What the industry believes
Ask anyone why backlinks disappear and you will hear that pages die: sites shut down, posts get deleted, 404s accumulate. I would have said the same thing before I looked at this, and a great deal of standard audit advice is built on it — monitor your referring pages for 404s, watch for dead domains, expect attrition as the web decays around you.
It is almost entirely wrong.
What actually happened
Majestic records, for each lost link, what the source page was doing at the moment the link was lost. Across all 645,202:
| What the source page did | Links | Share |
|---|---|---|
DownloadedSuccessfully — fetched fine, link no longer on it | 503,390 | 78.0% |
RedirectCanonical — canonicalizes elsewhere now | 124,861 | 19.4% |
HTTP_301_PermanentRedirect | 11,739 | 1.8% |
HTTP_302_307_TemporaryRedirect | 4,814 | 0.7% |
HTTP_404_NotFound | 4 | 0.0006% |
| Dead host, DNS failure or timeout, combined | 24 | 0.004% |
97.4% of lost links were lost while their source page was still perfectly reachable. Four links out of 645,202 were lost to a 404.
Links die because somebody edited the page, or a redesign dropped the link, or a CMS migration changed the canonical. The page carries on living without you in it.
What this changes about reclamation
Link reclamation is usually positioned as a marginal tactic — worth a pass once a year, unlikely to yield much, on the assumption that a lost link is a gone link. This data says it is systematically undervalued, for a specific reason: the prospects are alive.
When a link is lost to a 404, the opportunity is genuinely gone. When it is lost to an edit or a redesign — which is what happens 97.4% of the time — the page is still published, still ranking, and still has an editor who can be contacted. That is a materially warmer prospect than anything a cold outreach list contains, and the pitch is different in kind: you are not asking for a new link, you are pointing out that something was dropped.
In this profile, 717 referring domains at Trust Flow 21 or above have no live link remaining. That is a prospect list most organizations would pay a considerable amount to assemble, and it was sitting inside an export nobody had run.
What this changes about monitoring
A great deal of link monitoring is built around detecting broken links and dead referring pages. On this evidence, that approach would have detected four link losses out of 645,202 across twenty years.
Catching this properly means comparing referring-domain snapshots over time and flagging domains that have dropped out while remaining live. That is a different technical exercise from crawling for 404s, and most monitoring setups are not doing it. If your link monitoring has never alerted you to a loss, that is not necessarily good news.
The replacement rate this implies
Combine this page with the survival figures and you get a number you can put in a budget rather than a principle you can nod along to.
If the median referring domain stops linking after roughly three years, and the losses are dominated by edits and redesigns you cannot influence, then a profile of 1,000 referring domains sheds something in the order of 150 to 200 of them a year once it reaches steady state. Standing still means replacing those before a single new domain counts as growth.
That reframes two conversations. The first is what “pausing” link building costs — not a plateau, a decline on a schedule. The second is how to read a flat referring-domain count month over month: flat is not stagnation, flat is acquisition exactly cancelling attrition, which is considerably more work than the number suggests.
It also means a campaign should be judged on net domains, not on links won. A report showing forty new links in a quarter, with no mention of what was lost in the same quarter, is showing you one side of a ledger.
How to actually detect a lost link
If 404 monitoring will not find these, what will?
Snapshot referring domains, not links. Export the referring-domain list on a fixed schedule — monthly is enough — and diff consecutive exports. What you are looking for is a domain present last month and absent this month. That is the signal, and it is invisible to anything that crawls for broken pages.
Check the page before you conclude anything. A domain dropping out of an index is not proof the link was removed; it may simply not have been recrawled. Load the source URL and look. This is the step most automated alerting skips, and it is why most automated link-loss alerting has a reputation for crying wolf.
Triage by what the domain was worth. The reclamation figures on the audit page make the case: of 14,619 fully-lost domains in this profile, 717 were Trust Flow 21 or above. Chasing all 14,619 is a waste of a career. Chasing 717 is a project.
Approach them as a correction, not a request. The page still exists and somebody still maintains it. “You used to link to this and the link seems to have come off in a redesign” is a different email from “please link to me,” and it performs like one.
What this changes about risk
The largest single threat to your existing links is other people's redesigns. You cannot predict them, you cannot control them, and they will not tell you.
That is an uncomfortable conclusion because there is no tactic that addresses it. The only real defense is a replacement rate — treating link acquisition as an ongoing program rather than a project with an end date, precisely because the losses arrive on a schedule you do not set.
It also argues for diversification in a way that is usually stated vaguely. A profile concentrated in a few referring domains is not just fragile to a penalty; it is fragile to one of those sites hiring a new designer.
Questions
Could Majestic be reporting this wrong?
The field records the crawl result at the point the link was flagged lost, so it reflects what the crawler saw rather than an inference. A page could in principle have 404'd and been restored between crawls, but that would have to account for essentially the entire 97.4% to change the conclusion.
Does 'lost' mean the link was deliberately removed?
Not necessarily. 'Lost' means no longer observed in the index. Crawl scheduling and index policy contribute, and the redirect and canonical categories are often site restructures rather than anyone deciding to remove a link.
So should I stop monitoring for broken links entirely?
No, but change what you expect from it. Broken-link monitoring is useful for finding link building opportunities on other people's sites. It is close to useless as a way of detecting losses in your own profile, which needs referring-domain comparison over time instead.