What the SEO audit tells you
A page can only appear in search results if search engines are allowed to fetch it, are told to index it, and can work out what it's about. Small mistakes break that silently: a robots.txt line left over from a staging site, a noindex tag copied by a plugin, or a canonical link that points to an address that no longer exists. The page still looks perfect in your browser, so nobody notices until traffic drops.
The SEO audit fetches the page the way a crawler does and checks the technical signals one by one. You get a grade from A to F, a short verdict, and a fix for every problem we find, ordered by how much it matters.
What we check
- Status and redirects. The page's HTTP status after following redirects, how many redirects it sits behind, and whether the four entry points of the site (http and https, with and without www) all end on the same address.
- Indexing rules. robots.txt as Googlebot reads it, with the exact rule and line that decides, plus robots meta tags and the X-Robots-Tag header.
- Canonical. Whether the page has exactly one absolute canonical link, whether it points to the page itself, and, when it points elsewhere, whether that address answers.
- Content basics. The title and meta description and their length, the h1 and heading order, the language and the viewport tag.
- Links and tags. Internal and external links, images without alt text, Open Graph and Twitter tags, structured data (JSON-LD, microdata, RDFa), hreflang and a favicon.
- Sitemap. Whether the site publishes one, whether robots.txt lists it, whether it parses, and whether this page is in it. We read sitemap indexes and gzip-compressed sitemaps too.
- JavaScript. Signs that the page builds its content in the browser: almost no text in the HTML, an empty app root and several scripts.
Rendered mode
Many sites build their pages with JavaScript. Search engines can run scripts, but later and less reliably than they read HTML, and link previews and many other crawlers don't run them at all. Rendered mode loads the page in our sandboxed Chromium browser and shows, field by field, what changed once scripts ran. If the title, canonical or main text appears only after JavaScript, you'll see it in the comparison table, along with the number of requests the page made and how many of them went to other sites.
The browser runs in its own container with no network of its own: every request it makes is checked and fetched by the same safe fetcher as the rest of DNS Super, so a page can't use it to reach private addresses. Rendered audits run as jobs, at most two at a time.
How the grade works
The grade starts at 100 and loses the weight of each problem. Anything that keeps the page out of search entirely costs the most: a status other than 200, noindex, or a robots.txt block. Canonical problems, a missing title and a missing sitemap come next. Small issues such as a short title or a missing language cost a few points. Information findings, such as a canonical that deliberately points to another page, cost nothing. A is 90 and up; below 50 is F.
Reading the result
The result starts with whether the page can be indexed, then the canonical, the content basics, links and tags, the sitemap, and redirects. Each line has an icon and words, so nothing depends on colour. In rendered mode a table at the end lists each field that differs between the raw HTML and the rendered page. Every finding links to a fix guide with the exact steps.
The audit looks at one page. To find broken links across the whole site, use the broken link checker; to validate structured data in depth, use the structured data validator.