Cracking the TikTok Algorithm
Written by Ada Luo, Research Assistant in Cornell’s Social Media Lab and Information Science Major
TikTok isn’t the first social media platform to have an algorithmic system that tailors content to the user. YouTube recommends videos on their homepage and on the sidebar besides videos that the user watches. Instagram has an explore page that is full of pictures and videos of content that they think the user might enjoy. Although these algorithmic systems remain largely opaque, users start trying to crack the code of social media algorithms to improve the visibility of the content they created and to also get access to content they actually want to see.
According to Madrigal (2018) “YouTube heavily personalizes recommendations based on a user’s history”, such as likes or videos they’ve watched in the past. According to Canning (2019) “The Instagram algorithm is constantly learning from your behaviors on the app, like what accounts you follow, what posts you like and comment on, and what Instagram Stories and IGTV content you love to watch”. Viewers are also able to follow or subscribe to content creators that they enjoy watching. But how might creators reach these viewers in the first place? A possibility is by hashtagging or tagging their content so that people who are interested in the topics described can find them. On Instagram, there is a bigger hashtagging culture and many viewers actually browse through the hashtags to find content of interest. On Youtube, videos are tagged with the type of content it is, like gaming or fashion or cooking. Users may look through those topics starting from the explore page on Youtube.
A difference TikTok has from Youtube and Instagram is that there is an “unprecedented community effort by the platform’s creators and viewers to figure out exactly why some videos are recommended, and some aren’t” (Hale, 2019). The TikTok “For You Page” (FYP) is the default page of TikTok that showcases videos that the user might like. Although there is a separate page to see content just from followed creators, videos from creators users follow will also show up on the FYP. Because many creators want to expand their reach, many will test out different ways to present their content.
Before TikTok released the breakdown of their algorithm in June 2020, users have been experimenting with ways to hack it to boost their interactions. One of these strategies is adding various trending hashtags to their video captions. Some examples are “#foryoupage,” “#foryou”, or “fyp,” which directly refer to the FYP that creators want to get on. Other non-conventional hashtags that are popular lately are “#AmazonMusicJingleBellTok,” “#MacysGiftTok,” and “#ASOSChaoticToCalm.” These hashtags originate from advertisements made by companies and are then used on videos that may not be related to the advertised product. There is no evidence that hashtagging works but many videos have these hashtags and people keep on seeing videos with those hashtags on their FYP. Creators think that putting these hashtags helps with increasing their exposure.
Another theory is that “videos are deemed worthy of being included on For You if they perform well in a batch test” (Hale, 2019). Users theorize that a video is first shown to the creator’s followers and a small group of non-followers. If the video does not do well in this ‘batch’, meaning getting views but not likes or comments, then it won’t be shown to more people. If it does get interactions, then it will be showing up on more people’s FYP. This theory is also being investigated by audience members. They repeatedly tap the ‘Share’ and ‘Copy Link’ combination where they don’t have to actually share to someone but still increase the number of shares on the video. They also purposefully like videos with a small number of likes or comment things like ‘fyp’ on smaller videos even if they don’t necessarily like the videos but try to make it go viral.
In addition to trying to make videos go popular, many creators put in effort to make sure their videos won’t be suppressed. TikTok also implemented content moderation to censor videos that they think violate their community guidelines. This may include content with inappropriate language, hate speech, and many others. However, many creators notice that even though their content complies with the community guidelines, it still may be blocked by the moderation algorithm and they could become shadowbanned, which is “removal or suppression of content without the platform notifying the user that their content is in violation of any community guidelines or usage rules” (Cortés, 2020). To combat this, creators realized that they can censor words that may be flagged by TikTok, using non-alphabet characters to replace similar looking letters (‘death’ to ‘de@th’), using a combination of letters that sound similar to the original word (‘sex’ to ‘seggs’), combination of words that are synonymous with the original word (‘commit suicide’ to ‘commit un-alive’).
As TikTok grows and more content is being created, users will continue experimenting with the algorithm. Creators may be particularly interested in getting more interactions because of the TikTok Creator Fund, where creators can be paid for their content based on “a combination of factors, including the number of views and the authenticity of those views, the level of engagement on the content, as well as making sure content is in line with [TikTok’s] Community Guidelines and Terms of Service.” TikTok influencers are also making way for a new kind of celebrity: seemingly normal people, but with millions of fans and endless opportunities, that many may argue they do not deserve. Many of these influencers go viral out of nowhere, skyrocketing into fame without much warning. For young creators, this can pose issues of privacy and safety where everyone is watching their every move. Any mistake, no matter how small it is, is amplified and viewers will make sure they never forget it, sometimes in the most hateful ways.
Other creators that are interested in getting ahead of the algorithm are vulnerable populations that do not have a chance to voice their concerns regularly. Many creators are part of minority populations and post content on TikTok so that others can be educated about their culture and their struggles. They hope to gain support in their communities and reach audiences that would otherwise be clueless about them. They need to make sure their content isn’t lost in the sea of many others. This may be difficult when the TikTok algorithm itself is working against them. Their content may get flagged “because they are someone from a marginalized group who
is talking about their experiences with racism. Hate speech and talking about hate speech can look very similar to an algorithm” (Ohlheiser, 2021). These creators need to be more cautious about what they post, especially since there are higher stakes, like human rights, in their content.
Although there may never be a sure and definite way to make your account go viral, users are determined to make the most of it. However, more security and privacy guidelines should be in place to protect users from the potential risks introduced by social media’s algorithms, especially vulnerable populations.
Canning, N. (2019, May 22). The Ultimate Guide to Getting on the Instagram Explore Page. LaterBlog. https://later.com/blog/how-to-get-on-instagram-explore-page/
Cortés, M. S. (2020, July 17). What Is Shadow Banning & Why Are TikTokers Complaining About It? Refinery29. https://www.refinery29.com/en-us/2020/07/9901461/what-is-shadow-banning-tik-tok
Hale, J. (2019, August 15). TikTok Creators And Their Fans Are Trying To Crack Its Recommendation Algorithm (Report). Tubefilter. https://www.tubefilter.com/2019/08/15/tiktok-creators-and-their-fans-are-trying-to-crack its-recommendation-algorithm-report/
Madrigal, A. M. (2018, November 8). How YouTube’s Algorithm Really Works. The Atlantic. https://www.theatlantic.com/technology/archive/2018/11/how-youtubes-algorithm-really works/575212/
Ohlheiser, A. (2021, July 13). Welcome to TikTok’s endless cycle of censorship and mistakes. MIT Technology Review. https://www.technologyreview.com/2021/07/13/1028401/tiktok-censorship-mistakes-glitches-apologies-endless-cycle/
TikTok. (2020, December). Community Guidelines. Retrieved December 11, 2021, from https://www.tiktok.com/community-guidelines?lang=en
TikTok. (2021, March 25). TikTok Creator Fund: Your questions answered. Retrieved December 11, 2021, from https://newsroom.tiktok.com/en-gb/tiktok-creator-fund-your-questions-answered
