Portaltic.- A group of researchers has discovered 400 ‘herds’ (‘pods’ in English) on the social network Instagram, made up of groups of hundreds or even thousands of accounts that inflate likes and comments to trick the algorithm and improve its positioning.
The research, conducted by the Tandon Institute at New York University (United States), has analyzed 1.8 million posts from more than 110,000 Instagram accounts shared through Telegram channels by 400 account herds.
The herds of Instagram integrate a “robust and hidden ecosystem”, as the researchers have described, and are made up of an average of 900 accounts, although some have as many as 17,000 users and receive 4,000 messages a day.
These herds work through a technique called ‘reciprocity abuse’, where each member interacts with the content published by other members of the group to improve their results with the algorithm of the social network.
Until now, most attempts to fool automated social media systems have involved the use of automated bots and scripts, but companies are increasingly mitigating these attacks.
“Herds, however, involve people who operate manually, making them more difficult to detect,” as Janith Weerasinghe, one of the authors of the New York University study, explained.
70 percent of herd users managed to double the organic interaction levels of their posts and, on average, these users managed to multiply by five the number of comments on their Instagram accounts.
Often herds are focused on specific topics of interest and their publications are related to areas such as fashion, photography or entrepreneurship.
Also, the investigation indicates that the barrier of entry is low, since only 4 percent of the herds discovered asked their new members to previously have a minimum number of followers to be able to join the group.
It is also a technique that has been spreading in recent years. “The herds we discovered have grown at an accelerated rate in the past two years,” as Weerasinghe concluded.