Journal.
Short link syndrome
Listening to the Sitepoint podcast team discussing the rights and wrongs of URL shortening services, they missed – I think – the big issue.
URL shortening does what you might expect – it take a long website address, and turns it into a nice short one that is easier to share. Shortening isn’t a new thing, but Twitter’s 140 character limit and its own exponential growth has placed a real premium on the ability to shorten links. Consquently sites like bit.ly, tinyurl and ow.ly – and no, I’m not linking to them – are doing rather well and busily redirecting traffic all over the web. They give us convenience, we give them some rather rich traffic data. Everyone wins? Well, not so quick.
My concern with URL shortening is the end user. A quick glance at my own Twitter feed shows I’ve had over ten shortened links sent to me in the last couple of hours alone. Over the course of a day, a week, it’s a lot of links. Twitter’s move into the mainstream means that more and more of its users are not so technically confident or experienced as its early adopters. As URL shortening becomes a part of the mainstream experience of the web, the more I’m concerned about what these links teach us – to trust indeterminate links. This seems a recipe for disaster: users don’t know where they are going until they get there, and ‘there’ might not be particularly nice. Oh, and it’s a bit of an accessibility issue too.
What’s to be done? Short term, I’m with Jason Kottke’s proposal that Twitter should introduce a shortener of their own, but it is of course, a partial answer. It’s for publishers and publishing tools to create ways to responsibly link – something that the communities behind WordPress, Textpattern and others have reacted quickly to (yet another reason why I ♥ open source).
So, a plea. If you publish, create your own shortened URLs, and help your audience know where they are going. If your CMS or blog of choice doesn’t support it, lobby for it. If you want people to share links to your site – give them the canonical link that makes it easy for them. The future isn’t – or at least, shouldn’t be – third party url shortening.
I’m practicing what I preach here, so this post is available both at:
Further reading: Jeremy Keith tackles some of semantic and technical aspects of URL shortening with typical thoroughness, and Joshua Schachter explains why link rot matters.
Leave a comment
Farhan Lalji
1 July, 09:56 AM #
Nice post. I think the benefits outweigh the costs though, being able to distribute information with some note around the url in tweets.
What would be nice if url shortening services put in some kind of tag where on hover you could see the full url.
Also, if someone I follow sends me to crap pages, I probably won’t click on the link they lead to next time. But if many people link with the same description I’m more likely to trust it.
Byekick
1 July, 12:13 PM #
Thanks for commenting Farhan. Some services do offer a preview: adding ‘preview’ to the start of a tinyurl address does this for example. But it isn’t commonly known and rarely does it in a way that keeps with the intention of shorterning the link.
I did ponder the trust associated with the sender and didn’t explicitly cover this in my original post, but my problem with this is that it’s fine on Twitter where you are a follower and have implicitly made that trust relationship, but not so where you are following links from web searches that take the visitor to tweets.
These URLs are going to hang around for a long time…