In August 2010, Zhihu founder and CEO Zhou Yuan, led the original team of meta search and devoted himself to the entrepreneurship of Zhihu. In January 2011, Zhihu was officially launched, adopting an invitation-based registration system. In March 2013, Zhihu opened registration to the public. In less than a year, the number of Zhihu's registered users rapidly climbed from 400,000 to 4 million. As of July 2018, Zhihu has completed seven rounds of financing and entered the ranks of unicorns. As of January 2019, Zhihu's total number of users has exceeded 220 million. Speaking of Zhihu community, it is already a leader in this field. There are many reasons and methodologies for Zhihu to do a good job in community, but one of them that cannot be ignored is Zhihu's unique community cultural atmosphere. Just like the book "The Death and Life of Great American Cities" that Zhou Yuan strongly recommended, communities are like cities in different countries. For a city, the most important thing is not how it is decorated or modified, but how to create a cultural atmosphere. How to establish a protective atmosphere, determine a unique tone, and build an invisible moat for the content community is the key to whether a community can operate properly. Amidst the changes, What are Zhihu’s unchanging community governance principles? With over 220 million users, what are Zhihu’s unique methods for community operations? We interpret the principles, methods and practices of Zhihu community governance and explore the future development path of the knowledge question-and-answer community. It is divided into five parts: Over the years, Zhihu has been practicing the principles and methods of community governance. These principles and methods are not slogans or concepts, but valuable practices that result from constant trial and error, and summarizing experience. Is Zhihu a “slow company”? You must have heard the saying "Zhihu is a slow company". It depends on how you understand "slow". From 400,000 users in 2011 to 220 million users in 2019, the total number of Zhihu users has increased 550 times. Judging from the data alone, Zhihu is not slow at all. Guest Li Shanshan admitted that she understood why users had the perception of “slowness”. Zhihu encourages users to produce professional content, which often relies on the creator's accumulation in real life. Therefore, once Zhihu sets that professional content will be more encouraged, the operation team must be aware that the growth rate and creation frequency of professional creators will be relatively slow, which is an objective law. "Go to Zhihu and share your newly made story with the world" "Go to Zhihu and share the story you just made up with the world." Some netizens have used this sentence to make fun of Zhihu. Two years ago, Zhihu banned a user named "Pirate King Luffy". This user had answered a total of 244 questions from netizens on Zhihu, but the identity of each answer was different. This ban caused everyone to ridicule and pay attention to Zhihu's "making up stories". Zhihu is clearly opposed to "making up stories". The ban on "Pirate King Luffy" reflects the "serious" community governance principle. Guest Li Shanshan also shared the story behind the account closure. The operation team hopes that Zhihu is a real and professional Q&A community and that the content on Zhihu is trustworthy. When users open Zhihu, they will also assume that the experiences, lessons, and content therein all come from sincere sharing. Therefore, when a bad case like "Pirate King Luffy" appears, users will feel disappointed. How to deal with such a bad case? In the future, Zhihu will continue to spare no effort to adhere to the principle of "seriousness", increase supervision and review efforts, and fight against the behavior of "making up stories" to the end. She said that the Zhihu operation team will work with the majority of users to keep their eyes open and actively identify low-quality content such as "making up stories" to maintain a professional, serious and friendly knowledge sharing atmosphere for the majority of users. Avoid fan conflicts and strictly prohibit verbal abuse and attacks There have been conflicts among fans on Zhihu, with fans verbally abusing and attacking each other. Although this behavior destroyed the "friendly" cultural atmosphere of the community, the two groups of fans were powerful, bringing in a lot of traffic and even becoming a hot topic. Do you want traffic or a friendly atmosphere? During that war of words, there were many questions, answers and interactive behaviors. If we only look at the regular traffic, interaction and other data, this topic is a very positive thing. However, Zhihu still closed many related questions, banned many accounts, and deleted many answers. This is because: attacking each other is not something Zhihu encourages. According to the current community norms, if something is an unfriendly behavior, an act of mutual attack, or an act of provoking conflict, the operations team will quickly deal with it coldly, even if it means losing corresponding traffic. As the general manager of Zhihu community governance, guest Li Shanshan focused on introducing the three principles of Zhihu community governance: professionalism, seriousness, and friendliness. Zhihu’s community governance has shaped its unique cultural atmosphere and content tone. Guest Li Shanshan believes that when you are able to obtain certain information or make your information receive more attention, you gain some information power to a certain extent. The reason why every UGC community can stand out is that it must provide users with a new path to realize their information power. The reason why Zhihu has become Zhihu is that when it was first launched, it provided users with a brand new set of rules and a brand new path to realize information power. Zhihu adheres to professionalism, seriousness, and friendliness, encourages responsible speech, and respects diversity. At the same time, it also firmly opposes spamming, stupidity, rumors, and discrimination. On Zhihu, professional, serious and friendly knowledge sharing can bring you recognition and influence. The particularity of this path to realizing information power has shaped Zhihu's "product recognition". Community governance is not just empty talk, and these rules are not just sentimentality. Rules always require corresponding methods to implement and support them. In this sharing session, guest Li Shanshan introduced to us the three methods used by Zhihu in community governance: community norms, algorithm capabilities, and product mechanisms. 