Discipline is particularly important, and it is best to commit suicide now on a large scale.
Text: "Chinese Entrepreneur" reporter Kong Yuexin
Edited by Ma Jiying
Header photo by Deng Pan
In October, Zhu Xiaohu, the managing partner of GSR Ventures, went to the United States for a meeting, and his state surprised LPs. "It's not easy for you to have no gray hair. The other party said.
In an interview with China Entrepreneur in GSR Ventures' office, Zhu Xiaohu raised his arm and rubbed the top of his head, and he even felt that his hair was darker than before. One of the reasons may be that he is less anxious about the development of the invested projects: "As long as the company can make money, I won't be unable to sleep." ”
He remains sensitive to the wind and new trends. In April 2022, Fancytech founder and CEO Kongjie (pseudonym), who was trapped in Shanghai, could only communicate with Zhu Xiaohu online when raising funds, and after 15 minutes of the video, Zhu Xiaohu made an investment decision.
At that time, the popularity of AIGC had not yet risen, and Fancytech's main business was to help e-commerce websites aggregate content and improve brand conversion rates. After Fancytech accessed AIGC to achieve technical upgrades and accelerated performance rolling, Zhu Xiaohu chased 3 more rounds.
But to some extent, this sensitivity is in the form of non-consensus. For example, in the first half of this year, when most people were chasing large models, Zhu Xiaohu, who had been "silent" for a long time, suddenly appeared and poured "cold water" on everyone. In his public speech, he mentioned that ChatGPT is very unfriendly to startups, and everyone should give up financing illusions in the next two or three years, which immediately triggered a debate between Fu Sheng, chairman of Cheetah Mobile, and him.
Until now, Zhu Xiaohu still insists on his own point of view, and the large model is ultimately an opportunity for large manufacturers, and startups should focus on vertical scenarios such as sales and marketing. The same goes for investing, "I think you have to focus on what you're good at, because the market cycle is always changing."
AIGC and consumption are the tracks that Zhu Xiaohu has always focused on, but even if he is familiar with the field, his investment decisions have slowed down. This year, he spent more time inspecting the project, like in the past, after chatting with Cheng Wei for half an hour, the plot of making a decision to invest will hardly be staged again, "because no one grabbed (the project), this year I can watch a project for an average of three or four months to see if it is a 'good brand'."
From everyone "fast forward and rushing" when hot money poured in, to investing slowly and leisurely now, in Zhu Xiaohu's view, 2023 is the most comfortable year since he made an investment.
As of November 2023, Zhu Xiaohu's projects are only 1/3 of those in previous years, and there are still several projects under his inspection. "The sense of accomplishment of investors does not lie in making money, but in the development of the enterprise. ”
In an interview with China Entrepreneur, one of the words he repeatedly mentioned was "discipline". The specific actions that are in line with discipline are: reasonable control of scale, reasonable control of investment rhythm, and reasonable control of project valuation.
The following is a summary of the dialogue between "Chinese Entrepreneur" and Zhu Xiaohu, which is abridged.
Talking about large models: Those who chased large models in the first half of the year were more injured in the second half of the year
China Entrepreneur: What do you think about the impact on the technology cycle and the entrepreneurial cycle after the big model comes out?
Zhu Xiaohu: I think this is a good beginning of a new cycle, the key for entrepreneurs is what you want to do, just like the PC Internet and mobile Internet occupied 10 years each, and now the new AI era is opened, and there will probably be 10~20 years.
Source: Visual China
But it is definitely unrealistic for entrepreneurs to do mobile Internet iOS. There are many opportunities for the application of large models, and we have invested in many companies in the AIGC field and have done a good job.
China Entrepreneur: Does this change in the entrepreneurial cycle put forward new requirements for investors? Now that there are fewer entrepreneurs and they don't need so much money, how can investors find good projects in this environment?
Zhu Xiaohu: I think the core is the same as before, that is, to make independent judgments, and not to follow the crowd. In the first half of this year, everyone was chasing big models, playing a lot of money, and now they don't know what to do. The core issue is that (investors) should have their own independent judgment, where do you think the opportunities are? It must be very difficult for a startup to make a new iOS, and we have made similar mistakes on the mobile Internet, and we have to learn a lesson.
Many investors may not have such a deep memory of the story of the PC Internet and mobile Internet era, we have lost money, so I have a deep memory, I know where there are pitfalls, and we will focus more on some more creative companies or some points with more opportunities.
"Chinese Entrepreneur": What are these "pits"?
