Chinese Internet users are extremely unwilling to pay for content.

Do you know what Tencent Music did to make money in the past few years?

Live Reward!

In the past, more than half of Tencent Music's revenue was social entertainment, but Green Diamond was a small income. This is very weird, spend 10 yuan to buy a green diamond, you say you have no money. Swipe hundreds or thousands of dollars for female anchors, and brush them when they say so.

To make money on the Internet in China, if you want to make money by collecting subscription fees, it is really like selling a wooden comb to a monk. 3-5 years ago, almost all domestic companies that "sell members" were losing money: NetEase Cloud, iQiyi, Youku, Tencent, Himalaya, Soul, Tantan..... All losses.

At present, with years of efforts in the industry, long videos, music, and online articles have barely cultivated a little payment habit, but it is only very reluctant: the unit price of these domestic subscription members is much lower than the list price, and many of them are low-priced channel members, such as Xianyu buys, phone bill delivery, joint membership and so on. There are also a bunch of people who share an account.

At the beginning, domestic short dramas had to go through the paid model, and I suspected that this was the doctrine that foreigners wanted. Chinese users themselves do not pay, and your short drama is aimed at a group of people who are most unwilling to pay, how can you charge a subscription fee?

In 2023, I counted the payment rate of popular short dramas, and you look at this drama: starting to pay for 17 episodes, the playback volume of paid episodes has directly dropped by more than 95%.

I see media statistics that the payment rate of some short dramas at that time was less than 1/10,000.

As far as AI tools are concerned, who dares to charge, the decline in users will not be much better than this skit.

But do you think Chinese users have any advantages? Of course, there is, that is, it can be tolerated, and the tolerance for messy functions and messy advertisements is particularly high.

This is the interface of ChatGPT.

Here is the interface of the bean bag:

This is a huge plus. Because the Internet has two monetization methods: either let users spend money or let users watch advertisements. Although Chinese Internet users do not like to spend money, they love to watch advertisements.

Love to watch advertisements is value.

ByteDance is particularly accurate in grasping this point, let's go back to the short drama, and then ByteDance entered the market, bringing a free + advertising model, which is suitable for the physique of Chinese babies. As a result, the users of Hongguo short dramas repeatedly reached new highs, and soon rushed to the first place in the industry, crushing all paying players.

In the field of novels, the starting point has been engaged in a paid system for decades, and later as soon as Byte's tomato novel came out, it directly became the first place in DAU.

In our market, it is accurate to sell physical goods such as grain, clothes, and mobile phones, and if you want to sell virtual goods, it is not impossible, but there are only a few situations that are easier to sell:

First, colored. For example, paid tips.

Second, with vanity. For example, online game recharge.

Third, it can directly earn money, such as selling his football lottery treasure book, online earning channels, and postgraduate entrance examination materials.

AI tools obviously do not belong to these three points.

Of course, free doesn't mean bad.

I have a judgment: products such as Doubao, Quark, and KIMI will be more profitable than ChatGPT and Claude in the future (C-end).

Selling AI affiliates is a very bad business model. ChatGPT members have been sold very expensively, but they still can't support the cost, and the profit target has been supported until 2029.

As the world's No. 1 LLM, after 3 years of development, the breakeven is still far away, and there is no explanation except for the poor business model.

The development of large models to this day is very limited. You chase me between several head models, very much like the asset-heavy manufacturing industry of new energy vehicles - smashing money and people. This kind of business can't make a lot of money, today you are the global leader, and in ten years you may be bankrupt.

And Doubao is not the idea of making a tool product at all, it is an AI version of Douyin.

For example, if you click on one, you will enter the Douyin interface, and you can keep brushing. Doubao also has very complex community functions, I can't figure out the specific future form and model now, and it is estimated that the industry has not thought clearly, and the AI product form is still in a very early stage.

But one thing is certain: Doubao will become another Douyin and Toutiao-style content APP, and has a wealth of advertising monetization methods.

OpenAI

Gemini

And I think it will once again create a free business miracle like red fruit short dramas and tomato novels, bringing a little business model shock to OpenAI and Gemini.