How to build an operating system for community products

Community can be considered a form of Internet product with a very long history.

The earliest community was BBS, which was basically born together with Internet technology. Moreover, this product form has been constantly evolving with the development of the Internet. From the initial BBS to the current Zhihu, Weibo, and Bilibili, the community has always been a hot topic in the Internet field.

There are so many opportunities in the community sector. In addition to popular communities such as Weibo and Zhihu, some users in niche circles also have a strong demand for communities. This demand has also spawned a number of communities. For example, there are Banciyuan and Diyidan, which are specifically for the two-dimensional people; there are CSDN and Blog Garden, which are specifically for programmers; and there is TapTap, which is specifically for mobile game enthusiasts, and so on.

If we put aside these complex product forms and operating methods, then what is the underlying logic of community operations ?

This article will attempt to abstract a theoretical model of community operation. Starting from this theoretical model, the entire community operation system can be connected. The author believes that although the community’s product form and operating methods are constantly changing, the underlying logic will not change easily. In actual work, only by grasping this unchanging essence can we avoid getting lost in the various so-called "means".

Table of contents

What is a community?

1. Community and platform

2. Community and Society

Construction of community operation system

1. Community positioning and cold start

2. Community atmosphere

3. User incentives

4. Content Distribution

5. Community Governance

Conclusion

First of all, we need to clarify what the concept of community discussed in this article is.

What exactly is a community?

The community belongs to the category of platform. The community is not an organizational form of user relationships, but only a platform that carries such user relationships. In this sense, the community is different from general platforms and communities.

Community and Platform

The operation of platform products has a very interesting feature, that is, platform operation is a kind of "bilateral operation". In other words, platform operators have to face two completely different roles within the platform at the same time: "producers" and "consumers". This is completely different from "unilaterally operated" products such as education, games, and e-commerce.

Taking the education industry as an example, the courses sold by education companies are generally produced by lecturers within the company. Therefore, for education companies, the role of "producer" is undertaken by the official, and the operation of the education industry only needs to find enough students for these courses. Therefore, compared with platform operations, the situations faced by education industry operations are relatively simple, and the operations in the education industry are a kind of "unilateral operation".

But for a platform, the situation is different.

Take Didi Taxi as an example. For a taxi-hailing platform like Didi, during the operation process, it has to face two different roles: passengers and drivers on the platform at the same time. For a platform operator, it is not possible to introduce too many drivers when there are insufficient passengers, otherwise the drivers will not be able to find enough passengers and will quit the platform. It is also not possible to introduce too many passengers when there are insufficient drivers, otherwise the passengers will also quit the platform because they cannot get a taxi.

This is the mystery of platform operation.

There is no doubt that a community is also a type of platform, because community users can also be divided into two essentially different types, namely, "producers" of community content and "consumers" of community content.

The roles they play in the community are also very different. "Producers" are responsible for providing content to "consumers", while "consumers" are responsible for providing feedback to "producers".

The difference is that users come to the community to seek common ground, or to satisfy their own psychological needs for a sense of belonging. Therefore, not all platforms can be called communities. Only those platforms that are used to meet users' psychological needs for commonality (sense of belonging) can be called communities. For example, Taobao is an e-commerce platform, but we all know that Taobao is not a community because users go to Taobao to shop, not to find like-minded people.

Community and Society

A community is also different from a society. The essence of a community is a form of organization of social relations, and the community is the platform that carries this relationship.

The community as a form of relational organization has two prominent characteristics:

1. Centralization . In a community, there must be a KOL, or the group owner. This KOL or group owner is the center of the community and has the highest influence within the community.

2. Common hobbies or ideas . Community members share common hobbies and ideas, which is the inherent reason why a community can be gathered and established.

In fact, whether it is as big as a religion or as small as a group of square dancing aunties, they are essentially a kind of community. A community is essentially a relationship, which is completely different from a neighborhood. A community is a place that carries this relationship, and a community generally satisfies users' psychological needs for commonality through the community.

A high-quality community can often give rise to many communities, but once a community is formed, it may still exist even if the community collapses.

How to build a community operation system?

From the above discussion, we can understand that a community is a platform that can meet users' needs for commonality, or satisfy users' sense of belonging.

However, as a platform, the community itself does not have the function of meeting users' needs for common goals. The community achieves this function through social groups.

