the effect was the schema versus the visible text. Treat that as a real limit rather than promising the markup alone moves anything.
4. Do not expect this alone to move Google's AI Overviews. The one test that measured that specific outcome found no difference. Author signals may still be worth adding for other reasons, being crawled more, showing up in Bing-fed citations, being attributed correctly, but "will this get me into an AI Overview" is not a question this data answers yes to.

What we could not verify

Seer's write-up carries no visible publish date, not in the byline, not in the footer, beyond a general 2026 copyright on the site itself, so exactly when the test ran and how current the numbers are is something we could not confirm independently. The company's own authors describe the result as "positive but inconclusive" and say it needs deeper analysis, a caveat worth repeating rather than smoothing over. The sample is 123 pages on one agency's own blog, not a cross-section of the wider web, and the recognizability tiers thin out fast: the bucket that produced the standout 113% figure is only 15 pages. Nobody outside Seer has replicated any of these numbers, and no search engine has published its own account of whether, or how, it detects a byline at all.

Sources

The byline test, its methodology and all percentage figures are from Jamir Ong, Do Author Bylines Influence AI Visibility? Testing EEAT Elements in GEO, Seer Interactive, accessed September 2026, no publish date stated on the page. The definition of E-E-A-T and Google's statement that it is not itself a ranking factor are from Our latest update to the quality rater guidelines: E-A-T gets an extra E for Experience, Google Search Central Blog, December 2022.

Common questions