1. Community norms: Define user behavior and build the Zhihu Content Value Manual "Law-abiding" users have a weaker perception of community norms, but in fact, Zhihu has already integrated community norms into content recommendation, content circulation, content identification, and user identification. In 2018, Zhihu made a detailed redefinition of all encouraged, discouraged, and prohibited behaviors and content in the community, forming a "Zhihu Content Value Manual." For example: There will be some low-quality content in Zhihu’s health topics. The operation team needs to proactively identify and intervene to establish a processing system that can be executed and implemented on a large scale. To achieve this goal, the operations team will extract the features of low-quality content in health topics. One important feature of these low-quality contents is that they have no reliable sources. Therefore, in health topics, if an answer cannot be clearly found in core journals or other reliable research and sources, Zhihu's operation team will mark the answer accordingly. 2. Algorithm capabilities: 400+ cutting-edge deep learning models, efficiently identifying over 100 million pieces of content After formulating community norms and determining identification standards, the next step for the operations team is to batch define the over 100 million pieces of content in the community. The massive amount of content in Zhihu, which exceeds 100 million pieces, cannot be identified manually one by one, so the operation team must configure corresponding recognition capabilities for this. Zhihu has introduced more than 400 cutting-edge deep learning models and used AI to achieve human-machine collaboration. It has currently achieved a 95% coverage rate in the handling of negative content. In core identification categories such as violations of laws and regulations, negative emotions, and advertising diversion information, Zhihu's algorithm mechanism has performed quite well. In addition, Zhihu has also implemented algorithm recognition and coverage in many details such as originality recognition, risky image recognition, and personal consultation recognition. 3. Product mechanism: Based on community norms and algorithm capabilities, a salt value system is launched In its early stages, Zhihu did not set up a salt value system. This was because: at that time, there were very few Zhihu users, and the attention and behavior between users were enough to form a word-of-mouth understanding of the Zhihu community atmosphere. However, as Zhihu develops, the user base expands and the content becomes increasingly complex. When a new user enters the community, he must understand the community atmosphere and cultural tone of Zhihu. Therefore, Zhihu needs a "tour guide" to tell every new "friend" what kind of community Zhihu is. Based on this purpose, and also based on community norms and algorithm capabilities, Zhihu launched the salt value system and constructed 124 factors, which were ultimately implemented into five dimensions. Based on the consensus on community norms and public discussion order, Zhihu's operation team will make positive and negative judgments and weighted calculations on each user behavior, so that ultimately every "Zhihu friend" will have his or her own exclusive Zhihu salt value. The salt value will be continuously adjusted according to the behavioral changes of "friends", faithfully recording the growth of each "friend" in the community. The salt value system is divided into five dimensions: basic credit, content creation, friendly interaction, compliance with regulations, and community building. First, basic credit. If your persona is authentic and easily perceived by other users, your basic credit score should be relatively high. If you are often banned or punished for unfriendly behavior such as swearing, then your personal credit score should be relatively low. Second, content creation. On Zhihu, creation is the most direct business card of every "Zhihu friend". Your creation directly tells other users your expertise, your thinking, and your level. Third, friendly interaction. Fourth, comply with regulations. Fifth, community building. There are many users in Zhihu who may not create a lot of content, but have done other meaningful things for community building and have also made very critical contributions to the community. Such "Zhihu friends" deserve a higher salt value. Guest Li Shanshan said that the salt value system will be connected with Zhihu's recommendation system, weight system, and corresponding rights and interests system. For example, if the membership business department recently released a new product called "Yanxuan Membership", then all high-salt value users will receive one year of "Yanxuan Membership" for free. This is a basic application of the salt value system. Is the rationality of community governance led by the operations team questionable? Governing the community from a God's perspective, is this condescending attitude a bit too "arrogant"? With so many fields, so many topics, so many behaviors, and so many discussion scenarios, does the platform really have absolute professionalism? In actual operation, some "friends" will more or less question Zhihu's community governance: "Who are you to decide what is low-quality content? Do you have a product manager who graduated from medical school?", "I think the content of this answer is very stupid, so I expressed my opinion. Why do you think I am unfriendly?", "I think the story I wrote is very interesting and everyone likes to read it. Why do you think it destroys the order of the community?" Faced with everyone's confusion and doubts, guest Li Shanshan also expressed her own views. She pointed out that Zhihu's community governance is not a one-way centralized management, but a collaborative management that cooperates and co-creates with users. Users have an important voice in community governance and can deeply influence community governance through content production, content screening, and content consumption. 1. Relationship between the platform and users What is the relationship between the platform and users? To sum it up with a joke from guest Li Shanshan: I treat users like my first love, but users see me as a property manager , and often feel that this property manager is not doing a good job. However, Zhihu hopes to become a property owners' committee and work with users to formulate rules and fight for their rights. Zhihu is a UGC community. Users’ discussions and sharing create Zhihu’s core value, and the interactions between users define the core experience of other users when using Zhihu. The content is written by users and the approval is given by users. Zhihu only creates an opportunity for connection and an interaction scenario. The operation team's screening and judgment of content is ultimately based on every vote of approval or disapproval from all users. Therefore, users play a very important role in the community atmosphere of Zhihu. 