Content operation: content distribution model analysis!

In this article, the author will analyze with you: the advantages and disadvantages of the two content product distribution models, social distribution and algorithm distribution, as well as their respective scope of application. I hope everyone can get some inspiration from this.

In the era of mobile Internet with information explosion, how people can receive valuable information more efficiently has become an important issue. Content distribution, which connects content and content consumers, has become an important part of it.

The essential problems that content distribution needs to solve include two points: efficiently connecting people and information; filtering out valuable information so that the right people can see the right information.

Of course, factors that affect efficiency also include: content aggregation methods, content format itself, etc., which we will not discuss for now.

What is an efficient connection?

——By filtering out a small amount of information from massive amounts of information, the amount of information consumers receive can be reduced; the operating cost for users to receive information is reduced.

What is valuable information?

——The information topic is of interest to users; it has a certain information quality, that is, it is useful.

Currently, there are four main ways to distribute content products: algorithm recommendation, social recommendation, search, and editor recommendation.

This article mainly discusses the advantages, disadvantages and applicable boundaries of the two distribution methods: social distribution and algorithm recommendation.

1. Social Distribution

Social distribution relies on the relationship chain mechanism, and the people you follow determine what you can see. Generally, products will have a special "Follow" information flow to accumulate information about the users you follow, such as: WeChat's good-looking.

There are three benefits of social distribution:

1. Recognize the diversity of the world through friends, instead of being stuck in your own single preferences;
2. The foundation of content products is a group of content that can aggregate users. The user relationship chain is built based on the content, and it also reacts to the relationship chain. Birds of a feather flock together. Based on the content that friends are interested in, users are more likely to interact with each other, thus strengthening the relationship chain.
3. The influence of a single piece of content can be more easily amplified. When many friends are forwarding and commenting on the same content, you are more likely to check it out.

A product leader believes that social distribution can improve the quality of received content.

I think this is a false proposition.

First of all, the quality of content is constant and can only be intervened at the upstream of the content. Secondly, social distribution is based on relationship chain distribution, which means that if your friends send some low-quality content, you can still receive low-quality content.

The disadvantage of social distribution is that it will inevitably be marked by social factors. You may recommend content that strengthens your personal image rather than content that you think is particularly good simply from a content perspective. In addition, there is great social pressure, especially for social products among acquaintances. Influenced by the herd mentality, it is easier for false information to spread, and rumors are particularly evident in media products.

What products does social distribution apply to?

The main battlefield of social distribution is still social products, such as Momo, WeChat, and Soul. Secondly, it serves as a supplement to content products to strengthen social relationship chains.

2. Algorithm Distribution

Algorithmic distribution is a belief program that allows the machine to figure out your interests and preferences, and then push content to you, such as headlines.

There are five main ways of algorithm recommendation:

• Content-based recommendations: These are recommendations based on the user’s personal interests. Based on the historical behavior of individual users, the preference for content features is calculated, and then content that matches the user's feature preferences is recommended.
• Collaborative recommendation: This is a group-based recommendation. Recommendations are made based on user similarity, content co-occurrence, and by clustering users into different groups based on demographic characteristics.
• Extension recommendation: extensions based on user interests, content categories, etc.
• New and hot recommendations: recommendations based on the timeliness and popularity of global content.
• Environmental characteristics: recommendations based on region, time, scenario, etc.

The advantage of algorithm distribution is that it is easier for users to obtain valuable information.

Because recommendations are based on interests, users are more likely to be interested in them; the efficiency of content matching is further improved, because social distribution still has a certain delay, while algorithm distribution is more real-time. For example, for news information, you will only know that the user you follow has posted the content. For the algorithm, the news in this direction that you follow will be pushed to you.

The disadvantage is: it leads to the information cocoon effect, making it easy for users to fall into a narrow worldview. It will produce what users like.

In addition, the cost of controlling content is higher. The limitations of the algorithm are that it cannot judge the quality of content well, and the amount of recommended content is larger, so the cost of manual review is higher.

What products are algorithm recommendations suitable for?

High-turnover products with large content production volume, relatively simple content production, and short and fast content consumption are more suitable, such as news and information products and short video products. Algorithm recommendations based on user interests determine that such products are more suitable for users to consume multiple times to kill time.

For long videos, such as TV series, the production cycle is long and the production cost is high. In addition to user interest, it is more important to consider the character IP, publisher, etc., emphasizing the immersive experience of users after a decision.

The above are the main modes of content distribution. You are welcome to learn and communicate together.

Source: Miss Dong