How to operate short video content well?

As the combined daily active users of Douyin and Kuaishou exceed 700 million, short videos have become the most popular form of content delivery in this era because of their duration that is more in line with the mobile Internet and the rich categories derived from the short videos shot by everyone. In addition to Kuaishou and Douyin, wave after wave of short video product crazes have been spawned across the global mobile Internet.

I myself have participated in and led the incubation of several short video products from 0 to 1. Now it seems that short video products are content operation-driven products. The essential difference between the products is not the product functions themselves, but the short video content.

Here I will briefly summarize the content operations when I was working on short video community products and share it with everyone.

The goal of short video content operation

When short video community products are launched, the person who truly determines the core data of the product may not be the product manager, but the content operator.

When it comes to content operation, many people immediately start recruiting content creators, organizing activities, etc. But before doing this, we need to clarify the goals of short video content operations.

When it comes to goals, many operations students will count the number of videos uploaded each day, or the number of participants in an event. There is nothing wrong with these goals, but they are all one-sided.

My understanding of the goal of content operation, if put in one sentence, is "providing short video content that users like." In terms of data goals, it is the average number of video plays (VV) per person and the average playback time per person.

Providing short video content that users like is the foundation of the short video content community. If you can’t find content that users like, then core data such as user retention will inevitably not be good. Only when this problem is solved can the subsequent community be established. This is also why content operation is so important in short video products.

Short video content operation path

After clarifying the goals of content operation, how should we carry out content operation?

We can simply draw the process of content flow in the short video content community, as shown in the figure below.

Content flow diagram

From the figure, we can see that content circulation is mainly divided into three stages: content supply, quality control and content recommendation. At different product stages, the emphasis is also different.

(1) Cold start phase

During the cold start phase when a product is first launched, the most important goal is to quickly find content that users like. There should be many ways to find content that users like. Some people rely on experience, some people refer to competitors' products. Here I share a more commonly used method. Use testing to find out what users like.

We can start testing in the early stages of a product launch, or even before it goes online. We can define a rough range of content, and then through the delivery of advertising materials, we can know what kind of content will be liked by users.

After the product is launched and has a number of users, testing can be done inside the product. We can introduce the types of content to be tested during the content supply stage, and then ensure the number of content recommendations during the quality control stage.

The test usually lasts from a few days to a week. The final evaluation standard for whether the test content is what users like is to separate the data sheets of this batch of content and compare them with the average of the entire video. Generally, we will look at indicators such as the average number of video views, average CTR (click-through rate), and average interaction rate.

If it exceeds the overall mean or even reaches the 75th percentile level, this type of test content is undoubtedly liked by users, and the number of introduced content can be adjusted. Through this testing method, we can know what content users like in a very short time. After reaching reliable conclusions, these contents can also be fed back to the channel advertising side. When acquiring users, they can directly use this type of material, thus achieving a perfect closed loop.

(2) Stage of increasing content volume

When we find content that users like, the overall user data will generally improve significantly. At this time, the product will enter a faster growth period, and the overall number of users may reach a new level in a very short time. At this time, the problem faced by user operations is how to quickly increase the amount of content and ensure that there is sufficient content supply every day.

We can see from the content flow diagram above that there are three main channels for content supply: uploads by ordinary users, official production, and contracted producers.

In the early stages, when exploring content that users like, it is basically done by producing it yourself or signing contracts with producers. When it is necessary to focus on increasing the amount of content, you must still rely on ordinary users to produce it. This is also a key link in the establishment of a short video community. If the community has no producers, it will be useless even if it finds content that consumers like.

In terms of stimulating production among ordinary users, the following directions are worth exploring:

• Content guidance : We must clearly tell users what content we need. This mainly has two aspects: one is that the content in the content recommendation must clearly tell users what kind of content can be recommended. At the same time, during the review process, we strictly control the works posted by users. If they are not the content we need, we will not recommend them. Let users know what kind of works we need from both positive and negative incentives;
• Production incentives : Production incentives for users can initially be provided through some operational methods, such as regular themed content activities, which can motivate users to participate and upload videos through prizes. In addition to activity incentives, there are many product-side incentives, usually fans, content interaction incentives, etc. Going further is some monetary incentives, which can be tried at some appropriate product stages.
• Producer operation : In addition to the above methods, there is also a more important producer operation, which is generally done through contacting producers through operational activities, or inviting producers who currently perform well in the product to join the community, giving other ordinary users additional incentives that they cannot get in the community, and organizing this group of producers. They will become the core users of the product and ensure the content supply of the community.

