Open any backlink tool and the “lost” filter is never empty. Links you paid for, earned through outreach or picked up without trying are quietly disappearing, and the referring-domain graph that looked so healthy last quarter has started to sag. The instinct is to treat every lost backlink as damage and to chase every one of them. That instinct is expensive and mostly wrong.
The short answer: losing backlinks is normal and continuous. Pages get deleted, rewritten, redirected and deindexed, and tools also drop links from their own indexes for reasons that have nothing to do with your site. The large majority of lost backlinks either never carried much weight or are still being credited to you through a redirect. A small minority matter, and a share of those are self-inflicted: your own URLs changed and nothing was redirected. The job is to tell these groups apart quickly, fix what is in your control, and spend outreach only on the links worth having back.
This guide covers how fast links disappear across the web, the difference between a link your tool lost and a link Google stopped counting, the reasons backlinks decrease (including the ones you cause), a step-by-step way to investigate a drop, and a test for deciding which lost backlinks are worth recovering.
How Fast Do Backlinks Disappear? The Baseline Nobody Plans For
Link loss is not an anomaly you troubleshoot once; it is a background rate you should budget for. Two large studies put numbers on it.
Ahrefs, in a study first published in 2022, examined links pointing at roughly 2.06 million websites using its crawl data from January 2013 onwards. Over that nine-year span, 66.5% of those links had rotted (the linking page was gone, dropped or no longer linked) and a further 6.45% were returning temporary errors. Adding links that still existed but could no longer count for ranking purposes (pages set to noindex, pages whose canonical had changed), the share classed as lost reached 74.5%. The study also found that sites with more live links tended to have lost a larger proportion of them (Ahrefs link rot study).
Pew Research Center approached it from the page side rather than the link side. Sampling pages from the Common Crawl archive, it found that a quarter of all web pages that existed at some point between 2013 and 2023 were no longer accessible by October 2023. For pages that existed in 2013, the figure was 38%; even for pages that existed in 2023, 8% were already gone within the year. On the pages that survive, 23% of news pages and 21% of government pages contained at least one broken link (Pew Research Center, 2024).
Two caveats before you turn those into expectations. The studies measure different things with different definitions of “dead”, and the Ahrefs figures describe what one crawler could and could not find, not what Google counts. The usable lesson is the shape, not the decimal: over a multi-year horizon, losing more than half of the links you once had is the ordinary outcome, and loss accelerates on larger sites because larger sites restructure more. If your link acquisition is flat while attrition runs at this rate, your profile shrinks even though nothing “went wrong”.
Lost to Your Tool vs Lost to Google: Three Layers of “Lost”
The word “lost” hides three different events, and most bad decisions about backlinks come from collapsing them into one.
- The tool stopped seeing the link. The crawler was blocked, hit an error, could not render a script-inserted link, or pruned the page from its index to save space. Nothing changed on the web. Ahrefs labels most of these “Dropped” or “Crawl error”, and in its own study those two labels accounted for more than half of all lost links.
- The link changed on the web. An editor removed it, the page was deleted, the page now redirects, or the page was set to
noindexor given a different canonical. Something real happened, but not every version of “real” costs you anything. - Google stopped counting it. This is the only layer that can affect rankings, and no tool reports it directly. You infer it from the type of change and from what Google has said about how it handles each case.
The table maps the common tool statuses to those layers. The third column is an inference from Google’s published guidance and public statements, not a Google report.
