We get asked a lot what the difference is between China’s major tech hubs of Beijing and Shenzhen. Is Beijing or Shenzhen China’s Silicon Valley? Is Shenzhen really only about hardware?
Last week, the team was in Shenzhen for the recent TechCrunch conference. This was a great chance to try to find out. I spent it asking everyone I met: “What is Shenzhen? Is it only all about hardware?”
I learned that Haidian really is China’s Silicon Valley—a tech hub that emerged from universities and concentrates basic research, the country’s biggest VCs, and most of its brand-name internet companies. Shenzhen is something the US doesn’t have—it’s built on factories the way Silicon Valley and Haidian are built on universities.
The Chinese tech boom has built on this factory hub not only with more advanced products and manufacturing techniques but also, more importantly, e-commerce. By making the Chinese consumer the end point for most products, rather than container ships, e-commerce helped Shenzhen to outgrow its low-value export hub origins.
Bottom line: The Shenzhen hub reflects its factory roots—but factories are a good foundation for cutting-edge developments in fields including internet of things, industrial AI, and robotics. The city, however, is weak in basic research, education, software talent, and financing. The ambitious Greater Bay Area plan foresees a more Silicon Valley-like future, but don’t turn up your nose at manufacturing. The factories are a line to a future you can’t see from Palo Alto or Beijing.
The brand names: Looking at who’s headquartered in Shenzhen, there is certainly a hardware theme—with the huge exceptions of software conglomerate Tencent and financial services giant Ping’an. In no particular order, major companies with hardware roots include:
- Huawei (telecoms)
- ZTE (telecoms)
- DJI (drones)
- Aqara (smart home)
Ask someone in Shenzhen about Beijing, and a few ideas come up—basic research, branding, and “business model innovation.” It’s where you’ll find the most companies that compare well with US counterparts:
- Baidu (search, compare Google)
- Bytedance and Kuaishou (social media, compare Facebook and Twitter)
- Meituan and Didi (O2O services, compare Uber)
The tinkerer’s Mecca: The 1980s foundation of the Shenzhen hub was small factories that produced basic electronic components—and the famous electronics market at Huaqiangbei, where those components are traded in a seemingly endless sea of small stalls. The most famous depiction of the vibe of the place is Strange Parts’ make-your-own iPhone video.
Corporate supply chains have long since moved on from the digital bazaar, and the marketplace itself has diversified into areas like wholesale electronics.
But Trouble Maker CEO Henk Werner told us that the market is still essential for startups’ hardware prototyping. Product designers at his hardware accelerator—located on the seventh floor of one of the market’s buildings—have access to two resources available nowhere else in the world: immediate access to parts for prototyping, and market knowledge not taught in electrical engineer programs about what’s available, what’s cost effective, and what will fit into actual production processes. Unsurprisingly, other hardware accelerators, including HAX, are also located in the market. (Readers interested in hardware accelerators should also check out Shenzhen Valley Ventures).
Mike Reed, Mechatronics Engineering Lead at HAX warned that going to the market isn’t always the right approach to prototyping. Sheer size makes it hard to find things, and it can be more effective to use a trusted middleman. It also focuses on standard components—for brand-name or specialty products, Reed said, the answer is, as for everyone else in China, still Taobao.
‘Hardware is software’: Factories are to Shenzhen what Stanford is to Silicon Valley—they still shape its character, but they’re not the whole story. Thomas Goletz, co-founder and CEO of tech community builder MGI, said that the factory hub has drawn in software talent as it matures:
Shenzhen is not hardware, it’s manufacturing. When we talk about manufacturing, we’re talking about e-commerce; when you talk e-commerce, you’re talking software.
At a provincial level, the government of Guangdong expects the e-commerce sector to be nearly twice as large as electronics and IT in its current five-year plan, at RMB 7.3 trillion (about $1 trillion), vs RMB 4 trillion.
Anthony Lawrance, founder of regional news platform Greater Bay Insights, said that regional and national policy is pushing tech into currently low-end factories. Instead of moving textile factories to Vietnam and Bangladesh, Lawrance said, policy-makers hope to keep industry in the region with upgrades to efficiency and value-added.
