August 22, 2013 - Comments Off on An organised network: the method behind Twitter spam attacks

An organised network: the method behind Twitter spam attacks

A few weeks ago, a bunch of us at Bloom were incredibly excited to sit down to an episode of Channel 4’s Dispatches that we’d been eagerly anticipating for some time. Entitled “Celebs, Brand and Fake Fans”, the programme explored scamming in social media and highlighted how brands can use unnatural methods to grow their online engagement figures.

The programme raised the curtain on an issue that many working in the digital industry have recognised for some time – an issue that our insight and analysis team are fascinated by. It’s probably something you’ve had experience of too.

We’ll have all been to conferences where the conference hashtag has started to trend on Twitter.For some of these conferences, it’s not taken long for an influx of spam tweets to take over the hashtag and push everything from diet pills to porn sites in front of delegates hoping to share content from the conference.

Here at Bloom we’ve conducted some research into Twitter spam using Whisper, our earned media planning tool. Whisper measures influence in real time and it estimates that a pretty hefty thirty per cent of tweets are from fake accounts. This figure can sometimes range as far as fifty per cent-  estimates suggest that almost half of Justin Bieber’s followers are fake!

Our research used Whisper to visualise engagement on Twitter and found that often the spammers were not involved in the conversation, but existed on the fringe. They weren’t able to get a message directly to the people tweeting about a particular issue or on a hashtag, as those accounts didn’t follow the spammers. The intention most of the time was to trick people into clicking on spurious links.

What’s more, our research identified a consistent pattern emerging around spam attacks and fake accounts. The spammers’ network in the example we explored was highly organised and showed structured and prolonged growth. It has clearly been generated for a purpose, and likely by a computer.

We recently conducted a review of a digital conference that showed that over 30 minutes there were 750 tweets from 306 different accounts, at a rate of 25 tweets per minute. More than half the tweets were from spammers. Spam tweets often drive traffic to a single URL that directs traffic to different, seemingly random, end points. Often the sites are third-party ecommerce sites, and an affiliate referrer is always attached.

There were other clear patterns visible within the spam network from this conference.. For example, the accounts used to generate the messages were named after females with a number at the end of the account. The messages also tended to start with a hashtag.   Looking at different conversations, we found different classifications of spammers. We continue to see accounts made to look like a member of One Direction in our research!

On the surface, a Twitter spam attack may seem like a small nuisance, but there are important repercussions that need to be considered. Without a spam filtration system embedded within a social media listening tool, the tool is in danger of giving inflated figures that aren’t a true reflection of a campaign. If these figures are used by brands to make decisions about future campaigns, the spammers can change the numbers – so much so that the wrong decisions could be made, skewing return on investment.

It is vital that spam accounts and tweets be stripped out if any true understanding of social networks is to be created.

You can read more about our research and the Dispatches programme over on the Drum website.