What a ten-page SEO check can tell you

A ten-page check is a sample, not a site score

A bounded SEO check of about ten public URLs can tell you what those pages currently serve. It can show whether a URL returns the expected status, whether the title and main heading describe the page, whether a canonical points somewhere useful, and whether an important link is broken.

It cannot score the whole website. It cannot rank your market. It cannot tell you that a clean sample means the rest of the site is healthy. Treat the first ten pages as a work sample: enough evidence to choose one change, not enough to close the project.

SEO Dispatch's public check uses this same bound. The job is to find the first few issues worth reviewing, then stop. If you want the difference between a crawl score and a growth decision, read Site audit score vs ranking opportunity and the scoring methodology.

What the check can observe

On each sampled URL, a useful check looks for evidence you can reopen later:

1. Fetch and status. Does the public URL load? Does it redirect, and to which final URL?
2. Indexability signals. Is the page accidentally marked noindex, blocked, or canonicalised away from the URL you meant to check?
3. Page identity. Do the title, meta description, and visible H1 describe the same job?
4. Broken paths. Do in-page links return a useful destination, or do they strand a visitor?
5. Obvious markup mistakes. If structured data is present, does it match what the page visibly offers?

Those observations are concrete. You can screenshot them, write them down, and test them again after a change.

A ten-page sample is also good at exposing template problems. If three product pages share a missing title pattern, the fix is probably the template, not three one-off edits.

What the check cannot decide

A short crawl cannot tell you:

• whether a page will gain rankings, traffic, or AI citations;
• whether an unsampled URL is healthy;
• which query a page should target if the page itself is unclear;
• whether a competitor page is a better answer;
• whether a finding is still present after a later deploy.

Search Console can add demand evidence for URLs Google already knows. It still does not replace a live check of the page you changed. Use How to prioritise SEO fixes with Search Console when you have query and impression data. Use the ten-page sample when you need the current HTML and status.

How to use the first three findings

Do not treat the sample as a punch list you must empty today. Pick findings that have:

• a named URL or template;
• a user or indexing consequence you can explain in one sentence;
• a smallest change you can review;
• a way to check the live SEO change after you publish.

A useful first three often looks like this: unblock a public page that should be indexable, repair a broken link on a money or support path, and correct a title or H1 that describes the wrong job. Skip cosmetic length targets. Skip rewriting a useful heading only to insert a keyword.

If two findings describe the same template, record them as one change.

After the sample: choose one page to improve

Once you have a first fix, stay with that page long enough to verify it. Then decide whether the next work is:

• another template-level repair;
• an internal link from a related, already-useful page (internal linking for small sites);
• a Search Console review of pages that already receive impressions;
• a longer crawl inside a workspace, where more URLs can be sampled with the same evidence rules.

The ten-page check has done its job when you can name the next change and the test that will prove it on the live site. It has failed if it becomes a score you defend, or a backlog you never review.

Checklist

• Treat ten pages as a sample, not a grade for the whole site.
• Record the exact URL, status, and the element that is wrong.
• Prefer template fixes when the same issue repeats.
• Choose one change you can verify after publish.
• Do not use the sample to promise rankings, traffic, or citations.

Google's SEO Starter Guide and crawling and indexing overview describe the same limit: a crawler reports what it could fetch, not what a page will achieve in search results.