How to Find and Reclaim Broken Backlinks

· Updated

A broken backlink is an external link that still points to a deleted or mistyped URL on your site. The practical fix is not to redirect every 404. It is to find dead URLs that still have real, relevant links, verify that those links exist, and then choose the closest valid resolution:

  1. 301 redirect the old URL when the content permanently moved or has a close replacement.
  2. Restore the page when people still need the original resource and no equivalent exists.
  3. Ask the linking site to update the URL when its link contains a typo or the source page should point directly to the new location.
  4. Leave a true 404 or 410 when there is no relevant replacement and the backlink has no recoverable value.

That four-way decision is the entire workflow. The rest of this guide shows how to collect reliable inputs, prioritize without treating a vendor authority score as truth, implement the fix, and verify what actually changed.

Use this sequence for a first audit or a recurring monthly review:

StepActionRequired evidenceOutput
1Export broken targets and their referring pagesBacklink-tool export or a Search Console cross-checkRaw candidate list
2Verify every priority rowLive source page, live target response, link destinationConfirmed broken backlink
3Group rows by dead target URLNormalized URLs and unique referring domainsOne decision per dead page
4Prioritize recoverable valueRelevance, referral demand, number of distinct sources, link contextOrdered work queue
5Choose 301, restore, outreach, or no actionOriginal intent and closest current destinationDocumented fix
6Test and monitorHTTP response, final destination, analytics and tool recrawlVerified resolution

Start with the dead target URL, not with the authority score of the linking domain. One accurate redirect for a retired guide may resolve links from several sites at once. Conversely, a high metric attached to an irrelevant or removed source page does not create a useful recovery opportunity.

No single data source is complete. Ahrefs and Semrush maintain their own backlink indexes, while Google Search Console exposes a sample of links Google has discovered over time. Use the tool you already have, then verify the candidates on the live web before changing your site.

In Site Explorer, enter your domain and open Backlink profile > Broken backlinks. According to the Ahrefs Broken backlinks documentation, this report lists referring pages that link to 404 pages on the target site, together with the anchor and surrounding text. You can filter or sort the report and export the rows.

For page-level prioritization, open Pages > Best by links, apply the HTTP code: 404 not found filter, and inspect the links to each dead target. This groups the opportunity around the page you can fix rather than showing every backlink as a separate task.

Important limitation: the status reflects Ahrefs' latest crawl, not necessarily the response a person or Googlebot receives now. Recheck both the referring page and target URL before implementing anything.

Semrush: Filter Target URL Errors

In Backlink Audit, use the Target URL Error filter. Semrush defines this as a backlink whose target page was unavailable when Semrush last crawled it, usually because the target returned a 404. The filtered rows can be exported as CSV; see Semrush's Target URL Error documentation.

You can also open the Target Pages report and filter for broken targets. That view groups active, broken, and lost backlink data by destination page. Semrush explains that a target is counted as broken when the pages receiving its backlinks returned error codes during its reporting window in the Target Pages documentation.

Treat Semrush's Authority Score and toxicity labels as vendor-specific diagnostics, not Google rules. They can help break a tie after you have checked relevance, source quality, and real referral value, but they should not decide the fix by themselves.

Google Search Console: A Free but Partial Cross-Check

Search Console does not have a complete "broken backlinks" report. Use two exports and compare them:

  1. In Indexing > Pages, export the URLs under Not found (404).
  2. In Links > External links > Top linked pages, export the external-link data.
  3. Match the two exports by target URL.
  4. Open each candidate in the Links report to inspect linking sites and, when available, linking pages.

This method finds overlap between 404s Google encountered and the link sample shown in Search Console. It will miss some backlinks: Google's Links report documentation explicitly says the report is a sample, may include links that were later removed, and does not report whether a link uses nofollow.

Do not fix a URL merely because it appears under Not found. Google says most 404s are not worth fixing when the URL should not exist; redirect only when content moved or a clear replacement exists. The Google 404 guidance also warns against redirecting unrelated missing pages to the homepage because that can be treated as a soft 404.

Verify Candidates Before You Prioritize Them

Backlink indexes can be stale, servers can fail temporarily, and a source page can remove or change a link after the last crawl. For each candidate near the top of your queue, record four checks:

  1. Does the referring page load? A backlink on a source page that now returns 404 is not recoverable through changes on your site.
  2. Is the link still present? Search the page source or rendered page for your hostname and confirm the exact href.
  3. Does your target still fail? Check the response and follow any redirect chain.
  4. What did the link promise? Read the anchor and surrounding sentence. This tells you what destination would satisfy the visitor.

For a quick HTTP check, replace the example URL with your target:

curl -sSIL --max-time 15 "https://example.com/old-page"

Use a browser as well. A 200 response can still render an error message, and a bot may receive a different response because of a firewall or rate limit.

Keep candidate, confirmed, and resolved as separate statuses. This prevents a stale tool row from becoming an unnecessary redirect.

Prioritize the likelihood that a fix helps real users and consolidates a genuine page move. Do not sort once by DA or DR and call the result a strategy.

