A working link rot definition is short: a link has rotted when the URL it points to no longer returns the page, because the page was deleted or the whole site went down. It is different from content drift, where the URL still loads but now says something else. Both damage a reference. Only the first is easy to detect, which is why most measurements, including the one below, count only that.
The best recent measurement of scale comes from Pew Research Center’s report When Online Content Disappears, published on 17 May 2024 by Athena Chapekis, Samuel Bestvater, Emma Remy and Gonzalo Rivero, using data collected in 2023. Its numbers are easy to quote wrongly, so this post keeps each one attached to its scope.
Link rot vs content drift
Pew drew the line deliberately. It counted pages “as inaccessible only if they returned one of nine error codes that definitively indicate that the page and/or its host server no longer exist or have become nonfunctional”. Pages whose content “has been changed” were “beyond the scope of this research.” Pew calls its approach “the most conservative estimate possible”.
That matters for how you read every figure that follows. A page that now redirects to a homepage, or a domain that expired and now hosts something unrelated, is a dead reference to a reader but may not count as rot under Pew’s rule. Real decay is worse than the numbers.
How fast pages disappear: what Pew found
Pew sampled “just under 1 million webpages” from Common Crawl, “approximately 90,000 pages per year” from 2013 to 2023. Its headline: “38% of webpages that existed in 2013 are no longer accessible a decade later”. That figure is about the 2013 snapshot, not the whole web. Across all years, “25% of all the pages we collected from 2013 through 2023 were no longer accessible as of October 2023”.
The split is the useful part: “16% of pages are individually inaccessible but come from an otherwise functional root-level domain; the other 9% are inaccessible because their entire root domain is no longer functional.” Most rot, in other words, is a site deleting one page while carrying on. Pew also found that “even for pages collected in the 2021 snapshot, about one-in-five were no longer accessible just two years later.”
Pew then looked at links on three kinds of site. Here is every figure with its scope. Note the difference between the share of links that are broken and the share of pages that contain at least one broken link.
| Pew finding | Scope | Figure |
|---|---|---|
| Pages from the 2013 snapshot no longer accessible | Page-level, 2013 sample | 38% |
| Pages from 2013–2023 no longer accessible (Oct 2023) | Page-level, all years | 25% (16% page gone, 9% domain gone) |
| Pages from the 2023 snapshot no longer accessible | Page-level, 2023 sample | 8% |
| Links on government pages pointing to inaccessible pages | Link-level | 6% |
| Government pages with at least one broken link | Page-level | 21% |
| External links on news pages no longer accessible | Link-level, external links only | 5% |
| News pages with at least one broken link | Page-level, external links only | 23% |
| Wikipedia references no longer accessible | Link-level, English “References” sections | 11% |
| Wikipedia pages with at least one dead reference | Page-level | 54% |
Two more details are worth knowing. On government pages, “city government pages had the highest rates of broken links”, and 16% of the links it checked “now redirect to a different URL than the one they originally pointed to.” On news sites, traffic didn’t protect anyone: “25% of pages on news websites in the top 20% by site traffic have at least one broken link. That is nearly identical to the 26% of sites in the bottom 20%”. And “32% of links on news sites redirected”.
Link rot examples on your own site
Pew measured scale; it did not measure rankings, traffic or anything to do with SEO. What follows is our reading of what its findings mean for a site owner, not Pew’s. Rot touches you in two directions.
Outbound: the sources you cite
Every reference in an old post is a link that may die. If Wikipedia, with an army of editors, has a dead reference on more than half the pages Pew sampled, your three-year-old guides almost certainly have some. To a reader, a dead citation makes the claim it supports look unsupported. Our guide to broken links and SEO covers how to treat them.
Inbound: other people’s links to you
The 16% of pages deleted from working sites has a mirror image. When you delete or move a page without a redirect, every site that linked to it now carries a rotten link, and the value of that link is lost to you. Site migrations, CMS changes and content pruning are the usual causes. Our post on 404 errors and SEO explains which missing pages need attention and which can stay gone.
How to find and fix link rot
Treat it as maintenance on a schedule rather than a one-off clean-up.
- Crawl your outbound links. Any link rot checker or site crawler that reports external status codes will do. Export the failures to a spreadsheet with the page each one sits on.
- Replace, archive or remove. Find the source’s new location if it moved. If it is gone, link an archived snapshot or a different primary source. If neither exists, remove the claim or the link.
- Redirect your own deleted URLs. Point each to the closest equivalent page with a permanent redirect. Our post on 301 vs 302 redirects covers which to use, and redirect chains covers what goes wrong when redirects stack up over several migrations.
- Reclaim inbound links. Find links pointing at your dead URLs and either redirect the target or ask the linking site to update. Our guide to link reclamation walks through the process.
Rot elsewhere can also be an opportunity. If a resource in your niche has died and other sites still link to it, you can offer a working replacement; that is the idea behind broken link building. And if a page you need has vanished, our guide to recovering a website from the Wayback Machine covers getting the content back.
What the numbers can’t tell you
The limitation is in Pew’s own method. It counted only pages that return definitive error codes, so redirects to unrelated pages, soft errors and rewritten content are missing. Its news figures cover external links only, its data stops in 2023, and it says nothing about search rankings. Use it for scale, not as a measure of SEO damage.
Frequently asked questions
What is the meaning of link rot?
Link rot is when a link’s target URL no longer returns the page, because the page was removed or its site stopped working. Content drift, where the page loads but has changed, is a related but separate problem.
Has 38% of the web disappeared?
No. Pew found 38% of pages from its 2013 snapshot were inaccessible a decade later. Across all pages it collected from 2013 to 2023, the figure was 25% as of October 2023.
How common are broken links on Wikipedia?
Pew found 11% of references on the English Wikipedia pages it sampled were no longer accessible, and 54% of those pages had at least one dead reference.
How do I fix link rot on my site?
Crawl outbound links regularly, replace or archive dead sources, redirect your own deleted URLs to the closest equivalent, and reclaim inbound links that point at missing pages.
The takeaway Pages disappear steadily, mostly one at a time on sites that still work. Audit the links you cite, redirect the pages you remove, and quote Pew’s figures with their scope attached.
