Zhihu’s operational strategy

I recently updated the Zhihu app to 6.46.0., and found that this revision involves a lot of content (it may also be that I skipped N versions in the middle, resulting in some information gaps). I played with it for a bit and found it quite interesting, so I wrote this article to share my thoughts as a user on the adjustments in Zhihu's operational strategies in recent years.

01 Contents of this version iteration

1. Strengthen and highlight the subscription and entry of topic categories

This updated content introduction picture is placed directly at the first page of the app in the app store, which shows its importance.

At the same time, the number of discussions and readings for some hot topic categories and answers are highlighted respectively. Especially the number of readings. Previously, no one except the creator himself could see the relevant data. This is equivalent to making a "hot list" in the dimension of topic category. The previous Zhihu “Hot List” was only a question dimension. I guess this is to gradually shift the previous single product (single topic) operation strategy toward a category operation strategy.

2. Created a new Zhihu circle

I don’t know if this feature was inspired by “WeChat Circles”. But they are a little different. When viewing the content information flow of a circle you are not following, WeChat emphasizes the circle name, while Zhihu emphasizes the user name. In addition, various expression interactions have been strengthened, which is very rare. From a UI perspective, comments and messages have been weakened into a secondary form of interaction.

I think both in terms of presentation form and content tone, it is more like a combination of Douban group and Renren group. Its content is obviously disconnected from Zhihu's original positioning, and can be regarded as a product that seizes fragmented time after its transformation into "Bihu" and "Story Club".

3. Strengthened the live broadcast entrance

On the homepage, the UI levels for live streaming and content writing are listed side by side. The content of the live broadcast is basically consistent with the basic tone of Zhihu.

4. Actively display the user’s “follow” and “favorite” actions

What’s more interesting is that after clicking the “Message” button at the bottom, it will jump to the “Dynamics” page by default, rather than the real “Message” page. (It has not yet been tested whether fans can see the content if the user saves it to a private folder.)

Compared with the fancy UI operations, the underlying logic of this function has more flaws. Although in the past fans could also view the collections and content of those they followed on their homepage. However, the user experience is completely different when fans "actively check" the relevant information of the people they are following and when the platform actively "broadcasts" the actions of the people they are following to all fans through a "loudspeaker". Especially for the content in the "Collection" and "Follow" scenes, this function is comparable to Zhihu paparazzi.

For some followers, there are definitely scenarios where this feature is suitable. But I think we should at least give users a chance to choose. I am extremely pessimistic about this feature. Improving the experience of some users should not be at the expense of the legitimate needs of other users. In this regard, Zhihu should learn from Douban. When performing related operations on Douban, users will be asked to choose "whether to share to Douban Broadcast".

I guess Zhihu wants to learn from Zhang Xiaolong's spirit of "combining manual and system recommendations" and imitate the "Reading" function of WeChat public account tweets. However, users instinctively know from the literal meaning that "reading" is mainly for friends, while "collecting" and "following" are more for themselves. The scenarios are different, so it’s really inappropriate to force functions on them.

5. In addition to the above major updates, there are some details adjustments

For example, articles can be suspended on the side of the screen (refer to WeChat articles and mini-program suspension); for Zhihu paid members, the display of "reading" rights is strengthened.

Here, I have summarized the above minor updates as follows:

1. Content style extension. The platform has gradually expanded from an initial "knowledge-based" sharing and discussion community to an "all-type" content community. The value to users gradually changes from save time to kill time.
2. Content distribution logic iteration. Help the answerer to lead the flow in various ways. Based on algorithm recommendation + hot list recommendation, the logic of user recommendation and topic classification is superimposed.
3. Try different content presentation formats. In the past few years, based on the launch of live (audio live broadcast + recorded broadcast), Yanxuan column (paid answers), and short videos, we have further tried the reading and video live broadcast tracks.

02 What is the purpose of these revisions?

1. Anxiety about the efficiency of traffic distribution within the site

As early as December 2019, Zhihu officially launched such an activity: users can grab red envelopes from big Vs, and after grabbing them, they will automatically follow the big Vs. Ordinary users are also allowed to send red envelopes to others and gain fans at the same time.

Anyone who has worked in operations knows that this kind of fan buying behavior will obviously result in terrible fan retention. This time, such an activity was actually initiated by the official, which has actually exposed the difficulties in traffic distribution within Zhihu and the impatience of seeking medical treatment in desperation.

Question and answer content is a basic form of content community products. However, it is relatively rare to find a community like Zhihu that builds it as its core value and product presentation style. During its seed user acquisition phase, the scarcity strategy of limited invitation registration did attract a lot of high-quality knowledge-based answers.

However, the ceiling for the number of topics that take into account universality, professionalism, and timeliness is very low. The angles from which it can be disassembled are also limited. We are not debaters on "The Rap of China" who can analyze a topic from all kinds of tricky angles. For most questions, the person who answers first often explains almost all the angles for breaking down the individual questions. There is less and less room for latecomers to explore and dismantle. Professional writers may also have the opportunity to work on the planning and structure of the article, the choice of words and sentences, and the form of expression, and produce high-quality popular content through refined operations. But for most users, it is difficult to find room to break in.

From a supply perspective, question-and-answer content communities are divided into two types of content products: questions and answers. In addition to the dilemma on the answer side, the question side also has its own problems.

Zhihu platform regulations prohibit "one answer to multiple questions", which means that the same set of answer content cannot be applied to multiple questions. According to my own experience, this set of rules and algorithms is quite strict, and it is not something that can be avoided by simply changing a few sentences in the content. It requires users to make more substantial adjustments. Unless it is a professional team or individual, the marginal content production cost of "one answer with multiple questions" will not be much lower than purely creating an original content. This causes users to be confused about which similar questions to answer when their time and ability are limited.

