Why this signal matters for Google
To a machine, your name is just a string of characters. You sign forty articles as Marie Dupont. A reader knows instantly it is the same person. A search engine does not: it sees the same letters forty times, and those letters could belong to forty different people. Author schema markup solves that. It is a small block of code, invisible to the reader, written in JSON-LD (a data format Google reads inside the page), and it says: a person wrote this article, her name is Marie Dupont, her author page is here, and here is where to find her elsewhere on the web. The code does not create credibility. It makes credibility that already exists readable.
Google has human evaluators grade pages using a public manual, the Quality Rater Guidelines, or QRG. Section 2.5.2 tells them to find who is responsible for the website and who created the content on the page, before judging that content at all. When that search comes up empty, the page starts at a disadvantage. Section 3.3.4 goes further: it weighs the reputation of the author, not only the reputation of the site. Author schema markup serves both steps. It names the person responsible, and it says where that person's reputation can be checked.
The real work happens in the sameAs property. That is the line listing your public profiles: LinkedIn, your entry on a professional body's register, your author page at a publisher, your ORCID identifier if you publish research. Without it, your block only says this person is called Marie Dupont, which adds nothing to the byline a reader can already see. With it, an engine starts from your article, opens your LinkedIn profile, finds the same degree and the same job title, and concludes it is dealing with one person. One rule keeps this clean: put only live pages in sameAs, pages where your name actually appears. A link to a profile you abandoned six years ago works against you.
Generative engines need that connection even more. When ChatGPT, Perplexity or Google's AI summaries decide who to cite, they look for details they can confirm somewhere else. An article signed by an author tied to a consistent professional profile gives them something to check. An article signed by the editorial team gives them nothing. The gap widens on YMYL topics, the ones touching health, money or law: section 3.4.1 expects genuine expertise on those pages, and section 5.1 rates a page as low quality when the author lacks expertise in the subject. With no identifiable author, the evaluator cannot even ask the question.
The NeuroEEAT scoring grid (0-10)
Rate how far your articles lead back to a real person, from a name at the bottom of the page to markup confirmed everywhere else.
| Score | Description | Concrete indicators |
|---|---|---|
| 0-2 | No author | Articles carry no byline, or they are signed by the editorial team. Section 2.5.2 asks the evaluator to find who created the content: here, nothing turns up. |
| 3-4 | Name only | A name sits at the bottom of the article, with no author page and no link. The reader can check nothing, and the page code tells the machine nothing. |
| 5-6 | Author page, no markup | Every author has a real page with a bio, a background and a list of articles. None of it is marked up: engines have to guess that the byline and the page describe the same person. |
| 7-8 | Person marked up | A JSON-LD Person block names the author and points to their author page. sameAs is missing or thin, so the identity stays locked inside your own site. |
| 9-10 | Entity confirmed elsewhere | Person and sameAs tie the author to public profiles, and those profiles say what the site says. A human, an evaluator and an AI all land on the same person. |
How to audit it on your site
Five checks to find out whether your articles really lead to somebody.
- 01
Open one article and hunt for the author
Pick an article at random and read it like a visitor in a hurry. How many clicks does it take to learn who wrote it and what they know about the subject? If the answer is no name at all, or more than two clicks, the problem shows before you even open the code.
- 02
Check that every author has a page
One page per person, with a bio, a background and their articles. Without it, the markup has nothing to point at. It is also the first thing an evaluator reads when judging expertise (section 3.4).
- 03
Read the page code
View the source and search for Person or author. You can also paste the URL into Google's Rich Results Test, which is free and shows what it understood from your page. Very often the block is there but names the company instead of the person.
- 04
Test the sameAs list
Open every link in the sameAs property, one at a time. Each one has to land on a live page where your name appears. A deleted profile, a dormant account or a plain company homepage are links to remove.
- 05
Compare the site with the profiles
Put your author page next to your LinkedIn profile: name, job title, education, employer. Section 3.3.4 judges reputation on what outside sources say. Two different versions of you weaken both.
How to activate or improve it
Three levels, from a name at the bottom of the page to an author everyone can find elsewhere.
Give the article a name and a face
- Sign every article with a person's name, never with the editorial team or the company name.
- Build one author page per person: photo, two paragraphs of background, the subjects they cover on the site.
- Link the byline at the bottom of the article to that author page, as a real clickable link.
- Make sure the name on the site matches the name on your public profiles exactly, spelling included.
Put the Person markup in place
- Add a JSON-LD Person block to every author page, with name, url and jobTitle at minimum.
- On each article, fill the author property by pointing at that author page rather than retyping the name as plain text.
- Fill sameAs with two to five real public profiles: LinkedIn, professional register, publisher page, ORCID.
- Run every page template through Google's Rich Results Test and fix the errors it reports.
- Let your CMS generate the markup from the author record, so nobody retypes it on each article.
Keep the author entity alive over time
- Give each author a stable identifier in the markup (the @id property), reused across all their pages, so engines connect their articles to each other.
- Add knowsAbout, the property listing the subjects an author commands, with the topics they actually cover and nothing more.
- Get your authors cited elsewhere: interviews, guest posts, conference listings. sameAs is only worth as much as the pages at the other end.
- Put an annual review on the calendar: dead profiles to drop, job titles that changed, new authors to mark up.
- Write down who on your team updates the author record when somebody joins or leaves.
5 common mistakes to avoid
Sector specifics
Medical information site
An article about a treatment puts the reader's health on the line. Tie the author to their licence number and their public register entry, and point sameAs at those pages. Section 3.4.1 expects genuine expertise on these topics, and an unsigned article gives nobody a way to see it.
Financial publisher or advisory firm
Advice about money is judged on who gives it. An author tied to their registration, their firm and their past publications gives the reader enough to decide whether to follow them. The markup makes that same track record legible to engines at the same time.
SaaS blog or agency
Your articles are usually written by several people on the team. Give each of them an author page and a Person block, instead of one shared marketing account. That is what lets a prospect recognise, six months later, the person they read.
Local business or trade
Here the author is often the owner. Tie their author page to their Google Business Profile, their LinkedIn and any local press coverage. Customers check who is talking before they call, and engines walk the same path.
Signals to activate in parallel
Author markup is only worth as much as the identity it points to holds up elsewhere.
FAQ
Is author schema markup required to rank well?
No. No markup is required, and Google does not rank a page because it contains JSON-LD. Author schema markup does something else: it helps engines connect your articles to a real person. That connection is what feeds into how your content gets judged.
What is the difference between author schema markup and an author page?
The author page speaks to the reader, the markup speaks to machines, and you need both. Without an author page, the code points at nothing. Without the code, engines have to guess that the byline at the bottom of the article is the person described on that page.
What exactly belongs in sameAs?
Only public pages where the author's name appears: LinkedIn, a professional register entry, an author page at a publisher, a speaker profile, ORCID. Two to five live links beat ten when half of them lead to abandoned accounts.
Can the company be declared as the author of its articles?
Technically yes, but you lose ground doing it. Section 3.3.4 assesses the reputation of the person who created the content, and a company has no degree and no personal track record. Put the person in author and the company in publisher: both facts can sit there without competing.
How do I check that my author schema markup works?
Paste an article URL into Google's Rich Results Test: it shows what it understood from your markup, author included. Then open every sameAs link by hand and confirm each one loads a live page. Valid markup pointing at a deleted profile is still useless.
Anonymized case study
Before · signed articles, untraceable author
After · author marked up and checkable
Illustrative figures meant to show the typical progression of the signal. They do not come from a real client case and must be replaced with your own measurements before any external communication.