Short video account operation practice

We have talked about many methods of short video content operation before. In fact, the most important thing in content operation is the feeling about the content. The best way to exercise this feeling is to operate an account yourself, observe the changes in content, number of fans, and number of interactions under the account, and you will be able to feel the changes in community content. Here I will give you an example of my own practical experience in operating an account. I hope you understand that knowing is easy but doing is difficult. Many times you have to do it yourself to understand the details.

My original intention for creating an account was because I was unable to judge what kind of content users liked during the product development process. Many people talked to me about content tone and community tone, but it seemed that few people could clearly explain what tone was. Many people say that tone is the classification of content, but I am very skeptical about this statement.

So I wanted to find a methodology to explore what users like . Is there a way to quickly know what content users like and how to clearly mark this content? This is the original intention of doing this.

At the beginning, I thought about conducting experiments on Tik Tok or Kuaishou, but there was a problem. My content mainly came from the collections on various platforms, and was mainly video content that I collected (moved) myself. Kuaishou and Douyin have extremely strong content duplication checking strategies, and the content reserves of these two platforms are so large that almost all the content I upload will be checked for duplication, and I will hardly get any effective exposure.

So I looked at several other platforms. At the beginning of 2019, Xiaohongshu’s homepage feed was revamped, with a strong shift towards short videos, and their recommendations responded instantly. If you click on a few short video contents, almost the entire feed will turn to short videos. Therefore, I started uploading content on Xiaohongshu, and the number of views I achieved was much higher than that of Douyin and Kuaishou. At that time, Xiaohongshu's daily active users should have been more than 10 million, which was also a relatively large platform, enough for me to conduct content experiments.

(1) View content operations from the perspective of product development

Because the content of the products I make is mainly emotional, when I posted my works on Xiaohongshu, I also chose emotion as the main category. At that time, when I was searching for emotion-related accounts, there were already some accounts with thousands of followers.

After determining the general direction of the content, you will find that emotion is actually an all-encompassing category, including inspirational chicken soup, sad stories, love, family, friendship, etc., which can be subdivided into dozens of small categories.

So which category should we start with? This should also be the first question that all content operators need to answer.

When I answered this question, I returned to the idea of ​​product development. I regarded the content or this account as a product, so the question became three soul-searching questions about the product:

1. Who are your users?
2. What problem do you solve for them?
3. What are your advantages over other competitors?

First of all, who are the users? I had read some information about the user portrait of Xiaohongshu on the Internet at that time, so I positioned my target users in line with the main user portrait of Xiaohongshu, "young women in their 20s."

What problem does it solve? At that time, I thought that emotional content mainly solves the user's psychological needs for recognition, understanding, and resonance, which is essentially the social and respect needs in Maslow's hierarchy of needs theory.

Compared with some other similar emotional accounts, I found after looking at their account content that the content styles of many accounts are seriously inconsistent, and the emotions expressed seem to be messy. Some are sentimental, some are full of positive energy, some are for men, and some are for women.

Therefore, I felt that my advantage over other accounts at that time was that I must strictly control the style of the content and ensure the consistency of the content under the account. But this advantage can only be established after I have explored a content type.

(2) What do users really like?

After answering three soul-searching questions, I started uploading content on Xiaohongshu. The strategy at the beginning was very crude. I directly uploaded many popular emotional videos from Douyin and Kuaishou to Xiaohongshu. Later I found that the content that was popular on other platforms did not perform well on Xiaohongshu. At that time, I guessed that it was because the user groups were very different, which led to differences in their preference for popular content. Now that I think about it, it is very likely that they have already viewed these popular contents on other platforms.

After encountering this problem, I brought my starting point back to the perspective of my target users, young women in their 20s. What emotional problems will young women in their 20s have?

Most girls in their 20s are not married yet, and may have just had one or two relationships. In their 20s, it is basically the age for first love to break up. And at this stage, before finding a future spouse, the relationship is in a state of turmoil, full of nostalgia for the past, full of longing for the future, and full of anxiety and confusion about the present.

After thinking about these things, I made some adjustments in the direction of content selection. First of all, it has to be content for females, so I directly eliminate any content that encourages men and is from a male perspective, such as many videos and music that describe men getting hurt in relationships. Therefore, the content I chose is all about narrating the sadness, confusion, etc. after a breakup from a female perspective.