Primary Criteria

1. Destination match. Is there an existing page that fulfills the same purpose? An exact successor makes a redirect both useful and easy to justify. An unrelated homepage does not.

2. Live referring domains. Count distinct, verified sites still linking to the dead target. Ten links from one site-wide footer are not the same opportunity as editorial links from ten independent sources.

3. Topical and contextual relevance. Read the source page, anchor, and surrounding text. A contextual citation from a relevant article is usually more actionable than a scraped directory entry, regardless of a domain-level score.

4. Observed referral value. If analytics or server logs show visits to the dead URL from a source, prioritize it. Observed visits are stronger evidence than a third-party traffic estimate.

5. Business and user value. A dead documentation page, product comparison, research asset, or conversion-supporting guide may deserve attention even with few backlinks because visitors arriving there have a clear task.

6. Fix confidence and effort. An exact old-to-new URL mapping is a quick win. Rebuilding a resource with unclear historical intent is a larger editorial decision and should require stronger evidence.

Useful Tie-Breakers, Not Ground Truth

Third-party page and domain metrics can help order otherwise similar candidates, but each provider calculates them from its own index. Use them only after the primary criteria above. The same applies to estimated traffic and labels such as "follow" or "toxic": verify important links directly and consider why the link exists.

Copyable Prioritization Template

Create one row per dead target URL, then attach its referring pages in separate rows or notes:

FieldWhat to record
Dead target URLExact URL returning the error
Current response404, 410, 5xx, redirect, soft 404, or live 200
Original page purposeInferred from your CMS, archive, anchor, and link context
Verified live referring domainsCount after checking source pages
Best source contextEditorial citation, resource list, navigation, profile, scraper, other
Observed referral visitsAnalytics or server-log count and date range
Closest live destinationURL or "none"
Match qualityExact, same intent, partial, unrelated
Recommended action301, restore, outreach, leave 404/410, investigate
Owner and due datePerson responsible and review date
VerificationFinal status, destination, test date, notes

A simple queue can use P1 for verified links with an exact fix or observed visits, P2 for relevant links that require restoration or outreach, and P3 for uncertain or low-value candidates. This is a workflow label, not a claim about ranking impact.

Decision Tree: 301 vs Restore vs Outreach

For each confirmed dead target, follow this decision tree:

Did the original content permanently move to an equivalent URL?
├─ Yes → 301 redirect to that URL.
└─ No
   ├─ Is there a live page that satisfies the same user intent?
   │  ├─ Yes → 301 only if the match is genuinely close.
   │  └─ No
   │     ├─ Is the original resource still useful and supportable?
   │     │  ├─ Yes → Restore it at the old URL.
   │     │  └─ No → Leave a true 404 or 410.
   │     └─ Is the external link merely mistyped?
   │        ├─ Yes → Redirect the common typo and request an update.
   │        └─ No → No outreach needed unless the source sends users.
   └─ Does a high-value source need a direct, current URL?
      ├─ Yes → Add outreach after the technical fix.
      └─ No → Verify and monitor the technical fix.

The branches can overlap. A common typo can receive a permanent redirect immediately while you also ask the publisher to correct its link.

Fix 1: Use a Permanent Redirect for a Real Move

Use a server-side 301 or 308 when the old resource permanently moved and the destination is equivalent or very close. Google's redirect documentation says permanent redirects are a signal that the target should become canonical; temporary redirects do not send the same canonicalization signal.

A good mapping preserves intent:

Dead pageDestinationDecision
/guides/link-audit-2024/guides/link-audit with the current guideRedirect
/docs/export-csvNew documentation for CSV exportRedirect
/blog/link-auditHomepageDo not redirect; unrelated
Discontinued event pageGeneral blog categoryUsually leave 404/410 unless a true replacement exists

Do not promise a guaranteed ranking increase. Google states that permanent redirects do not cause a loss in PageRank during a properly implemented move, but search systems still need to crawl and process the URLs, and relevance remains essential. Read the broader 301 vs 302 guide before choosing a temporary response for permanently removed content.

After implementation:

  • Confirm the old URL returns one permanent redirect, not a 200 error page.
  • Confirm the final destination returns 200 and contains the expected content.
  • Update internal links and the sitemap so they point directly to the destination.
  • Avoid unnecessary redirect chains; see the redirect chain cleanup guide.
  • Keep the redirect documented so a future cleanup does not remove it accidentally.

Fix 2: Restore a Resource That Still Deserves to Exist

Restore the page at its original URL when no close replacement exists but the original resource still serves a clear user need. This is often the right decision for an evergreen guide, research page, downloadable template, or documentation that earned contextual citations.

Before restoring, establish what the page was:

  • Check your CMS history, repository, backups, and redirect log.
  • Review anchor text and surrounding copy on the verified referring pages.
  • Use the Wayback Machine to inspect your own archived page when available.
  • Confirm that your team can keep the restored information accurate.

Restoration is not permission to publish stale material. Preserve the original purpose and URL, but update facts, screenshots, product steps, sources, and calls to action. If the old topic is no longer safe or useful to support, a correct 404/410 is better than thin placeholder content created only to retain links.

