Dispatches raised the curtain on an issue that many working in the digital industry have recognised for some time. If you’ve ever attended an event that started to trend, you will have likely seen an influx of spam tweets pushing everything from diet pills to porn sites.
Spam tweets are sent from fake accounts that are created and suspended in a very short space of time. You can usually spot a fake account by the low level number of tweets and followers.
Bloom uses a tool called Whisper that measures influence in real time and it estimates that around 30 per cent of tweets are fake. Estimates have ranged as high as 50 per cent on high profile accounts such as Justin Bieber.
We can visualise engagement on Twitter and often the spammers are not involved in the conversation, but exist on the fringe. They aren’t able to get a message directly to the people tweeting about a particular issue or on a hashtag, as those accounts don’t follow the spammers. The intention most of the time is to trick people into clicking on spurious links.
Bloom researchers have identified a consistent pattern emerging around spam attacks and fake accounts. The spammers’ network is highly organised and shows structured and prolonged growth. It has clearly been generated for a purpose, and likely by a computer.
Fake Twitter accounts tend to have clear patterns. For example, the accounts used to generate the messages are named after females with a number at the end of the account. Next, the messages all started with a hashtag.
A recent review of a digital conference 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 end points, seemingly at random, often third-party ecommerce sites, and each time with an affiliate referrer attached.
On the surface, a tweet spam attack may seem like a small nuisance, but there are important repercussions that needs to be considered. Without a spam filtration embedded within a social media listening tool, the tool is in danger of giving inflated figures to the organisation using it. If these figures are used by brands to make decisions about future campaigns, the spammers can change the numbers so much that the wrong decisions could be made.
This has clear implications when judging the impact of a campaign, whether that be positive or negative. The data clearly shows that many organisations are over-estimating the reach of campaigns and this will clearly skew any 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.
Peter Laflin is head of data Insight at Bloom Agency, an integrated marketing agency based in Leeds, UK. Peter is interested in using big data to predict how consumers behave and how predictive modelling can be used to gain a commercial edge.
You will be sent a verification email. Click on the link in the email to post your comment.
Opinion, blogs and columnists - call them what you like - this is the section where people have something to say. You might agree or you might not - whatever opinion you have make your views known in comments. Views of writers are not necessarily those of The Drum. If you would like to contribute a comment piece, email your idea to email@example.com.