Link Reclamation: Lost Backlinks and Unlinked Mentions

In the reclamation audits we have run on sites older than five years, the recoverable pool has come in between forty and a hundred and forty referring domains: links the site earned and then lost to a migration that pointed everything at the homepage, a CMS change that dropped a path segment, or a page deleted because it was not converting. Roughly two thirds of that pool needs no outreach at all. The fix is a redirect or a restored page, and it ships inside a fortnight. Finding what was lost: Four sources, cross-referenced. Ahrefs Best by Links filtered to 404s and to 301s landing somewhere irrelevant. The Search Console crawl stats not-found report. Ninety days of server access logs filtered to external referrers hitting a 404, which catches links no backlink index has ever seen. And the Wayback CDX API to recover what the dead URL actually said. That last one decides everything after it. A link pointing at a 2019 guide to unfair dismissal timelines is not repaired by a redirect to a practice area page; Google reads that as a soft 404 and the link passes nothing. Restore or redirect: Where the archived page still holds an answer, we rebuild and republish it at the original URL with the content brought current and the revision date shown. Where the page was a product, an event or a role that no longer exists, the redirect goes to the closest page that delivers what the linking sentence promised. Redirect chains get rewritten at the origin rule rather than extended: we measure hop counts in Screaming Frog list mode, and anything past two hops is collapsed, because every hop is another place a future replatform silently drops the chain. The mentions half: Unlinked mentions come out of Ahrefs Content Explorer and a monitoring feed, then get filtered hard. We pursue a mention only where it sits in body copy on an indexable page and the publisher has a corrections process. Job boards echoing your own listing, press release syndication and aggregator reprints are skipped. The request goes to the writer with the exact sentence quoted and the URL that belongs inside it, and nothing else asked for. What we will not send: The other kind of broken link outreach, where you email strangers about a dead link somewhere on their site and propose a client page as the replacement, we do not do at all. The recipient can tell what it is, the reply rate does not justify what the volume does to your sending domain, and it is not reclamation. It is cold outreach in a favour's clothing, and it is frequently the reason a site's outreach replies had already dried up before we arrived. Who this is not for: A site launched in the last two years. There is nothing to reclaim, and an audit will tell you so at your expense. It is also wrong where nobody on your side can deploy a redirect map or restore a page, because the audit produces work for a developer and a finding nobody implements is worth nothing. What you get: A redirect map as CSV in a format your platform accepts, the restore-versus-redirect decision for every URL with the archived source attached to each, the outreach log for the mentions half, and a short migration note for your developers so the next replatform does not do it again.