We had a busy week at bit.ly. We announced our funding, decoded almost 24 million clicks, and saw some great posts on the subject of link shortening.
bit.ly is already starting to evolve beyond URL shortening but shortening is core to the work flow so here are our views. We believe URL shorteners are an important part of the emerging social distribution network. As Michael S. pointed out in one of the comments on a post — shortcuts of this kind have been used in the physical world for years (“second house on the left”, post boxes, call forward). Online they also have a long history — handles, redirected URL’s. Google also uses a form of shortening in their text ads.Some thoughtful criticism has been directed at the url-shortening industry as a whole. Here’s our perspective on a few of the key questions:
#1. Transparency of redirects. Joshua Schachter, citing his experience at Delicious, assumes that a “huge proportion of shortened” links are just a disguise for spam. This is not the case at bit.ly. Every link bit.ly shortens is passed through a batch of spam filters (SURBL , Safebrowsing , SpamCop , among others.) If a site is flagged as spam we present an interstitial page telling the end user that the link they are about to follow was determined to be spam by XYZ filter. A very small percentage of the bit.ly links are therefore spam. I hope none of you have ever seen such a page — I haven’t.
Sticking with the transparency question — there is also a UX aspect to transparency. In the past shortening links has been a tool for rickrolling etc. We have worked hard to make sure that this isnt the case with bit.ly — and we will continue to do so. Solutions in place to date include: Summize / Twitter Search inline preview, bit.ly Preview Plugin, a grease monkey scrip t to expose the underlying link etc… The preview is available in the bit.ly API we hope that people will use it like Twitter search, Power Twitter and others do.
#2. Centralization, Permanence and Reliability. These questions are related. When we architected bit.ly we developed it in such a fashion that there are many layers of redundancy to the redirect process. Should our web site go down — the redirects will still work. While not as decentralized as DNS this is step towards decentralization and an important part of reliability of our service. Similarly — we intend to offer a way for users to backup name/value pairs for their shortened URL’s for export. More to come on this subject, in a while. We have an awful lot of work to do right now.
#3. Frames. We don’t like them. They interfere with the user’s browsing experience, see this video for why and how. From About.com forward, frames have tended to create more problems than they have solved for users, and obscured the destination URL in the location bar. bit.ly wont use frames, ever, as they are not good for users or publishers.
Some Links from this week of interest, in long form:
Danny Sullivan’s review of shorteners: http://searchengineland.com/analysis-which-url-shortening-service-should-you-use-17204
Josh’s article: http://joshua.schachter.org/2009/04/on-url-shorteners.html
Dave Winer’s article from 2007: http://www.scripting.com/stories/2007/11/20/solvingTheTinyurlCentraliz.html
Dave Winer’s article from 2009, rerun with edits and fresh comments: http://www.scripting.com/stories/2009/03/07/solvingTheTinyurlCentraliz.html
Marshall’s article on funding and the vision: http://www.readwriteweb.com/archives/look_out_tinyurl_bitly_gets_hot_silicon_valley_h.php
Om’s article: http://gigaom.com/2009/03/31/why-bitly-could-upstage-digg/
David Weiss article: http://unweary.com/2009/04/the-security-implications-of-url-shortening-services.html
Rafe Colburn: http://rc3.org/2007/11/20/urls-by-design/
Kellan’s blog: http://laughingmeme.org/2009/04/03/url-shortening-hinting/
Conversation on YC: http://news.ycombinator.com/item?id=545565
Mashable piece on 5 reasons why: http://mashable.com/2009/04/05/url-shorteners/4 years ago