Emirjon Senja
TikTok is still there...
The government's decision to shut down the TikTok platform appears to have had the opposite effect, at least for now, if we refer to statistical data on video viewership. Data published by Pikasa Analytics for the period March 6 to 20, which coincides with the week before and after the decision to close, shows a 14% increase in video views, despite a slight 3.3% decrease in posts. This showed that the closure was largely ineffective in terms of reducing content consumption.
While the TikTok ban did not significantly impact the platform's viewership, it appears to have impacted user behavior on other platforms.
Instagram saw a 16% increase in interactions during the week after the ban, despite a slight drop in the number of posts.
This phenomenon seems to have emerged also due to the period before the official launch of the election campaign, where a large number of candidates for MPs, including those from the majority, chose this platform to communicate with citizens, despite the ban.
"Naked" in front of algorithms
A study on the behavior of social media users in Eastern Europe, published by the Thomson Foundation, shows that in most cases, the audience makes a passive consumption of content on various platforms such as Facebook, Instagram or TikTok. Such behavior confirms the fact that it is the content that chooses the audience and not the audience that seeks the content.
The study talks about "passive browsing," where users continuously browse videos without having a clear idea of what they want to see.
This behavior is especially reinforced after the introduction of the “reels” format, first by Instagram and then by Facebook, while it has been a permanent practice of TikTok. In this way, the audience is completely dependent on the selection that the algorithms make for them, completely exposing them to the classic “symptoms” of social networks related to the way the algorithms target the audience.
In social networks, algorithms are systems designed to keep the user on the platform as long as possible, favoring the distribution of content that causes an emotional response and, consequently, high engagement.
As a result, algorithms not only select the content that appears to us while using social networks, but they also shape the way we perceive reality.
Moreover, algorithms significantly reduce exposure to opposing opinions , contributing to ideological division, misunderstandings between different groups, and further promoting the polarization of society, especially during periods of social tension, uncertainty, or elections, as in our case.
During the recent elections in Albania, this behavior was most evident in the high number of interactions on social media pages or profiles with an editorial policy strongly supporting one or the other political wing. This resulted in citizens being exposed primarily to one-sided narratives, without the opportunity to compare information from different sources.
Here we enter the terrain that in technical language we call "echo chambers", which are nothing more than virtual spaces or "rooms" where, because users only follow sources that publish content that confirms their beliefs and interact only with similar individuals, they create a completely closed reality by ignoring any alternative stance or idea.
In this new, disconnected reality, opposing ideas are ignored or demonized. Not coincidentally, the discourse on such online sites was harsher during the electoral campaign and elections, compared to other media outlets that have a more balanced editorial policy.
This is the first step in the big challenge that social networks present for the way algorithms target their users. Since algorithms are designed to keep users engaged for as long as possible, being inside these “echo chambers”, they are constantly “bombarded” with the same content that they have interacted with before, and now they have not only lost the chance to have content on their “wall” that contradicts their ideas, but they also move into the second phase of manipulation by algorithms: the creation of “filter bubbles” or otherwise, “algorithmic bubbles”.
In both cases, the user has the illusion that they are being informed, when in fact they are losing the opportunity to create a complete and balanced view of certain developments.
META vs. TikTok: How platforms influence the distribution of disinformation content
Facebook and Instagram, as part of the META group, use algorithms built on social interactions and user history, weaving together existing connections and previous user behavior.
Essentially, META algorithms favor content that comes from friends, pages, or groups that the user has interacted with most frequently in the past and prioritize posts that have received likes, comments, or shares within a given social circle.
Meanwhile, TikTok's content recommendation logic is quite different from META's, raising greater challenges in terms of what the audience consumes there. TikTok focuses primarily on the content itself and the relationship the user has had with this type of content, regardless of whether or not the user follows the profile that uploaded the video.
TikTok's algorithms analyze each video's watch time, quick interactions, and other details like description, audio content, and hashtags to target the video's potential audience. This makes TikTok extremely efficient at identifying and distributing content that has high potential to go viral.
While Facebook primarily aims to strengthen existing relationships through personalized content within a closed network, TikTok creates an open and decentralized experience, where any video can be shown to millions of users, regardless of who uploaded it.
This fundamental difference in distribution logic gives TikTok a much faster pace for content dissemination and a greater potential for virality, making it a much more dangerous platform for the distribution of disinformation content.
Promoting polarization
By analyzing how social networks target users, we can say that Facebook users are more at risk of becoming part of "echo chambers", while TikTok users are more likely to enter "filter bubbles", consuming similar content from different creators.
In the context of the Albanian election campaign, a typical case that confirms how social media algorithms help spread misinformation was the one about a rumored sexual affair of a parliamentary candidate, shared just two days before election day. The post received almost double the interactions on the TikTok profile of one of the most followed pages in the Albanian space, compared to those on Facebook, until the day it was deleted.
Given the way the messages were conceived during the electoral campaign, it seems that through Facebook and Instagram, the parties aimed to consolidate their electorate, while on TikTok they experimented with more informal videos , targeting young voters.
In an environment where content distribution is driven by algorithms that personalize content, digital education becomes essential in order to strengthen users' abilities to understand, analyze, and evaluate how social networks work and how and why a certain post appears on their profile.
This is especially important in the context of societies with low levels of trust in institutions, such as ours , where the risk of being manipulated by disinformation campaigns is even greater.
