Website Health Check: A Plain-English Guide to Finding What Needs Fixing

What is a website health check?

A website health check is a structured review of whether a site can be accessed, crawled, indexed, understood and used as intended. It brings several kinds of evidence together:

• a crawl of the site's internal links and pages;
• Google Search Console indexing and performance data;
• analytics and conversion data;
• manual checks of important page templates and user journeys;
• business knowledge about which pages and actions matter most.

The aim is not to achieve a report with no warnings. Websites contain exceptions, and many crawler notices are informational. The aim is to find material problems, understand their cause and make the site measurably better.

An SEO health check concentrates on search access and signals, but a complete site-quality review also considers usability, accessibility, content accuracy and whether important actions work.

What should you prepare before checking a site?

Define the purpose and scope

Record what the website needs to achieve and which sections are included. A small brochure site, an ecommerce store and a publishing platform have different critical paths.

List the pages that matter most, such as:

• the homepage and main category pages;
• important product or service pages;
• high-traffic landing pages;
• contact, signup and checkout routes;
• legal or support pages required by the business.

This list prevents a sitewide average from hiding a serious problem on one valuable page.

Record recent changes

Note migrations, redesigns, template releases, domain changes, CMS updates and large publishing projects. Dates help you connect new crawl patterns or traffic changes with plausible causes.

Establish a baseline

Save the first crawl rather than treating it as a disposable scan. Record the crawl date, scope, settings and important counts. The baseline lets you distinguish long-standing conditions from new regressions.

Can important pages be found?

Begin with discovery. A page cannot earn search visibility or help a visitor when no useful route leads to it.

Check navigation and crawl depth

Confirm that important sections are reachable through ordinary HTML links. Pages that require an internal search, a form submission or an interaction a crawler cannot perform may be difficult to discover.

Review pages several clicks from the homepage. Depth is not a ranking score by itself, but an unexpectedly deep page can reveal weak navigation or missing contextual links.

Find orphan and weakly linked pages

Compare the crawl with XML sitemaps, analytics, Search Console and CMS exports. A URL present in those sources but absent from the internal crawl may be orphaned.

One incoming link is not always enough. Check whether the link appears in a useful context and whether visitors can understand why they should follow it.

Inspect the XML sitemap

A sitemap should contain preferred, indexable URLs that return successful responses. Remove redirects, errors, non-canonical variants and pages deliberately excluded from search.

Can search engines index the right pages?

Discovery and indexing are separate. A crawler may reach a URL that the site then tells search engines not to index.

For important pages, check:

• the HTTP status is 200 OK;
• robots.txt does not block necessary crawling;
• no unintended noindex directive is present;
• the canonical points to the intended URL;
• the page is included in the appropriate sitemap;
• the content is available in rendered output when JavaScript is involved.

Also review the opposite problem: low-value filters, search results, duplicates or staging pages being left indexable.

Use Search Console to compare the site's declared signals with Google's processed view. A crawl shows what the site currently publishes; Search Console shows what Google has encountered and indexed. A regular website health check should include indexability and broken internal links.

Do the pages and links work?

Review response codes

Group URLs by status rather than looking at errors one at a time.

A successful status does not prove the page is healthy. Soft 404s, empty templates and error messages can still return 200 OK.

Test internal links

Fix broken links at their source. When a close replacement exists, update the link directly rather than relying on a redirect. Check navigation, breadcrumbs, footer links, related-content modules and links inside body copy.

Test important actions manually

A crawler cannot complete every user journey. Submit forms, use site search, add products to a basket, request password resets and test contact routes where relevant. Check both successful and error states.

Are page signals clear?

Search engines and visitors use page labels to understand what a URL offers.

Review:

• unique, descriptive title tags;
• one clear primary heading where appropriate;
• headings that reflect the structure of the content;
• meta descriptions that set an accurate expectation;
• canonical URLs that agree with internal links and sitemaps;
• descriptive link text;
• structured data that matches visible content.

Duplicates need interpretation. Repeated titles on a paginated series may require a different treatment from identical titles across unrelated service pages. Inspect patterns by template before editing individual URLs.

Are the pages useful to visitors?

Technical access is necessary, but it does not make a page worthwhile.

Ask whether each important page:

• answers the need implied by its title and links;
• is accurate and current;
• makes its main action clear;
• works on small screens and with a keyboard;
• uses readable text, meaningful headings and useful image alternatives;
• avoids intrusive elements that block the main content;
• gives visitors a sensible next step.

Use analytics to find exits, failed conversions or unusual behaviour, but do not treat one metric as a verdict. A contact-confirmation page may have a high exit rate because it completed its job.

How should you prioritise the findings?

Rate each issue across four questions:

1. 01
   Impact

   What happens to visitors, search engines or the business?

2. 02
   Scale

   How many pages or journeys are affected?

3. 03
   Confidence

   How strong is the evidence that this is a real problem?

4. 04
   Effort

   How difficult and risky is the fix?

Prioritise root causes. One faulty template may create thousands of symptoms, while one broken link on a critical conversion path can matter more than a large low-value count.

Run a website health check with Clearcrawl

How do you verify the work?

Verification is part of the fix.

• Recrawl the affected scope using comparable settings.
• Confirm that the intended pages, not merely the issue count, changed.
• Test a small sample manually.
• Inspect important URLs in Search Console when indexing was involved.
• Monitor search and conversion trends over an appropriate period.
• Record anything deliberately accepted so it is not reopened every month.

A falling warning count is useful only when the underlying outcome improves.

Common website audit mistakes

Frequently asked questions