explainers

Link Rot: What It Is and How to Prevent Dead Links in 2026

Link rot is the slow death of hyperlinks as the web changes. Learn what causes it, how much of the web is already gone, and the practical fixes that keep your links alive.

Team U2L • 20 min read

You bookmark an article you loved. Two years later you click it. 404. That is link rot. It sounds like a nerdy internet complaint, but it is one of the quiet reasons the web feels less trustworthy every year. When Pew Research studied it in 2024, they found that 38% of pages that existed in 2013 were already gone, and a quarter of every page on the web created between 2013 and 2023 had disappeared by the time they finished counting.

If you publish anything online, run marketing campaigns, cite sources in a blog, print QR codes, or share links from an app, link rot is coming for your work whether you think about it or not. The links you shipped last year are quietly dying in the background. Some of the biggest names on the internet, including Google itself, have wiped out entire link infrastructures with a single deprecation notice. This guide covers what link rot actually is, why the web breaks itself, what it costs you, and the specific things you can do to keep your own links from becoming another 404 statistic.

Link rot is the gradual process of hyperlinks on the web becoming broken as the pages, files, or servers they point to move, get deleted, or shut down. According to Pew Research, roughly 38% of webpages from 2013 are no longer accessible today, and 23% of news pages contain at least one broken link. The best defence is a stable redirect layer, 301 redirects for anything you move, regular link audits, and archiving important sources with the Wayback Machine.

Link rot (also called link death, link decay, or reference rot) is the phenomenon of hyperlinks on the web ceasing to work over time because the destination page has been moved, renamed, deleted, or the entire host has gone offline. It affects citations, backlinks, bookmarks, and any URL shared in a place that outlives the source it points to.

Table of Contents

Link rot is what happens when the destination of a hyperlink stops existing at the address the link points to. The link is still there, sitting in your article or your email or your database, but clicking it lands on a 404, a hijacked domain, a login wall, or a completely different page than the one that originally lived at that URL. The web treats a URL as a promise. Link rot is the web quietly breaking that promise.

The term dates back to the mid-1990s, coined by web developers watching their carefully curated resource lists degrade. It got a formal Wikipedia entry, an academic literature, and even legal definitions when courts started worrying about citations to statutes that no longer resolved. And yet, three decades later, the problem is worse than it has ever been. That is not because engineers got lazy. It is because the web scaled faster than the discipline required to maintain it.

There are a few flavours worth naming, because they show up in slightly different ways:

  • Hard 404s. The classic. Page gone, server returns "not found," done.
  • Domain expiration. The whole site vanishes because nobody renewed the domain. Sometimes it goes to a parking page. Sometimes it goes to something a lot worse.
  • Silent redirects. The URL loads, but you land on a homepage or a store, not the specific piece of content you wanted. Technically alive, functionally dead.
  • Content drift. Same URL, completely different content. You cited a paragraph, now the page is an ad. Reference gone.
  • Access rot. The page is up but paywalled, geoblocked, or gated behind a login that did not exist when you first linked to it.

All of them count as link rot from the reader's point of view. All of them break the same trust.

How Bad Is It, Really?

The Pew Research Center's May 2024 study is the most rigorous data we have, and the numbers are worse than most people guess. Their researchers pulled samples from Common Crawl for every year between 2013 and 2023 and tried to load each page. Here is what they found:

  • 38% of pages from 2013 were no longer accessible by October 2023.
  • 25% of all pages created between 2013 and 2023 were already gone.
  • 23% of news pages contained at least one broken link.
  • 21% of government pages had the same problem.
  • 54% of Wikipedia articles had at least one dead reference link in their citations section.
  • Nearly one in five tweets disappeared within months of being posted.

Sit with those numbers for a second. More than a third of the web from just a decade ago is gone. Government pages, the ones that are supposed to be the most durable digital records in society, are broken more than one time in five. Wikipedia, which most of the internet uses as a source of last resort, has broken sources on more than half its pages. Fast Company summarised the wider picture bluntly: the internet is littered with dead links.

Independent audits paint an even bleaker picture in specific corners. Ahrefs studied over two million domains and found that 66.5% of links to pages created in the previous nine years were already dead. That specific number depends on how the sample was drawn, but the direction is consistent everywhere anyone looks. The web is decaying faster than it is being maintained.