Zhihu is a community, and every question and answer in Zhihu is a kind of community;

Tieba is a community, and each bar, such as D bar and World of Warcraft bar, is a kind of social group;

Weibo is a community, and the big Vs on Weibo and their fans are a kind of social group.

For a community, the established community is its most core asset. Therefore, the foothold and ultimate goal of the community operation system should be to ensure the establishment, improvement and development of the community ecology within the community.

You can start from these five aspects to build an operating system for community products. These five links are community positioning and cold start, community atmosphere, user incentives, content distribution and community governance.

1. Community positioning and cold start

The problem to be solved in this link is to clarify what kind of community to establish, what kind of people this community is for, how to obtain the first batch of seed users in the community, and how to complete the cold start through seed users. Many of the issues involved in this link are strategic and fundamental. If there is a problem in this link, especially if there is a problem with the community positioning, then no matter how well the following four links are done, it will be of no help.

2. Community atmosphere

The problem faced in this link is what kind of community atmosphere should be created in the community, what kind of tone and temperament should be created, and what kind of community culture should be nurtured. The important goal of the community cold start mentioned above is to complete the initial establishment of the community atmosphere.

3. User incentives

After the community atmosphere is formed, the problem of user activity must be solved. In other words, what means can be used to keep users in the community and continue to contribute content to the community.

4. Content Distribution

How to distribute the content contributed by users so that it can be fully consumed and bring enough feedback to the content creators? Such feedback is undoubtedly an important factor in motivating community creators to continue contributing content.

5. Community Governance

Once the above four links are successfully resolved, the community may usher in a considerable wave of user growth. So, in the face of the ever-increasing number of users in the community, how can we regulate the users so that the good atmosphere in the community is not destroyed and the community is always on the track of healthy development? This is the problem to be solved in this link.

These five links together form a community operation model, and their interactive relationship is as follows:

Next, we will give a detailed introduction to these five links respectively.

1. Community positioning and cold start

Community positioning and cold start are the first steps in community building. As the name suggests, this step includes two levels: community positioning and community cold start.

1) Community Positioning

The positioning of a community involves the strategic direction of the community. To some extent, the positioning of the community directly determines whether the community product can be successful. If there is a problem with the positioning, no matter how hard the community tries, it will be useless.

The field that the community is targeting must have a sufficient number of people and demands, and it is best to be close to money, which is more conducive to subsequent commercialization. Many people dream of building a small and beautiful community specifically for a niche circle, but the reality is often that these "small and beautiful" communities are extremely difficult to survive, because a community can only succeed if it can form a scale effect in terms of the number of people, which is determined by the platform attributes of the community.

Of course, many communities will target niche circles in the early stages. For example, Zhihu’s early users were concentrated in the Internet field, and the users tended to be high-end; Bilibili was initially positioned in the field of two-dimensional animation.

However, when the community develops to a certain extent, it will inevitably break through the original circle and carry out the so-called "expansion of the circle". Taking Bilibili as an example, many old users of Bilibili have complained that Bilibili has changed and is no longer the small site it used to be. There is a question on Zhihu: Why do you hate Bilibili more and more?

There are 3,822 answers in total, most of which are written by old users of Bilibili, and some of them have received tens of thousands of likes.

However, this phenomenon is actually an inevitable result of the growth of Bilibili users. The good atmosphere of the community before was based on a small number of people and a single user group. When the number of people expanded, it would inevitably have an impact on the original community culture. If Bilibili chooses to stick to the small circle of the two-dimensional world, it will inevitably bring great restrictions to the company's valuation and commercialization.

2) Community Cold Start

The main purpose of a community cold start is to acquire a group of seed users and allow them to complete the initial construction of the community atmosphere.

Therefore, for community operations, the key is to clarify the following three questions:

1. What kind of atmosphere does the community want to create?

2. What kind of users are capable of building this kind of community atmosphere?

3. What kind of content these users need to contribute to create the atmosphere?

Only by clarifying these can we understand what kind of seed users the community needs and what kind of content these users need to contribute.

During the cold start period of the community, user registration is usually by invitation.

The benefits of the invitation system are obvious. On the one hand, it can make users cherish their accounts more and prevent spamming and advertising. More importantly, it can limit the users and content of the community to a controllable range, so as not to cause too much pressure on community governance in the early stages of the community. Community operators will then have ample time to improve content distribution and governance functions based on user feedback, and prepare for subsequent community registration.

