A broken link itself is a link pointing to a non-existent place on the network, whether it’s an entire site, a separate page, or a specific file. If the entire Internet is compared with a city map, then a broken link can be presented as a non-existent house number on the map. That is, on the map there is a house, and in life it is not.
In technical words, a link is a single resource pointer (from English. URL is Uniform Resource Locator). And if this very resource does not exist, then the reference is called a bat.
The reasons why broken links appear online are several. Among the most common are information obsolescence, technical failure or human factor.
The Internet is like a huge living organism that constantly lives and develops. And the page that existed yesterday could be deleted today. There are many reasons for this: the page was decided to delete for one reason or another; the site could change its structure so that the page to which the link leads has changed its address and is no longer accessible by the former link. In the end, the site could have simply ceased to exist, and links to it could have stayed.
Technical failure means incorrect publication of a link, for example, on a forum. Some forums shorten long links and there are times when the user tries to post the correct, working link, and the result is a beat.
Finally, the human factor is when a user when placing a link manually types it instead of copying. In case of a typo, you get a broken link leading to a non-existent site. By the way, there are people who by means of special software scan the Internet for the presence of such links, which lead to other domains, for the purpose of their subsequent registration.
Search engines, such as Google or Yandex, scan the Internet with links, so if you are the owner of the site, you should watch out, so that there are no broken links on your site. Broken links lead “to nowhere” and a large number of such links on the site can lead to some sanctions on the part of search engines, such as downgrading in search output on one or another requests.