How Chinese social media users are fighting their platforms

If you spend any time on Chinese social media, you're almost certain to come across a strange phrase: “Send you trending.” This phrase is ubiquitous in reports about serious social injustices, such as companies cheating their customers. It's shorthand for a movement, users' way of encouraging the rapid spread of information to bring about social change.

The way it works is relatively straightforward. Most Chinese social media sites maintain a list of “trending” topics. In theory, this is a collection of the most popular and discussed keywords at any given time. These are selected and displayed according to the platforms' often opaque algorithms. Although the lists are usually dominated by entertainment news and gossip, they are still subject to a certain degree of user control, and topics with enough likes and comments can quickly shoot up the charts. Once they're there, they're seen by all users of the platform and can generate significant public attention and discussion.

That is, the spontaneous behavior of netizens commenting “Send you to trending” can be seen as a kind of resistance to the existing sorting logic of the algorithm through collective action. At the heart of this resistance is the struggle for visibility, where different interest groups seek to increase their public visibility and social discursive power.

Let's look at a typical case. On May 7, 2022, a user posted a video on the microblogging site Weibo accusing a Nankai University faculty member of having an inappropriate relationship with a female student. The video attracted a large amount of traffic to the comments, and support quickly spilled over from the original post to the university's official Weibo account. Soon, keywords such as “Nankai University” and the faculty member's name began trending. Posts with related terms quickly exceeded 100 million views and ranked around 11th on Weibo's trending list.

On May 8, these terms were removed from the trending section and Nankai University's official Weibo account closed comments on their posts. Two days later, Nankai University told a China News Weekly reporter that it was investigating the incident and would look into the matter. This quickly led to the appearance of a new sentence – “Nankai University responds to the signed accusations against its teacher” – on Weibo's trending page before that tag too was removed on May 11.

But the case didn't end there. After activating comment protection on May 13, Nankai University announced the results of its investigation, which included penalties for three faculty members. As of May 15, the hashtag combining Nankai University and the accused teacher's name had been viewed more than 220 million times and generated over 700,000 interactions – numbers that largely came about despite the platform's interference.

To be successful in this battle, tactics must be constantly refined, and social media users are constantly developing new approaches to trick the algorithm. One such tactic is “encryption,” in which users create encrypted communication methods to prevent their messages from being automatically filtered out by algorithmic detection technologies. These include using abbreviations (such as “NK” for Nankai University), inserting punctuation into sensitive words (such as “sex/ual assa/ult”), using special phrases or symbols (such as breaking Chinese characters into their component parts), or sharing text screenshots.

Another tactic is “appropriation,” in which users seek algorithmic visibility by actively clicking, sharing, commenting, adding hashtags, and paying for memberships to leave a trail of data for the algorithm to analyze.

Additionally, users can achieve more productive results by employing tactics such as “relay” and “bricolage.” Relaying refers to the posting of the same text by several different users, where users demonstrate a spontaneous connection and sense of community. While the platform sometimes deletes accounts that share information, users usually encourage each other to keep doing so, creating a sort of collective identity that, no matter how unstable or imagined, unites participants and builds momentum.

“Bricolage,” meanwhile, is evident in users’ creative use of platform features, such as taking over the comment sections of independent news reports or linking related topics to events trending during the same time period. During the Nankai University incident, for example, netizens linked the university to topics such as Mother’s Day and the 2008 Sichuan earthquake. An original post with several hashtags read, “Today is Mother’s Day and I’m so angry at your university that I can’t sleep… What are women in 2022? Must every victim be humiliated like this before being heard?” Online posts like these are intended to create more space for discussion and social support for related issues.

Overall, algorithmic resistance is a form of self-organized, collective action by grassroots users committed to correcting platforms' unfair visibility policies and demanding social justice. Of course, this type of spontaneous resistance suffers from a lack of sustained critical reflection, and users have neither the motivation nor the ability to engage continuously. Netizens' comments are often irrational and emotional, resembling a cathartic reaction rather than a meaningful, coherent outpouring. Some even contain violent remarks that deviate from the original intention of fighting for justice and personal dignity.

In conclusion, ordinary internet users participate in algorithmic resistance because they are dissatisfied with online visibility controlled by platform algorithms. However, their method of resistance is to actively participate in the competition for algorithmic traffic, thus becoming even more integrated into the existing algorithmic sorting mechanism, which then leads to even more traffic on the platforms. Ironically, the biggest beneficiaries of algorithmic resistance may be the platforms themselves.

Translator: David Ball; Editor: Wu Haiyun; Portrait artist: Wang Zhenhao.

(Header image: Ramas/VectorStock/VCG, reworked by Sixth Tone)