Twitterbots

17th March, 2014

Recently we have been detailing social media sources and their use to market researchers. Our previous blog entries have covered use and methodology, briefly touching upon utilising social media to provide quantitative data in place of formalised questionnaires. Most social media networks work by numbers: your collection of Twitter Followers, the number of Facebook Likes or the number of Pinterest Pins. These are all useful gauges of overall interest in a sector or specific brand, and along with monitoring the numerous mentions that certain entities receive across social media networks one can begin to deliver a picture of current market situation.

As briefly mentioned in our previous post many social media networks still, inevitably struggle with spam. Spam is big money to whoever produces it, clearly providing some monetary return otherwise we would notice a decrease in false operatives. Facebook has moved swiftly to remove all forms of spambot that could compromise its network. Since its IPO in 2012, there has been a notable decrease in all forms of spam. That is not to say that the spammers are not evolving, but the privacy and security settings of Facebook are now more robust than ever in the quest to apprehend any potential negative factors that could impact its share price. Twitter completed its IPO in 2013 though has had varying success dealing with spambots.

Spambots are automated pieces of code that assist with large scale spam operations, usually creating fake accounts and utilising them as the front of their operations. This practice is particularly apparent on Twitter where surprisingly large markets exist to supply one thing: bots. The bots, once programmed usually latch onto popular search terms (hashtags) and follow the thousands of legitimate accounts interacting with these terms. Whilst most people won’t notice that their Twitter feed has become infested with spambots – one here, one there – the number of bots in operation can be problematic, especially to market researchers attempting to understand their market situation.

Spambots – or Twitterbots specifically – can be used to bulk out an account to promote its social search standings. A small business or an individual can quickly propel themselves up the rankings through sheer weight of numbers as social search, in most cases, does not take into account the substance of an account relying only on the numbers inserted into its algorithm. Businesses and retailers are unlikely to deploy large networks of spambots but nonetheless, due to the bots search and follow algorithms many become entangled in the network, unintentionally sending click now posts or nonsensical gibberish to followers. Again, these are the large networks.

Other spambots target specific search terms and latch onto any account containing their keywords – inflating the overall output of tweets, the number of individual accounts searching for the term and the number of individuals interacting with a specific hashtag. So it is worth considering that if a relatively obscure branch of key words is returning seemingly inflated statistics it could well be the case.

As with all social media trends, this could be set to change. Where Facebook has blocked their spammers, Twitter is moving to alter their terms and conditions in an effort to battle the Twitterbots (under current T&C’s it is not illegal to utilise botnets) but many individuals and business are understanding the vast nature of social media. One study last year monitored 900 shopping portals, noting that 62% preferred a Pinterest Pin button rather than the Facebook Like. Google+ and Badoo also report increases in numbers; will the bots soon become a spammy social media issue also afflicting these networks?