On-Page SEO Auditor Guide
What this auditor verifies
The On-Page SEO Auditor examines the HTML that a page actually returns and separates source-level findings from live server evidence. It checks titles, meta descriptions, indexing directives, canonical URLs, headings, main-content depth, repeated paragraphs, placeholder language, images, links, social metadata, JSON-LD syntax, mobile viewport settings, document language, and other basic markup signals.
Live URL mode also verifies the HTTP response, final URL, HTTPS use, redirect behavior, content type, X-Robots-Tag, robots.txt access, response size, and a basic fetch-time measurement. When a website returns a login wall, rate-limit response, access-denied document, or browser-security challenge instead of the requested page, the auditor reports the result as inconclusive and does not score the substitute HTML.
How the score is calculated
The score is deterministic: identical input produces identical output. Source points are assigned across metadata, indexability, canonicalization, headings, content quality, images, links, social metadata, structured data, and technical markup. Live URL mode combines that source score with verified server checks. Critical findings apply score caps so a strong collection of minor passes cannot hide a noindex directive, invalid canonical, blocked response, placeholder content, or severe on-page duplication.
A score of 100 is reserved for pages that pass every applicable automated check without warnings. It is not a Google ranking prediction. Search engines use many signals that this browser tool cannot measure, including off-page authority, historical behavior, query intent, rendering at scale, and proprietary ranking systems.
Content-quality checks
The auditor does more than count words. It isolates the main content from navigation, forms, related-tool cards, and code samples, then measures exact repeated paragraph blocks, repeated sentences, and highly similar paragraph pairs. This helps identify pages that appear long only because the same explanation was copied into multiple sections.
It also looks for unfinished placeholders and production-oriented wording that should not appear in public copy. Topic coverage compares the page title and H1 with the main text so a generic article cannot receive full credit merely because metadata is complete. These checks are evidence-based heuristics; the report shows the detected counts and percentages so an editor can review the underlying reason.
Metadata, canonical, and indexability
The metadata review verifies that one usable title and one meta description are present, checks practical length ranges, and identifies duplicate tags. Canonical analysis resolves relative URLs, checks protocol and host, compares the canonical with the audited URL, and flags fragments or multiple canonical declarations.
Indexability checks parse robots and Googlebot meta directives as a combined instruction set. Contradictory index/noindex or follow/nofollow values are treated as critical. Live mode separately inspects X-Robots-Tag and the longest matching robots.txt rule available to Googlebot or the wildcard user agent.
Headings, links, images, and social tags
The heading review expects one nonempty H1, detects skipped heading levels, finds repeated heading text, and compares the H1 topic with the title. Link checks distinguish internal and external HTTP links from empty, JavaScript, data, or malformed destinations. The auditor also calls out generic anchor text that gives users little context.
Image checks distinguish a missing alt attribute from an intentional empty alt value used for decoration. Open Graph and Twitter/X checks measure core tag coverage, validate social-image URLs, compare og:url with the canonical page, and detect duplicate or conflicting social fields.
Structured data and technical markup
Every JSON-LD block is parsed as JSON and checked for recognizable Schema.org context and type information. Structured data is optional, so a page is not penalized simply for omitting it; however, invalid markup reduces the score when it is supplied. The technical review also checks the HTML5 doctype, character encoding, responsive viewport declaration, language attribute, and HTTPS page URL.
Using Live URL mode
- Enter the complete public URL, including HTTPS when available.
- Add a focus phrase only when you want exact phrase-placement checks.
- Click Run Live SEO Audit and wait for the verified response.
- Review the source score, live score, coverage, HTTP result, redirects, and robots status together.
- Open the detailed report and confirm each warning against the published page.
- Correct genuine issues, deploy the update, and run the live audit again.
Using HTML Source mode
Choose HTML Source Audit when a remote site blocks automated access or when you are reviewing a page before deployment. Paste the complete document rather than a partial head or body fragment. The page URL is still required because the auditor uses it to resolve relative links and evaluate canonical and social URL alignment.
Source mode cannot verify HTTP status, redirects, response headers, robots.txt, or live response timing. The result is therefore labeled as a source-readiness score rather than a verified live score.
Understanding passes, warnings, and critical issues
A pass means the supplied evidence met the rule shown in the report. A warning identifies a condition that deserves review but may be intentional in a specific context. A critical issue is a condition that normally prevents an indexable page from being considered ready, such as noindex, contradictory directives, severe duplicated content, invalid canonical markup, or unfinished placeholder text.
The report includes category scores and detected values so the final number is explainable. Do not change a valid implementation solely to chase a score. Review the page purpose, accessibility needs, legal requirements, and editorial intent before applying any recommendation.
When to run the audit again
Repeat the audit after changing metadata, page copy, templates, structured data, redirects, hosting rules, a CDN, or an SEO plugin. A deployment system can alter final HTML or headers even when the editor changes only one field. Auditing the published URL confirms what visitors and crawlers can actually receive.