Looking at Chinese tech companies' expansion overseas through the Manus craze: Can Singapore be the best first stop?

Recently, Manus, an AI Agent company that gained global popularity through a demonstration video, announced that it had relocated its headquarters to Singapore and made significant adjustments to its domestic team. Only core technical staff were moved to the headquarters, while the rest of the employees either remained in China or were laid off. This move not only highlights Singapore's unique advantages in coping with international political and investment pressures but also provides the latest evidence for the strategy of "using Singapore as the first stop for going global."

In today's increasingly complex globalization era, how can Chinese technology companies go global? Should they first establish a foothold in the domestic market and then expand overseas, or directly target the main battlefields in Europe and the United States? In this speech, Guo Wei, the founder of Beluga Open Source and a senior open-source evangelist, will personally share his real experiences of polishing products and expanding into the global market after deeply cultivating the Chinese market. He will also share his insights on going global based on his on - the - spot research in Singapore and the United States.

Through a comparative analysis of Path A (China → Southeast Asia → Global) and Path B (United States → Global), we will delve into key issues such as software product localization, PMF verification, brand building, and market connection, helping Chinese technology entrepreneurs find suitable "ways out" and "paths forward." If you are starting a business or your company is at a crossroads in product globalization and market selection, the insights in this article will provide valuable references for your decision - making.

At the AICon Global Artificial Intelligence Development and Application Conference hosted by InfoQ, Guo Wei, the CEO of Beluga Open Source and the former president of the Beijing chapter of TGO Kunpeng Club, gave a special speech titled "Should Chinese technology companies use Singapore as the first stop when going global?" He shared a practical roadmap from the "Jinggangshan" of Chinese software to globalization, teaching us how to make correct market choices and organizational focus when resources are limited.

Highlights of the content

Instead of simply believing that "the US market is the largest," understand the differences in the difficulty of PMF verification in different regional markets and improve the ability to implement localization in a global perspective.

Learn how to build a product - oriented company in China through real - world cases and use non - traditional approaches such as open source and communities to build brands and user ecosystems.

The following is the transcript of the speech (edited by InfoQ without changing the original meaning).

1 Globalization Path Selection: A vs B

On the road of global entrepreneurship, I have profound experiences and lessons to share. Many people are enthusiastic about starting global businesses when starting a venture, but only find out after entering this field that so - called globalization cannot be achieved overnight. It requires localization in each country. From the product development stage, the needs of different countries vary greatly.

Take the recently popular large - model applications in China as an example. For products like Dify, its ecological connection in China needs to be integrated with local platforms such as Doubao, while in overseas markets, it needs to be integrated with products like ChatGPT. The preferences and usage habits of product users also differ.

This difference is not only reflected in the technical ecosystem but also in the product usage methods. Take the Southeast Asian and Japanese markets we are exploring as an example. The Japanese market has much higher requirements for documents and product details than the European and American markets. In China, users have relatively less strict requirements for the interface, while Japan is more demanding in this regard.

In terms of marketing, different countries have different approaches. In China, we are used to participating in various industry conferences such as QCon and AICon. However, in the United States, in addition to offline activities, more online traffic promotions are needed. In the Japanese market, since most local enterprises conduct business through agent channels, we need to cooperate more with agents.

Moreover, the compliance requirements of each country are different. China has various compliance requirements such as independent innovation in information technology, the United States has its own industry - specific compliance standards, and Europe has many compliance regulations such as GDPR. In terms of personnel management, for start - up enterprises, at the beginning, founders or co - founders must be present in the target country in person. Ordinary employees have difficulty mobilizing company resources to develop the local market.

Therefore, we cannot start a full - scale global business at the beginning. Take Dify as an example. Its founder also needs to travel frequently between China and the United States. Generally speaking, for consumer - oriented products, the process from 0 to 1 in a country may take about two years; for enterprise - oriented products, it takes two to four years to deeply cultivate a country's market. This means that we should clearly recognize that globalization cannot be achieved overnight and requires a lot of time and effort.

In the development path of global enterprises, I think there are mainly two paths to choose from. The first path is to start from the Chinese market, then gradually expand to Southeast Asia, Japan, Europe, and finally enter the US market. The second path is the opposite, starting from the US, Europe, Japan, and Southeast Asian markets and finally entering the Chinese market.

