Why links break: the common causes
A link is only a saved address. It doesn't follow the page around. So when a page changes address, or disappears, every link pointing at the old address goes dead.
What does it mean when a link "rots"?
Over time, hyperlinks tend to stop pointing at the file, page or server they first targeted1. People call this link rot. It happens because the thing was relocated to a new address or became permanently unavailable1.
It's common. One study from Pew Research found that 38% of pages from 2013 had gone missing by 20231. Nobody has to do anything wrong for a link to die. Time does it.
Which causes are most likely on my own site?
Here's how a typo or a move without a redirect can play out in real life.
- You changed a page's address. Renaming "services" to "what-we-do" breaks every link to the old name.
- You rebuilt or restructured the site. Moving pages into new sections can change many addresses at once1.
- You deleted a page. A past event, a retired product or an old offer. Anything linking to it now leads nowhere.
- A typo went in. One wrong letter in a link you typed, or in an address someone shared.
- You changed domain or platform. A new host or website builder often builds addresses differently.
- A file was moved. Images, PDFs and downloads break the same way pages do.
What about links on other people's sites?
Those break too, and you can't edit them. A site you linked to may have rearranged itself or closed. The link on your page is then dead, even though you did nothing.
Others may be linking to you. If you move a page, their links break, and visitors arrive on your error page. That's one more reason to plan changes before you make them.
How do I stop links breaking?
Two habits help.
First, keep addresses stable. Link rot can be avoided by designing stable and clean addresses1. Pick short names that won't date, and think twice before putting a date in one.
Second, when you must move a page, set up a permanent redirect, known as a 301 (moved permanently)2. It sends both visitors and search engines to the new address. Skip the redirect and you get the error described in what a 404 error is.
To find the damage already done, see the page on checking a whole site for dead links. If a page is gone for good with no replacement, the guide to not found versus gone codes explains the choices3.
A last tip: keep a simple list of every page you publish, with its address. When you rename or remove something, you'll know straight away which links and pages might be affected.
Reference list
- Link rot, Wikipedia
- Redirects and Google Search, Google Search Central
- 404 Not Found, MDN Web Docs