Operate short video community from 0 to 1!

As a hot product direction in recent years, short videos have spawned well-known products at home and abroad, such as Douyin, Kuaishou, Meipai, Weishi, Tiktok, and Kwai. As an upgraded carrier of pictures and texts, short videos have attracted a large number of user groups due to their low consumption threshold and stronger entertainment. At a time when customer acquisition costs are skyrocketing and many products are struggling to grow, short videos have become the darling of many companies' projects. The author has experienced two short video projects and recorded some experiences to share with you.

As the name suggests, short video community has two meanings. One is short video, and the other is community. It represents content consumption and user community. To build a short video community, the ultimate goal is to do well in both directions.

• From the perspective of content consumption, the core goal is to increase the average consumption time per person (i.e., average vv per person). The higher this figure is, the more popular the content on the terminal is among users.
• From a community perspective, the core goal is to increase various interaction indicators (like rate, comment rate, sharing rate, etc.) and attention indicators, which means that users are truly participating in the platform and connecting with other users.

Both of these will eventually be reflected in the daily active user level (DAU) and retention data.

After clarifying the goal, the next step is to break it down. Products cannot be achieved overnight. They need to be built in stages and steps. It takes a certain amount of time to complete the interim goals and move closer to the ultimate goal.

Therefore, at different stages, there will be different goals and things that need to be focused on. In the projects I have experienced, short video communities generally go through several stages, which will be described in several articles below.

The first stage is the initial volume of the product.

1. Starting quantity

For a new project, unless it is a company priority or strategic level, the primary consideration in the early stages must be how to quickly increase volume. Short videos are carriers of content, so there are generally two ways to increase volume: content consumption and content production tools .

1. Content consumption

That is, attracting users to consume by providing content that suits their preferences. In the past few years when the domestic short video market was still emerging, many products used this method to attract users, such as the curious and rural content on Kuaishou, the trendy content on Douyin, and the funny videos on Neihan Duanzi. As long as these videos are watermarked with the platform and released to the target user group, they will immediately attract a large number of users.

However, as various categories of content are aggregated on oligopolistic platforms, it is becoming increasingly difficult to introduce users through content consumption, because everyone has discovered that these contents can be consumed on platforms such as Douyin and Kuaishou, and the consumption efficiency is higher. The new platforms have no barriers to content, so the way new short video platforms initially increase their volume is more towards another direction, namely, content production tools.

2. Content production tools

That is, the tools used by users to create short video content. In a narrow sense, it often refers to shooting tools, and in a broad sense it refers to all tools that can help produce short video content (shooting, beauty, stickers, templates, special effects, etc.). A large number of short video communities will choose to enter the market through production tools in the early stages because of the following three advantages:

1. The starting price of the tool is relatively cheap, which is suitable for situations where funds are relatively tight in the early stage;
2. Tools naturally have the property of attracting traffic, are suitable for dissemination, and can easily attract a large number of natural users;
3. The R&D cost of tool capabilities is low, the cycle is short, and they can be quickly launched, which can create enough time for the product to lay the foundation for subsequent planning.

In the early days of Kuaishou, which was still in the Kuaishou GIF era, it gained a large number of users by making it easy to make GIF images. Overseas short video app Likee relied on special effects editing tools to achieve a daily active user base of 0-1 million.

Biugo short videos, which belong to the same series, also achieved early user acquisition overseas through the tool of creative video templates. The most suitable way to enter the short video business and take the business from stage 0 to stage 1 is to use content production tools that are distinctive and attractive to users.

2. Product, Operation and Buying Volume

Now that the method of increasing volume has been clarified, the next product focus will be to continuously launch easy-to-use, fun, innovative and interesting video production tools during the volume-increasing stage.

Whether it is providing exquisite special effects, convenient and simple video editing tools, interesting interactive video shooting methods, or carefully prepared beauty filter effects, the premise of these tools is to lower the threshold for creators, so that they can produce relatively exquisite video content, and these contents must be suitable for dissemination and can bring unexpected natural user growth.

At this stage, operations are more responsible for tool operations. The focus is to continuously use the tools provided by the product to guide users to play as much as possible, while providing more creative and material updates to ensure that users continue to be interested in the tools provided by the product in the early stages. There are three aspects that can be implemented:

• Update tool materials on a daily basis to ensure the freshness of the tools for users;
• Cooperate with product managers to maintain sensitivity to outstanding competitive products and technologies on the market, and innovate new tools and gameplay in a timely manner;
• Through operational activities and other means, guide users to use tools and create content in a targeted manner.

Another focus of work is buying volume .

User buying means spending money to place advertisements on channels to attract users to download products. In China, advertisements are generally placed through DSPs or app store channels such as Guangdiantong and Pangolin, while abroad, they are mainly placed through Google Admob or Facebook advertising platforms.

In the early stages of a project when funding is limited and product quality needs to be improved, the budget for buying traffic will generally not be too much, and it is more about maintaining a certain level of users.

At this time, the cost of releasing materials using content production tools is generally low, which is very suitable for this stage, and most of the users attracted also like or want to use the tools (otherwise they would not click on the ads and download the App).