Zhu Xiaohu: It's very simple, in the early days of the mobile Internet, there were many companies that were very popular at the beginning, such as Android Optimizer, which had hundreds of millions of users at that time, but who still uses Android Optimizer now? No.
For a simpler example, there was a company that made flashlights on mobile phones back then, and today iOS and Android systems have their own (flashlights).
Now similar to the early days of the operating system, the early stage of the large model may not take care of many small functions in the corners, leaving opportunities for startups at any time, but later it will free up its hands and can immediately make up for these things, and this part of the startups will be gone.
There was a previous article mentioning that GPT-4 launched a package of tools, and a bunch of startups were destroyed. I don't think [startups] should be under direct fire from large models, otherwise it will be difficult.
For example, the programming ability of large models can be increased by 50%, that is, after it automatically codes 50%, it will leave a bunch of bugs, although it helps the company reduce the labor cost of junior programmers, but it needs more senior programmers to debug (debug and troubleshoot). A lot of startups are heading towards using AI to optimize their code, or better debug, and that's just a short-term opportunity. If GPT-5 comes out next year, the code generation efficiency of large models may be increased to 90%, then this branch is in danger again.
Therefore, I think that entrepreneurs should not run directly on the large model (track), stay away from it, and at the same time, use it to seize customers and realize the closed loop of user data, which may be a better defense.
"Chinese Entrepreneur": I discussed with Fu Sheng in the circle of friends before, do you think there is a clear conclusion?
Zhu Xiaohu: To be honest, for startups, I don't think there is much chance for the large model itself, first, the cost is very high, and second, there is no data and scenarios, so it cannot be optimized.
Therefore, I think that large models are ultimately opportunities for large manufacturers, and startups should focus on the application layer that they are good at such as sales and marketing, and there are still more opportunities in various vertical scenarios.
China Entrepreneur: What are the specific vertical scenarios?
Zhu Xiaohu: First, it is difficult for large model companies to obtain vertical industry data. Second, they need to be able to close the loop of data so that they can continuously improve their vertical models. Third, we must be able to continue to maintain the connection with customers, because it is impossible for large manufacturers to knock customers one by one. These three points are particularly important.
In other words, adding AI on top of the original business is the most reliable and relatively easy, but it is difficult to add something to the AI model or product first.
China Entrepreneur: How do domestic entrepreneurs or large factories think about this?
Zhu Xiaohu: We've been talking, and big factories are really concerned about where their boundaries are. Some big companies are better, and they don't want to grab too many startup opportunities. But I think that whether it is China or the United States, it is unrealistic to rely on the pity of others, and you must have your own value to customers and large factories in order to truly find your own base.
China Entrepreneur: What do you think of the current progress of the commercialization of domestic large-scale models? What stage are you in?
Zhu Xiaohu: Now I'm holding a hammer to find a nail. To be honest, by the end of this year, I think that several large models in the head can basically achieve the 3.5 level, and there is hope to reach 4 in November next year. However, after OpenAI's GPT-4 comes out, the commercialization scenarios (in the United States) that it can really land in are also very limited, and it should be similar to China.
China Entrepreneur: In which areas are you more optimistic about its commercialization?
Zhu Xiaohu: At this stage, text creation and other projects are definitely the easiest, such as helping to write some advertising copy, work summaries, speeches, etc., and they can be perfected relatively well; secondly, generating photos, such as Midjourney, is doing a good job; in China, we feel that it is better to generate short videos, because China has a lot of short video data, the United States does not, and all their e-commerce websites, such as Amazon, are almost still based on pictures. In the past three years, all e-commerce companies in China have basically shifted from pictures to short videos, so all traffic on e-commerce websites today is prioritized for short videos, not photos.
China has a lot of short video products and data, and you can use this data to train your own vertical model, so Fancytech has done a very good job in product transformation.
China Entrepreneur: Is this a big enough investment opportunity and money-making opportunity for investors?
Zhu Xiaohu: We hope that when we start investing, everyone thinks that it is a niche market, so that big platforms will not pay attention to it, or will not spend heavy troops to do it. After a few years, when it grows up, there will be more space.
"Chinese Entrepreneur": Although AI is popular this year, most of the head projects can't get in, and everyone may still be waiting and seeing small projects, so they won't invest in AI at the end of the year, and will start to invest in the upstream and downstream of cars, such as new energy and carbon peaking and carbon neutrality.
Zhu Xiaohu: New energy has also plummeted.