Nine times out of ten, link rot is not an accident. It is a predictable consequence of one of a small number of decisions:

Someone rewrote a URL. A team migrated their CMS, cleaned up their permalink structure, or just decided a URL "read better" a different way. If they did not set up 301 redirects, everything that ever linked to the old URL now breaks. We wrote a whole guide on why permalinks need to be permanent for exactly this reason.

A page got deleted. Company shut down a product, publisher removed an article, moderator deleted a post. If nothing takes the old URL's place, the URL becomes a 404 immediately.

A domain expired. Someone forgot to renew, or a company folded, or a hosting bill went unpaid. Sometimes the domain gets snapped up by a squatter and starts serving something completely unrelated - which is worse than a 404 because it might harm your readers.

A platform shut down a feature. This is the modern version, and it is the one that catches even careful teams off guard. Google killed the goo.gl URL shortener. Google killed Firebase Dynamic Links. Twitter's t.co shortener has had multiple wobbles. When platform-owned link infrastructure disappears, everything downstream disappears with it.

A CMS update changed everything. WordPress permalink settings, Shopify URL structures, Ghost slug rules - one checkbox in the wrong place can rewrite thousands of URLs in a single deploy.

HTTPS migrations without redirects. Sites that moved from http:// to https:// and did not properly redirect the old scheme still leak traffic to broken URLs a decade later.

Sitemap/robots.txt mistakes. Not link rot exactly, but functionally similar - the page exists, but you have told search engines to forget it, and now nobody can find it.

The unifying pattern: link rot happens when someone changes the URL side of the equation without controlling the link side. Anywhere your links live outside your reach, they are one careless deploy away from breaking.

If you publish or market anything, link rot is quietly draining value in ways that never show up on a single dashboard. A few of the biggest costs:

Lost SEO. Every backlink you have earned pointing at a URL that now 404s is dead weight. Google will eventually drop those links from its index, and the ranking authority they were passing evaporates. Broken outbound links on your own site also hurt - Google's own guidance treats them as a quality signal, and enough of them look like a maintenance problem.

Broken trust. A user who clicks a link on your site and lands on a dead page will not blame the destination. They will blame you. Publishers who let outbound links rot become less useful over time, even if their original content was good.

Lost conversions. Every campaign URL that dies mid-campaign is a bucket of unattributable traffic. If your affiliate link 404s after a domain change, you stop earning. If your QR code on a printed poster points at a broken URL, thousands of scans go to a dead end.

Wrecked citations. Academics, journalists, and anyone else building an argument on top of other people's work loses the ability to prove their sources ever said what they said. Content drift and hard 404s both destroy the citation chain.

Analytics gaps. Your click-through data becomes noisy when links die halfway through their useful life. A URL that got 10,000 clicks in month one and zero in month six is either a viral flameout or a broken link, and if you cannot tell which, you cannot learn anything from it.

Support load. Someone always emails to say "your link is broken." Every one of those messages is a small tax on your team's time.

A few high-profile deaths are worth knowing about, because they show how link rot can hit even sophisticated teams:

goo.gl (Google URL Shortener). Google stopped creating new goo.gl links in 2019, then turned off redirects for existing links on August 25, 2025. Anything that had ever been printed, embedded, or shared as a goo.gl link went to 404 overnight. We covered replacements in our Google URL shortener alternatives guide. The lesson: if your links live on infrastructure you do not own, treat them as leased.

Firebase Dynamic Links. Google's mobile deep-linking service shut down the same day as goo.gl. Apps that used FDL for share buttons, email onboarding, and post-install routing saw those flows silently break. Teams that had built years of growth on top of it had to migrate in a hurry. Our Firebase Dynamic Links alternative roundup walks through what filled the gap.

Yahoo Groups, GeoCities, Google Reader. Entire communities and archives went dark when their host platforms shut down. Every link into those archives died with them.

News archive purges. Multiple major publishers have quietly deleted decades of archives after acquisitions or paywall changes. Every citation that pointed into those archives is now a dead reference.

The unifying pattern: platform decisions kill links faster than any technical failure ever will. If the entity that owns the URL decides tomorrow that they do not want to keep serving it, every link that depended on it dies.