Why are there only these two paths? Because in the early stage of entrepreneurship, we can only choose one country as the starting point, and the market of this country must be large enough, with sufficient opportunities and growth space. Globally, currently only the United States and China meet these conditions.

Although the Japanese and European markets seem to have great potential, they have their own particularities. The Japanese market is very different from the US and Chinese markets. If our products are customized only for the Japanese market at the beginning, they may not be suitable for the US and Chinese markets. Most customers in the European market have a similar acceptance of new technologies as those in North America, but their customer groups and market understanding are not as active as the emerging customer groups in the United States and China.

For start - up companies, finding early adopters is crucial. These early customers need to have a broad technological vision, be able to accept advanced technologies, and iterate products with us.

Currently, only the customer groups in the United States and China can meet these requirements. They are in the first echelon both in terms of paying customers and open - source users. Therefore, when choosing the starting point, we don't need to consider other countries because the development of other countries in the fields of AI and data lags behind that of the United States and China.

Why can't we directly enter the US market after completing the Chinese market? Generally, there are certain differences in product - market fit (PMF) between the Chinese and US markets. After completing the Chinese market, we need to gradually strengthen our connection with overseas markets and then march towards the US market.

Otherwise, we may fall into the dilemma of fighting on two fronts. On the contrary, if we start from the US market and then enter the Chinese market, there will be another logic. The US market is relatively a "simple mode," with abundant funds and friendly customers. The Chinese market is relatively more complex, with more difficult scenarios.

If we directly enter the Chinese market after completing the US market, we may get stuck in a quagmire and slow down the development pace. Therefore, for many entrepreneurs, after completing the US market, they often choose to expand into the European, Japanese, and Southeast Asian markets first and then consider entering the Chinese market.

Actually, we also have relevant experience and lessons. At the beginning, we tried to invest resources in both China and the United States at the same time, but soon found that the core resources on both sides were insufficient, resulting in a wavering PMF.

Take Apache DolphinScheduler as an example. In the United States, its biggest competitor is Apache Airflow, and both are scheduling tools. In Asia, DolphinScheduler ranks first, while in the Americas, Airflow dominates. In order to compete with Airflow, we tried to make DolphinScheduler compatible with workflow code and generate various processes with Python, which actually destroyed the original advantages of DolphinScheduler.

We spent about half a year contacting and communicating with US users, and found that fighting on two fronts was not as good as focusing on one front. Later, when developing Apache SeaTunnel, we learned from the lesson and found that if we wanted to target the US market, the drag - and - drop operation mode in Asia might not be suitable for North American users. North American users prefer to use code.

Therefore, when incubating Apache SeaTunnel, we completely adopted a code - based approach and successfully won many users and customers in the US market, such as JPMorgan Chase, the largest commercial bank in the United States. This shows that when doing PMF, we cannot waver. If we have adopted the Asian system, it will be different from the North American market; if we aim at the North American market, we should not take care of the Chinese market because these two systems are completely different, and we must make a firm choice between the two.

The US market is a market with high input and high output. Medium - level input may lead to no output. We once invested a team of two or three people in the United States for marketing and community building, but failed to meet the targets for several consecutive quarters.

This is not because they didn't work hard, but because locally, the number of marketing personnel of competitors like Airflow far exceeded ours, and their investment was also much higher than ours. The US market is a very fair market. To do well, the investment must be sufficient. If the investment is small, it may be as if there is no investment at all. It is better to operate remotely from China.

Take TiDB as an example. It took root in the US market for four years. At first, it was just a side - job. Later, it found that it didn't work, so it increased the investment, from less than ten people to dozens of people, and now it is close to 100 people. Only then did the market gradually develop.

Therefore, the US market is a market that requires full - hearted investment, not a place that can be done casually. If the investment is less than 10 people, I think it is very difficult to get any output. Small investment is basically equivalent to no investment. This is the lesson we learned from fighting on two fronts.

Chairman Mao once said, "Concentrate superior forces to annihilate the enemy one by one." This also applies to business competition. In a specific field, we must ensure that our resource investment exceeds that of our competitors so that we can stand out and make breakthroughs in the fierce market competition.