I think AI is still non-consensus to this day, and there are very few investors who really have the ability to think independently. In the first half of this year, a bunch of people went to invest in large models, and the valuations were very expensive, what is the value of large models? Today, every large model is about the same, and they all say that they can score 95 points.
Now is a very good time to explore good projects with high cost performance in non-consensus.
China Entrepreneur: So are the top large-scale model startups still sought after?
Zhu Xiaohu: It was sought after in the first half of this year, and everyone regretted it in the second half of the year.
China Entrepreneur: For entrepreneurs, now can only embrace AIGC?
Zhu Xiaohu: AIGC must be embraced. In another two years, any company will not be competitive without AIGC. Companies including consumer and home appliances need to consider using AIGC to reduce costs and increase efficiency.
Now it is very easy to reduce costs and increase efficiency with AIGC. For example, we invested in a company with more than 1,000 employees, and the human resources department has 4 people whose responsibility is to answer employees' questions about the company's welfare. This kind of internal training of the robot is very easy, as long as the company's welfare policies are fed to the robot training for 1 week, and then the employee asks the company any welfare questions, the robot can answer. Others, such as game image generation, are done with AIGC, which can reduce costs and increase efficiency.
Talking about entrepreneurship: "less money" is a double-edged sword
China Entrepreneur: At the end of the year, how many projects do you expect to invest in this year?
Zhu Xiaohu: It's almost 1/3 of previous years, not that there are few projects, many projects are in the observation period, and I have a good number of projects that are still being investigated, to see if or when to make money.
In the past, investors and entrepreneurs paid money without meeting, but now it is impossible to happen. It's like going back fifteen or sixteen years, and I could watch a project for three or four months. I will observe whether the forecasts given to me by entrepreneurs every month are consistent with the real data of each month. After observing for three or four months, what he said every month was very reliable, and I was willing to pay for it.
China Entrepreneur: Has GSR changed or adjusted the pace of fundraising and investment in the past two years?
Zhu Xiaohu: We are a new fund that was raised at the beginning of last year, so it's okay.
In the past two years, we have communicated with LPs, and the most important thing is that the secondary market has fallen too much. Many funds have a ratio of primary and secondary markets, and the proportion of the secondary market has fallen a lot, and the proportion of the primary market has risen too high, so the mismatch of the primary and secondary markets is very serious. In this case, it is necessary to wait until the secondary market comes back before re-allocating the primary market.
Second, the United States now has more than 5 points of risk-free interest, and since the money stored in the bank has a demand interest of more than 5 points, why should investors invest in the illiquid primary market? This will affect the entire market, so that the amount of funds willing to be allocated to the primary market will be relatively affected.
Finally, there is the problem of liquidity, the part of the money that I like to invest in the Chinese market the most is still locked up and not cashed out, so I can't come back and reinvest.
All of this limits the recycling of funds. Similar to many RMB funds, they don't see the DPI (cash return), the money can't go back, and it is more difficult to come up with new money. The same is true for the US dollar. He needs to see that China's money can flow again, that some good companies can exit, and that money can come back to China. So I feel that the market will not start to pick up until at least 2025.
China Entrepreneur: From this point of view, the current environment has many challenges for entrepreneurs?
Zhu Xiaohu: Yes, "less money" is a double-edged sword, and it is a good thing for people who really start a business, because there is less noise and fewer competitors. However, to VC companies, they will definitely not be able to get money now.
China Entrepreneur: Do you give more advice to entrepreneurs this year than in previous years?
Zhu Xiaohu: Last year, the companies that listened to our advice were all doing well.
China Entrepreneur: What is your advice?
Zhu Xiaohu: Last year, it was to optimize cost reduction and efficiency increase. The companies that optimize the cost reduction and increase the efficiency are all profitable this year, and there is no one who is not profitable. There are still a lot of entrepreneurs who have listened to the advice, and everyone is very disciplined, so they are all making money this year.
"The coldest is not winter, but spring, when the snow melts. "This year is the spring of snow, if the startup was ready last year and lowered the cost, they will live well this year. If you didn't reduce the cost last year, you will be very tired this year, and then you will have to reduce the cost repeatedly.
China Entrepreneur: After so many years of investing, how do you feel different this year?
Zhu Xiaohu: This year is the most comfortable year for me to invest, and I really go back to 15 years ago. I can see again and again if this card is really good.
China Entrepreneur: This year, RMB LPs are more active in the primary market, does GSR consider increasing the scale of RMB funds?
Zhu Xiaohu: That's why it's very important to talk about discipline, and it's best to commit suicide on a large scale now.