Most of the fixes are unglamorous. That is fine. Link rot is a hygiene problem, not a hero problem. The teams that avoid it are the ones that treat URL stability as a discipline instead of an afterthought.

Design permalinks like you mean it. Pick a URL structure at launch that you can live with for a decade. Short, descriptive, no dates you might regret, no auto-generated IDs, no session parameters baked into the path. Then keep it. Every URL change is a small emergency.

Set up 301 redirects for anything you move. A 301 tells search engines and browsers that the content has permanently moved, and Google passes the majority of the ranking signal through. Our guide to 301 vs 302 redirects explains when to reach for each and why the wrong choice can silently cost you rankings.

Never delete a URL cold. If you have to retire a page, redirect it to the closest equivalent or a topic hub. A redirect to a helpful page is infinitely better than a 404, even years after the original content is gone.

Version your APIs and public URL patterns. If you have any URL structure that other people depend on (an API, an embed, a callback), version it explicitly and keep old versions working long past the deprecation date.

Run a broken-link audit at least quarterly. More on tools below. The point is to make it a habit, not a one-time cleanup.

Warn users before a URL changes. If you have to rename something high-traffic, tell people in advance. A blog post, a banner on the old URL, an email to subscribers. Change is expensive whether or not you announce it, but not announcing it makes it more expensive.

Preventing your own URLs from rotting is only half the fight. Every article you publish that cites something on someone else's server is a link waiting to break.

Archive important citations. The Wayback Machine can snapshot a page on demand. Paste the URL into web.archive.org/save/, and you have a permanent copy. When the original inevitably breaks, you can update your citation to point at the archived version instead. Perma.cc, run by Harvard Law School, does the same thing with a stronger institutional preservation guarantee for academics and journalists.

Prefer authoritative sources. A citation to a government page, a major publisher, or a well-funded institution is less likely to rot than one pointing at a random Medium post. This is not fairness, it is math - the domains that will still exist in five years are the ones that already survived the last five.

Wrap outbound links in your own redirect layer. This is the underrated one. If every outbound link you publish goes through a redirect you own, you can fix broken destinations without touching the original article. When the underlying source moves, you update the redirect target once, and every historical reference across your entire site starts working again.

Track what you link to. Keep a list. Even a simple spreadsheet with URLs and last-checked dates puts you ahead of most publishers.

Here is the strategy that punches above its weight. If you own the domain your links live on, you own their fate.

Say you publish an article that cites example.com/study/2023-report. Two years later, example.com restructures their site and the report moves to example.com/research/2023/quarterly. Every article you wrote citing the old URL is now broken. Painful.

Now imagine you had wrapped that citation in a short link you controlled: u2l.ai/2023-report. The published version of your article still points at your own short link. When the underlying URL changes, you go into your dashboard, update the destination, and every article across your entire site that used that short link is fixed in one edit. Nothing to republish, no editing dozens of posts, no readers landing on a 404.

That is the case for using a redirect layer for anything you cite or share widely, and it is why we built U2L AI's URL shortener around dynamic destinations by default. Every short link is a redirect you can point somewhere else at any time. You can change the destination without changing the shared link, which is the exact defense link rot requires.

The same trick makes QR codes durable. A static QR code encodes the destination URL directly, so if that URL breaks, the QR breaks with it, printed materials and all. A dynamic QR encodes a short link, and the short link's destination is editable forever. Restaurants change menu PDFs, event organisers update schedules, product packaging keeps pointing at a live URL through multiple website migrations - all because the link the QR encodes lives on infrastructure they control. Our dynamic vs static QR code explainer breaks the trade-offs down in more depth.

A quick word on choosing a shortener responsibly: use one that has been around long enough to trust with your redirect layer, and prefer providers that offer a custom domain option so that even if you switch tools later, your links keep working. Free shorteners hosted on someone else's domain are exactly the kind of borrowed infrastructure that killed goo.gl users. Our roundup of the best URL shortener platforms covers what to look for.

