This year is already halfway through. Compared with the 21 versions released in the whole of last year, this year is indeed a lot less. Will it be uncomfortable to slow down all of a sudden? Don't rush, take your time, only when you really slow down can you really get up fast. In the year-end summary of last year in 2021, what did our CSDN App do? There are many students who have put forward various suggestions. I hope you can talk about [what kind of CSDN App you want] this year.
Let me briefly talk about what we did in the first half of the year.
mobile editor
I won’t go into details about the functions, but the native editor is definitely much better than the bad H5 embedded editor before. If you have used it, you can definitely feel it. Of course, many students complain about incomplete functions, not as many functions as on PC, this is inevitable. We keep some frequently used functions in the App while keeping it as simple and concise as possible, so that the editing on the mobile terminal is not so complicated.
Tell me why you do it. I am very entangled with this function that consumes relatively more research and development resources. I know that writing articles on the mobile terminal will not be the choice of most people. I know that many people will not use it after it goes online. It's just the castration of the PC version. Both the performance of the data and the suggestions of my colleagues are telling me that I shouldn't do this thing with a low input-output ratio, but it is still a refactoring. I have to admit that sensibility has a slight upper hand when making decisions, just like the phrase "We have heard a lot of truths, but we still have a bad life", knowing that we can't do it. Although only a small percentage of users are using the mobile editor, I don't want you to make do with it. Since you are willing to write something on the CSDN platform, I hope that the tools you use can be easier to use and not so bad. This is actually the original intention of refactoring the mobile editor.
other
There are still many iterative optimizations of existing functions, such as: news, comments, live broadcast, news, search, negative feedback, skill tree, creation center, etc. I will not elaborate on them one by one. Students who have already used it will definitely feel the optimization After the experience, students who have not used it or don't really need it can have a taste. In addition to functional improvements, we have also made many improvements in product performance. Now you can clearly feel that there will be almost no crashes, the page (native) will open faster, the content browsing will be smoother, the installation package will be smaller and smaller, etc...
Above~ If what you want to see is what the app did in the first half of the year, then you can stop here.
If you are willing to listen to me talking about something else, please continue to scroll down...
I just said that our pace in the first half of the year was very slow. Indeed, every version we have made this year will be very cautious and will take more time to think. Think about what is CSDN's positioning for App, think about what we want to do, and think about what users think of App.
At present, there are several points that are relatively clear. In the eyes of the vast majority of students, App is a search tool, part of which is used as a management tool, and part of it is used as a content community. On the other hand, the current product structure design does not match the distribution of user usage, such as [Recommendation] on the homepage. I have been thinking about what content should be pushed to meet the needs of users. Is the content pushed according to the portrait really what the user wants to see? At least from the data point of view, the current recommendation does not meet the needs of users.
I have some ideas to discuss with you. According to the current distribution of users' usage of App, App may have a structure of [Search > Recommendation > Follow], which is more like a search tool product rather than a content consumption product that we have always defined. So for search products, the main content consumption scenario will happen on the search result page, and recommendation is more suitable as a content distribution channel for auxiliary search .
For example: after you search for "how to install docker" and read the search content, returning to the recommendation stream will push you "how to use Dockerfile to make a mirror?" "teach you to use docker to run Python programs" and so on Content, will this be more in line with your expectations?
In addition, some content that is more suitable for reading on the mobile terminal and can convey the brand value of CSDN will be introduced in the recommendation, such as professional articles/solutions in frontier fields, industry news alerts, etc. Such a combination of content may be an ideal [recommendation]. If more and more users read the content in the recommendation after adjusting according to this idea, and the number of articles read increases, then we should be doing the right thing. .
Let's look at this from another perspective. Is [Recommended] content distribution a good thing for creators? Can creators accept different strategy changes? From a utilitarian point of view, most creators do it for traffic and influence. But the absolute recommendation distribution will cause relatively large fluctuations in traffic. The creator may have many fans, but the articles written may have only a few readers. Isn't this embarrassing? Therefore, creators with many fans may not like the feeling that they have no control over traffic, and hope to increase the proportion of distribution by paying attention to it. Creators with few fans will hope to do a better job of recommending cold start and pay attention to the proportion of distribution reduce.
So will users like absolute referral distribution? The recommendation distribution will have a high degree of matching with the user's interest and browsing behavior history. If you have a well-recognized creator on CSDN and only want to see Ta's updates, it may be more suitable to follow the channel. As mentioned above , the app may have a structure of [search>recommend>follow], then the distribution of follow channels will not be a major scenario, the overall penetration rate is relatively low, and whether follow channels meet this requirement What about the needs of some users? This is also a question mark.
From what we just said, we can lead to another question [Follow], why do we need to pay attention? We can define two types:
Social relationship concerns: I know this person well
Subscribe Relationship Follow: I endorse this person's content
It can be seen that these are completely two sets of different logics. The bottom layer is focusing on people/focusing on content. So for a vertical content consumption product, which one has more weight in product design? In fact, this is essentially a question of "fan value". What do you want your fans to be like? What do you want your fans to expect from you? This is also a question we are exploring recently.
In a community, supply and demand, key people, and common values are the three elements we usually think of. How to make more people able to create and want to create? How to let more people have good content consumption? How to establish a head author with strong appeal and strong demonstration effect? How to create a demonstration effect of strong identity? How to create shared community values? These will be the direction we will continue to explore and move forward.
say something off topic
It may be that the recent temperature is too hot, and the layoffs of major factories have made everyone very anxious. After finding that there is no solution to anxiety, they lie down in anxiety. I have talked with many friends, and everyone expressed a question, "In this era when everything is flat, what is the point of our hard work?" Years of work experience have already smoothed the edges and corners when we first came out of school, and we have long forgotten that time. Set ideals and aspirations. Now that digital media is highly developed, TikTok, Weibo, Xiaohongshu...the accurate content recommendations of each platform make us feel that the whole world is pleasing us, and we all live in an extremely self-centered world. They will naturally avoid things that are slightly difficult or "uncomfortable", and everyone will naturally lower their desires. Tencent advocates "Technology for Good", so is this kind of "pleasing" good or bad? Are we going to live in our own cocoons or go out and accept the voices of the outside world?
I hope that we are not only ready to welcome the joy of victory, but also have the courage to face accidents and failures at any time~~
Share with you!
Related documents: Mid-year inventory｜In the first half of 2022, what problems did CSDN help users solve?
Here is a small survey, I hope everyone can participate