3. Content Reserve & User Account System

Although the main goal in the early stage is to make tools in order to increase the volume, the ultimate core is to create a short video community , so we must prepare for this goal in advance. There are two main aspects to focus on. The first is the video content reserve, and the second is the user account system .

1. Video content reserve

As a short video platform, it must provide a large amount of content to users for consumption. Therefore, a complete set of links from content supply to content consumption must be opened in advance. Simply put, it is to run through the process of " content production-content understanding-content review-content distribution ".

First, content production is the upstream of the entire chain, and generally there are two main sources: external crawling and on-site UGC (user-generated content).

In the early stages of the product, unless you recruit or spend money to sign a team with video production capabilities (and the number must be very large), the quality of UGC content is basically very poor and not consumable. A large amount of meaningless content is a portrayal of UGC at this time. Even if the operation team guides users to produce through operation means, it is still not feasible compared to the content of mature video production teams.

Therefore, at this stage, the main method is to crawl content as the main method and use UGC content as the supplementary method. On the premise of ensuring that the content is consumable, the platform's users are trained in advance to become mature video content producers.

Content crawling is a huge project, which involves what specific content to crawl, which platforms to crawl, how to solve the anti-crawling mechanisms of other platforms, etc. It requires a dedicated team to complete this task.

Looking at it in more detail, issues such as the clarity, scale, content category, whether watermarks are removed, how to crawl regularly, and the warehousing process after crawling are all issues that need to be resolved at this stage. Only when the first step is successful can subsequent content supply be less prone to problems.

From a domestic perspective, Douyin, Kuaishou, Bilibili, Zuiyou, Weishi, and Meipai are relatively direct short video content platforms, while abroad, there are ins, kwai, tiktok, likee, etc. (YouTube has the richest content, but because it is owned by Google, it will lead to the removal of the App if it is discovered, which is extremely risky). These can all be used as targets for the initial content cold start, but it should be noted that content crawling involves legal risks, so they need to be avoided.

Second, after the content is produced, it needs to be understood . In the early days of the editing system, content understanding mainly relied on content editors. Today, in the era of content explosion, it relies more on machines for understanding, which requires a mechanism that allows machines to quickly understand the content.

In the early stages of the short video project, this work mainly relies on manually establishing a reasonable labeling system and relatively accurately labeling the content to help the machine understand. At this stage, the more detailed the labels are, the better, because the more detailed the labels are, the more accurately the labelers can define the video content through the labels.

Generally, labels are classified and layered according to different dimensions. Classification includes author level, content attributes, security risks, etc. Layering is based on the detailed breakdown of each level of labels. For example, content attributes will be broken down into first-level labels (such as dance, singing, talent show, comedy, animation, etc.). Based on the first-level labels, a large number of second-level labels will be further split out, and these second-level labels will directly define the content of the video. At the same time, there will be priorities between labels, because labels are essentially the reason why users consume this content. For example, the label of beauty will be better than singing, and users will pay attention to beauty first, while the label of singing is less important than beauty.

Third, after the content is understood, it must be reviewed. Vulgar, pornographic, or value-incorrect content must not be distributed . Otherwise, the platform will be sanctioned by the government, and the phenomenon of "bad money driving out good money" will occur, causing the quality of platform content to decline sharply and destroying the content ecology of the entire platform.

Fourth, content ultimately needs to be distributed for consumption , and the mainstream approach is to establish a recommendation algorithm system and let the algorithm distribute content. In the early days, when there was no user behavior and data accumulation, the ability of tags was used extensively to understand and distribute content. As more and more user behavior data was collected and recommenders knew more about users, more models and algorithms were used to push content. Therefore, the important part in this part is to obtain user portraits and behavior data, to clearly understand the video content, and to match users with content.

2. User Account System

How to get users to leave more information is an issue that every product needs to pay attention to. If it is narrowed down to one question, it should be how users log in to this community product and build their own accounts.

The sooner users are made aware of this, the better it is for the product. Because once users form habits, it will take a lot of effort to make subsequent changes. Therefore, the issue of "getting more users to log in" must be kept in the mind of the product manager in the early stages, and more users must be allowed to log in and leave their accounts through product functions in a subtle way.

In general, this system may not be perfect in the early stages, but it must at least be able to run smoothly, allowing users to develop the habit of content production and consumption in the early stages, laying the foundation for later product transformation.

4. Existing Problems

In the first stage of the short video community, the strategy is generally to increase the volume by focusing on content production tools and supplemented by content consumption. Although this stage seems to be making tools, a lot of work is in preparation for subsequent content consumption.

Since tools naturally have the problem of "high short-term retention and low long-term retention", when the user scale reaches a certain stage, product transformation is imminent, that is, it is necessary to begin to shift the focus from content production tools to content consumption communities.

The short video products in the first phase are equivalent to "saving the country by taking a detour". They enter the market through tools and simultaneously cultivate users' habits of content consumption and content creation. When the product is relatively stable and the data is growing, you can start to consider preparing for the second phase.

Author: Yin Gu's Tavern

Source: Silver Ancient's Tavern