1 framework and 3 models for UGC community operations!

Community is a good product, but it is also a complex product. The author of this article starts from the definition of UGC community circles, combines his own thinking, and uses three models to sort out and analyze the community circles. Mainly: interactive equilibrium model, pyramid model and viral transmission model, which I would like to share with you.

For apps with community attributes, the most commonly used community component is the circle. Another community component that is easily confused with the circle is the topic. In this article, we will discuss the circle component.

From the beginning of human existence, humans have lived in groups and needed to socialize. By forming circles through socializing, everyone can jointly resist risks and survive more easily. Socializing can give us a sense of security and satisfaction, so we generally have a natural need for socializing and for circles.

Circles and communities have similar meanings, but the concept of community has now been narrowed down by the micro-business economy, which is not conducive to my discussion.

The bond of a circle can be built through some common point. The more primitive way is blood relationship. Later, as the scope of activities expanded, there were relationships based on hometown and alumni. When people no longer had to worry about food and clothing, there were relationships based on interests and ideologies.

The circles of Internet communities are nothing more than the inheritance of traditional social circles, moving the social scenes from offline to online, but with some new things, making social actions simpler and lower-cost, making it easier for socializers to dress themselves up, and extending the social scope to a wider range.

In the BBS era, each page of a BBS site often formed a circle in itself. I spent some time on the Bicycle section of the school's BBS. Some people would post about their experiences of riding thousands of miles alone. Active people on the board would often organize activities such as riding a hundred kilometers to a certain place to drink a bowl of chicken soup. When it came to getting their bikes repaired, they would go to Master Yan at the school's repair shop.

On Renren, each school was a circle. People would apply to add alumni from the same school who had many mutual friends. By posting opinions, quotes, beautiful pictures, or wonderful articles, they would get their posts forwarded and viewed, thus increasing the number of friends. At that time, having 1,000 friends would make you an opinion leader in the school, and having 5,000 friends would make you famous and well-known across the country.

Information will spread quickly within the circle through forwarding by friends. I remember that once there was a conflict between students and the school, the principal urgently announced on the BBS that he would hold a meeting to communicate with the students. The news immediately spread on Renren. Although the notice of the meeting was a temporary decision and it was only an hour from the announcement of the news to the start of the meeting, 600 people still flocked to the scene to watch.

1. Three models for understanding UGC community circles

Let's give a loose definition of a circle:

(1) A group of people gathered around a commonality;

(2) Treat each person as a node, and the nodes in the circle are interconnected;

(3) There are long-term and frequent interactions within the circle.

Based on this definition, we can cite three models to study circles:

Model 1: Interactive equilibrium model (3)

From the consumer's perspective, there must be enough information for consumers to consume. From the producer's perspective, there must be enough consumers to consume the information produced by the producer. Specifically, the updates posted by users bring enough feedback, such as reading, liking, commenting, sharing, and private messaging. There is enough new content to read in the user's timeline or information flow every day, and enough content to elicit user feedback, such as reading the full text, liking, commenting, sharing, and private messaging.

Model 2: Pyramid Model (2)

In a circle, the influence of members is often stratified. As the influence of members decreases, the number of members that a member can influence also decreases. For example, in a certain community APP, the members of a circle can be simply divided into three levels: A, B, and C. The influence of level A can cover 10,000 people, the influence of level B members can cover 100 people, and the influence of level C members can cover 5 people. At the same time, the retention of each level B member requires the influence of 10 level A members, and the retention of each level C member requires the influence of 5 level B members. So, if we develop 1 million C-level members, there may be 10,000 B-level members and 100 A-level members. By maintaining these 100 A-level members and 10,000 B-level members, we can support the stability of 1 million daily active users.

Model 3: Viral transmission model (1)

The natural expansion of the circle depends on word-of-mouth transmission. On average, an old member can attract the attention of m people outside the circle. On average, n people outside the circle who notice the circle decide to join it. So, on average, an old member can attract k=m×1/n people to join the circle. k>1 means that the circle can continue to expand naturally without external force.

For example, if a circle can attract the consensus and joining of 1% of people, then on average each member needs to attract the attention of 100 people to attract the consensus and joining of a new person. This is how members continue to join the circle naturally.

2. Application of the interactive equilibrium model

We can test the effectiveness of the interactive equilibrium model from different data perspectives.

