Evaluate observable experience, expertise, authoritativeness and trust signals across every page.
36 active of 41 published rules · each active rule is versioned and backed by evidence. Planned rules are on the roadmap.
Article-type content should identify its author. None was found in JSON-LD, meta tags, rel=author, or a byline.
What it checks
Looks for an author in JSON-LD, meta[name=author], and rel=author on article-type pages.
Confidence
High — derived from evidence quality, not asserted.
No publication date could be discovered for this article-type page.
What it checks
Looks for a publication date in JSON-LD datePublished, <time datetime>, and article:published_time.
Confidence
High — derived from evidence quality, not asserted.
The article states a statistic or figure but links to no external source that supports it.
What it checks
On article-type pages containing a percentage or currency figure, checks for at least one outbound external link.
Confidence
Medium — derived from evidence quality, not asserted.
The article shows a publication date but no modification date, so readers cannot tell whether it has been kept current.
What it checks
On dated article pages, checks whether a modification date (article:modified_time / dateModified) exists.
Confidence
High — derived from evidence quality, not asserted.
The structured modification date is later than today — a data defect that undermines trust in the page’s dates.
What it checks
Compares the modification year to the evaluation year; flags future dates.
Confidence
High — derived from evidence quality, not asserted.
The article names an author but links to no dedicated author page, reducing how verifiable the authorship is.
What it checks
On article pages with an author, checks for a rel=author link or an /author/ profile link.
Confidence
High — derived from evidence quality, not asserted.
The page has Article JSON-LD but no author property, a machine-readability gap for search engines and AI crawlers.
What it checks
When Article/NewsArticle/BlogPosting JSON-LD is present, checks for an author property.
Confidence
High — derived from evidence quality, not asserted.
The byline uses a non-personal placeholder such as “Admin” or “Staff” instead of a real person, signalling low individual accountability.
What it checks
Checks whether the detected author name is a generic placeholder (admin, staff, editor, team, …).
Confidence
High — derived from evidence quality, not asserted.
The page contains affiliate-pattern links but no disclosure text — a compliance and trust risk.
What it checks
Detects affiliate-pattern links (e.g. tag=, /go/, amzn.to) and looks for disclosure wording nearby.
Confidence
Medium — derived from evidence quality, not asserted.
The page uses phrasing like “our data” or “our study” but links to no source that lets a reader locate it.
What it checks
Detects proprietary-claim phrasing and checks for at least one outbound external link.
Confidence
Medium — derived from evidence quality, not asserted.
No page describing who runs the site was discovered — foundational to organisational trust.
What it checks
Looks across the crawled set for an About page by URL pattern or link.
Confidence
Medium — derived from evidence quality, not asserted.
No contact page or method was discovered, so users and search engines cannot reach the site owner.
What it checks
Looks across the crawled set for a contact page by URL pattern or link.
Confidence
Medium — derived from evidence quality, not asserted.
The site never declares an Organization schema entity, so its business identity is not machine-readable.
What it checks
Checks whether any crawled page includes Organization JSON-LD.
Confidence
High — derived from evidence quality, not asserted.
No page explains how content is reviewed, fact-checked, or corrected — a standard publisher trust signal.
What it checks
Looks for an editorial / corrections / review-process page across the crawled set.
Confidence
Medium — derived from evidence quality, not asserted.
Baseline legal/trust pages (privacy policy or terms) are missing entirely.
What it checks
Looks for a privacy policy or terms page across the crawled set.
Confidence
High — derived from evidence quality, not asserted.
No registered business name (Ltd, Inc, LLC, …) was found anywhere, making the organisation feel less accountable.
What it checks
Scans crawled page text for a legal-entity suffix (Ltd/Inc/LLC/GmbH/…).
Confidence
Low — derived from evidence quality, not asserted.
The Organization entity exists but does not cross-reference social/Wikipedia profiles via sameAs, weakening entity verification.
What it checks
When Organization JSON-LD exists, checks for a sameAs property.
Confidence
High — derived from evidence quality, not asserted.
The site does not provide an llms.txt file, an emerging convention for guiding AI crawlers. Informational only.
What it checks
Fetches /llms.txt at the site root and checks for a 200 response.
Confidence
High — derived from evidence quality, not asserted.
A “By …” byline is present but the structured data declares no author, so authorship is only inferable from text.
What it checks
Detects a text byline and checks whether Article JSON-LD declares an author.
Confidence
Medium — derived from evidence quality, not asserted.
A page on a sensitive (health/finance/legal) topic shows no certification, accreditation, or review marker.