| What the tool reports | What actually happened | Likely effect in Google | Worth acting on? |
|---|---|---|---|
| Dropped / crawl error | The tool could not fetch or chose not to keep the linking page | Usually none; Google crawls independently | Verify manually; otherwise ignore |
| Link removed | Page is live, your link is gone | Lost once Google recrawls the page | Yes, if the page was worth having |
| Not found (404/410) | The linking page was deleted | Lost; Google removes 404 pages from its index over time | Rarely; the host has to restore the page |
| 301/302 redirect | The linking page moved | Often preserved if the destination still links to you; lost if it does not | Check the destination page |
| Noindex | Linking page excluded from search | Weakens over time; Google has said long-term noindex pages eventually stop having their links followed | Only if the page is important and the tag looks accidental |
| Not canonical | Page now points to another URL as canonical | Usually consolidated onto the canonical version | Check whether the canonical page links to you |
| Your target URL returns 404 | Your own page moved or was deleted without a redirect | Lost: Google does not count links to URLs that do not resolve | Yes, immediately; entirely in your control |
Three of the Google-side inferences are worth sourcing. On redirects, Google’s Gary Illyes stated in 2016 that 3xx redirects no longer lose PageRank, and Google’s documentation describes permanent redirects as a signal that the destination should be treated as canonical (Search Engine Roundtable; Google Search Central). On dead targets, John Mueller wrote in 2024 that if links point to URLs that return 404 on your site, “they’re already dropped” and “if there’s no indexable destination URL, there’s no link” (Search Engine Roundtable). On noindex, Mueller explained in 2017 that a page carrying the tag for a long time is eventually removed entirely, after which its links are no longer followed, making a long-standing noindex, follow effectively the same as noindex, nofollow (Search Engine Roundtable).
Why Backlinks Decrease: Who Caused It
Lists of “reasons you lost backlinks” usually mix causes you cannot influence with causes you created. Sorting by who is responsible is more useful, because it tells you where the fix lives.
Losses caused by the linking site
Editors prune old posts, consolidate articles, swap a citation for a newer source, or adopt a policy of stripping outbound links added by past contributors. Sites redesign and drop pages that did not fit the new template. Businesses close and domains lapse. Some of this is editorial judgement about your page; most of it is housekeeping that was never about you. Ahrefs’ breakdown put outright link removal at 34.2% of lost links and deleted linking pages at a further 4.11%. Guest-post and sponsored placements are especially exposed here, because they sit on pages nobody at the host has a reason to maintain. Our guide to vetting guest post sites covers the signals that predict whether a host keeps its archive.
Losses you caused
This category is under-reported because tools attribute the loss to the linking page, and because the people who break URLs are rarely the people watching the backlink graph. Typical triggers:
- A migration or redesign that changed slugs, folder structure or the trailing-slash convention without a complete redirect map.
- A blog post deleted during a content audit because it had little traffic, with no one checking whether it had links.
- A product or pricing page retired when the offer changed.
- Redirects switched off after a platform change, often a year or more after the original move.
- Parameter or protocol changes that leave old links pointing at a URL that no longer resolves cleanly.
Every one of these produces a link that still exists on the web and counts for nothing. Google’s site-move guidance recommends keeping redirects in place for as long as possible, and in any case for at least a year, precisely so that links from other sites pointing at old URLs can be reassigned (Google Search Central). The practical reading is that a redirect for a URL with external links should be treated as permanent infrastructure, not a temporary courtesy.
Losses nobody caused
Backlink indexes are finite. Crawlers drop pages they consider low value, time out on slow hosts, get blocked by firewalls and bot-protection services, and cannot see links that are inserted by JavaScript after the page loads. A “lost” link in this category is a reporting artefact. It will often reappear on a later crawl, and its disappearance tells you nothing about Google, which crawls the web with its own budget and its own index. If you have ever wondered why backlinks do not show in Search Console while a tool lists them, this is the same mismatch viewed from the other side.
How to Investigate a Drop in Backlinks
Work through these in order. The first two steps resolve most alarms without any outreach.
- Look at referring domains, not raw backlinks. A single site that linked to you from its footer or sidebar can account for thousands of “backlinks”. When it removes that one element, the backlink count collapses and the referring-domain count moves by one. Judge the drop on domains, and on domains that were indexed and relevant.
- Read the reason column and the crawl dates. Filter out dropped and crawl-error rows for now. Sort what remains by the strength of the linking page, however your tool measures it, and by whether the link sent any referral traffic.
- Check your own side before anyone else’s. Export the target URLs from the lost list and test their status codes. Any 404, 410 or redirect chain ending in an error is a self-inflicted loss and the cheapest to fix. Do the same for your most-linked pages in Search Console’s Links report, which shows what Google credited to you historically but has no “lost” filter of its own.