Generally speaking, for a circle on a certain platform, the circle will be full of vitality, with a large amount of content emerging and the daily login rate of users remaining at a high level only when the number of members in the circle reaches a certain critical value.

The Mustard Campus APP was once called the APP version of Renren. Users also formed circles based on their respective school identities and can post updates and socialize. After observation, it was found that on the eve of the shutdown of Mustard Campus, when the number of registered users of a school on Mustard Campus reached 5,000, the daily active users of the campus circle were relatively high; and when the number of registered users of a school on Mustard Campus was less than 1,000, the daily active users of the campus circle were almost 0.

The principle behind this is that when the number of users in the campus circle is less than a few thousand, new users cannot add enough friends, resulting in insufficient content on their timeline, and therefore no desire to continue logging in. However, old users in the circle have no motivation to post content again due to the lack of interaction in the content they post, thus the circle falls into a vicious cycle and the daily activity level soon drops to zero.

Therefore, there is a trick on Weibo and Twitter. In the early stage of new user registration, they recommend dozens of popular bloggers for users to follow. In this way, the content on the new user's timeline is more sufficient, and users are more likely to be retained.

Therefore, the sharp drop in daily active users during the years of Renren's decline occurred during the graduation season. At the end of June, on the one hand, graduates' willingness to publish content decreased significantly, and on the other hand, the content on the timeline decreased sharply, so users were lost collectively.

WeChat’s public data shows that WeChat users spend 30 minutes a day browsing their Moments. Provide enough content so that users can browse for 30 minutes every day, which can be used as experience points.

When ByteDance's Feichat was promoting its product, the community did not prepare enough content, and as a result, it used up all its traffic at once. Although more than 1 million users flocked to the community in a short period of time, everyone found that the content in each circle was seriously insufficient, so they never came back after just a few visits.

The promotion of WeChat video accounts starts with internal testing, allowing some users to obtain the authority to publish video accounts, and then opening up traffic to a limited extent so that video publishers generally have the motivation to continue publishing. When the content is sufficiently accumulated, more traffic will be gradually introduced, and more creators will be allowed to join. This ensures that at each stage, consumers have enough content to watch and producers have enough consumer feedback to motivate them to continue producing.

So we often say that products that make a lot of noise at the beginning fade away quickly. Duoshan and Feiliao made the same mistake.

Do you still remember this hotly debated issue? "Has China's social sector reached its end?" Well, the person in charge of Duoshan and Feichat seems to be the same person.

Regarding the application of the interactive equilibrium model, we will extend it a little further to inspire everyone's thinking.

Why did the updates on QQ Space change from having no reading count mark to having reading count mark, and how did the reading count statistics change from counting only those who read the original text to counting only those who pause briefly on the update?

Why is the number of readings on Weibo counted as a brief pause on the post, rather than a certain number of seconds of reading?

Why the CEO of Jike was so diligent in giving likes to users that he got tendonitis.

3. Application of the Pyramid Model

All circles have a pyramid structure. Everyone has the desire to climb up in certain circles to gain recognition. Some circles have clear hierarchies, while the hierarchies in some circles are deeply hidden.

Some business associations rank you based on how much capital you have or how much power you hold. Some circles rank people based on the number of their fans. In the minds of ordinary people, people are ranked according to some labels. 985 has become a label, ByteDance has become a label, and product manager has become a label. Therefore, if you want to get a higher level for yourself, you have to work hard to get more and more effective labels, and there are even some behaviors of creating labels.

For the platform, it is necessary to consider what kind of hierarchical system to establish so that the process of users climbing the hierarchical level is more beneficial to the platform, and how to maintain the hierarchical system so that this hierarchical system can gain deeper recognition.

In Honor of Kings and Peace Elite, the level is not a one-time thing. The level will be regressed to a certain level every season. To return to the previous level, you need to continue playing games and win more games. This forces users to stay active. In order to make this level system more recognized, levels are not obtained by how long you play, but by indicators such as your win rate and the number of enemies you kill. These indicators can prove your gaming ability.

Although the level system of Peace Elite imitated that of Honor of Kings, it was not recognized at the beginning. In Honor of Kings, only players of similar levels can be matched in ranked games, which forces players to increase their levels if they want to play with people of higher levels, and players have to get close to their friends' levels if they want to play with friends.

