Zhang Genghua, Source Code Capital: The Answer to Going Global Is North America

By Muxin Xu

In a March 29 article published by An Yong Waves ("暗涌Waves") titled A Globalization Guide for Chinese Founders, we noted two core hotspots in the North American market right now — cross-border e-commerce and AI. The aggressive expansion of China's "Little Four Dragons" of cross-border e-commerce in recent years, combined with the explosive rise of AI as the venture capital world's hottest theme, means that despite an increasingly strict policy environment, North America remains the most closely watched region for Chinese investors focused on globalization.

In 2015, Genghua Zhang began paying attention to going global, and his first stop was a research trip to North America. Eight years later, as an executive director at Source Code Capital, he returned.

In the intervening years, investment directions had shifted repeatedly. In 2015, North America was still the origin of "Copy to China," and globalization opportunities mainly appeared in business model innovation. Starting around 2017, the sequence flipped to "Copy from China" — mature domestic models proliferated into emerging markets like India and Southeast Asia, giving Zhang the chance to invest in Zomato (India's Yelp equivalent) and Grab (the region's essential ride-hailing app). When COVID hit, the cross-border e-commerce market heated up rapidly, and Zhang pivoted to catch export-oriented projects like PatPat (maternal and infant e-commerce) and Urbanic (a fast-fashion brand targeting India).

In Zhang's view, the shifting winds of globalization investment over the past decade share one core driver: the wave of business model innovation brought by mobile internet technology. Going forward, opportunity will be granted by AI. This is Source Code's current answer in the globalization space: North America, AI.

When An Yong Waves reached out to Zhang about "globalization," he was at CES. While most Chinese investors around him saw spreading opportunity, this veteran of the going-global track also spotted hidden risks in what is now the biggest trend — for instance, how the emergence of phenomenon-level companies like Shein and Temu has actually squeezed opportunities for new entrants.

But just as going global has become a "must-do" rather than a "nice-to-have" for companies, looking beyond borders for commercial opportunities ahead of entrepreneurs has become a "must-do" for investors too.

Below are key points from Zhang:

1. Going global and AI are the two consensus themes in China right now. Going global is mandatory, not optional. The reason can be seen in what happened to Japanese companies after their "Lost Three Decades": many went abroad, and the share of international business rose quickly.

2. Every shift in Source Code's globalization investment focus has been based on a three-dimensional choice: region, sector, and stage. Weighing these three factors produces an optimal solution. Today, we believe that optimal solution is: North America, smart hardware.

3. From an investment perspective, since the US entered its rate-hiking cycle, capital costs have climbed higher and higher. Many startups that previously generated no profit and survived on heavy financing now face an increasingly harsh funding environment. The Silicon Valley Bank incident in early 2023 was the concentrated explosion of this tension. "Funding difficulty" will persist for some time, and investment institutions are paying more attention to the profitability potential of projects.

4. In North America, Chinese companies' advantage lies in direct-to-consumer personal-use scenarios. These products are "harmless" — low political sensitivity — because they help American consumers improve personal life efficiency (say, not having to sweep floors themselves). But if these same companies move into B2B scenarios, like commercial floor-cleaning robots, they hit sensitive political issues, because such products directly replace local American workers.

5. Opportunities in AI hardware have actually been around for a while — robotic vacuums, 3D printing, and so on. It's just that with the emergence of large language models, people are now thinking about new opportunities at the intersection of AI hardware and LLMs. At the product level, there are traffic-entry-type products represented by smart glasses and smart earbuds. On the underlying technology side, some small language models that Meta is developing could be loaded directly into these two product categories, combined with edge computing for real-time interaction without needing connectivity. If these two categories truly become global next-generation traffic entry points, it could be a hundred-billion-dollar opportunity.

6. Comparing investment tendencies at home and abroad: while China is pouring massive resources into hard-tech sectors, overseas institutions — North America especially — are increasingly "moving from solid to virtual," preferring soft-tech startups like AI software. They've discovered they can't compete with China on hardware mass production and supply chain infrastructure.

7. I once invested in several great companies in Southeast Asia's emerging markets — Grab in Southeast Asia, Zomato in India. But looking back, these projects succeeded because they caught the mobile internet wave. That was a global-level wave that created opportunities in every region, not something unique to Southeast Asian markets.

8. There's a perpetual debate in going-global circles: invest in a pure overseas/external-cycle company, or one that balances internal and external cycles? I prefer companies that are global from day one. For example, Source Code invested in a light electric commercial vehicle company called Fest. The founder is Chinese but based in Turkey; headquarters are in Singapore; and senior management barely includes any Chinese. A company with this profile can integrate smoothly into overseas dollar circles and raise money from the Middle East, the US, and Europe.

9. In 2023, the emergence of phenomenon-level Temu and Shein was actually not good news for startup cross-border e-commerce players. That's because full-service托管 (full-service merchant models) is the trend, but behind full-service is the platform replacing the entrepreneur's work. Previously, entrepreneurs found factories, helped with product design and market research, then sold on AliExpress. Now Temu can do all of that, so factories can connect directly with Temu — no need for "middlemen taking their cut." This resembles the domestic situation after platform monopolization: space for new entrants gets squeezed.

10. In North America, a significant amount of consumption is returning to offline scenarios. First, because online channel traffic is in a "transitional gap" — traffic costs on the previous mainstream platforms (Google, Facebook) keep rising, while new platforms like TikTok are still in early commercialization stages — so entrepreneurs find offline offers better return on investment. Additionally, large offline retail chains still best fit local consumption patterns. Consumption experienced brief rapid onlineization during COVID, but has gradually returned offline since the pandemic ended.

Image source: ic photo