- Open the linking pages that remain. View the page source and search for your domain. If the link is there, the tool is wrong, or the link has moved into an element the crawler cannot read. If it is gone, use the Wayback Machine to see what the page looked like when the link was present, which tells you whether it was a deliberate edit or a wholesale rewrite.
- Check for ranking movement on the target pages. Compare positions and impressions for the linked pages across the window in which the links disappeared. If nothing moved, the loss is real but not urgent, and your recovery effort should be sized accordingly. If a page that depended on a handful of strong links has slipped, that page goes to the top of the list.
Keep notes as you go. A monthly record of which domains were lost, why, and what you did about it is more informative than any single snapshot, and it feeds directly into the kind of SEO reporting that stakeholders can act on.
Does Losing Backlinks Hurt Rankings?
Sometimes, and far less often than the size of the drop suggests. Three facts explain why.
First, link value is unevenly distributed. Link-based ranking systems weight a link by the strength of the page it comes from, so a profile of a thousand referring domains typically owes most of its authority to a small fraction of them. Losing a hundred weak links from the tail changes little; losing three of the strong ones can be visible. Which is why step five above matters more than the headline count.
Second, many lost links were already being ignored. Google has said repeatedly that it discounts links it considers spam, and that links to pages that do not exist are not counted at all. A tool records those as losses; Google never recorded them as gains.
Third, as the table above shows, redirected and re-canonicalised linking pages generally carry the signal with them, so a share of what the tool calls lost is a bookkeeping change rather than a change in what Google can see.
Where loss does hurt is predictable: a page whose rankings rested on a small number of strong, relevant links; a migration where external links now hit error pages; and a sustained period in which attrition outpaces acquisition, so that competitors who keep earning links overtake you slowly rather than suddenly. None of these announce themselves. They show up as gradual decline that is easy to misattribute to an algorithm update, which is one more reason to keep the monthly record.
Which Lost Backlinks Are Worth Recovering? A Three-Question Test
Recovery takes someone’s time, usually yours and an editor’s. Spend it where three conditions hold at once. This is a working test rather than a rule from Google.
- Was the link doing anything? The linking page was indexed, topically related to yours, and the link was a normal followed link in the body of the content. Bonus if it sent visitors. A link from a page Google never indexed, or one that carried a
nofollow,sponsoredorugcattribute, was already contributing little, and getting it back is not a priority. - Is recovery actually possible? If your own URL changed, yes: add the redirect and the link is recovered without contacting anyone. If an editor removed the link but kept the page, recovery is possible and depends on a polite, specific ask. If the linking page was deleted or the domain is gone, there is nothing to recover; the right move is to earn a replacement elsewhere.
- Is the cost proportionate? One short email to a site you have a relationship with is cheap. A multi-step chase of an editor at a publication that pruned your link on purpose is not, and it often ends in a worse outcome than silence.
Applying the test to a typical lost list produces a familiar pattern: a large pile of dropped and crawl-error rows to ignore, a modest set of your own broken URLs to redirect today, a short list of editorial removals worth one message each, and a residue of deleted pages to write off and replace. Work it in that order: cheapest and most certain first.
How to Recover Lost Backlinks
1. Redirect your own broken URLs
For each target URL on your site that returns an error, redirect it with a permanent (301 or 308) redirect to the page that now serves the same purpose. If no equivalent exists, redirect to the closest relevant page rather than the homepage; a redirect that lands users somewhere unrelated is a poor experience and a weak signal. Point the redirect directly at the final URL to avoid chains, and keep it in place indefinitely. Then recheck your redirect inventory after every platform change, because this is exactly when old rules disappear.
2. Ask for reinstatement when a link was edited out
A reinstatement request works when it is specific and easy to act on. Name the article, the sentence where the link sat, and the exact URL you would like restored. If the page you were linked from has been updated and your original page is now dated, offer the newer version instead; editors remove links to stale content for good reasons, and a fresher destination is a better argument than a reminder that the link used to be there. Do not imply an obligation, and do not send the same message to a site that removed the link deliberately as part of a policy change. The principles in our guide to link building outreach apply here with one advantage: you are asking someone who has already chosen to link to you once.