You cannot fix what you cannot see. A few tools worth knowing:

  • Screaming Frog SEO Spider. Desktop crawler, free up to 500 URLs, paid above. The classic. Crawls your whole site and flags broken links.
  • Google Search Console. Free. The "Pages" report surfaces 404s Google has found while crawling. Not exhaustive, but it is the source Google itself is looking at.
  • Ahrefs Site Audit / Semrush Site Audit. Paid, but part of most SEO stacks already. Both catch broken outbound and internal links on a schedule.
  • Dead Link Checker. Free web-based tool. Good for quick spot checks on a single URL or a small site.
  • Broken Link Checker (WordPress plugin). Runs continuously against every link in your posts. Convenient if you are already on WordPress.
  • Wayback Machine. Not a broken-link finder exactly, but the last line of defense once you find one. Search for the dead URL and grab the most recent snapshot.

Set a recurring calendar reminder to run one of these quarterly. Then actually run it. The gap between "we should audit our links" and "we do audit our links" is where most rot lives.

Print makes link rot brutal. You cannot recall a business card, a poster, or a product label after it ships, so any URL you print is stuck for the life of the physical medium. If that URL breaks a year later, everyone who scans the code lands on a 404 and blames your brand.

This is the case for dynamic QR codes on absolutely everything you print. A dynamic QR points at a short link, and the short link's destination is editable. Your printed materials keep working through website redesigns, product renames, campaign endings, and even total domain moves, because the QR itself does not care what URL the short link resolves to today.

If you are printing QR codes at any real scale - packaging, signage, business cards, event materials - treat static QRs as a print-and-pray anti-pattern and reach for dynamic instead. This is one of the small operational choices that pays back for years. Our guides on editing a QR destination without reprinting and how to create a dynamic QR walk through the mechanics.

Frequently Asked Questions

Link rot is what happens when a hyperlink stops working over time because the page it points to has been moved, deleted, or taken offline. The link itself is still there, but clicking it lands on a 404 or a completely different page than the one it originally pointed to.

According to Pew Research's 2024 study, 38% of webpages from 2013 are no longer accessible, and roughly 25% of all pages created between 2013 and 2023 have already disappeared. Even Wikipedia, one of the most maintained sites on the web, has broken reference links on 54% of its pages.

Yes. Backlinks pointing at 404s stop passing ranking authority, and Google treats excessive broken outbound links on your own site as a quality signal. Regular link audits and 301 redirects for anything you move are the two most important defences.

On your own site, yes. A 301 redirect tells search engines and browsers that a page has permanently moved, and Google passes the majority of the ranking signal to the new URL. Setting up 301s for every URL you retire is one of the highest-leverage things you can do to prevent link rot on infrastructure you control.

Use a crawler like Screaming Frog, Ahrefs Site Audit, or Semrush Site Audit for a full sweep, and Google Search Console to see what Google itself has been hitting. For quick spot checks, Dead Link Checker and the Wayback Machine's URL search both work in a browser with no install.

It can, if you choose the wrong provider. Free shorteners run on someone else's infrastructure have shut down before (goo.gl in 2025 is the highest-profile example). Pick a shortener with a durable business model, ideally one that supports a custom domain so your links keep working even if you switch tools later.

A dynamic short link is a redirect you own. If the destination URL ever breaks, you update the destination once in your dashboard, and every place that link was shared, printed, or embedded starts working again with no republish. Static QR codes and hard-coded URLs have no equivalent escape hatch.

What is the Wayback Machine and how does it help?

The Wayback Machine is a free archive of the web run by the Internet Archive. It periodically snapshots public pages, and you can also request an on-demand save. When an outbound link on your site rots, you can replace it with the most recent Wayback snapshot so the citation still resolves.

Link rot is one of those problems that never announces itself. You do not notice it until traffic quietly drops, a reader emails to complain, or a campaign underperforms for reasons you cannot pin down. The teams that stay ahead of it treat URL stability as a policy, not a project - permalinks designed to last, 301 redirects for every move, quarterly audits, and a redirect layer wrapping the outbound links they care about. That last piece is where a shortener you control turns from a nice-to-have into insurance.

If you want a redirect layer you own and dynamic QRs that outlive whatever you print them on, create a free U2L AI account and start moving your important links behind short URLs you can update whenever the web underneath them changes.

Ready to try U2L AI?

Free forever plan. No credit card required.

Get Started Free View Pricing