How to Find and Fix 404 Errors

What is a 404 error?

A 404 error is an HTTP response that says the server could not find a resource at the requested URL. The website still responded, but the specific page, image or file was not available.

You might see a 404 when:

• A page has been deleted.
• A URL changed without a redirect.
• An internal or external link contains a typo.
• A product, article or file was unpublished.
• A CMS creates a link to content that no longer exists.
• Someone requests a URL that was never real.

A proper 404 response is often the correct outcome. If a page has gone permanently and there is no useful replacement, the server should say that it cannot be found. Search engines can then stop treating the URL as a live page.

What is the difference between a 404 and a soft 404?

A normal 404 page returns the HTTP status 404 Not Found. The visible design can still include navigation, search and helpful links, but the server response remains 404.

A soft 404 usually returns a success response such as 200 OK while showing an error, an empty result or a page with little meaningful content. The words “page not found” may appear on screen even though the server tells crawlers that the request succeeded.

This mismatch can confuse visitors and search engines. Google may classify the URL as a soft 404 and exclude it from search.

Soft 404s can also occur when an otherwise useful page fails to load its main content. A broken database connection, missing JavaScript resource or empty product template may leave Google with a page that looks like an error even though it returns 200.

Use the response and the actual content together:

Do 404 errors harm SEO?

The existence of 404 URLs does not by itself damage a website’s rankings. Google advises that most 404s can be ignored when the URLs genuinely should not exist.

They become worth fixing when they interrupt something useful:

• Internal links: visitors and crawlers follow a link from your own site into an error.
• XML sitemaps: a sitemap lists a URL as a page you want crawled, but the URL returns 404.
• Moved content: an old URL has links or traffic and a clear new replacement, but no redirect connects them.
• Important assets: missing images, scripts or documents break the page or its meaning.
• Repeated patterns: a template, navigation rule or deployment creates broken URLs across many pages.

Do not measure success by making every 404 disappear from every report. People mistype addresses, other websites link incorrectly and crawlers revisit old URLs. Focus on the 404s your website creates, links to or still presents as important.

How to find 404 errors on your website

Crawl the website

A site crawl follows internal links and records the response for each URL. This is the fastest way to find 404 pages that visitors and crawlers can reach from your live content.

Review both sides of the problem:

• The broken destination returning 404.
• Every source page linking to that destination.

The source is often where the practical repair belongs. Correcting a mistyped link may be better than redirecting a URL that never should have existed.

Check the XML sitemap

An XML sitemap should contain current canonical URLs that you want search engines to consider. Remove 404 URLs from it, then update the system that generated them so they do not return on the next publish.

Review Google Search Console

The Page indexing report can show URLs Google found that returned 404 or appeared to be soft 404s. The examples may include historic addresses and links from outside your control, so do not treat the list as a demand to repair every URL.

Inspect priority examples to see when Google last crawled them and whether the current response has changed.

Check analytics, logs and external links

Analytics can reveal error URLs that still receive visits. Server logs can show repeated requests from users and crawlers. Backlink data can identify old URLs with valuable external links that may deserve a relevant redirect.

Combine those signals with the crawl. A 404 with no links, visits or intended content is different from a former product guide that still attracts visitors.

Check this in Clearcrawl

Clearcrawl helps separate isolated old URLs from live internal problems. You can trace a broken destination back to its source pages, decide on the right repair and run a fresh crawl to confirm that the route now works.

Which 404 errors should you fix?

Use the URL’s history and intended outcome to choose the action.

A redirect should take the visitor to a page that satisfies the same or a very similar need. Sending every missing URL to the homepage creates a confusing experience and can be treated as a soft 404. If a missing URL has a close replacement, learn when a redirect is appropriate; if the source link itself is wrong, follow the broken-link repair steps.

How to fix 404 errors

1. 01
   Confirm the exact URL and response.

   Open the URL and check its HTTP status. Pay attention to the protocol, hostname, capitalisation, path, trailing slash and file extension. Similar-looking URLs can behave differently.

   Confirm whether the page is truly missing or whether the server, CMS or application route is failing unexpectedly.

2. 02
   Decide whether the content should return.

   Restore the page when it was removed by mistake or still has a useful purpose. Keep the original URL where possible, especially when it already has relevant links or traffic.

   If the page no longer deserves to exist, identify whether a close replacement is available. Do not create a redirect simply to make the error count smaller.

3. 03
   Apply the appropriate repair.

   • Restore: publish the intended content at the same URL and return 200 OK.
   • Redirect: use a permanent server-side redirect when the content has genuinely moved.
   • Correct the link: change internal navigation or page content to point directly to the right destination.
   • Retire: return 404 or 410 when there is no suitable replacement.

   Avoid redirect chains. Update internal links to point straight to the final page even when a redirect protects old external routes.

4. 04
   Clean up supporting signals.

   Remove retired URLs from XML sitemaps, canonical tags, hreflang annotations and structured data. Update navigation, related-content blocks and templates that still generate the broken link.

   For redirected content, list and link to the final canonical URL rather than the old address.

5. 05
   Improve the error page for visitors.

   A custom 404 page can help someone recover. Include clear site navigation, a search option when useful and links to popular or closely related sections. Explain plainly that the requested page could not be found.

   Keep the correct 404 status. A polished design should not turn the response into 200 OK.

How to verify the repair

Check the final behaviour rather than assuming that a saved rule worked.

• Open the original URL and confirm the intended response.
• If redirected, follow the route and make sure it reaches one relevant 200 OK destination without a chain or loop.
• If restored, confirm that the complete content loads and the page is indexable when appropriate.
• Run another Clearcrawl crawl and check that internal broken links have cleared.
• Confirm that retired URLs no longer appear in the XML sitemap.
• Inspect important repaired URLs in Search Console after Google has had time to recrawl them.

Historical 404s may remain in Search Console for a while. Judge the repair by the current response and fresh crawl evidence before waiting for Google’s report to update.

Common mistakes

Frequently asked questions