Meta Tags Analyzer — what the page tells Google before it renders
Most meta-tag problems are boring and expensive: a duplicated title, a canonical pointing at the wrong URL, an accidental noindex, an og:image that is not there. This tool reads the HTML the server sends and shows you what Google gets. It cannot see the version Google chooses to display, because Google rewrites titles it thinks are weak. That gap is the whole reason to read the raw HTML yourself.
What it reads
Title, meta description, canonical, robots, Open Graph, Twitter card, headings, JSON-LD presence, and images without alt text. Server HTML, not the rendered DOM, so it shows exactly what a crawler receives on the first pass.
Google rewrites titles and descriptions
The title you write is a candidate, not a guarantee. Google rewrites when the title is too long, keyword-stuffed, boilerplate, or a poor match for the query. The title-link documentation is the honest source; the fix is to write one clear title that matches the page, not to game the rewrite.
The one-H1 rule
One H1 that names the page. More than one is not a penalty, it is a lost hierarchy signal. Zero is usually a template bug. The H1 should say what the page is for in the same words the reader would use.
Canonical mistakes
Missing, pointing at http while the page is https, missing or extra trailing slash, or pointing at a paginated page without a plan. A self-referencing canonical is the default; anything else needs a reason you can explain.
Noindex accidents
A staging tag pushed live is the most common self-inflicted deindexing. It is also the easiest to check: one meta tag. If a page is missing from Google, read the robots meta and the X-Robots-Tag header before you rewrite anything.
What it cannot see
The rendered DOM, the title Google chose to display, whether the page ranks, or whether it is indexed. Those need a renderer and Search Console.
The pre-publish check
Five fields, once per page, before any publish: title, description, canonical, robots, og:image. One minute, every time. It catches the expensive mistakes and it is boring enough to keep doing.
What the table says
| Tag | What it should be | The fix if it fails |
|---|---|---|
| Title | One title, ≤60 characters, matches the page | Rewrite the title, do not add a second one |
| Description | One 140–160 character promise | Write the promise, not a keyword list |
| Canonical | Self-referencing, absolute URL, correct protocol | Point it at the page you want indexed |
| Robots | index,follow unless there is a reason | Remove noindex before launch |
| Open Graph | og:title, og:description, one absolute og:image | Use a real image URL, 1200px wide |
| Headings / schema | One H1; JSON-LD that matches the page | Fix the hierarchy, then the markup |
Questions people actually ask
Does Google use my title?
Sometimes. It is a candidate. Google rewrites titles that are weak, too long or mismatched with the query, so the title you write is not always the title that appears.
Why does it differ?
Because Google may rewrite it, and because templates or plugins sometimes print a second title or a different canonical. This tool shows the raw HTML so you can see which.
Is this a crawl?
No. It fetches one URL and reads its server HTML. It does not crawl the site, render JavaScript, or check indexing.
Related tools
Fix it, do not just read about it
Run the free 40-point scan on the page that earns money first. If the list is short, fix it yourself; I mean that. If it is forty things, the Fix Sprint is $1,500 for one site up to 50 pages, with a report card and a 14-day recheck. No ranking guarantee.
Run the free 40-point scan See the fix service