Some people say that Douyin is currently the lowest-cost marketing platform and the last traffic depression.
If you, like me, have contact with many people who want to do and are doing Douyin, you must be familiar with the following questions:
- Has the Douyin been restricted recently? My previous video playback was over 10,000, but now it only has three digits?
- Now Douyin doesn’t feel as easy as it used to be. A popular video of mine has only gained a few thousand fans...
- Why do people’s videos with similar content become popular, and no one cares about mine?
Obviously devoted a lot of time and energy, but failed to achieve the expected explosion and growth. What is the reason behind this?
To solve the above problems, proceed from the following two aspects:
- Douyin's algorithm and recommendation mechanism
- How to make high-quality content that users love
When many people used to do official accounts, the number of fans was an extremely important thing, because with fans, the articles posted on your official account have the opportunity to be read and shared. But inside the vibrato, the system according to the characteristics you publish short video content, the video automatically "share" to which people think will like this video .
For ordinary people, this is the greatest opportunity during the bonus period. Even if your account does not have a fan, the video can still be shared by the system and seen by thousands of people.
The algorithm used by Douyin is a " traffic horse racing mechanism ", in simple terms: first a small amount of traffic is given to the new video that is arranged, and then various indicators (such as like rate, broadcast completion rate, etc.) are compared based on similar videos. After winning the PK of the data and similar videos, gradually give more traffic.
In this way, as long as the various indicators of the video are good enough, new traffic can be continuously obtained. If it is compared in a certain PK link, then the platform will no longer give a larger amount of traffic.
So, what kind of video can break through on Douyin and become popular?
2
If you frequently use Douyin, you will find that the types of accounts recommended by Douyin can be basically divided into two types:
The first type, the obvious feature is that the number of likes for popular videos may be hundreds of thousands to millions, but when we click on his homepage, we will find that this type of account has actually uploaded many works, but Most of the videos are tepid and don't have too many likes, and the video you see is just one of his few hot videos. The total like-to-follow ratio of this type of account is mostly below 10:1, and the user's evaluation of the video content is too unstable, and it depends on luck to a certain extent.
But there is also a problem with vertical accounts: most of the videos’ likes are not particularly high, they may be in the range of several thousand to hundreds of thousands, and there may be videos with millions of likes occasionally, but not many.
The reason is that when the system gives traffic that meets its target population, its video content will spread quickly, but when the traffic given exceeds its target population, the data is not necessarily good, and it will eventually fail to win from the traffic PK, so as to get more Large Flow.
For example, if you are making a tutorial video for using an Apple mobile phone, then if the video quality is good, your video can be quickly disseminated among Apple users, but if you give you the traffic of Android users, then the various indicators of the video will be Will be pulled down, the system will not give you more Android traffic.
So the question is, how does Douyin know which videos to recommend to which users?
3
Here is a knowledge point: collaborative recommendation.
We said that in addition to the content itself, the collaborative recommendation of the Douyin platform algorithm is based on the user's own recommendation. Based on the analysis of the similarity between different users, the exploration capabilities of the recommendation can be expanded. For example, Zhang San likes technology, finance, and sports, and Li Si likes technology, finance, sports, and health content. Then, the algorithm will try to push health information to Zhang San. Because in the recommendation system, Zhang San and Li Si are similar people. As a result, Zhang San can receive health information that was not originally in his interest.
For the new account, the first batch of Douyin traffic will be given to people near you, friends in your address book, and people who might be interested in the video. For mature accounts, the recommendation of the system is to recommend the content you are interested in to people with similar user characteristics , and then decide whether to enter the next batch of recommendations based on data feedback.
4
A user may like a funny video you accidentally shot, and will spread it, but his actions to pay attention to you must be based on one of the most basic needs: paying attention to you is valuable to him. This puts forward requirements for continuous content output capabilities.
How to bring users more creative and continuous output? This problem has bothered many friends.
Although short video is a new gadget, it still can only rely on content to get traffic. The source of inspiration for the content is a very easy way-from the official account material.
The content of the official account can detonate the circle of friends, which is much more difficult than detonating on Douyin. The difficulty lies in that one requires a certain user base, and the other requires user action to forward it to spread. TikTok recommending content to users does not need these two points at all, so content that can be detonated in Moments must also be exploded on Douyin.
If you sort out all the explosive content that has exploded in the circle of friends over the years, and make a table, this is a good material toolbox, from qq space to Weibo, from Weibo to WeChat, and then to Douyin. .
One of the benefits that the Internet brings to us, in any field, if you really want to learn and pay, there will be professional people who will guide you quickly, learn 70% of the necessary knowledge in this field, and rely on us for the remaining 30%. Think, practice.