Having recently started to use Twitter, I’ve noticed that a lot of posts use a URL-shortening system because of  Twitter’s 140 character tweet limit. I’m already familiar with TinyURL.com as it’s useful for Usenet and email (and offline use, such as in print) when you need to shorten unfeasibly large URLs. TinyURL is fine for most uses, as it’s short enough to fit on a single line (the key requirement for email and Usenet) and also fairly memorable (very useful for offline use). But Twitter users (Twitterers?) tend to prefer alternative URL shorteners such as bit.ly and is.gd, since they’re shorter than TinyURL and size matters on Twitter. In fact, the growth of the new ultra-short URL shorteners seems to have been driven primarily by Twitter, since they rarely seem to get used anywhere else.

Bit.ly, in particular, seems designed primarily for Twitter (or Twitter-like uses). Unlike the others, its unique code (the bit after the slash) is case-sensitive, which makes it pretty useless for offline use as it’s much harder to memorise but is ideal for Twitter as it means it can fit more codes into fewer characters than the others. For non-web online use, such as Usenet and email,  it’s neither better nor worse than TinyURL – the extra reduction in length isn’t significant, but neither is the case-sensitivity (since it’s almost always going to be either clickable or copied and pasted rather than retyped). It’s clear, therefore, that bit.ly has a fairly narrow focus, as its only advantage over the more well-known TinyURL is its suitability for Twitter.

The interesting thing about all the URL shortners is how they’re funded. Is.gd doesn’t seem to have any revenue stream at all; it appears to be a personal project funded by one individual. TinyURL has adverts on the setup page as well as a Paypal “donate” button. I don’t know how much it actually earns, but given that URL shortening isn’t exactly rocket science I suspect that it covers its costs. Bit.ly, though, is incorporated as a separate entity and is backed by a range of investors (including O’Reilly). That must mean they’re expecting it to earn money eventually, even if it isn’t yet.

However, I have a big problem with the whole concept of using URL shortners on the web. The whole point of the web is that hyperlinks don’t need to expose the underlying URL to the reader within the text. They’re visible to the browser, and the reader can see them in the status bar, but you don’t normally see them in the link. If I link to my own site, I’d usually use something like “Mark’s Musings“, not “http://mark.goodge.co.uk“. The full URL is there, but it’s inside the code, not in the text you read. If I want to link to a photo in my photo gallery, I can do it just the way I have done there – I don’t need to use hacks like calling it http://bit.ly/4k1WCU. The only reason hacks like that are necessary in Twitter is because Twitter doesn’t allow you to enter HTML directly into the status – instead, it looks for something beginning with “http://” and then converts it to HTML for you.

There are other good reasons for not liking URL shorteners, too. They are too easily used for spam and other nefarious purposes, they break the logical link between URL and content, and if the underlying service breaks down then so do all the links. These are possibly acceptable risks for cases where URL shortening is the only practical solution, such as offline or text-only email/news posts. But they’re not acceptable risks for the web, where URL shortening simply isn’t needed.

So, it seems to me that a reliance on URL shorteners is a serious flaw in Twitter. Even if Twitter doesn’t want to allow users to enter raw HTML, there’s no reason why it can’t provide something equivalent to BBCode (the pseudo-HTML markup often employed by forums to allow restricted use of HTML within posts). Or Twitter could simply exclude URLs from the character limit. Or it could come up with a Facebook-style “link” option that’s separate to the tweet. There are many possibilities, some of which may be more practical than others, but almost any of them would be better than the current, rather broken, system. Twitter is a web application, so it should behave like the web, not like email or instant messaging. 

But what happens to bit.ly if Twitter do fix this flaw? Their entire business model flies out of the window. Without Twitter, bit.ly loses one of its biggest advantages over TinyURL, as well as their biggest source of custom. And their other current advantage (click tracking) is one that TinyURL could easily replicate if they wanted to. If bit.ly isn’t currently earning money (and it isn’t), it’s even less likely to do so in a future with a better Twitter.

But then, Twitter isn’t earning money either. At the moment, it’s funded by investors. But sooner or later, it will have to do one of three things: either generate a significant revenue stream, sell out or shut down. The last of these is probably unthinkable for its users. But the first is probably unrealistic, at least in the immediate future. And, unlike the URL shorteners, Twitter is a complex system that needs a lot of backend power. So it’s costing a fair bit to run, and the burn rate must be pretty high too. The only realistic proposition for securing the future of Twitter, therefore, is for it to be sold to some other organisation that can both support the short-term funding needs and build it into a long-term structure that generates revenue – or that can afford to run Twitter as a non-revenue earning operation in return for access to the huge amount of data that Twitter has on its users. (Actually, there’s another option: Twitter monetises itself through a combination of paid-for premium services and/or selling access to their data to third parties. But would the users stand for that?)

Whatever happens to Twitter, one thing is pretty certain: it’s going to have to change. So business models built on supplying a service to users of Twitter as it currently is are, to say the least, precarious.  And Twitter itself is quite possibly a lot more precarious than its owners would have us believe.

So, having dissed the new breed of URL shorteners, what about the other part of this post’s title? The connection, I hope, should be pretty obvious. The success of Google in turning a loss-making startup with no visible means of support into one of the biggest revenue earners on the Internet has prompted a whole load of other projects that aim to acheive similar results. But it’s turning out to be a lot harder than expected. Facebook has a significant negative cashflow, and probably will have for some time. MySpace figures are harder to come by, but media reports say it’s still below target. Pick almost any web-based startup with a similar profile, and the story is the same.

The original dotcom bubble was inflated primarily by ISPs and online stores (think boo.com). As with Google today, the success of one prominent market leader (Amazon) led others to try and copy their formula. But, for most of them, it didn’t work. Most of the profitable online stores now aren’t the ones that blindly tried to copy the sucess of Amazon, they’re the ones that learned from the failures of the wannabes.

It doesn’t take a genius to draw a parallel with the current situation. Web 2.0 has fuelled a boom in websites that look pretty and have all the necessary geek-chic, but earn very little real money. Most of the big ones have enough money for now, at least in the short term. But the credit crunch will seriously affect their ability to raise further capital when what’s currently in the pot runs out. I’ve picked on bit.ly, because of its combination of serious backers and yet no visible means of earning revenue, but it’s by no means alone. The fact that peple think URL shortening is a business model worth investing in is merely a symptom of a much wider over-optimism in the sector. In the next few years, we’re likely to see a pretty big shakeout of the bubble 2.0 websites, and it’s not at all implausible to think that some of the big names could be among those to fail.