In order to make it easier for players to match, Peace Elite does not require players of similar levels to play ranked matches. This has led to a decrease in the value of the level system and insufficient motivation for players to strive for levels. Peace Elite then uses some strategies to strengthen the ranking system, such as displaying a list of popular players at the beginning of the game.

In a circle, users will choose to stay in the circle because of some highly influential people. Selecting, cultivating and maintaining highly influential people becomes the platform's job.

For example, what motivates me to log on to Weibo every day is to see what opinions several people with A influence level have expressed, and what moods several people with B influence level are in. If I think I won’t see these people’s updates on Weibo, then the frequency of my logging into Weibo will be significantly reduced.

In the process of user growth, the UGC platform must maintain the stability of the pyramid structure. If the number of ordinary users of the platform increases rapidly, but there is a serious shortage of users with influence A and users with influence B, then ordinary users will most likely leave soon after they come. This appropriate ratio is different for different platforms.

Let users anticipate what level they can climb to and what kind of rewards they can get, which determines what they are willing to pay.

Therefore, the Douyin platform needs to create some amateur bloggers with 20 million fans, and some celebrity bloggers with 20 million fans. The higher the upper limit is, the more motivated those who are climbing will be. The more high-level bloggers earn, the more motivated climbers become.

As for Douban, it is rare to see bloggers with 100,000 followers. Ordinary users understand that it is unlikely for them to reach 100,000 followers on Douban. It is also rare to see Douban bloggers making any money. Ordinary users understand that it is unlikely to make a living from Douban. Those potential bloggers lost the motivation to work hard on Douban.

This aspect is actually the user growth system and user incentive system. I may discuss this in a separate article later.

IV. Application of Viral Propagation Model

This model tells us that increasing the probability of audience attention and the probability of audience action, as well as increasing the initial transmission traffic, can improve the effect of user fission.

Different banners, or different themes, have different probabilities of attracting the audience's attention. For example, portrait photography may attract the attention of 30% of people, but python may only attract the attention of 1% of the crowd. Then it will be much more difficult for us to promote the Python circle than the photography circle.

Since the probability of audience attention is not evenly distributed, finding local areas with a high probability of audience attention can improve the effect of user fission.

Reducing the cost of action and increasing the motivation for action are two ways to think about increasing audience action. For example, in an essay writing campaign, if the text requirement is reduced from more than 2,000 words to 140 words, the cost of action will be greatly reduced. By changing the reward of the activity from community peripheral dolls to matching with a destined person, the user's motivation to act may be enhanced.

Testing the conversion rate with a small amount of traffic first, and then deciding whether to invest in larger traffic to expand the impact of the event, can make better use of traffic. After all, even free traffic in your own community faces pressure in terms of the number of times it is used. Reducing disturbance to users can create greater potential for more valuable activities.

Although the thinking framework is simple, many community products still make many basic mistakes.

Some products, regardless of whether the circle can survive, just apply the circle list of competing products and build the same circle. As a result, the traffic of each circle is very limited, the number of posts in each circle is very small, and users do not know the value of this product. That’s what Feichat is all about.

Zhihu later also set up circles, based on universities, to guide users to join university circles. However, the traffic entrances of the circles were hidden deep and the traffic was low, with only a few thousand people in each university circle. This reduced the probability of users reading circle content and also reduced the probability of users posting content. As a result, Zhihu's university circles died down.

Zhihu’s interest groups are not doing very well either. Although some interest groups can gather a lot of people, the traffic entrances of the circles are still hidden deep, making it inconvenient for users to read and post content in the circles. For example, although there are 40,000 members in the Python circle, there are only 100 visitors and 10 posts every day, so there is insufficient motivation for users to come every day. The only end of this circle is death.

We immediately created some circles, but we also made the mistake of being too aggressive. When we had only 2 million daily active users, we pushed for the establishment of circles in each university. As a result, traffic was dispersed, and one university circle had only a thousand people. Even in the ideal situation of 1% posting volume, it was difficult to maintain the improvement of the circle with 10 posts a day.

postscript

This article only provides a thinking model to help community operators think about new operating strategies. Since the situations in different UGC communities vary greatly, each specific strategy is only effective for specific situations and specific stages, so this article does not intend to discuss specific strategies. The purpose of this article is just to stimulate discussion and we welcome more in-depth discussions.

Author: Jiang Liu

Source: Jiang Liu