Here's the scenario. You and your team work hard to use a new piece of collaboration software that offers some unique features that are suited either to the work at hand or that appeal to your sense of novelty. The tool works great and you get into it, using some of the more complex parts of the system to great advantage. Then the system's owners send out the dreaded self-congratulatory message saying that they have been acquired by a big company, or some beleagured product manager announces that their system is being sunsetted. You need to move.
Moving turns out to be hard. Once upon a time when the only thing you needed to move was electronic mail, there were only a handful of email file formats, and though they were incompletely specified you could readily move among them. Moving between news readers was just a matter of reusing your .newsrc file. Life was simpler then (cue sad violins).
The situation now is much more complicated. If you are extremely lucky, the system you have has an export file that is the same as its import file, and someone has built a faithful clone of your original service for you to migrate to. That was the case in my move from Delicious to Pinboard. If you're keeping a weblog, though, things can get really unhappy really fast, and a friend to recently migrated from Posterous to WordPress went through gyrations that I'm not sure that I would have tolerated.
The fundamental problem, I suppose, is that the more innovating the service the less likely that its data is structured so simply that you can drop it in to someting else. Brand new user interfaces, the ones that attract traffic because of the ever present need for novelty, also have novel data structures. Watch out when getting into something new that you may find it easier to start than to stop.