Take the Chinese market as an example. Many enterprises have taken root and developed here. The same is true in the United States. The ToB market in the United States has broad prospects and great potential. The competition in each ToB market is very fierce, and the investment cost is also high.

If you want to gain a share in this market, you must reach the same level of resource investment as your competitors, or even exceed them. Only in this way can you have the possibility of gaining an advantage in the competition and winning market share.

2 The "Hell Mode" and Strategic Value of the Chinese Software Market

As a Chinese person, I know that deep in everyone's heart, they want to make a name for themselves in the domestic market first. However, for those of us in the software industry, the reality is not so. Many people think that there are few opportunities in the Chinese software market. Objectively analyzed, the Chinese software market is indeed not mature enough.

This is not only reflected in the eyes of investors but also in the communication with many executives of overseas companies in China and domestic peers. The consensus is that the Chinese software market is not mature enough.

In terms of overall revenue, the Chinese software market accounts for only 2% to 3% of the global market. Although China has a large population base and seemingly a vast market space, from a global perspective, in terms of US dollars, the share of the Chinese software market is negligible.

From the perspective of entrepreneurs, the Chinese software market is characterized by "less money and more work," with various requirements emerging one after another.

At the same time, the customer group in China is not large. Many people may think this is inconsistent with the fact, considering that there are many central and state - owned enterprises in China. However, it should be clear that central and state - owned enterprises need not only pure software but also overall solutions and projects. The part that can be truly converted into pure software revenue is very limited.

In the field of pure software, based on my experience, the price of general software does not exceed 3 million RMB. Once it exceeds this price, it is no longer pure software but requires customization, implementation, and on - site services.

3 million RMB is equivalent to 400,000 US dollars. In the United States and overseas markets, the purchase of software for 400,000 US dollars, even the subscription fee, is not considered high. From a global perspective, the average customer price in the Chinese software market is relatively low. And due to the relatively low maturity of software enterprises' own software, the satisfaction of customer needs is relatively low, and the corresponding customization requirements are serious.

If it is open - source software, the competition is extremely fierce. Others may use the open - source version to provide similar services; if it is not open - source, other commercial companies will compete with you on price. In addition, domestic cloud providers will maliciously lower prices. A project originally worth millions may be reduced to hundreds of thousands or even tens of thousands.

Overall, the Chinese software market is objectively not mature enough. To tell a joke, someone said that doing software in China is like starting an electric vehicle business in Africa. You painstakingly build a top - notch Tesla, only to find that all your competitors are second - hand Volkswagen Beetles from the 1990s.

Although these second - hand cars are old, they are cheap and can meet customer needs. Moreover, there are no charging stations in Africa, only gas stations. The entire market environment is incompatible with electric vehicles. This is the current situation of software entrepreneurship in China.

However, every coin has two sides. Although the US market is relatively mature, not all entrepreneurs can easily enter it. The US market requires high input and high output, and may require dozens of times more investment than the Chinese market. If your financing amount is limited, you may not be able to bear the cost of the US market. Although the Chinese market is not mature enough, it also has its own unique advantages.

First of all, the Chinese market has great potential, like an uncultivated wasteland. If your core technology is strong enough and leading globally, then in the Chinese market, customers will come to you actively. In China, if you can achieve technological leadership in a certain field, it is easy to stand out.

Secondly, the competition in the Chinese market is relatively small. Compared with the United States, the number of software manufacturers in China is limited, and most manufacturers lack core competitiveness. If your product is excellent enough, it is easy to gain an advantage in the Chinese market.

Thirdly, another unique advantage of the Chinese market is the large number of developers with relatively high basic levels. China is in the period of demographic dividend, with a large number of developers whose levels are not inferior to those overseas. The promotion speed of open - source software in China is extremely fast, far exceeding that in the United States and other countries.

This provides a good foundation for the development of Chinese software. If the open - source model is adopted, Chinese software enterprises can quickly accumulate users and developers in the domestic market and form a strong community support. Then, they can promote these open - source projects to overseas markets through the commercial version, forming a unique competitive advantage.

Fourthly, another advantage of the Chinese market is the support of national policies. The existence of "independent innovation in information technology" provides protection for domestic software enterprises, enabling them to avoid the impact of mature foreign software.

This provides an opportunity for Chinese software enterprises to develop from 0 to 1.