In proofing a manuscript that’s examining online engagement for a 2009 London protest, they mention that the Facebook event page (right) is still up and active, with all the comments visible and able to be analyzed, while the actual organization’s website (left) is dead, vanished into an incomplete archive.org snapshot, with all its user-generated content gone as well. (The researchers saved their own copy of the website at the time.)
The idea that something posted online lives forever is massively wrong; in fact, its life depends on the type of content and where it’s posted, all of which determine how much interest other web actors will have in preserving it. (Not to mention how much accompanying contextual information they will be interested in preserving.) Ask anyone who’s tried to write a history of some part of the web more than 2 years old.