Community positioning and cold start are crucial to the development of the community, because most of the issues involved in this link are strategic and fundamental. If this link is a great success, the community is basically half successful.

Zhihu, which is now familiar to us, is undoubtedly the hottest community at present. The author is also an old user of the Zhihu community. If we look back at Zhihu's development history, we will find that Zhihu's invitation system lasted for three years, which means that the cold start of Zhihu's community lasted for three years. The large amount of high-quality content accumulated by Zhihu in these three years laid the foundation for Zhihu's unique community atmosphere and became Zhihu's most important moat.

Zhihu's success is due to the favorable factors of time, place and people, which is undoubtedly difficult to replicate. Even if today’s community entrepreneurs have the patience and courage that the Zhihu team had at that time, it is still a big question whether the capital behind them will give them such sufficient time again.

2. Community Atmosphere

Weibo and Zhihu are both communities, but we can clearly feel that these two communities are different. This different feeling is more caused by the difference in community atmosphere.

It's just like when we interact with people in life, we feel that people are different. Some people are introverts, some are extroverts; some are emotional, some are rational; everyone holds different values, but in the end, only those who are compatible with us in all aspects will become our friends.

The same is true for communities. The community atmosphere is the temperament and tone of a community. Only if this temperament and tone match ours can we become loyal users of this community.

A good community atmosphere is self-growing.

Because under the constraints of the community atmosphere, users generally only contribute content that fits this atmosphere, which undoubtedly promotes the growth of this atmosphere in reverse. On the other hand, a good community atmosphere can also screen out users. It can screen out users who truly recognize this atmosphere and eliminate those who do not fit in with this atmosphere, thereby making the user group in the community purer. This also reversely promotes the further self-growth of the community atmosphere at the user level.

When the community gradually forms some exclusive discourses for community users, community culture is formed. Community culture is the ultimate form of self-growth of community atmosphere.

What is the exclusive discourse of community users?

For example, when we hear words like "I'm in the United States, just got off the plane" and "Thank you for the invitation", we may smile knowingly, because we all understand that these memes come from the Zhihu community; when we mention "It smells so good", "High energy warning" and "Oli Gei", we also know that these are memes from Bilibili.

These memes are the exclusive language of Bilibili and Zhihu users. This kind of exclusive discourse from different communities actually has a great impact on our daily lives, and these influences break through the original circles of the community. We may not be loyal users of Bilibili, but we may frequently use the "It smells so good" meme in our daily chats.

Community culture is the ultimate form of self-growth of community atmosphere. The community atmosphere is introverted and generally only affects users within the community. Some high-quality communities will also have a certain impact on the circle in which the community is positioned.

Community culture is purely extroverted. It spreads outward through the exclusive discourse of community users. This kind of dissemination can break down the barriers between circles and exert influence on users outside the circle.

When the community atmosphere evolves into community culture, the community can be considered truly mature, and this stage is often also the stage when the community seeks commercial monetization. Community culture is undoubtedly the most important moat for a community.

Since community atmosphere is so important, how do we create it?

You can start from two aspects.

Zhihu’s famous slogan: Be serious and you will win; Bilibili’s mascot: a cute little TV; and Weibo has an impressive logo with big eyes.

These slogans, logos, mascots and other cultural symbols are all conveying the concept of the community to users. Even the community’s overall copywriting style, fonts, main colors, and functional interactions should be consistent with the community atmosphere, because these elements themselves are the carriers of the community atmosphere.

What is more important than this is the community content. Community operations must ensure that content that fits the community atmosphere can be effectively output and distributed.

For a startup community, it is necessary to complete the initial construction of the community atmosphere through the content contributed by seed users during the cold start phase. For communities at this stage, the most important thing is to figure out what kind of users contribute what kind of content to create the community atmosphere you want, so that you can build it. Only by making this clear can this atmosphere be implemented. Not just a concept in the head.

As for a relatively mature community, they generally have relatively complete mechanisms for user incentives, content distribution and community governance to ensure the healthy development of the community atmosphere. These will be discussed in detail below.

3. User Incentives

The problem that user incentives need to solve is how to keep users in the community and motivate them to continuously contribute content to the community. This is also an important aspect to consider in community operations.