At the beginning, I found that there were a lot of pure text PPTs with music, so I chose to upload a lot of this content. But the strange thing is that there are some contents that I think are good, but the performance is not good, while there are some that I think do perform well.

During this period, I often uploaded about 5 videos a day. Generally, 1 or 2 of them would have a few hundred views, and the others would have very few. After more than a week, the difference between good and bad videos in the first few dozen videos became increasingly obvious.

I tried to summarize the types of videos that performed better, because the video content was presented in the form of PPT with music, and those that performed better didn’t seem to have any particular commonalities. Although I didn't know why, I still deleted all the content with poor performance to ensure that the account was filled with good content. I thought there must be some commonality, but I just hadn't discovered it yet.

(3) Inspiration from Dong Qing’s famous quotes

During the subsequent exploration process, there was a very obvious change. One day I uploaded a few sentences narrated in Dong Qing's voice. I think some of them were not said by Dong Qing, but as long as they were signed with Dong Qing, the overall playback effect was greatly improved. Later, I found that videos featuring celebrities such as Ma Yili, Yao Chen, Shu Qi, and Yang Lan also performed well.

Careful readers should be able to understand the "mystery" here. These female stars have one thing in common: "very independent." This seems to be a common appeal among young women on Xiaohongshu, who all hope to be more independent. At this point, I recall that Xiaohongshu is a shopping platform that promotes products. The women who are active on this platform are likely to have certain material pursuits, and therefore they will inevitably pursue independent existence. This seems to explain it, of course these are just my own guesses after the fact.

After getting this revelation, the content I chose became more focused. I deleted a lot of videos that emphasized the pain and suffering after a breakup, and only chose videos that emphasized female independence, such as how to make the other person regret after a breakup, how to be a better person in life, etc.

Since then, the number of views of such videos has almost quickly reached a new level. I even did a very dangerous experiment. I accidentally came across a video that described how women should be "bad girls" in relationships. The video emphasized that women should be dominant in relationships and men are just "playthings." After this video was released, it broke the record of 1 million views, and the number of likes, shares and comments all broke records. There was a lot of discussion in the comments.

"Bad Girl" theme content

Number of likes, comments and reposts of "渣女" themed content

(3) What exactly is tonality?

After doing it for a while, I have a deeper understanding of my original intention. In fact, it is just a question: what exactly is content tone?

I think simple content classification is not the right tone, because it is an objective classification from the perspective of content management that is separated from the classification knowledge of users and scenarios, and its significance is retrieval. However, many detailed categories may be impossible to define. For example, the earliest music playing platforms classified music into classical, pop, Hong Kong and Taiwan, R&B, etc. However, the classification that truly conforms to user habits should be like the NetEase Cloud Music playlist, "Songs suitable for listening to while running", "Top 50 must-have songs for self-study", etc.

In many short video production tools today, when adding music, many are still classified according to music type, but now I have seen that some platforms have begun to use this scene-based classification method, such as music suitable for shooting food, music for shooting scenery, etc.

My understanding of tonality is still to think about it from the product perspective. Who uses it in what scenario and what problem does it solve? This is the tonality of the content. For example, the content tone of my Xiaohongshu account is to meet the desire of young women to be more independent after setbacks in their relationships. Later, in the short video community I created, I also used this definition to find content with the target tone.

(4) There are some fun things

During the operation of the account, some interesting phenomena were encountered. The title of the video is very important. It is a great way to promote user comments. If the title is a declarative sentence, the effect will definitely not be good, but if it is a question, such as "Do you still remember him?", "Will you shed another tear for him?", etc., these titles with strong questions will generally receive better comments.

In addition, when choosing a video cover, a cover with a beautiful female image will definitely bring more clicks, while a cover that is just a landscape or meaningless objects will have a much worse effect.

During the operation of the account, as the number of fans increased, it now has more than 50,000 fans. I received a lot of emotional consultation questions from fans, which is smaller than the target user portrait I had expected. Some of them are still in middle school. Maybe it was a failed secret crush, a rejected confession, or they rejected someone else's confession under pressure. It made me feel personally that "love" is a bitter thing. For some of them, I didn’t know how to comfort them so I didn’t reply. For others, I did reply, hoping that my reply could help them solve their inner problems.

postscript

This article is a bit long, but there are still many details in content operation that have not been explained clearly, and there are also many questions that I am thinking about now. If you are also doing this or are interested in this area, let’s communicate together.

Author: Nancun Xiaofu

Source: Nancun Xiaofu