TikTok's content recommendation algorithm attracts more attention than the technologies of competitors like Facebook, Instagram, or YouTube. Here are some reasons why.

Algorithm

Algorithms are considered core to ByteDance's overall operations. Reuters, citing sources, reported that ByteDance would rather shut down TikTok than sell it.

China made changes to its export laws in 2020, giving it the power to approve any export of algorithms and source code, increasing the complexity of selling applications.

Academics and former company employees reveal that it's not just the algorithms, but also the combination with the short-form video format that has made TikTok a global success.

unnamed file 61.jpg
The secret algorithm plays a crucial role in TikTok's popularity. Photo: Reuters

Before TikTok emerged, many believed that the technology connecting users' social relationships was the secret to a successful social media app, typified by Meta's Facebook and Instagram. However, TikTok has demonstrated that an algorithm, driven by an understanding of user preferences, can be far more powerful. Instead of building its algorithm on a "social graph" like Meta's, TikTok executives—including CEO Shou Zi Chew—say their algorithm is based on "interest signals."

Although competitors also possess similar preference-based algorithms, TikTok can enhance the effectiveness of its algorithm with short-form video formats, notes Catalina Goanta, associate professor at Utrecht University. "Their recommendation system is very popular. However, what really makes TikTok stand out is its design and content," she says.

The short video format makes TikTok's algorithm much more flexible and even capable of tracking changes in user preferences over time, down to specific time intervals throughout the day.

Rapid data collection

In addition, the short video format allows TikTok to learn about user preferences much faster, said Jason Fung, former head of TikTok's gaming division.

"Because they are short videos and have a small file size, you can collect data about user preferences much faster than on YouTube, where the average video is under 10 minutes," he shared. "Imagine you're collecting data about users on average every 10 minutes compared to every few seconds."

Positioning TikTok as an app built for mobile devices from the start gave it an advantage over competing platforms, which had to adapt their interfaces from desktop to mobile. Furthermore, entering the short video market early gave TikTok a significant edge in the initial stages. Instagram didn't launch Reels before 2020, and YouTube introduced Shorts in 2021. Therefore, both lagged behind TikTok by many years in terms of data and product development experience.

Explore the content

TikTok also frequently suggests content that is outside the user's interests, something the company's leadership has repeatedly stated is necessary for the TikTok user experience.

A study, published last month by researchers from the US and Germany, showed that TikTok's algorithm "exploits user preferences in 30% to 50% of suggested videos" after examining data from 347 TikTok users and 5 automated bots.

"This finding suggests that TikTok's algorithm chooses to recommend a large number of discovery videos in order to better infer user preferences or maximize user retention by suggesting more videos outside of their (known) interests," the researchers wrote in the study titled "TikTok and the Art of Personalization."

Gather people into groups.

Ari Lightman, a professor at Carnegie Mellon University, points out another effective tactic TikTok has used: encouraging users to form public groups through hashtags. Through this, TikTok can more effectively learn about users' behavior, interests, affiliations, and ideologies.

If TikTok is ultimately banned in the US, according to Lightman, while American tech giants are certainly capable of replicating TikTok with their own products, replicating TikTok's user culture is a daunting task.

China's advantage

TikTok's recommendation algorithm is largely derived from Douyin, an app launched in 2016. While ByteDance often emphasizes that TikTok and Douyin are separate apps, a Reuters source said the two algorithms remain similar to this day.

In return, Douyin's AI is bolstered by ByteDance's leveraging of cheap labor in China. TikTok's parent company hires numerous people to carefully tag all content and users on the platform.

Yikai Li, a manager at advertising agency Nativex and former executive at ByteDance, said: "Around 2018 and 2019, Douyin made an effort to tag every user. They would manually tag every video, then they would tag users based on the videos they had watched." This tactic was also applied to TikTok.

Although hiring people to tag data is a common and important activity for AI companies today, ByteDance adopted this strategy early on. According to Li, tag classification is labor-intensive, giving Chinese companies an advantage due to their abundant and cheaper workforce compared to North America.

(According to Reuters)