(1) Choose questions that have high popularity but already have highly upvoted answers - unless the answer is from a big V, forget it.

(2) Choosing questions with low popularity is a waste of time unless the topic becomes popular for some reason later.

(3) You can only reply to questions that have received some attention and whose top-rated answers have not yet been completely outdated by the Matthew Effect. But grasping this scale has become a metaphysics.

It's difficult! If we continue to develop along the path of the "professional question" community from the question and answer sides, it will be a dead end.

So now there are more and more contents such as "witty jokes", "storytelling", "showing off", and "fishing". Extending to various subcultures with low cognitive thresholds is a natural and inevitable choice. But this will further increase the traffic distribution and burden within the site.

No one is a living Lei Feng, and no one will continue to contribute high-quality content without compensation. It’s just that everyone has a different definition of reward, such as money, connections, honor, confidant... and so on. If the traffic distribution mechanism does not iterate along with the growth rate of content, and content contributors cannot obtain corresponding compensation through exposure, loss is inevitable sooner or later.

Zhihu has also made various attempts before:

• Zhihu LIVE: audio recording/live broadcast product. Unfortunately, the recorded broadcast has been blocked by leading platforms such as Duoduo and Himalaya. Audio live streaming has also been replaced by video live streaming.
• Zhihu’s idea: Similar to the Zhihu version of Moments, it may be helpful for distributing traffic to existing fans. However, it is between WeChat Moments and Weibo and is a semi-closed ecosystem. For non-fan users, the path to content acquisition is very long.
• Column: A WeChat public account similar to the Zhihu version. It stands out in the entire Q&A content community ecosystem. And it is different from Zhihu's idea, it is long content. The reading threshold is further raised. Therefore, unless it is some well-known big V or corporate account, the exposure efficiency is not as good as the conventional question-and-answer format. Even for these two types of accounts, you must decide whether to use "column" or "answer" to output the specific content. It is a great test of the degree of experience accumulation.

Overall, most attempts were not successful, so we tried to improve traffic distribution efficiency through content classification aggregation and user manual recommendations.

2. The battle for “user time” in the era of existing users

On a global scale, the trend of content becoming "nipple-pleasing" is inevitable, and there is nothing we can do about it. "Learning" itself is anti-human, not to mention that as professionalism deepens, the number of users who can feed a single piece of content will definitely decrease. Therefore, regardless of what Zhihu's original intention was, joining the entertainment army (at least the entertainment and fragmented form of knowledge sharing) is an inevitable choice of history. Otherwise, it will become increasingly difficult to compete with Douyin, Kuaishou, Taobao, Douban, etc.

After a wave of aggressive promotion, the short video project that we tried last year at least has a foothold. In terms of user experience, it has also changed from Douyin's public domain feed to the browsing method of "question → answerer's corresponding video → answerer's other videos". Like Kuaishou, it emphasizes the exposure logic of the private domain dimension of a single answerer.

The extension from short videos to live streaming is in line with the current hot topics in the past two years. However, it seems that the traffic distribution based on live broadcast products, including keyword search, is not very powerful in the current version. It is expected that after the live broadcast MVP product is successfully verified, the related traffic distribution functions will follow.

As for e-book products, they have been proven to have great potential by the god-level product WeChat Reading. In theory, they can effectively offset the negative impacts such as homogeneity, joke-like nature, and emotionality brought about by the sinking of mainline product content.

3. The squeeze from other track players on their home turf

No matter what the initial positioning of each current Internet product is, as long as it develops to a certain level, it will instinctively grab market share from products in other tracks. Didi Takeout, Meituan Taxi, Xiaomi Youpin, JD.com... are all products developed under this logic.

In the field of knowledge content, Zhihu has developed a PGC-style question-and-answer product called "Dedao Jinnang", and Bilibili has upgraded "Classroom" to a first-level category. Regardless of whether it succeeds or not, it will have some impact on Zhihu's positioning in the minds of users. Therefore, Zhihu will also try to grab market share in the WeChat official accounts, WeChat Moments, Douban groups and other ecosystems.

03 Zhihu’s struggle is very similar to Douban’s seven or eight years ago

Seven or eight years ago, Douban launched a "family bucket" series on the app and also made many inexplicable products in its business line. There are too many articles online that explore Abei’s idealism and dilemma in trying to balance “our spiritual perspective” and traffic monetization in Douban’s operations.

From flourishing to stagnant, perhaps this is the inevitable fate for most content-based products.

In order to extend the life cycle of content products as much as possible, in addition to content quality, the following conditions are also required:

1. Content area positioning : Users have a sustained interest in topics and are willing to discuss and share different opinions due to information asymmetry.
2. User growth : There are viable channels for community class advancement. The class pyramid from the bottom up is: Lurkers → Ordinary Users → Active Users → Small Scale Users → Big Scale Users
3. Traffic monetization : For users in the middle and top of the pyramid, there are effective monetization methods that match the investment.
4. New user guidance : In order to counter the Matthew effect of traffic among users in the middle and top of the pyramid, there are practical means of supporting primary users.

The playing style is not planned, but tried out. As long as there are no operations that violate basic logic and cognition, I believe Zhihu can still be revived even if the road is long and difficult.

Author: The Helmsman of the Carp

Source: The Helmsman of the Carp

Related reading:

Creative analysis of advertising on Zhihu platform!

Practical operation! 4 tips for attracting traffic from Zhihu!

Zhihu account anti-ban rules and product promotion guidelines!

Zhihu’s traffic pool!