0-100 score + A/F grade on 7 E-A-T signals objectively detectable from HTML: author, dates, authority citations, schema, publisher transparency, social presence, HTTPS.
Experience, Expertise, Authoritativeness, Trustworthiness, the 4 dimensions Google's Quality Raters use to evaluate a page (Search Quality Evaluator Guidelines, public document). E-A-T is not a direct algorithmic score, but feeds several signals: content quality for YMYL (Your Money Your Life), weight of review aggregators, editorial authority signals. In practice, a bad E-A-T profile means an indirect penalty on sensitive queries (health, finance, legal).
7 objective signals detectable from HTML: (1) identified author, rel="author", Person schema, byline classes (20 pts); (2) publication / modification dates, schema datePublished+dateModified, time[datetime] (15 pts); (3) citations to authority sources, .gov, .edu, Wikipedia, AFP, INSEE, etc. (20 pts); (4) Article schema JSON-LD (15 pts); (5) About / Contact / Legal page (10 pts); (6) social presence, sameAs + social links (10 pts); (7) HTTPS (10 pts). Total /100.
Several E-A-T dimensions are impossible to score from HTML: external reputation (brand mentions, third-party reviews), the author's real expertise (CV, publications), editorial quality of the content (Google uses a language model for that), behavioral signals (dwell time, pogo-sticking from SERP). This tool only measures what is objectively declarable in HTML, and many pages already fail on these basics.
E-A-T matters most on YMYL queries (health, finance, legal, safety). On less sensitive topics (cooking, lifestyle, mainstream tech), an average E-A-T score gets by. But the market is evolving: since the Helpful Content Update 2022 and its iterations, Google uses E-A-T as a proxy for "site worth recommending" on more and more verticals. Score < 50 means concrete risk at the next core update, even outside YMYL.
Our network catalog is browsable without signup. Publisher pricing shown, no commission, no middleman.