TikTok and Network Effects
https://newsroom.tiktok.com/en-us/how-tiktok-recommends-videos-for-you
TikTok has emerged as a competitive social media platform this past year placing its name among other platforms such as Facebook, Twitter and Instagram. The app features many new features other platforms lacked such as a For You page. The For You page contains videos a user would mostly enjoy viewing specific to their interest. TikTok has gained popularity for the algorithm responsible for this. People wondered how videos are recommended to a user and pop up on their For You page. TikTok has a “number of factors” that are taken into account while recommending videos to a user. The first factor is “User interactions” which include “videos [users] like or share, accounts [users] follow, and comments [users] post.” If a user interacts with a video, they are more likely to view more videos similar to it on their For You page. The second factor taken into account in recommending videos for a user’s For You page is “Video information.” This factor looks at certain features of a video a user views such as its “captions, sounds, and hashtags” and uses them to suggest videos with similar content. The final factor used in the recommendation system is the users’ “Device and account settings” like “language preferences” and such. These three factors together work to populate a user’s For You page with videos of which they have a high likelihood of interest.
If we consider a user posting a TikTok video as a product being sold by the user to other users for views and likes, then this market has a strong network effect. A network effect occurs when a good’s value increases as more people use it. In class we discussed how Facebook becomes more valuable as more users create accounts as this creates more interactions. It is kind of a similar situation with TikTok, when a user posts a video the more people that view the video the more chances of those users interacting with the video and sharing it to even more people. One way the recommendation system populates the For You page of a user is by looking at the accounts they follow. If one user likes a video that video has a chance of popping up on the For You page of the followers of that user. Due to this if a user posts a video and it gets a lot of views and likes they have a higher chance of getting even more likes and views. Therefore in this presence of a Network Effect a user should aim to get as many views and likes as possible after posting to surpass z’ (where z is equal to the fraction of users that viewed and liked the video) to create an “upward pressure” on the views and likes of the video. If z is between 0 and z’ fewer people will view the video creating a “downward pressure” and making the video flop.