Crawling and Indexing

Fixing Broken Links and Error Pages During an Audit

How to locate dead links and error pages on your site, and how to decide whether each one should be repaired, redirected or simply removed.

On this page
  1. Why broken links matter
  2. The common error types
  3. Where to find them
  4. Decide what to do with each one
  5. A useful error page
  6. Prioritize
  7. After the fixes

A broken link points to a page that no longer works. For visitors, it is a dead end that can feel careless and may send them away. For search engines, a few missing pages are normal and harmless, but a large number of broken internal links wastes crawling and can leave important pages poorly connected.

Not every error needs fixing. The skill in an audit is telling which ones do.

The common error types

Different status codes mean different things:

  • 404 (not found). The page does not exist at this address. This is the usual “broken link” result.
  • 410 (gone). The page was removed on purpose and will not return. Search engines treat it much like a 404, sometimes dropping the page a little sooner.
  • 403 (forbidden) and 401 (unauthorized). Access is blocked or requires a login.
  • 5xx errors. The server failed. These point to a server problem rather than a missing page, and persistent ones deserve prompt attention.
  • Soft errors. A page says “not found” to visitors but returns a normal success code. These are covered separately, but you should watch for them during the same review.

Where to find them

Use several sources, because each sees something different.

  1. A site crawl. A desktop crawler can follow your internal links and report every link that leads to an error, along with the page it was found on. This is the most useful source for internal links.
  2. Search tools reports. Coverage or indexing reports often list URLs that returned errors when search engines tried to fetch them. These can include old addresses that other sites still link to.
  3. Analytics. Look for pages that visitors reached and that show your error page, if the error page can be identified by its title or address.
  4. Server logs. If you can read them, logs show which missing addresses are actually being requested, by people or by crawlers.
  5. Your own outbound links. A crawl can also check external links. Pages elsewhere move or disappear over time, so outbound links go stale.

Decide what to do with each one

For every broken address, ask two questions: where is the link coming from, and does anything useful still exist for this topic?

If the link is on your own site and the target page exists at a different address, or the link was mistyped, fix the link itself. This is the cleanest outcome, because it avoids an extra redirect and keeps internal links tidy.

Redirect the old address

If the page moved or was replaced by something equivalent, set up a permanent redirect from the old address to the closest matching page. This helps visitors who arrive from bookmarks or other sites, and it passes along much of the value those links carried. Redirect to a genuinely relevant page. Sending every missing page to the home page is generally a poor experience and may be treated by search engines as a soft error. The mechanics are covered in the article about redirect chains.

Let it return an error

If a page has no equivalent and no one needs it, it is fine for it to return a 404 or 410. Search engines expect this. What matters is that you remove internal links that still point to it, and that the error page itself is helpful.

Bring the page back

If a removed page was still useful, had good links pointing to it, or receives steady visits, consider restoring it. Sometimes a deletion was a mistake.

For outbound links, either update to a working replacement, find an alternative source, or remove the link and adjust the sentence. Do not leave visitors with dead references.

A useful error page

Even a correct 404 should help. A good error page tells the visitor plainly that the page was not found, keeps the normal site navigation, offers a search box or links to main sections, and, crucially, returns an actual 404 code. A friendly design that returns a 200 code creates soft errors.

Prioritize

Not all broken links are equal. A sensible order is:

  1. Server errors on important pages.
  2. Broken links in the main navigation, footer and templates, since they appear on every page.
  3. Broken links on high-traffic or high-importance pages.
  4. Old URLs that still have external links or steady visits, which are redirect candidates.
  5. Everything else, in batches.

For large sites, grouping by pattern helps. A single template mistake can create thousands of broken links, and one fix removes them all.

After the fixes

Recrawl to confirm the errors are gone and that no new redirect chains were introduced. Keep a record of the redirects you added, because they should usually stay in place for a long time. Broken links appear again whenever content is moved, so checking for them belongs in your regular maintenance routine, and it matters especially around a site migration, when many addresses change at once.

The complete guide to crawling and indexing

Crawling and Indexing Explained for Site Owners
Read the guide

Keep exploring

Guides from other topics

All topics →