2. Users circulate through content. Deeply participate in community governance In community governance, Zhihu has always been committed to amplifying the influence of professional, serious and friendly "friends". At present, users mainly participate deeply in community governance through content circulation links such as content creation, content screening, and content consumption . 1. Content Creation First, users deeply participate in community governance through content creation. Zhihu starts with content creation, but its core scenario is question and answer, which is a discussion context. Asking questions is the beginning of all discussions. Whether an issue is worth discussing and whether it can get a lot of attention is an important part of community governance. 2. Content screening Secondly, users can deeply participate in community governance through content screening. Zhihu's content screening is highly dependent on users and has two core concepts: sorting algorithm and professional weight. When a question is asked, there are often many corresponding answers, and the order is determined by the votes of each user for and against. In addition, there is an algorithm of professional weighting behind the votes for and against. Zhihu believes that the judgment of users who continuously create high-quality content in a certain field is more meaningful. So, when they vote yes or no within their field, it has a more significant impact on the ranking. 3. Content consumption Finally, users participate deeply in community governance through content consumption. From 400,000 users in 2013 to 220 million users in 2019, from millions of dollars in angel round financing from Innovation Works to $270 million in E round financing, Zhihu's community evolution journey has been long and arduous, yet steady and firm. 1. The past: New York, 1970-1990 In 2013, Zhihu had only 400,000 users, and the startup team decided to open registration to the public. In order to welcome new users and maintain a healthy discussion order, the operation team has done a lot of work, such as: closing questions and suggesting modifications, optimizing sorting algorithms, improving professional weights, etc. Afterwards, Zhihu also released a convention for healthy discussions, hoping to use these rules to clearly tell users how they should discuss on Zhihu. Some time ago, guest Li Shanshan saw an old document, which was a summary written by a colleague at that time. This colleague described Zhihu in 2013 like this: We are like New York from 1970 to 1990. Guest Li Shanshan said that people who work in the community particularly like to use the development and evolution of the city to benchmark their own work. New York from 1970 to 1990 was in a period of population explosion, accompanied by an increase in crime rates and social disorder. During that period, New York invested a lot of manpower and material resources in urban infrastructure. When the infrastructure was complete enough, New York experienced a renaissance. This situation is very similar to Zhihu during its rapid growth period. It is no wonder that colleagues compared Zhihu at that time to "New York" from 1970 to 1990. 2. Now: Across different economic environments “Big City Complex” In 2019, Zhihu has more than 220 million users, 130 million existing answers, and billions of discussions on more than 200,000 topics every day. The overall atmosphere of the community is very active. Therefore, in reality there is no city with such a scale that can compare with Zhihu today. There are so many topics on Zhihu, and so many users from all over the world, from different backgrounds and professions, actively discussing each topic. If we must use a city to represent such a complex scenario, it should be a "big city complex" spanning different economic environments. 3. Future: Be more professional, serious and friendly. More diversified and commercial In the last part of "In-factory Performance", guest Li Shanshan also shared with us Zhihu's future development vision. First of all, Zhihu will continue to adhere to the principles of professional, serious and friendly community governance. Secondly, as Zhihu's user base becomes more and more diverse, the gains that users hope to gain from sharing on Zhihu will also become more and more diverse. Based on these diverse changes, Zhihu will update the corresponding rules and regulations. Finally, Zhihu will actively promote commercialization. For example, many content creators have gained recognition and influence on Zhihu. Can the platform give these users some monetary rewards? How to create opportunities and possibilities for users to earn profits? This does not conflict with the pleasure brought by sharing itself, and it also ensures that the content shared by users remains professional and credible. Related reading: 1. Zhihu’s operational strategy: What are the chances of fighting against content “entropy increase”? 2. Zhihu promotion and operation: How to become a Zhihu celebrity with millions of likes? 3. How to use Zhihu promotion to efficiently attract targeted users? 4. Zhihu’s 4 quadrants and 2 rivals! 5. Zhihu: Douyin and Kuaishou cannot represent all short videos 6.How do keep, Zhihu, etc. conduct user operations? Author: Source: |
>>: Xiaohongshu Promotion Method | Is the Grass-Planting List Still Reliable?
In the process of website optimization, the const...
Baidu drop-down box: It is mainly used when searc...
The registration process for an Apple developer a...
1. How long does it take to make money? How much ...
Wenchang Tower is one of the common Feng Shui orn...
In large-scale marketing activities, community op...
As the capital of our country, Beijing is the cit...
Yesterday (May 6), the trend of apps being remove...
Starting a business requires costs, and mini prog...
Recently, the weather in many parts of my country...
Your colleagues will always be your best teachers...
Copywriting is an indispensable basic task for op...
When talking about information feed ads , we have...
In 2020, the popularity of Douyin live streaming ...
I believe that every event operator wants to plan...