BLOOMINGTON, Ind. — A study conducted by researchers from the University of Southern California and Indiana University found that as many as 15 percent of active users on Twitter — or 48 million accounts — are actually robots.
Twitter “bots” are essentially accounts that are run remotely by software to send out automated messages and tweets. They exist to increase follower counts and make new connections without any effort on the part of an individual. Some of them even respond to tweets.
To determine how many of these bots exist in the Twitter universe, the researchers came up with 1,150 different features that set them apart from the average Twitter user. Those features included information about the user or the username, measurements surrounding tweets, such as the time between them, and also the sentiment and tone behind them.
The team then scoured the site to get an estimate of how many Twitter user displayed those features. They determined that between nine and 15 percent of users are bots, but think the estimate is on the higher side since some of the high-functioning bots might have actually slipped under the radar and been mistaken for real users.
“These complex bots may be active on Twitter, and therefore present in our datasets, and may have been incorrectly labeled as humans, making even the 15% figure a conservative estimate,” the paper, “Online Human-Bot Interactions: Detection, Estimation, and Characterization,” states. It adds that some accounts may be “cyborgs,” a human-bot hybrid that uses automation under the supervision of a human.
“Some have been allegedly used for terrorist propaganda and recruitment purposes. It remains unclear how these accounts should be labeled, and how pervasive they are,” the authors write.
The researchers classified three different types of accounts that included outright spammers, self-promoters, and accounts that post content through the help of connected applications.
Twitter estimates the number of bots far lower: between 5% and 8.5%, according to a report the company released to the US Securities and Exchange Commission in 2014.
Whether bot accounts exist to cause problems or bring positive impacts to the social network, they make it a bit more complicated for brands and companies to accurately assess how far their engagement reaches on the social media site. When they’re engaging with a bot as opposed to a real potential human customer, they may be wasting more time and money than they actually realize.