Outreach is appropriate when the linking page contains a typo, names an outdated product or document, or sends meaningful referral traffic and would benefit from a direct current URL. It is usually a supplement to a technical fix, not a reason to postpone one you control.

Use a short, specific message:

Subject: Broken link on [page title]

Hi [name],

Your page [source URL] links to [broken URL] in the section about [context].
That URL now returns 404. The current page is [correct URL].

Could you update the link when convenient? I checked the replacement and it covers the same topic.

Thanks,
[name]

Do not claim a partnership, traffic benefit, or result you cannot verify. One polite follow-up is enough. If you are looking for new opportunities on other sites rather than recovering links to your own pages, use the separate broken link building workflow.

A Reproducible Typo Example

Ahrefs documented a referring link that accidentally included a closing parenthesis in the destination URL. This is a useful example because the technical distinction can be checked without inventing recovery results.

As checked on August 24, 2026:

curl -sSIL --max-time 15 "https://ahrefs.com/site-explorer)"
# HTTP/2 404

curl -sSIL --max-time 15 "https://ahrefs.com/site-explorer"
# HTTP/2 200

Ahrefs describes the original typo in its broken-link repair guide. The evidence supports two actions: redirect the mistyped URL if you control the destination site, and ask the linking publisher to remove the extra character. It does not prove that a redirect produced a specific traffic or ranking gain, so none is claimed here.

This is the standard to use in your own audit: preserve the source URL, exact href, response headers, chosen fix, and verification date. Record measured outcomes later instead of estimating them in advance.

Verify the Fix

Verification has three layers.

1. Technical Verification

Test every changed URL immediately:

curl -sSIL --max-time 15 "https://example.com/old-page"

Record the first response, each redirect hop, final URL, and final status. Also open the old URL in a browser to confirm that a visitor reaches the intended content.

After the tool has had time to recrawl:

  • Re-run the same Ahrefs or Semrush export and compare by target URL.
  • Inspect important URLs in Search Console to confirm Google can fetch the redirect or restored page.
  • Check that a restored page is indexable and internally linked.
  • Treat a row disappearing from one vendor report as tool evidence, not proof of a ranking change.

There is no universal two-week or eight-week guarantee. Crawl and processing time varies by URL and site. Google notes that even site moves are processed per URL and can take weeks; monitor the actual responses and reports rather than setting a promised recovery date.

3. Outcome Verification

Use a consistent before-and-after window and record only observable measures:

  • Referral sessions to the restored or redirected destination.
  • Search Console impressions and clicks for the destination page.
  • Confirmed live referring pages.
  • Conversions attributable to those visits, if your analytics setup supports it.

Annotate the implementation date. Rankings and traffic change for many reasons, so describe correlation honestly rather than assigning every movement to a redirect.

Most recovery work can be avoided with a URL-change process:

  1. Export URLs that have verified external links before a redesign or migration.
  2. Create an old-to-new mapping and review partial or unrelated matches manually.
  3. Implement permanent redirects before launch.
  4. Update internal links, canonicals, and sitemap entries to the final URLs.
  5. Crawl the old URL list after launch and preserve the redirect map.
  6. Recheck external-link targets monthly or after major URL changes.

Use the pre-redesign link audit for a migration checklist and automatic broken link monitoring for recurring site checks. For Google's view of missing pages, see the Search Console 404 guide.

FAQ

No automatic penalty exists merely because external sites link to your 404 pages. Google says 404 responses do not harm a site's indexing or ranking when the URL should not exist. Recovery matters because a useful old URL may still receive visitors or have a relevant replacement, not because every reported 404 must be eliminated.

No. Redirect only when there is a close replacement or the content genuinely moved. Restore a still-useful resource when no equivalent exists. Leave a true 404/410 when the page is intentionally gone and has no relevant substitute.

Is DR, DA, or Authority Score the best priority metric?

No single third-party authority metric is a Google ranking metric or a complete measure of value. Prioritize verified source pages, topical relevance, distinct referring domains, observed referral visits, user value, and the quality of the destination match. Use vendor metrics only as tie-breakers.

No. Its Links report is a sample and may include historical links that no longer exist. Cross-reference the 404 and external-link exports, then verify the source and target live. Use a dedicated backlink index when you need broader discovery.

Should a permanent fix use 301 or 302?

Use 301 or 308 for a permanent move. A 302 or 307 is for a temporary redirect. Google follows both, but a permanent redirect is the appropriate canonicalization signal when the old URL will not return.

Review them after any migration, redesign, bulk deletion, or URL-structure change. A monthly check is a practical starting point for an active content site; a smaller stable site may need less frequent review. The useful trigger is change volume, not an arbitrary universal schedule.

Pavel Molyanov

Pavel Molyanov

Creator of Broken Link Checker

Content marketer with 10+ years of experience. Founder of a content marketing agency. Writing about SEO, content workflows, and website maintenance.

Broken Link Checker

Check Your Links in One Click

Broken Link Checker finds broken links and redirects on any page or across your whole website, and works in Google Docs and Sheets. Try 3 checks free, with no signup or card required.

More on broken links, SEO, and web maintenance.