TikTok is another short video social APP under ByteDance that recommends algorithms. This short video social APP has inherited the mantle of Toutiao and carried forward the recommendation algorithm once again.
The Tiktok algorithm is essentially:”
Recommend content of interest to users from an aggregated content library.”
If there is a formula to sum it up, it is:
User tag Content tag intelligent matching = intelligent algorithm
User label:
The first is the user tag. From the moment you register your account on TikTok, you will click on something you are interested in every day. You will like it, comment on it and even forward it
Let’s say you like short videos about cars. You like and comment on short videos about cars. You like short NBA videos, you like them, you comment on them and you retweet them.
Tiktok will analyze all your user behaviors and habits according to its system to judge your preferences for such content. Tiktok will give you a user label based on your behavior track, scene and time.
For example: a 27-year-old college undergraduate who likes cars and basketball
According to the information you fill in your TikTok account and your usual user habits, for example, if you like automobile videos for a long time, you often click “like” or comment frequently on TikTok, you will be given the next TAB
I like to watch content about cars. If you like NBA videos, like and comment a lot, and watch them for a long time, then TikTok will also give you the next hashtag.
Content Label:
Each account will label your content, which is convenient for TikTok to match more accurate users to you in the future
So when you learn Tiktok courses and make accounts, the first step is to locate, find a vertical subdivision of the field, and constantly output the vertical field of valuable.
Only when it is vertical and subdivided, can TikTok give you more accurate content labels.
If you post a short video about cars today and a short video about NBA tomorrow, TikTok will not be able to judge your content label at this time, and the traffic he pushes to you will be inaccurate, resulting in poor overall reading and liking data.
Intelligent matching:
The next step is intelligent matching. Tiktok will recommend different content to each user according to their usage habits, usage scenarios and time.
To put it simply, Tiktok’s intelligent matching is to recommend the high-quality works that users like to watch to users. Only when you know his algorithm and produce content according to his algorithm, can your account get the recommendation of large traffic.
Tiktok has two huge databases, one is the user database and the other is the content database. This intelligent matching mechanism of Tiktok is based on the scene, time and usage habits of the user, to retrieve the content he likes from the content database and recommend it to the user!
For example: when you get up in the morning, the first thing you might do is pick up your mobile phone and check Tiktok. At this time, Tiktok might recommend news videos to you.
When you get home and lie in bed at night, Tiktok may be recommending videos about cars that you are interested in.
Tiktok, a set of artificial intelligence algorithm mechanism, is constantly evolving and iterating in the process of matching data to users, and it is more and more able to understand the mind of users. Therefore, you will find that there is a habit of scrolling Tiktok that you cannot stop.
Horse racing mechanics
In Tiktok’s algorithm, the regional hub, an account with 10,000 followers and an account with 1 million followers. They have an equal chance of getting high traffic as long as you have good content.
The number of users who like the auto field every day is constant, let’s say 10 million users, and the auto field account will produce 10,000 short videos every day.
The 10 million traffic will be allocated to the 10,000 accounts. If each account releases a work on the same day, TikTok will allocate different traffic according to the data feedback of the work.
If your work has a high completion rate, a high liking rate, a high number of comments and a high sharing rate, it will recommend you a large amount of traffic and push you to the top. Tiktok is using this set of horse racing mechanism to screen quality content and then recommend it to users who like to watch videos in related fields
This case only gives you the tip of the iceberg of how Tiktok can make huge profits. There are more interesting ways. In fact, I think any industry can use short videos to do it again in the future!
Clothing short video,
Food short video,
Beauty short video,
Baby products short video,
Daily necessities short videos
Training short video
In 2020, TikTok will be a huge traffic entry point. In the following articles, I will analyze how traditional industries should transform by combining TikTok short videos and play short videos together.
