Free tool
On-Page SEO Checker
Check one page against title, heading, metadata, link, and image observations, with optional keyword placement.
The check could not be completed
Page check results
Checks that look fine
0
Worth reviewing
0
Problems
0
Notes
0
Heading outline
What This Tool Does
The on-page checker fetches one page and reports each on-page observation as its own check: indexability, metadata, heading structure, visible text, links, images, social tags, and technical basics. Given a target keyword, it also reports where on the page that term appears.
There is deliberately no overall score. A single number needs a weighting that you cannot see and did not choose, and it hides the individual findings that tell you what to change. Counts per severity are given instead, and every check stays visible, including the ones that do not apply.
How to Use It
Enter one page address. Add a target keyword only if the page genuinely has one; leaving it empty simply skips that section rather than guessing.
- Start with the indexability group. If a page declares noindex or canonicalises elsewhere, nothing further on the page will change its position, and the rest of the checks can wait.
- Read the heading outline on its own. If it does not describe the page, the structure is not doing its job, whatever the individual checks say.
- Treat the keyword placement section as a question about whether the page is really about that term, not as a list of places to insert it.
- Check a competing page you do not own too. The observations are the same, and the comparison is usually more informative than either page alone.
How the Tool Calculates or Retrieves Data
The page is fetched once from this website's server, and every redirect hop is validated before it is followed. All observations come from the HTML that arrived: nothing is requested from a third-party SEO data provider, and no figure here is an estimate.
The word count covers visible text only. Script, style, template, SVG, and frame content is removed before counting, and words are matched in a way that does not assume a Latin script, so a page in Arabic or Japanese is counted rather than reported as empty. Links are classed as internal or external by comparing the link host with the host that served the page, so a link written as a relative path counts as internal.
The thresholds used for title length, description length, and the short-page prompt are fixed, published in this page, and versioned with the result. Where a check cannot apply — a page with no images, or no keyword supplied — it is marked not applicable rather than passed or failed.
Understanding Your Results
Problems are things that will misbehave as written: no title, conflicting metadata, a page served over plain http, a noindex you did not intend. Items worth reviewing are conventions the page departs from, which is often the right call once you know you are making it. A long title is not an error, it simply means the end of it may not be read.
The short-page prompt is the check most often misread. There is no minimum length that search engines require, and a short page can be the complete answer to a short question. It is flagged so you can confirm the page says what it needs to, not so you can pad it.
Keyword placement results are observations about text, matched case-insensitively. None of them is a confirmed ranking factor, and repeating a term to satisfy a checklist makes a page worse to read, which costs more than the placement gains.
Frequently Asked Questions
Because any score needs a weighting, and a weighting you cannot see is a judgement made on your behalf. A score also hides its inputs: a page can score well while carrying a noindex that makes every other check irrelevant. Severity counts and a visible list of checks tell you the same thing without the false precision.
There is no required number. The short-page prompt exists to make you look at the page, not to set a target. A definition page of eighty words can be complete; a padded page of two thousand is not better for having been padded.
No. The title and the H1 are worth getting right because they tell a reader and a search engine what the page is about. The rest are observations. If a term does not fit naturally in alt text or a URL, forcing it in makes the page worse.
Usually because the page builds its content with JavaScript after loading. Only the HTML the server sent is counted here. That is also roughly what a crawler sees on its first pass, so a large gap between the two is worth knowing about.
No. It checks one page thoroughly and says nothing about the rest. Site-wide questions — duplicate titles across templates, orphaned pages, crawl depth — need a crawl, not a single page fetch.