September 16, 2020

The Infamous TikTok Algorithm

In efforts to show full transparency, TikTok has revealed more information regarding its infamous algorithm. TikTok’s algorithm uses machine learning to learn about user interests and feed them that same type of content for maximum engagement. They source videos that are similar or liked by people with similar user preferences.

Here’s a breakdown of what takes place:
1. When users open TikTok for the first time, they are shown 8 popular videos that differ in trends, music, and topics. Based on what you interact with, they’ll continue to show you similar videos.
2. With each video that is shown, more information (interests) are determined and a user is shown videos in that same realm. The information captured includes details such as captions, hashtags, and sounds.
3. Once enough data is collected, TikTok is able to map a user’s preferences in relation to similar users and group them into “clusters” to further feed you the content you want to see.
4. Machine learning allows TikTok to serve videos to clusters who have similar interests and engage with the same type of content.
5. One thing you will notice is that TikTok tries to ensure you don’t see the same video twice. They also make a conscious effort to avoid redundancies such as using the same music or videos from the same creator over and over again.

Why does this matter? How TikTok chooses your next video to watch is a large part of why the US government is concerned with the app’s Chinese parent company and their relation to the Chinese government. Although TikTok is trying to be more transparent of how they operate, some people are still concerned about security and privacy. What are your thoughts on how TikTok targets users?

Source: https://bit.ly/32AglTy

-Nic Mercado
Director of Operations & Strategy