E-E-A-T Trust Signals Checker
Check what a reader, and a search quality rater, can actually find about who runs your site and who writes for it. Observations with evidence, no made-up score.
Trust signals the checker looks for
Trust pages are found from the site's own links, the way a reader would find them. Nothing is guessed from common paths.
- Trust pages
- About, Contact, Privacy, Terms, Editorial Policy and Corrections, plus an authors page, a disclaimer and an affiliate or advertising disclosure when the site links to them.
- Confirmed or likely
- A trust page is confirmed only when the link says what it is and the fetched page's title or main heading agrees. Otherwise it is shown as a likely match for you to review.
- Contact routes
- What the contact page really offers: email links, phone links, a form, and a postal address when the markup declares one.
- Visible bylines
- Whether each sampled article names its author where readers look for one: a "By" line, a rel="author" link or an author page link beside the article.
- Structured authors
- The author of the article's own Article or BlogPosting markup, compared with the visible byline so a mismatch stands out.
- Reviewers and editors
- "Reviewed by", "Edited by" or "Fact checked by" credits in the text or in the markup.
- Published and updated dates
- Dates from the markup, the article meta tags and labelled visible text, compared after time zones are normalised.
- Citations
- Links out from the article's own text, grouped by the kind of site they reach. Navigation, footers, share buttons and sponsored links are counted apart.
- Author and reviewer profiles
- Up to five profile pages are opened to see what they declare and show about the person.
- Publisher identity and profiles
- Whether the organisation name in markup, site name tags and policy pages agrees, and whether the declared sameAs profiles answer.
How to check your site's trust signals
Enter the home page
That is where most sites link to their about, contact and policy pages. Any public page works, but the home page gives the fullest picture.
Let it read the site
The home page, the trust pages it links to, up to five articles and a few author profiles are fetched, two at a time, following robots.txt.
Go through What Was Observed
Each line says how many sampled pages showed the signal. Open a line to see the exact page and where on it the evidence was read.
Use the checklist for the rest
Some trust signals need a person to judge, such as real team photos or first-hand experience. They are printed as a checklist, not scored.
A run makes at most 40 requests and stops starting new pages after about 90 seconds. Anything skipped is listed with the reason.
How to read the trust report
Every figure is "of the pages read", never "of the site". Five sampled articles show a pattern, not a census.
- Site Transparency
- How many of the six expected trust page types were confirmed, and which ones are only likely matches.
- Likely match, review
- The link suggests a trust page, but the page itself did not confirm it. Open it and check it is what it claims.
- Sampled Editorial Pages
- One row per article with its byline, markup author, reviewer, dates and citation count, so gaps across the sample are easy to spot.
- Dates: Mismatch, review
- The markup and the visible date disagree. The tool does not decide which is right; you do.
- Publisher Identity
- Consistent, possible mismatch or not enough evidence. Different names for the publisher in different places are flagged for a person to look at.
- Declared Profiles
- Reachable, redirects or could not verify. An answer means the address exists, not that the profile belongs to you.
Trust gaps it finds, and how to close them
Policy pages exist but nothing links to them
Fix: Add About, Contact and Privacy links to the footer of every page. A page readers cannot reach does not help them trust you.
Articles with no visible author
Fix: Show a byline near the headline that links to an author page with a short, real biography.
The byline and the markup name different people
Fix: Make the CMS write the same author into both. Often an old plugin or a default admin account sets the markup.
The updated date in markup is newer than anything visible
Fix: Show a "Last updated" date on the page and change it only when the content really changes.
A contact page with only a form
Fix: Add at least one direct route, such as an email address or phone number, and a postal address if you have one.
Organisation markup with a different name from the site
Fix: Use one publisher name across Organization markup, the og:site_name tag and the About page.
Who checks trust signals
- Publishers in health, finance and legal topics, where readers need to know who wrote the advice.
- Affiliate site owners making sure disclosures and author pages are in place.
- SEO consultants preparing a quality review after a core update.
- Small businesses checking that their contact details and company name are consistent.
Why there is no E-E-A-T score
Google does not compute an E-E-A-T number. It is a concept from the guidelines its human raters follow. This tool reports what it saw and where, so you fix real gaps rather than chase a figure someone invented.
Trust Signals FAQ
Short answers about the Trust Signals.
Experience, Expertise, Authoritativeness and Trustworthiness. It comes from Google's search quality rater guidelines and describes what raters look for, not a ranking factor with a score.
Usually the home page, since that is where sites link to their trust pages. The articles it samples are picked from the site's own links, then from the sitemap if links run out.
In a fixed order, never at random: article cards and editorial addresses first, then headline-like links, then the sitemap. A page counts as an article only on combined evidence such as article markup, a byline and a date.
No. Matching words such as "PhD" in prose is unreliable, so qualifications are not read from body text. Author pages are reported for what they declare and show.
The link pointed to it as a trust page, but its title and main heading did not say so. It may be fine, or it may be a different page. It needs a look.
21 items across identity, creator expertise, editorial process, originality and commercial transparency, such as real team photos, a corrections policy and disclosed affiliate deals. It is not scored.
Yes. Pages the site asks crawlers to avoid are not read, and any crawl delay is honoured. If robots.txt cannot be read, only the page you entered is fetched.
Related SEO tools
Schema Intelligence Analyzer
Score the structured data already on your pages.
1 free run a dayOn-Page Audit Pro
70 checks on one page, scored against the basics.
1 free run a daySite Audit Pro
Crawl up to 100 pages and get a technical health score.
1 free run a daySchema Generator
Valid JSON-LD for 17 schema types from a form.
Free, unlimited