For a startup community, especially one that is still in the cold start-up phase, user incentives are not a critical issue, because community operators at this stage generally have QQ groups of seed users, through which one-on-one communication between officials and users can be achieved. In fact, much of the content in the community at this stage may even be contributed by official personnel themselves.

However, once the community has passed the initial stage, it is time to start designing user incentive mechanisms at the product level.

In order to effectively motivate users, many communities will design a complete user incentive system at the product level, such as a points system, a medal system, etc. However, the prerequisite for this user incentive system to be effective is that users must have a strong sense of identity and centripetal force towards the community. Therefore, user incentives are closely related to the community atmosphere. If the community does not have a good atmosphere and users do not have a sense of identity with the community, then some medals, privileges and points in the community will be worthless to users.

The most effective and sustainable means of user motivation is to allow users to receive feedback. Feedback generally takes the form of following, commenting, liking, collecting, and forwarding.

There is a saying in "Social Dividend": Interaction is the simplest and most direct motivation. Users will decide to stay or continue to contribute based on the interaction the product brings to them. This is the same principle that makes games addictive.

When a user creates a work, if he can immediately receive feedback from other users, then this process is equivalent to a player killing monsters and getting equipment in a game. Although the process of killing monsters may be dull and boring, the feedback of the equipment can make the player endure it all.

What's more, whether or not equipment will drop and what kind of equipment will drop are all random. This kind of random feedback will undoubtedly be more stimulating to users. The same principle applies to motivating users to contribute content.

In fact, many things in life can be understood from the perspective of feedback.

We all know that we should develop good living habits, which is very important for happiness in life, such as going to bed early and getting up early, avoiding staying up late, exercising regularly, and controlling diet. But it’s hard to stick with it because the feedback cycle is so long. We may have to persist for a year before we see the changes brought about by healthy living habits, but the feedback from bad living habits is often immediate. As ordinary people, it is actually very difficult for us to overcome this addiction to immediate feedback that is rooted in human nature.

Going back to the discussion above, the same logic applies to motivating users to contribute content. Timely feedback is so important for community operations that there are even products such as fan robots, comment robots, and like robots on the market.

I don’t know if you have ever encountered this situation. When we just register a community, we will immediately receive attention from some users, and when we publish some content, we will immediately get likes and collections from some users. If you encounter this situation, don't be too happy too soon, as it is very likely that the community has deployed robots.

For a startup community, it is necessary to adopt a community robot, but this is not a healthy and sustainable means for users to get feedback. So what means should we use to get feedback from users?

The answer is content distribution, which we will talk about next.

4. Content Distribution

The goal of content distribution is to find the right readers for the content and to find content that readers like. Simply put, the content distribution mechanism plays the role of a content intermediary. This part solves the problem of connecting people and content.

The importance of content distribution to the community is self-evident. We can take a closer look at the community operation model diagram below.

Content distribution is the only key link that is closely related to the other three links.

It can even be said that as long as there are no problems with community positioning, cold start, and content distribution, this community is basically sufficient to achieve a virtuous cycle of community development.

There are two dimensions to evaluate the quality of a content distribution mechanism, namely the efficiency and accuracy of content distribution.

Efficiency of content distribution

The efficiency of content distribution refers to how many pieces of content a content distribution mechanism can distribute per unit time.

When a user creates a piece of content, what he needs is timely feedback. The key to timely feedback lies in the word "timely".

If a user works hard to create a piece of content, and it is distributed through a content distribution mechanism, but only receives feedback from other users a week later, then this content distribution mechanism can hardly be called an efficient content distribution mechanism. Because the feedback cycle is too long, users may have lost their original desire to receive feedback after a week.

Accuracy of content distribution

The accuracy of content distribution refers to the ratio of the number of feedbacks received to the number of users after the content distribution mechanism distributes a piece of content to a certain number of users at the same time.

Making content interactive is the ultimate goal of content distribution. The generation of interaction can at least illustrate two points: first, regardless of whether the interaction is positive or negative, the content at least touches the readers; second, the author receives feedback from readers.

Therefore, once the interaction occurs, we can assume that this content distribution mechanism has found the right readers for the content, or that it has found the right content for the readers.

Based on the two dimensions of content distribution efficiency and accuracy, we can simply classify common content distribution mechanisms. Depending on the distribution platform, content distribution can be divided into off-site distribution and on-site distribution.