What it checks
Detects YMYL topic keywords and checks for a trust/certification marker.
Confidence
Low — derived from evidence quality, not asserted.
The article has links but none point to an external domain — no independent third-party sourcing.
What it checks
On article pages that have links, checks the outbound external-link count.
Confidence
High — derived from evidence quality, not asserted.
The page contains a blockquote with no visible source or attribution.
What it checks
Detects blockquotes and looks for a cite element or nearby attribution.
Confidence
Low — derived from evidence quality, not asserted.
The content references a study, report, or research but links to no source to locate it.
What it checks
Detects study/report phrasing and checks for at least one outbound external link.
Confidence
Low — derived from evidence quality, not asserted.
Sponsored-content indicators are present but no disclosure was found — a compliance risk.
What it checks
Detects sponsored-content phrasing and looks for a disclosure statement.
Confidence
Medium — derived from evidence quality, not asserted.
Content that reads as time-sensitive has no recent modification date (two or more years old, or none).
What it checks
Detects time-sensitive phrasing and checks the modification year against the evaluation year.
Confidence
Low — derived from evidence quality, not asserted.
Structured data declares a publication date but no date is shown to readers on the page.
What it checks
When JSON-LD has datePublished, checks for a visible date in the page text.
Confidence
Medium — derived from evidence quality, not asserted.
Sensitive-topic content credits no expert reviewer separate from the author — no independent sign-off.
What it checks
Detects YMYL topic keywords and looks for a “reviewed by / medically reviewed” byline.
Confidence
Low — derived from evidence quality, not asserted.
Content that states figures never quotes an outside source directly (no blockquote).
What it checks
On pages with a numeric claim, checks for at least one blockquote.
Confidence
Low — derived from evidence quality, not asserted.
A cited external source link returns an error status, so the citation no longer resolves.
What it checks
Fetches the page’s external links (SSRF-guarded, capped) and flags any returning 4xx/5xx.
Confidence
High — derived from evidence quality, not asserted.
The linked author page exists but its biography is too short to establish real credentials.
What it checks
Fetches the author profile page (SSRF-guarded) and counts the bio words.
Confidence
High — derived from evidence quality, not asserted.
The author page has a bio of sufficient length but states no qualification, role, or experience.
What it checks
Reads the fetched author page for credential/role/experience wording.
Confidence
Medium — derived from evidence quality, not asserted.
The author page links to no social profile, ORCID, or personal site — no independent identity signal.
What it checks
Reads the fetched author page for an outbound verification link (social/ORCID/etc.).
Confidence
Medium — derived from evidence quality, not asserted.
An expert-review byline links to a reviewer profile that returns an error status.
What it checks
Fetches the reviewer profile link (SSRF-guarded) and flags a 4xx/5xx response.
Confidence
High — derived from evidence quality, not asserted.
Pages carry author bylines but no author profile page exists anywhere on the crawled site.
What it checks
Combines byline detection with a site-wide search for an /author/ page.
Confidence
Medium — derived from evidence quality, not asserted.
A contact page exists but offers no visible email or phone — forms alone are weaker/less verifiable.
What it checks
On a crawled contact page, checks for a mailto/tel link, email, or phone number.
Confidence
Medium — derived from evidence quality, not asserted.
The editorial-process page shows no modification date, raising doubt about whether the process is still current.
What it checks
On a crawled editorial/corrections page, checks for a modification date.
Confidence
Medium — derived from evidence quality, not asserted.
One author name resolves to multiple profile pages with disagreeing bios.
What it checks
Needs a cross-page author-profile graph comparing multiple author pages.
Status
Published in the catalogue; not yet evaluated by the scanner, so it never produces a finding.
Name/Address/Phone details appear in multiple places but do not match.
What it checks
Needs structured NAP extraction and comparison across pages.
Status
Published in the catalogue; not yet evaluated by the scanner, so it never produces a finding.
The clickable anchor describes something different from the linked page’s actual title.
What it checks
Needs fetching every link destination and comparing titles — deferred for cost.
Status
Published in the catalogue; not yet evaluated by the scanner, so it never produces a finding.
Over half of external links point to low-authority or unrelated domains.
What it checks
Needs a domain-authority / topical-relevance data source (e.g. Moz/Majestic). Cannot be judged honestly from the page alone.
Status
Published in the catalogue; not yet evaluated by the scanner, so it never produces a finding.
A statistic’s citation is rel="nofollow" and points to a non-authoritative domain.
What it checks
Needs a domain-authority data source. Cannot be judged honestly from the page alone.
Status
Published in the catalogue; not yet evaluated by the scanner, so it never produces a finding.