Link reclamation repairs backlinks you previously earned by fixing a broken destination or asking a publisher for a specific correction. Start by proving the original link, then check the referring page and destination separately. A missing row in a tool is a reason to investigate, not proof of removal. The useful result is a working, relevant route for readers; restoring a link does not guarantee restored rankings or a particular transfer of search signals.
Use the free backlink research guide to assemble an initial inventory, and the backlink monitoring hub to organize observations over time. This guide covers what to do with a suspected break once it has a concrete referring page and destination.
the reclamation workflow
- record the original relationship. Save the referring page, linked URL, surrounding passage and date of the observation. A referring domain alone is a research lead.
- check both ends of the link. Inspect the referring page and your destination separately. Record the current link, HTTP outcome and final destination, keeping inaccessible cases unresolved.
- choose the smallest useful repair. Restore useful content, redirect a permanent move to an equivalent page, or request a precise correction from the referring publisher.
- verify the reader journey. Open the referring page again, follow the link and confirm that the destination answers the cited question. Record what changed without promising ranking recovery.
Keep this repair queue separate from new prospecting. An article that once linked to your study has an existing citation to investigate. An article that merely mentions your brand might be worth approaching, but you are asking for a new editorial decision. A competitor backlink is another acquisition lead. Our guide to finding link prospects with ai addresses that research task; it should not become evidence that someone owes you a replacement link.
establish what existed
Make one record per referring-page and destination pair. Save the exact referring URL, the old destination, anchor text, surrounding sentence, observation date and evidence source. An older page capture or a dated page-level export can support the history. A domain-level graph relationship cannot, by itself, identify the sentence or URL that needs repair. Do not fill those gaps by guessing which article looks most relevant.
Next, open the referring page and search for the destination and the cited passage. Record whether the link remains, points somewhere else, or cannot be checked. A login wall, blocked request or server failure limits what you know. Google documents different crawler outcomes for HTTP errors and redirects; a failed fetch is not equivalent to inspecting a successful page response and finding the citation absent.
Check your destination independently. Record its HTTP response, any redirects and the final page. Read the content rather than stopping at a successful response. A page can load while the table, download or explanation cited by the publisher has disappeared. Conversely, the old URL may already lead to the correct replacement while a monitoring report still flags a change. These require different actions.
Add a short diagnosis in plain language: for example, “source still links to the retired guide; our destination returns an error.” Record the check date and who controls the affected side. That sentence makes the next action clearer than a generic lost-link label, and lets another person repeat your check without needing your interpretation of a dashboard score.
choose a repair from the evidence
The following matrix is our operational decision aid, not a Google ranking rule. Select a row only after checking the page-level facts. Prioritize cases where the original reference remains useful, the failure is reproducible and a responsible person can make the repair.
| observed case | appropriate action | evidence needed to close |
|---|---|---|
| source link remains; your page has an accidental error | restore the original useful page or fix its serving problem | the old URL loads the resource and the source link reaches it |
| source link remains; content permanently moved | redirect the old URL to the genuinely equivalent replacement | the redirect reaches the replacement and the cited material remains available |
| source link contains a confirmed URL typo | request a correction to that exact destination | the edited link points to the intended resource |
| publisher removed the link but retained the citation | ask whether restoring the useful reference fits the article | a current page inspection shows the restored link, or the editor declines |
| source cannot be accessed during the check | retain an unresolved status and check later when access is available | a readable source page supports a more specific diagnosis |
| your old resource is obsolete with no equivalent | consider rebuilding it if useful; otherwise explain the obsolete reference | the reader is no longer sent to misleading or missing information |
| brand mention exists, but no earlier link is established | move it to a separate acquisition queue | the request is described as a new link opportunity, not restoration |
Do not rank this queue solely by a third-party authority score or the number of repeated links. A relevant citation sending readers to a missing technical resource may deserve attention before a large collection of template links. Consider the usefulness of the destination, the clarity of the failure and the effort required. Avoid paying for restoration as an SEO tactic: Google's spam policies identify manipulative link practices, and reclamation is not a justification for manufacturing endorsements.
repair content you control
If a deploy accidentally removed an otherwise useful page, restoring that page can be the simplest repair. Confirm that the restoration includes the cited material, not just the original title. Keep the source URL stable where practical. A reader following a reference to a checklist should arrive at the checklist, not an empty landing page explaining that your site has changed.
When content has permanently moved, Google recommends a permanent server-side redirect for that move. Map the old URL to a destination that genuinely serves the original purpose. Verify the final location and avoid an unnecessary chain of intermediate addresses. A redirect response is a technical observation; it is not proof of indexing or a promised ranking result.
