There has been a lot of praise for bit.ly, the new entrant in the URL shortening field created by betaworks. Now I do on occasion have a need for URL shortening for either tweeting or emailing, but other than that I consider the practice harmful (it should not even be necessary for email if email clients weren’t so awful).
Why do I consider URL shortening harmful? First, because good URLs convey a ton of information. For starters, there is the domain itself, which gives me a good idea of what I will be looking at if I follow the link – amazon.com, probably a product, readwriteweb.com, probably a technology blog post, etc. Well over 50% of the time I choose not to click on something simply based on that heuristic alone (which is also why I think it’s a terrible idea to use Javascript or other hacks to try to suppress the browser from showing a link address on a hover).
Second, because it creates a completely unnecessary and centralized layer on top of a beautiful distributed system known as DNS. If bit.ly or any of the other services goes down, none of these URLs are resolvable. Even when the services are up the resolution requires an extra roundtrip when otherwise resolution could often be local due to DNS caching.
I can probably think of a bunch of other reasons, but these two alone strike me as sufficient reason to avoid URL shortening whenever possible.