I need to reasonably control the scale, reasonably control the pace of investment, and reasonably control the valuation of investment, which is particularly important.
China Entrepreneur: In terms of exit, how do you think the organization has performed this year?
Zhu Xiaohu: I think the core is to look at profits, when capital is the coldest, only every enterprise has profits, can I fall asleep. Exit is a matter of course, as long as every business can make money, when the United States starts to cut interest rates, there will naturally be exit channels.
China Entrepreneur: The state of you and the GSR team seems to be a bit unusual or special in the current primary market, and you are not particularly anxious.
Zhu Xiaohu: First of all, we basically don't have peak projects. Although I also invested at the peak, I still need to reflect on it, but compared to our peers, our highest percentage is very small, so most of our companies are still profitable today. We now have a company that makes a profit on bank interest, and he gets too much money at the peak, and he can get 5 points of interest in the bank, and I don't have to worry about whether his company can survive.
At least for me, it's relatively comfortable now, I didn't go in during the peak period, and now I can watch it slowly during the trough period, and then I can find a good company and make a move.
China Entrepreneur: How does GSR's exit performance compare with previous years this year?
Zhu Xiaohu: Let's look at one indicator now - profit. A few months ago, I reported to RMB investors that in the first half of last year, the top 10 companies of RMB funds lost more than 100 million yuan, and in the first half of this year, the cumulative profit was nearly 200 million, from a loss of 100 million last year to a cumulative profit of 200 million.
China Entrepreneur: When do you think there may be a high point in terms of exits?
Zhu Xiaohu: The Fed's starting line to cut interest rates is the core indicator. Before the dollar starts cutting interest rates, I don't think any large-scale exit is realistic.
Talking about consumer investment: more than 15 times PE, you can't see it
China Entrepreneur: Are there any changes in the project areas inspected by GSR this year?
Zhu Xiaohu: We always focus on two areas, AIGC and consumption.
China Entrepreneur: It seems that it is difficult to find good projects in both areas right now, especially consumption.
Zhu Xiaohu: There are still good projects, but the key is whether (it) needs financing. Valuations are too low now, and a lot of good projects don't want to be financed. Therefore, the two parties must have a certain fate, for example, the founder may indeed need funds, or need the endorsement of well-known investors.
"Chinese Entrepreneur": Now the consumption potential of the sinking market is very strong, many big brands are going down, and there are many brands that specialize in expanding chains in the sinking market, what do you think of the opportunities for these brands?
Zhu Xiaohu: We have always been very optimistic about China's overall consumption, including the GDP growth in the third quarter, which is mainly driven by consumption. A few months ago, I also told LPs in the United States that China has a population of 1.4 billion and a middle class of 300 million, which is the second largest consumer market in the world, which is comparable to no other country. With 1.4 billion people eating and drinking every day, this market is very large.
Now that entrepreneurs have fewer competitors, if they provide consumers with really good products and good prices, the consumption rate will be very high. China's e-commerce market has risen by 30% this year, and the entire market has risen by 30% this year under the condition that last year's e-commerce was good and high-based, mainly Pinduoduo rose by 60%. There must be a brand in such a large plate that can run out, and it is more cost-effective than the original.
In addition, the chaining rate in Chinese mainland is much lower than that in Japan, so the chaining of the sinking market is bound to be a great trend.
Source: Visual China
China Entrepreneur: In the consumer sector, how do you choose investment projects?
Zhu Xiaohu: Let's look at two logics, one is to make a new product or a new brand, whether the single product can achieve a scale of 10 billion, the ceiling must be high enough, and the second is to do a sinking chain, whether the category of the chain is standardized enough to open 10,000 stores.
In addition, all consumer companies today have to talk about profits and grow reasonably while ensuring profits, so the growth rate of some consumer brands may not be as fast as everyone thinks, but it is healthier.
Chinese Entrepreneur: Is the main reason why you are pursuing profits so much now is because you don't have so much financing?
Zhu Xiaohu: First, there is less money that can be invested outside, and secondly, the valuation system is completely different from before, and money has become more expensive. It is not necessarily worth it to go to financing again, it is better to pursue profits and rely on your own (earned) money to develop.
China Entrepreneur: How much has the valuation system of the consumer industry changed compared to the original?
Zhu Xiaohu: 2021 was the highest peak, when the valuation was 10 times the sales in the next 12 months.
"Chinese Entrepreneur": Sales in the next 12 months are also virtual.
