Reclaiming Links Lost to Migrations and Rebrands

On a manufacturer's domain that has been through an acquisition and a replatform, I typically find 40 to 120 referring domains pointing at URLs that now return 404 or land on the homepage. The homepage case is the common one and the expensive one: a bulk 301 from every dead URL to the root is treated as a soft 404, and the link stops there. This is the piece of link work where most of the result comes from your own server rather than from anyone's goodwill, which is why it is the first thing I run for any company older than about fifteen years, and why I will tell a five-year-old DTC brand not to buy it. How the audit runs: Export of referring pages from Ahrefs filtered to 4xx and to redirect chains, cross-checked against Search Console's not-found report and, where you can give me them, a sample of server logs, because logs surface the URLs that still receive traffic and no tool lists. Then Screaming Frog in list mode across every referring page, to confirm the referring page itself is still live and the link is still on it. A third of them usually are not, and emailing about a link that no longer exists is how a link builder loses a publication's attention. Then the Wayback CDX API to recover what each dead page actually said, because a redirect that does not match the anchor's promise is a redirect Google discounts and a reader resents. The order of operations: Fix on your side first. Where a discontinued machine series holds thirty editorial links, it deserves a live page describing what replaced it, not a redirect into a category listing. Where a page can be mapped, it gets a specific 301 to the closest equivalent, one at a time, by hand. Only what cannot be resolved internally becomes outreach, and by then the list is short enough to write properly. Unlinked mentions: Google Alerts and Talkwalker, plus a standing set of site-restricted queries across the Italian trade press, plus the print-to-web archives where a lot of Italian coverage of manufacturers actually sits. Mentions get reported with the sentence and the outlet, and the request to link is only made where a link genuinely helps that reader. Nofollow mentions are reported separately and never counted as links, because they are still read by people who buy from you. What I refuse: No blanket redirect to the homepage to make a recovery report look complete. No emailing a webmaster about something a 301 on your own server solves. No counting a link twice when it was already resolving through a chain and I only shortened the chain. No chasing links from domains that have become link farms since they published — those are listed as leave alone, with the evidence, and occasionally one is worth disavowing. Who this is not for: Domains under about eight years old with no migration history: there is nothing to recover and I will say so after an hour of looking rather than sell you an audit. Companies that cannot make redirect changes, or whose platform is controlled by an agency that bills per ticket, because the fix has to be deployable. Sites mid-replatform: wait until the new URLs are stable, or better, call me before the migration, when this work costs a fraction of what it costs afterwards. What you receive: One CSV: every referring URL, its current status, the recovered original target, the action taken, and the outcome. Plus the count of what is unrecoverable and why — publication closed, page gone, editor left — because that number tells you what a future migration would cost you.