3. Offer a replacement when your page is the problem
If the link was removed because your page was deleted or is no longer accurate, the fix is on your side first. Restore or update the page, then let the editor know it exists again. Pages that attract links deserve maintenance for that reason alone, and a content refresh of a well-linked post is often cheaper than earning the same links again.
4. Replace what cannot be recovered
When the linking page or site is gone, treat the loss as a gap in the profile and fill it the way you would fill any other: with new placements on relevant, indexed pages. Set the target by referring domains lost, not backlinks lost, so that a single footer-link removal does not trigger an oversized campaign. If you work with an agency, ask how placements are verified after they go live and what happens when one disappears within the first months; our guide to choosing a link building agency lists the questions worth asking.
How to Prevent Avoidable Link Loss
You cannot stop other sites from editing their pages, but most self-inflicted loss is preventable with a few habits.
- Treat URLs as commitments. Before any page is deleted or moved, check whether it has external links. Make that check a required step in the content-audit and migration process, not an afterthought.
- Maintain a living redirect map. Store old-to-new URL mappings in version control or a shared document, and re-test them after every CMS, host or theme change.
- Review referring domains monthly, not backlink counts daily. Daily counts are dominated by crawler noise. A monthly review of lost referring domains, with reasons, catches the losses that matter without generating false alarms.
- Verify new placements at 30 and 90 days. Paid and guest placements are the most exposed to early removal, because the host has no editorial stake in keeping them. Checking at fixed intervals tells you which hosts keep their pages and which do not, which should inform where you buy or pitch next.
- Keep linked pages worth linking to. Editors remove links to pages that have gone stale, broken or off-topic. The pages that earn links are the ones to keep current.
FAQs
Is it normal to lose backlinks?
Yes. Large-scale measurements suggest that, over several years, most links to a site stop working or stop counting, and that larger sites lose a bigger share. A steady trickle of lost links is the ordinary condition of a backlink profile. What deserves attention is a loss of strong, relevant referring domains, or a cluster of losses that coincides with changes to your own URLs.
Why does my backlink tool show a link as lost when it is still on the page?
Because the tool lost the page, not the link. Crawlers get blocked, hit temporary errors, drop pages from their index, or cannot see links that JavaScript inserts after load. The link usually reappears on a later crawl. It has no bearing on whether Google still sees the link, since Google crawls independently.
Should I redirect every 404 page that has backlinks?
Redirect the ones with links from real, indexed pages, and send each to the closest live equivalent rather than the homepage. Pages with links only from spam sources can be left as 404s; Google has said that links to pages that do not exist are simply not counted, so there is nothing to protect. The effort is small either way, so when in doubt, redirect to a relevant page.
Does Google Search Console show lost backlinks?
Not directly. The Links report is a historical sample with no lost filter, and links Google found in the past can remain listed after they are removed. You can approximate losses by exporting the report at intervals and comparing them, but a dedicated backlink tool is the practical way to track which links disappeared and why.
How quickly do I need to act on a lost backlink?
Broken URLs on your own site should be redirected as soon as you find them, because every day they stay broken is a day the link counts for nothing. Editorial removals are less time-sensitive; a considered request within a few weeks is fine. Links that a tool merely dropped need no action at all unless a manual check shows the link is genuinely gone.
Conclusion
Backlinks are a depreciating asset. The web forgets pages at a steady rate, editors tidy up, and your own site changes under you. A falling count is not a crisis; an unexamined one is a missed opportunity, because the most valuable fixes (redirecting your own broken URLs) cost almost nothing and the least valuable ones (chasing links a crawler merely dropped) cost the most. Separate what your tool lost from what Google lost, repair your side first, ask once and politely for the editorial removals that mattered, and replace the rest. When the gap needs filling with new placements, confirmed live and reported clearly so you can track them, Ranking Frontier’s link building services are built for that.
