It is said that necessity is the mother of all invention, and that's certainly true of URL shortening services. The rise of the status update means that there's no room for an extraneous character, and that has in turn led to the rise of URL shortening services that make sure the links shared in status updates don't take up any more room than needed.
As these services have grown in popularity, it's not uncommon to see shortened URLs used in places where there's really no need for them, from blog comments to emails.
But URL shorteners aren't perfect. There are security risks that arise with their use, and there's no guarantee that the link you create today will work for all eternity. Now, there might be another reason to think twice before using a URL shortening service: they're often a performance impediment.
That's according to web monitoring company WatchMouse, which took a look at the uptime and performance of a variety of URL shorteners ranging from independent URL shortening powerhouse bit.ly to the YouTube-operated youtu.be. Of the 14 services tested from WatchMouse's 44 global monitoring locations, only two (goo.gl and twt.tl) had uptime of 100% during the testing period, and 12 of them added half a second or more to page load time. Facebook's URL shortener, fb.me, added a whopping 2+ seconds to the total page load time.
A couple of interesting findings:
- International users often get the short end of the stick (no pun intended). WatchMouse says that "only a few of the URL shorteners optimized their name servers (DNS) for international use".
- You don't always get what you pay for. bit.ly's paid service, bitly.pro, offers features that aren't available to users of bit.ly gratis but WatchMouse actually discovered that bitly.pro is slower than the free version.
Obviously, it shouldn't come as a surprise to publishers that putting a URL shortener between users and a website is liable to decrease how quickly they get to their final destination, and may increase the frequency at which users are not being able to get to there at all.
But I do think many publishers underestimate just how important a half a second can be, or what a few fractions of a percent of additional downtime can mean to user acquisition or revenue. Unfortunately, the big problem here is that publishers can't stop users from linking to their websites using URL shorteners. Even for those publishers who roll their own URL shorteners, there's no way to ensure that users link to them using it.
The cat is out of the bag and all publishers can hope for is that URL shortening services will get better over time, and that the best will rise to the top.
Photo credit: chrisdlugosz via Flickr.
Reader comments (7)
10:57AM on 19th March 2010
Can anyone recommend a URL shortener that is secure, keeps all the URL's it produces and doesn't impede page load time?
Tech Reporter at Econsultancy
2:49PM on 19th March 2010
Emma,
That's asking for a lot. In terms of page load time, you will always have some delay with a URL shortener as the user will have to resolve two addresses - the URL shorteners and then yours once the URL shortener issues a redirect.
1:54AM on 20th March 2010
Quick correction! Watchmouse didn't know who was using bit.ly Pro and who wasn't. Turns out some of the fastest shorteners are bit.ly Pro whitelabels. Also, bit.ly Pro is a free service.
10:44AM on 20th March 2010
Snipurl has been around since 2001 and never given me any such problems. It's fast. Now, author, find a more meaningful topic to write about :)
Managing Director at Ergo Digital
11:16AM on 20th March 2010
I concur with most of the comment above. This fits very neatly into 'We're Watchmouse and we've come up with some really unmemorable stats because we need a bit of PR for our services'.
12:57AM on 21st March 2010
Econsultancy use shorteners on their Twitter buttons? I agree with your post, just wondering why your team doesn't practise what you wisely preach. I think if publishers such as yourself, used naturally smaller URLs which aren't full of surplus characters (like this very post: "http://econsultancy.com/blog/5614-another-reason-to-avoid-url-shorteners-they-re-slow" doesn't need 5614, another, avoid, they, re, slow) then users would be more than happy to display the original, more efficient URL. I use the shorteners for one thing — stats. When I RT this post, how will I know it's what my followers wanted from me, if not for bit.ly? I am very open to being shown up for my tech-naivety on that one!
CEO at SciVisum.co.uk
10:12AM on 26th March 2010
it's always easy for folks to measure the wrong stuff. Numbers are not everything!
The closer you make your measurements to 'do what the customer does' the better... and this wasn't very close.
Here's a question - there's a LinkedIn poll today about network measurements - normally the preserve of the tech team, but I wonder if the marketing folks here would sometimes want an easy dashboard to look at.
For the occassions when the team say ' it wasn't slow because of us govnr, it was the networks fault' !
The poll is the linkedin homepage, down the bottom under Polls:
See "Agnes Gough has created a new Poll: Do you use network monitoring tools across your complete IT network (including bandwidth utilisation and application usage) "
Log in to post a comment