How the SEO Analyzer works
RankCortex downloads the page once, the way a search engine crawler would: plain HTML, without running JavaScript. It reads the head, the headings, every link and image, the structured data and the visible text, then runs about 30 deterministic checks.
The same page always gives the same result. No AI model is involved and nothing is sent to a paid SEO API.
What it checks
- Technical SEO: HTTP status, redirects, HTTPS, indexability (robots meta, X-Robots-Tag, canonical), mobile viewport, language, character set, HTML size, hreflang, URL structure and server response time.
- On-page SEO: title tag and meta description (presence, length, duplicates), a single clear H1, and the heading hierarchy.
- Links: internal and external links, empty or generic anchor text, javascript: links, nofollow, sponsored and ugc.
- Images: alt text, width and height (layout shift), lazy loading and image formats.
- Structured data: JSON-LD blocks, their syntax and schema.org types.
- Social metadata: Open Graph and Twitter/X card tags.
- Content: visible word count, reading time and text-to-HTML ratio.
About the RankCortex SEO Health Score
The score adds up weighted checks in seven categories: technical SEO 25%, on-page 25%, links 15%, images 10%, structured data 10%, content 10% and social 5%. A passed check counts in full, a warning counts half, a failure counts nothing.
It measures technical health. It is not a Google ranking score, and no tool outside Google can reproduce how Google ranks pages.
When a site blocks the analyzer
Some sites protect their pages with Cloudflare, a WAF or a CAPTCHA. RankCortex respects robots.txt and never bypasses bot protection or logins. When a page can't be fetched, open it in your browser, view the source (Ctrl+U, or Cmd+Option+U on a Mac), copy it, and paste it into the Paste HTML tab. You can also upload a saved .html file.
Frequently asked questions
Is the SEO Analyzer free?
Yes. It runs without an account. To prevent abuse there is a limit of 20 URL checks per minute per network. Pasted HTML has a higher limit.
Does it render JavaScript?
No. It reads the HTML the server sends, like most crawlers do on their first pass. If your content only appears after JavaScript runs, the word count will be low. That is a useful warning in itself.
Why does it say the page couldn't be fetched?
The server refused the request (for example HTTP 403 or a Cloudflare challenge), robots.txt disallows RankCortexBot, or the site didn't answer in 15 seconds. Paste the HTML source instead: the analysis is the same.
Do you store the HTML I paste?
No. Pasted HTML is analyzed in memory and discarded. URL results are kept in a short cache (10 minutes) so repeated checks are instant.
Is this the same as a full site audit?
No. This checks one page. To audit a whole site, sign in to RankCortex and import it from Google Search Console.