1) Off-site distribution

Off-site distribution is relatively simple. As the name suggests, it means distributing content in the community outside the site. Common off-site distribution channels include the community’s official WeChat, official Weibo, official social groups, etc.

Many communities will distribute their content outside the site, and Zhihu is undoubtedly the best among them.

The content promoted on Zhihu’s official WeChat and official Weibo are basically the best content within the community. Zhihu has also launched an APP called "Zhihu Daily", which selects 5 pieces of content from the community every day and puts them on "Zhihu Daily". Zhihu even printed the high-quality content within the community into books and sold them on e-commerce platforms.

Zhihu "The Art of Money"

Off-site distribution has never been the main way for a community to distribute content, because the efficiency and accuracy of off-site distribution are extremely low. However, off-site distribution can bring users strong incentives and a sense of honor, which is unmatched by on-site distribution. Therefore, off-site distribution often serves as an additional supplementary means of on-site distribution, and it is more suitable to be made into an intermittent, fixed operational activity.

2) Distribution within the site

In-site distribution is the highlight of content distribution, and the types of in-site distribution are also varied and diverse. Generally speaking, there are only two most common ways of on-site distribution: social distribution and algorithmic distribution.

Generally speaking, most communities will adopt both distribution methods at the same time to ensure that community content can be fully consumed within the site.

Social distribution and algorithmic distribution of Weibo

1. Social Distribution

The most typical example of social distribution is Weibo. When we follow a Weibo celebrity, the updates of this celebrity on Weibo will be displayed in our information flow. The same is true for Zhihu. When we follow a Zhihu user, the updates of this user on Zhihu will also be synchronized to our information flow.

Social distribution relies more on active screening by users to achieve the circulation of community content. Judging from the efficiency and accuracy of content distribution, social distribution can basically reach a passing level, but it is far from excellent.

However, social distribution is a standard feature of a community, because it is social distribution that defines the community attributes of a platform, gives the community the attribute of "social", and is the intrinsic reason why a community can be called a "community". If a community does not have social distribution, then at best it can only be considered a content aggregation platform.

2. Algorithm Distribution

Algorithmic distribution is currently the most popular way of content distribution, especially personalized recommendation algorithms.

Nowadays, various communities, whether mainstream or not, are adopting algorithmic distribution, which is an inevitable result of technological development. Because measured by the efficiency and accuracy of content distribution, algorithmic distribution has reached unprecedented extreme levels in both dimensions.

Algorithmic distribution relies on machine algorithms to achieve distribution. This distribution method is completely automated, and the amount of content that can be distributed per unit time is almost infinite. And for the currently popular personalized recommendation algorithm, it relies on the user's behavioral data in the community to determine the recommended content. In the initial state, the accuracy of the content recommended by the personalized recommendation algorithm may not be high, but once the user portrait is perfected, the accuracy of the recommended content will reach a terrifying level.

Therefore, if there are conditions to deploy algorithm distribution in the community, especially personalized recommendation algorithms, try to adopt them. This is the survival of the fittest brought about by technological development.

When other communities are using algorithmic distribution, if you are still only using traditional social distribution, or more traditional chronological distribution, then at least in terms of content distribution methods, you have already lost to other communities. Unless the content distributed by your community is relatively special, it is not suitable for algorithmic distribution.

However, for a startup community, there are generally no conditions for algorithm distribution, especially personalized recommendation algorithms. In addition, the number of users and user data in a startup community are very limited, so it is difficult for personalized recommendation algorithms to play their due role.

So, how to solve this dilemma?

Next, I will share with you an algorithm distribution method that is very friendly to the startup community and more practical.

The general logic of this algorithm distribution method is to first establish a content library in the community background, and the content distributed by the algorithm is selected from this content library. The community's operations staff is responsible for daily updates to the content library: adding some high-quality content to the library every day and removing some content with poor data.

There are 4 rules for algorithm distribution, as follows:

1. Randomly select content from the content library for display;

2. If a user blocks a tag, when retrieving content from the content library, the content with the user-blocked tag must be blocked;

3. Content that users have browsed will not be displayed repeatedly;

4. If the user has browsed all the content that can be presented according to the above rules, then Rule 3 will become invalid and the content will be redistributed according to Rules 1 and 2.

