Recovering the links you earned and then lost

The largest recoverable block on most sites we look at is not on anybody else's server. It is in your own redirect map. After a replatform, somewhere between a fifth and a half of the external URLs pointing at a site land on a 404, a chain of three or more hops, or a blanket redirect to the homepage that passes nothing you should plan around. That part is recovered by editing rules, with nobody contacted and no goodwill spent, which is why we do it first and why it is the cheapest work in this listing. Order of work, cheapest first: One. Reconcile three sources: the Search Console links export, lost referring domains in Ahrefs over 24 months, and server log 404s that carry an external referer. The log file finds links the tools never crawled, and on older Czech domains that is a meaningful share. Two. Map each dead target to its closest live equivalent. Not the homepage. Where no equivalent exists and the dead page earned real links, we recommend rebuilding the page rather than redirecting it somewhere convenient, and we say which pages are worth the rebuild. Three. Collapse chains to a single 301, kill loops, and find redirects that cross protocol and host twice, which is the standard artefact of two migrations layered on each other. Four. Decide on deliberate deletions. A discontinued product line that carried links from twenty suppliers is a decision, not a cleanup task, and it is yours to make with the numbers in front of you. Then, and only then, outreach: Five. Unlinked mentions, found through the brand string in Ahrefs Content Explorer and Talkwalker Alerts, then filtered twice: pages with any organic traffic at all, then pages where adding a link genuinely helps that page's reader. What survives both filters is usually a short list, and a short list is what gets a reply. We approach the author, in their language, not a generic editorial inbox. Six. Broken-link building on other people's 404s. We do very little of it and here is why: you are asking an editor to repair a page they have forgotten exists, in exchange for pointing at you. It converts poorly and it spends a contact we would rather spend on something the editor gets value from. Where we do use it, it is on pages the editor has updated in the last year, which is a small set. What we will not do: Send templated broken-link mail at volume. Scrape resource-page lists and mail them. Push redirect rules to your production server ourselves unless your engineers specifically ask us to, because a redirect map is a change with blast radius and it belongs in your release process. Count a recovered link before it re-appears in your own Search Console data. Not for you if: You have never migrated, never rebranded, never deleted a section, and your brand is not mentioned anywhere you are not already linked. There is nothing here to reclaim and the honest answer is that a link gap analysis or placement work will serve you better. Say so on the enquiry and we will tell you which one. What you receive: A redirect map as CSV with source, target and the referring domains each row recovers, in the format your stack ingests: an nginx map file, a Cloudflare bulk redirect list, or a platform CSV. A ranked rebuild list where redirecting is the wrong answer. The unlinked mention list with the two filters applied and the ones we contacted marked. And a re-measurement at 60 days against the same Search Console export we started from, so the recovery is a number you can check rather than a claim we make.