Dissecting the “algorithm logic” of Douyin and Kuaishou

This article introduces the algorithm mechanisms and challenges of the two major platforms, Douyin and Kuaishou, and explores how to optimize the content ecology and achieve higher-quality content delivery by integrating just content algorithms and ESG concepts.

Nowadays, Internet platforms are deeply connected to people’s lives. People browse information, find like-minded people, and shop on various platforms. Virtual digital communities provide almost everything people need in life today, and the driving force behind ensuring that everything runs in an orderly manner is “algorithms”.

How should we understand "algorithm"? The algorithm is like an intelligent "content dispatcher". It determines which content will be pushed to which users based on user behavior data and content features.

For example, on the TikTok platform, the algorithm extracts content that users are interested in and pushes it by analyzing browsing history, likes, comments, favorites, and other behaviors; on Kuaishou, the algorithm not only considers user interests, but also pays attention to social relationships . If you often interact with certain creators, then the "interest community" composed of similar users is more likely to be pushed to you.

However, with the surge in the number of creators and users of Douyin and Kuaishou, the "content competition" enabled by algorithms has become increasingly fierce, so how algorithms create content value and how algorithms optimize the content ecosystem have become key considerations for platforms.

01 Tik Tok trapped in the "information cocoon"

People who are familiar with the Douyin algorithm have more or less heard of a term - decentralized algorithm. That is to say, Douyin will provide content based on user interest preferences, rather than having a centralized editorial team decide the flow of content.

As a result, Douyin has become the most typical "label" to "label" platform. As a creator, the platform will form a "creator label" based on the published content and recommend it to users with the same interests; as a user, the platform will form an "interest label" based on browsing preferences, directly realizing the connection between creators and users.

The advantages of this algorithm are obvious:

On the one hand, facing the different needs of massive users, tags amplify the personalized "content experience".

The most popular short video category on Douyin is the food track. Under the #美食 tag alone, the video playback has reached an astonishing 198.67 billion times. In order to enrich the food short videos, they can be subdivided into different forms such as food reviews, food tutorials, eating broadcasts, and store visits, giving users more personalized choices.

On the other hand, labels are more conducive to making each vertical category more "vertical" and easier to spread.

In July this year, the number of views of topics related to "Shi Tiesheng" on Douyin exceeded 3 billion times, driving the sales of his book "I and the Altar of Earth", ranking first among the most popular classics on Douyin, and also making him one of the most popular writers on Douyin this year, returning to the public eye. Against this background, Douyin held an offline sharing event titled "He Once Again Stirred the Strings of a Generation - Returning to the Altar of Earth, Rereading Shi Tiesheng", connecting humanities and social science enthusiasts through platform data and user preferences, and Douyin's "humanistic atmosphere" continued to heat up.

However, the dilemma of the Douyin algorithm has gradually become apparent. Because the algorithm emphasizes personalized recommendations, users are firmly trapped in the "information cocoon". Once an interest tag is formed, the algorithm will continue to push content that matches it, making the information that users are exposed to become monotonous. Just like a user who only pays attention to funny videos, his recommendation page is almost filled with funny content, and high-quality content in fields such as culture, technology, and art rarely has a chance to enter his field of vision. This relative limitation of information is described by young people as "there seems to be a wall between the Douyin you play and the Douyin I play."

In essence, the "algorithm value" of Douyin lies in accurately meeting the personalized needs of users, while the "content value" is reflected in providing rich, diverse, in-depth and high-quality content. However, under the current algorithm mechanism, the two cannot maintain a balance. Although Douyin regulates traffic distribution through algorithms, distributes traffic based on the comprehensive performance of accounts, avoids excessive concentration of traffic on top accounts, or encourages the creation of high-quality content through hot topic challenges and "short and direct linkage", these measures have not fundamentally broken the shackles of the "information cocoon".

Because the underlying logic of the algorithm has not changed, it still prioritizes satisfying the personalized needs of users, but relatively sacrifices the diversity and comprehensiveness of the content. As content demands become increasingly diversified and refined, if Douyin wants to achieve a balance between "algorithm value" and "content value", it may need to explore a diversified content algorithm model.

02 "Local Kuaishou" urgently needs to be explored in depth

Compared to Tik Tok's algorithm that emphasizes personalized content recommendations, Kuaishou's algorithm focuses more on "social relationship" links and user interactions. In layman's terms, it places more emphasis on each individual.

The favorable impact of Kuaishou's algorithm is that it covers a wider age range and is closer to the sinking market of third- and fourth-tier cities; the form of content creation is also closer to rural social life. Therefore, many creators from rural areas and small towns use Kuaishou to share daily life, talent shows, etc. Some rural creators attract fans by showing the natural beauty of the countryside, farming life and folk culture, allowing more people to appreciate the real rural landscape.