Zhu Xiaohu: Yes, it was predicted. The general assumption is that it can grow by 2 times, so the final valuation may be 20 times the company's sales this year. The current valuation is 10 times this year's profit, do you say that the valuation system is not much different?
China Entrepreneur: Why is there such a change? How was this standard established? Who first proposed it, or was it a consensus formed by everyone?
Zhu Xiaohu: The market bubble is too big at the peak, and everyone is chasing those excellent consumption projects, so it is 10 times PS, and all consumption can be done again at that time.
Now, on the one hand, RMB funds cannot be invested in consumption, because they can't withdraw from A-shares, so everyone can only look at profits and pursue annual dividends, in this case, they can only chase 10 times PE.
"Chinese Entrepreneur": Are you more cautious about investing in consumer projects now?
Zhu Xiaohu: I can only give 10 times PE now, whoever wants to take it, if you don't want to, you don't have to talk about it, more than 15 times PE, you can't see it, everyone doesn't waste time. Really good projects, and entrepreneurs want to develop, there are still opportunities to raise funds, but it is true that they should be more cautious in their development.
Talking about going global: The hard work of Chinese entrepreneurs is unmatched in the world
China Entrepreneur: What do you think of Chinese companies going overseas now?
Zhu Xiaohu: There are two core points, one is that China's supply chain advantage is too strong. As long as anything has hardware in it, China's [supply chain] advantage is unmatched, at least 50% cheaper than anything else.
Second, the hard work and fighting spirit of Chinese entrepreneurs themselves is not found in other countries in the world.
Some time ago, India's richest man called on India to work 70 hours a week, and was scolded by Indians.
I went to the United States for an annual meeting in October, and a dollar LP who had voted for us for more than 10 years had a big "996" on his notebook, and he said that it was the first time he had heard 996. I think the hard work of Chinese entrepreneurs is unmatched in the world, and I think that's very important.
Combined, Chinese startups are too powerful to go overseas. Nowadays in the global market, basically Chinese and Chinese "fight", do social, short video to PK with TikTok, do e-commerce to compete with SHEIN, Temu, do games, mihayou, Tencent, Netease are competitors, Southeast Asia, Africa, South America in the field of financial technology, half are Chinese.
Now the warehousing and logistics robots in Indonesia are also Chinese, and Chinese e-commerce companies have to make 300 million orders every day, and 300 million singles of tested robots are invincible in the world. Where else has such a large burst of orders, including lawn mower robots and pool cleaning robots.
China's innovation ability is invincible, for example, Americans like to hunt, shotgun sight is such a small thing, there is a Wuhan company, you can achieve 2 billion revenue, two or three billion profits, the scope is also high-tech. Its laser shines out and can stun the deer. All these things are Chinese, and it is impossible for the Americans to make them.
I think the supply chain is now transferred out of the lowest level, such as the clothing industry chain of Lululemon in Bangladesh and Sri Lanka. Basically, it is difficult for hardware to go to sea, they are all Chinese supply chains, and it is impossible to go to Southeast Asia.
China Entrepreneur: What is the current development trend of GSR investment companies going overseas? In particular, some consumer brands have become more popular in Southeast Asia.
Zhu Xiaohu: In the enterprise service industry, I usually recommend going to Singapore and the Middle East first, and then see if there are opportunities in Europe and the United States. The Middle East and Southeast Asia are relatively receptive to Chinese products, and it is up to the Middle East and Singapore to improve their English proficiency, otherwise the English proficiency of Chinese companies is a big problem. First go to Singapore to practice for half a year, and the founder's English level can communicate with the local people.
As for consumer companies, I suggest that they can directly consider entering Europe and the United States.
Chinese Entrepreneur: This year, Middle Eastern capital is very active in investing in China, as a VC, how do you see this change?
Zhu Xiaohu: The Middle East has its own demands, and he hopes that industries can complement each other, or counter-invest. I think they are learning from some of the Chinese government's anti-investment requirements, such as that if they want to vote for you, they need to go back to the Middle East to build factories to promote local employment and economic development. Therefore, it must be in line with his strategy to have a chance to get their investment.
China Entrepreneur: Are domestic investment institutions likely to consider doing some investment projects to cater to their demands?
Zhu Xiaohu: The key is to see whether your strategy is in line with their demands. Although we also have some LPs in the Middle East, we may not take the initiative to spend more time and energy in the Middle East, which is not our priority at the moment, and we are still more focused on the Chinese market. What's more important is that Chinese entrepreneurs go overseas, not only to the Middle East, but also to the world.
News Hotline & Submission Email: [email protected]