SEO webpage analyser: the eight checks on a single page, in order

An SEO webpage analyser fetches one URL and reports what a search engine can see on it. There are eight layers to check, and they belong in this order: can the page be crawled and indexed, does it render, what does the content cover, how are the headings and title written, where do the internal links point, do the images and structured data hold up, how fast and stable is it, and how does it compare with the pages currently ranking. You can run every one of those for free with Google Search Console, PageSpeed Insights and the Rich Results Test. Paid analysers add a score, bulk checks and history.

The order matters more than the tool. A page with a beautifully optimised title that Google has not indexed is worth nothing, and most page level analysers present their findings alphabetically or by their own scoring instead of by what blocks the page.

The eight layers

Layer	What you are checking	Free way to check it
1. Indexability	Status code 200, no noindex, canonical pointing at itself, not blocked in robots.txt, present in the sitemap	URL Inspection in Search Console, which shows Google's own view of the page
2. Rendering	That the main content and links exist for a crawler, not only after JavaScript runs	The rendered HTML and screenshot in URL Inspection
3. Content coverage	Whether the page answers the query fully, including the sub questions the top results answer	Read the three pages ranking above you and list what they cover that you do not
4. Title, description and headings	One h1, a descriptive title under about 60 characters, headings that read as an outline, a description that earns the click	View source, or any free on page checker
5. Internal links	Links in from relevant pages with descriptive anchors, and links out to your own related pages	A site search for the page's topic, or a crawl of the site
6. Images and structured data	Sized and compressed images with real alt text, and valid markup for the page type	Rich Results Test and the Schema Markup Validator
7. Speed and layout stability	Largest Contentful Paint, layout shift, and what the first screen costs to load	PageSpeed Insights
8. Competitive comparison	Format, depth, freshness, author credibility and links, against the pages that rank	By hand, and no tool does this properly

Work in this order

1. Anything that stops the page being indexed. Fix it the same day. Nothing else on the list matters until the page can be found.
2. Anything that breaks the page for a visitor. A broken layout on a phone, a form that fails, a cookie banner covering the content.
3. The first paragraph and the title. These are the edits that move a page most often, and they take ten minutes.
4. Content gaps against the top three. Usually one or two sections that every ranking page has and yours does not.
5. Internal links into the page, added from your strongest related pages with descriptive anchors.
6. Technical housekeeping. Images, markup, the remaining speed items.

Reading the findings an analyser gives you

Page analysers are built to produce a list, so they report everything they can measure. These are the findings to read carefully rather than act on mechanically:

• Keyword density. There is no target percentage to hit. Use the phrase where it reads naturally, including once in the title, once in the first paragraph and in a heading, then write for the reader.
• Word count targets. Length follows the question. A definition page answers in three hundred words and a buying guide needs two thousand. Compare with the pages ranking, not with a tool's default.
• The score out of a hundred. Useful as a to-do list, and it has no relationship to position. Read the findings underneath it and judge each one.
• Meta keywords. Google does not use them. An analyser flagging them is reporting a field nobody reads.
• Exact match anchor text everywhere. A tool that wants your internal anchors to be identical commercial phrases is giving you a pattern to avoid.

A fifteen minute manual pass

This is faster than most software and it finds more. Do it on your five most important pages.

1. Open the page in Search Console URL Inspection. Confirm it is indexed and read the rendered HTML.
2. Search Google for the query the page targets. Read the top three results properly, not their titles.
3. Write down three things those pages cover that yours does not, and one thing yours does better.
4. Reread your first paragraph. Does it answer the query in two sentences? If not, rewrite it now.
5. Check the title against the query, and the description against the reason somebody would click.
6. Scroll the page on a phone. Note anything that pushes the answer below the first screen.
7. Run PageSpeed Insights and note only the items affecting the first screen.
8. Add two internal links into the page from relevant existing pages.

When a paid analyser is worth it

• Bulk. A hundred pages at once, which is a crawler rather than a single page tool.
• History. Seeing that a page was fine last month and is not now.
• Content scoring against the live results, where a tool pulls the terms and subtopics common to the ranking pages. This genuinely saves time on a long brief.
• Team workflow. Assignable issues, so findings become tickets instead of a PDF.

For the categories of tool and what each one owns, see SEO audit tools. For the site wide version of the same checklist, see the technical SEO checklist.

This page covers analysing a single page and acting on the result. For whole site crawling and the tool market, see the two guides above. For the internal link structure that decides which of your pages rank at all, see internal linking.

Where people get stuck on SEO webpage analyser

Why does my page score well and rank badly?

Because the score measures the page in isolation and ranking is a comparison. The page ahead of you answers the question better, has more or better links, or matches the intent more closely. Read it side by side with yours and the reason is usually obvious in five minutes.

How long should a title tag be?

Write it to be read, then check it is around sixty characters so it survives truncation. Put the distinctive words first, and keep the brand name at the end or leave it out on long titles.

Do I need one h1 per page?

One is the cleanest structure and it describes the page in the reader's words. Then use h2 and h3 as a genuine outline, because that outline is what both readers and assistants scan.

Does the analyser see the same page Google does?

Not always, if your page relies on JavaScript. That is why URL Inspection is on the list: it shows the rendered HTML Google holds for your URL, which settles the question.

How often should I reanalyse a page?

Check your top pages quarterly, and immediately after a redesign, a platform change or a drop in impressions. Pages decay quietly, and a quarterly pass catches it while it is cheap to fix.

What single change moves pages most often?

Rewriting the first paragraph so it answers the query directly, and the title so it matches what was asked. It costs ten minutes and it is the edit that works most reliably across sites.

Check nine: what an assistant says about this page

The eight checks look at the page itself: title, headings, links, speed, markup. Check nine looks at the page's reputation in a system you cannot crawl. Ask an assistant the question the page was written to answer, and see whether it quotes you, quotes a rival, or writes an answer with no source you recognise.

AI Knows Us runs that ninth check for a list of pages and queries, so it can sit beside the analyser instead of replacing it. The first pass costs you a few minutes and one browser tab.