*The content distribution method introduced above is just a basic framework. If a startup community is interested in adopting this distribution method after reading this article, it must be improved based on the specific situation of the community.

Strictly speaking, this distribution method is not algorithmic distribution, because it relies too much on the manual operation of the content library by community operators. The operator's taste for content is the key to whether this distribution method can achieve results. Fortunately, this content distribution method is simple and easy to operate. For start-up communities that do not yet have all the conditions, it can also be a reference for content distribution.

V. Community Governance

Community governance is a key link that must be paid attention to once a community reaches a certain size. In the internal organizational structure of many large communities, community governance is the responsibility of an entire department.

Community governance is closely related to community atmosphere. When a community develops to a certain scale, there will inevitably be some users with "bad intentions" sneaking into the community. Their purpose of coming to the community is either to cheat or to advertise. They are destroyers of the community atmosphere. If they are left alone, the community atmosphere will inevitably collapse gradually, which will have a fatal impact on the community.

The premise of community governance is to have a clear and thorough understanding of the community atmosphere. On this basis, most communities will compile a community management code of conduct and place it in an inconspicuous location within the community. Although the location is not very noticeable, it is indeed the basis for community governance. It is this regulation that enables the community atmosphere to be implemented in community governance and can be regarded as the basic law of the community.

From the perspective of community governance types, community governance can be divided into: official governance and community co-governance.

Some start-up communities mostly adopt official governance methods in community governance. Because the number of users in the community at this stage is very limited, it is sufficient to rely on manual review and governance by community operators. However, with the development of the community and the expansion of the user scale, it is impossible to rely solely on official governance. At this time, a common way of community governance is community co-governance, that is, allowing users to participate in the governance of the community. Many communities have adopted a community governance approach of co-governance, for example, Zhihu’s “Public Arbitration Hall” and Zuiyou’s “Audit Area”.

Due to space limitations, I will not give a detailed introduction to Zhihu's "Public Arbitration Hall" and Zuiyou's "Review Area". If you are interested, you can go to these two apps to learn more about them.

Zhihu's Public Arbitration Hall

The rightmost review area

The benefits of community co-governance are obvious. In addition to mobilizing the power of the masses and improving the efficiency of community governance, if the mechanism of community co-governance is designed to be very interesting, it will also increase user participation, enhance community cohesion, and even create a new way for users to interact with content in the community.

In terms of community governance methods, most communities adopt algorithmic governance and design an incentive system at the product level that can have a restraining effect on users.

Algorithmic governance is one of the community governance methods adopted by most communities. When the content of the community reaches a certain scale, it is no longer possible to rely on operators to manually identify low-quality content. At this time, corresponding algorithms can be configured for the community to replace manual content identification. This is almost standard for all large communities.

In addition, some communities will reduce the probability of users violating regulations by designing an incentive system to constrain user behavior. The most classic one is Zhihu’s salt value system.

Zhihu Salt Value System

Zhihu's salt value system uses five dimensions to make positive and negative judgments and weighted calculations on each user's behavior, ultimately giving each user an exclusive "salt value". Zhihu's salt value faithfully records the growth of each user in the community.

Conclusion

Community operation is a very interesting topic. As mentioned at the beginning of the article, community operation is a "bilateral operation". A notable feature of this type of "two-sided operation" product is that it will be difficult to operate in the early stages of the product, because it has to solve the problem of which came first, the chicken or the egg, and it also requires a good operating rhythm. It cannot introduce too many content creators when there are insufficient content consumers, nor can it introduce too many content consumers when there are insufficient content creators.

In short, the early stages of a community are very difficult, and many communities die during this period.

However, once the difficult start-up period is over, the community is able to achieve self-growth. This self-growth often brings some surprises (or shocks) to the operators, because many things that happen in the community are not under the control of the operators, but are brought about by the natural growth of the community.

Community operation is more like building a city. The operators are only responsible for building the city’s infrastructure and maintaining urban security. As for other aspects, the operator is more like a bystander, observing everything happening in the city. This process of observation will allow the operator to have some new understanding of relationships, human nature, and society. This is a completely different experience from the "unilateral operation" products.

If you are also interested in community operations and have certain insights, you are welcome to leave your ideas in the comment area.

Author: Su Tanbin

Source: Su Tanbin (sqypggs)