This makes Kuaishou increasingly closely tied to the label of "local".

Kuaishou is like a colorful "rural culture garden". When people mention Kuaishou, the first impression they get is the down-to-earth title "Lao Tie", followed by the popular sports such as Village Super and Village BA in recent years, or the large number of live broadcasts on agriculture, rural areas and farmers in Kuaishou. Therefore, this rural impression deepens the algorithm logic, allowing Kuaishou to have a large number of "rural elements" sections.

However, Kuaishou also faces some challenges in the process of deeply binding with the "local" label. On the one hand, a large amount of local content may cause users to be aesthetically fatigued; on the other hand, a large amount of local content may make it difficult for other types of content to be fully developed. For example, the content in the fields of science and technology, culture, and art is not widely disseminated on Kuaishou, which may also limit the expansion of Kuaishou's user base.

So, how does Kuaishou use algorithms to create content value and optimize the content ecosystem?

One way is to make "local" more profound, which is also the content transformation idea that Kuaishou is doing. Kuaishou can use the precise analysis of algorithms to explore the content with depth and connotation in rural culture. For example, craftsmen like "Grandpa Amu" who inherit ancient folk skills and women like Xiajie in northern Shaanxi who show the folk life in the northwest have extremely high content value. At the same time, through algorithms, creators are guided to innovate content forms and quality on the basis of maintaining local characteristics, and use more vivid narrative methods to make the impression of local no longer limited to backward perceptions.

Another way is to find local materials through algorithms and make "local sense" a program brand of Kuaishou. Fei Xiaotong's "Rural China" uses popular and concise language to summarize and analyze China's grassroots society, and comprehensively shows the face of China's grassroots society. Kuaishou can explore the influence of local cultural content in the contemporary context of China. This approach has actually been presented. "500 Hometowns" is a program formed by Kuaishou focusing on China's rural areas, where different people tell their stories about their hometowns. For the Kuaishou content ecosystem, telling the value of rural content is undoubtedly the most important link next.

03 The advent of the “Just Content Algorithm”

As the two top platforms in China, whether they are facing the "information cocoon" or continuing to deeply explore the "local culture", they are all telling one thing: it is not realistic to rely solely on existing algorithms to create "content value". The platform urgently needs to integrate a "just content algorithm" to help optimize the content ecology.

The so-called "just content algorithm" is, in simple terms, to focus more on screening and disseminating high-quality content on the basis of existing algorithms. This is also the practice of the "ESG concept of the entertainment media industry". From the S (social responsibility) dimension, good content is a good algorithm. Pushing high-quality content to users is actually a manifestation of the platform's practice of social responsibility; and from the G (governance) dimension, using good algorithms to allow good content to emerge is itself an important part of "content governance". Therefore, promoting the platform to integrate the "just content algorithm" and "entertainment media ESG concept" is a drive upgrade for the platform's content value.

Take the "Douyin Selection" launched by Douyin as an example. This is a positive attempt. "Douyin Selection" recommends those truly valuable and meaningful short videos on humanities and social sciences to more users through a series of screening criteria. These contents cover many fields such as knowledge popularization, culture and art, and social welfare. It will serve as high-quality content to meet higher-quality user needs.

Kuaishou also has ESG-oriented actions. For example, in terms of "driving employment", Kuaishou provides job-seeking channels for job seekers through the "short video + live broadcast" format. In 2023, the total number of "Kuaipin Live Broadcast" sessions was nearly 5.5 million, and the number of online jobs in the manufacturing industry doubled. By the end of 2023, the Kuaishou platform had driven a total of 40.22 million employment opportunities, and also created more than 30 new occupations such as anchors, assistants, and product selectors, demonstrating Kuaishou's attitude towards social responsibility.

It is worth noting that in the process of integrating the "Just Content Algorithm", the platform needs to clarify the standards for high-quality content. High-quality content must not only have a certain degree of professionalism and artistry, but also conform to the mainstream values ​​of society and be able to convey positive energy. In addition, the platform needs to establish an effective feedback mechanism. In particular, users, as recipients of content, their feedback is crucial to optimizing the "Just Content Algorithm".

Only by incorporating the "Just Content Algorithm" into algorithm iterations and continuously practicing the "ESG Concept" can platforms such as Douyin and Kuaishou truly maximize the "content value" and avoid falling into the algorithm's rut.

Author: Fenwei This article is written by the author of Operation Party [Media No. 1], WeChat public account: [Media No. 1], original/authorized to be published on Operation Party, and any reproduction without permission is prohibited.

The title image is from Unsplash, based on the CC0 protocol.