Twitter’s spam problem

April 6, 2008

I’ve been using Twitter for a while now. I mainly keep to myself, following a few other Melbourne people and a few other bloggers and personalities I’m interesting in hearing from. You can find my Twitter page here, if you’re interested. However, recently, there has been a growing problem, one that seems to infect each new web service in one way or another: SPAM.

Twitter spam is becoming a real problem these days. On average, I have 2-3 new follower notification emails a day for Twitter accounts that are obviously auto-following everyone they can and contain little more than a couple of links that there’s no way I’m going to click on. By my definition this is spam and abuse of the system, but not everyone seems to agree: this conversation at Mashable is interesting, as there seems to be a meeting of people who use Twitter quite differently, with not everyone deciding that these accounts are spam.

So what’s the answer?

I’m sick to death of receiving these emails and having to vet each request when I know that 95% of them will be rubbish. However, I don’t want to miss out on being notified when I have a legitimate new follower, so turning off notifications altogether is out.

One of the giveaways of a spam account is the high number of followings and the lopsided following/follower ratio, such as in the image here. I doubt this can be used as a blanket rule though, as there are legitimate users with huge numbers of each (Scoble, as usual, is through the roof). Although I personally question the utility of following so many people, it can be a legimate use of Twitter, so blocking users based on followings or ratios is out too.

Perhaps the easiest way to block a lot of automated following is to implement a CAPTCHA-style check when following someone. This should knock out a lot of the automated following while leaving the human followers only slightly inconvenienced. This isn’t without its issues either, though. For example, one of Twitter’s biggest features is how easy it is to use via mobile phone. Simply sending ‘Follow John’ to Twitter via SMS is all you need to start following someone. Implementing a CAPTCHA system here simply won’t work, unless you want to start using MMS which seems like a very bad idea.

Twitter is great. Its ambiguous nature means that it can be a lot of different things to a lot of different people. However, the ease of use that comes with this also means that it would be difficult to police and to implement any kind of filtering system that will block spammers. I can’t see a clear way of securing the service against these users without negatively impacting on users doing the right thing. I guess I’ll keep on deleting notification emails and ignoring spammers manually. Hopefully, Twitter will one day hit on a solution that keeps everyone happy, except the spammers.

Posted in Online