Do not send every retired URL to the homepage. A generic page rarely replaces a cited dataset, how-to guide or product manual. If the cited resource has no successor, decide whether recreating it would still help readers. Otherwise, leave an honest retirement outcome and explain the situation to the publisher. An irrelevant destination can make the link technically clickable while leaving the original problem unresolved.
request a specific editorial correction
Contact the person responsible for the article when the fix belongs on their site. Use an editorial contact, named author or clearly relevant support channel. Include the exact passage and a checked destination. Do not require the recipient to search their entire site or accept a vague assertion that their links are broken. Explain your relationship to the resource so the request is transparent.
This copyable template is for a verified move, not an unlinked mention. Replace every bracketed field with evidence, remove anything you cannot establish and adjust the request if the publisher intentionally removed the citation. Keep the existing anchor when it remains accurate; do not turn a correction into a demand for commercial keywords or an unrelated product endorsement.
subject: broken reference in [article title] hello [editor name], in [referring article URL], the paragraph beginning “[short passage]” links “[existing anchor]” to [old destination URL]. that URL currently returns [verified outcome, checked on date]. the original [resource name] has moved to [equivalent destination URL], which contains [specific material cited in your article]. would you consider updating that one destination? the existing anchor can stay if it still describes the resource accurately. thank you, [name and relationship to the resource]
Offer the correction rather than treating it as an obligation. The publisher may have changed their editorial policy, rewritten the article or judged the resource outdated. If they decline, record that outcome and stop. If there is no reply, retain the case as requested rather than recovered. A message sent successfully does not prove that anyone read it or changed the page.
an illustrative case from flag to closure
Consider this hypothetical case; the domains, paths and dates are illustrative, not customer results. An older capture of publisher.example/research-notes shows a link to resource.example/guides/citation-checklist. A later inventory flags the relationship for investigation. That flag starts a check; it does not establish why anything changed.
On the illustrative check date, the publisher's article still contains the same citation and old URL. The destination returns a 404. The resource owner finds that a site migration moved the complete checklist to resource.example/resources/citation-checklist. Reading both the saved description and replacement confirms that the checklist still answers the cited question. This is a permanent move with an equivalent destination, so the matrix points to a repair on the resource owner's site.
The owner configures a permanent redirect and checks the old URL again. It now leads to the replacement checklist. They follow the link from the publisher's article and confirm the same reader journey. They also send a concise editorial correction using the template, so the publisher can link directly if they choose. The redirect repair can be recorded as verified while that separate editorial request remains pending.
If the publisher later updates the URL, a fresh inspection supports closing the editorial request too. If they never reply, the owner should not claim an editorial change. If the replacement had been only a homepage, the equivalent-content test would have failed and the proposed redirect would need reconsideration. None of these observations establishes a ranking increase; the documented outcome is a repaired reference.
verify and close the case
After any change, repeat the original failing journey. Open the source, inspect the destination it currently names and follow it through to the final content. Record the check date, response outcome, relevant passage and person who made the repair. Keep technical repair, editorial update and unresolved access as separate statuses. That prevents a completed redirect task from silently becoming a claim that a publisher changed their article.
Keep counts at the same unit. Repairing one destination can help several referring pages, but it remains one destination repair with several verified source journeys. Do not inflate that into several newly earned endorsements. For periodic follow-up, preserve your original observation and compare later checks; a new inventory can suggest another review without overriding current page evidence.
crawlgraph can help build a referring-domain research list from its available Common Crawl graph data. It does not replace inspecting the exact source page and destination for this workflow. Check current plans and export access if you need an inventory for a larger review. Keep acquisition research, such as competitor opportunities, separate from this repair ledger so your outreach accurately describes what you found.
faq
is an unlinked mention link reclamation?
For this workflow, reclamation means repairing a relationship that previously included a link. A brand mention with no evidence of an earlier link is an acquisition opportunity. Keep it in a separate queue and do not tell a publisher that they removed something you cannot establish.
should every broken backlink redirect to the homepage?
No. Choose an equivalent destination when content permanently moves. A homepage usually does not answer a specific citation. If no useful equivalent exists, consider restoring the original resource or asking the publisher to remove an obsolete reference rather than inventing a replacement.
does reclaiming a link guarantee its previous rankings?
No. A working link and a useful destination are observable outcomes. Indexing, signal handling and rankings are separate outcomes that cannot be guaranteed by a redirect or editorial correction.
documents crawlgraph backlink workflows, open-data methods, and the limits of each report.
plus one when a new common crawl release lands. that is all.