Additionally, Goletz said: “Hardware is software.” IoT, drones, and robotics are all driven by integrated software, and this software is often developed in Shenzhen. Werner said that companies like Aqara—whose products are sold as Beijing-based Xiaomi’s smart home line—and Tuya (headquartered in Hangzhou) are the future of Shenzhen. Rather than assembling other companies’ designs, these all-in-one firms design and produce complete IoT products lacking only a brand name, capturing far more of the profit.
Nonindustrial giants: Shenzhen’s two biggest companies, Ping’an and Tencent, are not grounded in manufacturing, and Lawrance said that each has created a cluster of its own. Ping’an, originally an insurance company, has pivoted to fintech, he said; its fintech-focused incubator program is “churning out companies.” Tencent, he said, is concentrating AI talent in its own R&D, as well as co-sponsoring an AI park with the Zhuhai government. Chance Jiang, China CEO at collaborative product development platform Wikifactory, mentioned another Tencent-affiliated cluster in nearby Guangzhou, where software companies specialize developing mini-programs for Wechat (headquartered in the city).
Where are the universities?: Beijing’s tech hub spilled out of the south gate of Peking University; PKU and Qinghua remain key drivers of tech in the city, conducting basic research and fueling research-intensive business like computer vision (Mengvii, Sensetime) and the famous Beijing Microsoft Research Lab. But the whole province of Guangdong has only three national key universities—compared to Beijing’s 27—and none in Shenzhen. The government has tried to fill the gap with satellite campuses of international, Hong Kong, and top Chinese universities. But sources agreed that these do not draw top talent.
A 2018 report on AI development by a Qinghua research institute found that Beijing dominates academic talent, with more than twice as many leading researchers as runner-up Xi’an. Shenzhen doesn’t even rate a mention in the report’s academic top ten (although the report does find that Huawei leads the corporate tables).
Turmoil in Hong Kong, Lawrance said, could be an opportunity to shift resources across the border. All of the city’s top universities have Shenzhen campuses, and many students at the hardest-hit (and now closed for the semester) Chinese University are mainlanders and have evacuated to Shenzhen.
If it can’t graduate research talent, the city also hopes to buy it—like in many Chinese city governments caught up in what local media call a “talent war,” Shenzhen’s pays people with advanced degrees to move to city—with subsidies ranging up to RMB 1.5 million for Phd holders, and higher for experienced researchers.
Financing: Shenzhen has a healthy VC network and its very own stock exchange, but many Shenzhen tech companies opt out. Jiang said that south Chinese entrepreneurs often prefer to bootstrap to medium size rather than trading equity for growth—the typical local tech firm, he said, is built around a sales opportunity and grows off revenue without ever going to VCs or the markets. A few strike a rich vein and grow big, but most are content at the middle.
From an official perspective, Lawrance said, this is a weakness—the state wants champions, not small-is-beautiful SMEs. Officials have laid out plans to encourage more companies to raise money for growth.
The Shenzhen VC world is smaller than Beijing’s—and nearly all USD funds with a China presence are in the capital—but some sources described VCs who have experience at companies like Huawei and are more patient on long hardware development times. Yongxi Li, investment manager at Tamarace Capital, told TechNode that:
Beijing venture capital has a state-owned background, and likes heavy capital investment and rapid IPO listing. Shenzhen venture capital is looking forward to the development of intelligent manufacturing projects because of the developed manufacturing industry in the pearl river delta.
Other sources, however, said over Wechat that Shenzhen VCs could be even more impatient—many have individuals as limited partners, who demand returns on their money in as little within five to seven years.
What about the GBA?: Shenzhen is at the heart of one of Beijing’s favorite initiatives—the Greater Bay Area, meant to knit 11 cities and a population of 80 million into one border-spanning supercity. Policy details are thin, and mostly cover transportation infrastructure. But Lawrance said the plans speak to a very familiar question Shenzhen’s future: Is it only about hardware?
Shenzhen, he said, is getting expensive, and many factories have to move out. The question is whether they stay in the region—plans call for them to move to Dongguan and the suburbs of Guangzhou—or move out of China entirely. Policy-makers, Lawrance said, don’t want to follow the Silicon Valley path, which left hardware behind—they want to keep it while mastering software.
For now, Shenzhen’s tech hub exists on all levels of the value chain at once. But to keep moving up, it will need to solve the financing, research, and education problems—and to avoid losing its hardware edge, it will need to figure out how to square rising salaries and rents and low-value production. I don’t know if we can imagine it doing both.