Real business stories and field notes from Shenzhen.
Founders, operators, factories, markets, and the decisions shaping how business gets built across Shenzhen and the Greater Bay Area.
FOUNDER STORIES
View all-
She Spent 15 Years Inside One of China's Most Competitive Markets. Here's What It Actually Taught Her.
Robin LuoLiang Xiaohua didn't plan to build a business in Shuibei. But 15 years in, she's still here, and most people who started when she did are not.
She Spent 15 Years Inside One of China's Most Competitive Markets. Here's What It Actually Taught Her.
Robin LuoLiang Xiaohua didn't plan to build a business in Shuibei. But 15 years in, she's still here, and most people who started when she did are not.
-
From Power Systems to Smart Security: How Leo Built a 20-Year Business in Shenzhen
Robin LuoLeo’s 20 year journey spans from power systems to smart security exports. Quietly adapting to market shifts, he now focuses on overseas growth balancing industrial roots with evolving consumer needs. A grounded story of persistence and reinvention from Shenzhen.
From Power Systems to Smart Security: How Leo Built a 20-Year Business in Shenzhen
Robin LuoLeo’s 20 year journey spans from power systems to smart security exports. Quietly adapting to market shifts, he now focuses on overseas growth balancing industrial roots with evolving consumer needs. A grounded story of persistence and reinvention from Shenzhen.
-
In a City That Doesn’t Sleep, He Built a Brand Around Rest
Robin LuoIn fast-paced Hong Kong, David built a sleep brand inspired by his mom’s insomnia and his own restless nights. What began as a quiet product test is now a growing global business, built slow, clear, and strong, without sacrificing rest.
In a City That Doesn’t Sleep, He Built a Brand Around Rest
Robin LuoIn fast-paced Hong Kong, David built a sleep brand inspired by his mom’s insomnia and his own restless nights. What began as a quiet product test is now a growing global business, built slow, clear, and strong, without sacrificing rest.
Hardware & Supply Chain
View all-
This Engineer Left DJI to Help Hardware Startups Survive the 0–1 Phase
Robin LuoIf your hardware idea is still on paper or halfway through a rough prototype, Carson Lin has a message for you: “The most expensive problems in hardware aren’t in R&D or manufacturing. They’re in all the wrong assumptions made in between.” Carson spent 7 years at Foxconn, Huawei, and DJI, working across the full NPI pipeline—from product definition and EVT/DVT builds to production ramp and post-launch support. Now based in Shenzhen and Dongguan, he runs a small technical supply chain team that specializes in the hardest part of the process: turning ideas into real, testable, buildable products. Why Thinking About Mass Production Too Early Can Kill Your Hardware Product Many founders think about production too early and validation too late. According to Carson, that’s where most hardware projects start to go sideways. “We tell every client: if your EVT or DVT stage isn’t painful, you’re probably doing it wrong.” He’s not exaggerating. Carson’s team spends most of their time catching problems that should’ve been discovered months earlier: Designs that skip DFM (Design for Manufacturing) Products with vague or incomplete reliability standards Early builds with no structured change log or review checkpoints And when those problems only show up during tooling or mass production, the costs multiply. What Successful Teams Do Differently From the projects that went smoothly, Carson sees five key patterns: 1 They study how users actually use the product. Not just what users want, but how they behave. “Use scenarios change everything,” he says. “People abuse hardware in ways you never expect.” 2 They treat EVT/DVT as failure-finding missions. The goal isn’t just to build something—it’s to break it. Especially in reliability testing, Carson’s team pushes hard to simulate edge cases early. 3 They document everything. Every design change. Every supplier tweak. Every BOM update. Without this, “a single missed update can create weeks of chaos.” 4 They pause to review before every stage transition. No known critical issue should cross from DVT to PVT. No exceptions. 5 They build trust between R&D and NPI. “Engineering and manufacturing speak different languages. NPI is the translator. If that bridge breaks, the whole thing breaks.” Why Carson Built His Own CNC and Pilot Line One of the most surprising things about Carson’s team is that they’ve invested in their own CNC shop and pilot-scale test line. “It wasn’t a strategy, it was survival.” Early clients were mostly startups. They needed fast, iterative prototyping. But traditional factories wouldn’t cooperate without firm volume forecasts. After too many delays, Carson decided to build in-house. Now, they can turn around small-batch mechanical samples in 1–2 days, with full control over scheduling, process tuning, and IP confidentiality. “We only do one product in the workshop at a time,” he says. “That’s how we protect client privacy and flexibility.” They Don’t Build Their Own Hardware And That’s the Point Interestingly, Carson’s team has no plans to launch their own products. “We want founders to trust that we’re here to help—not compete. That’s why we don’t do our own hardware. Instead, we’re building some general-purpose testing modules that can lower setup costs for others.” What Overseas Teams Often Miss Carson’s been working more and more with international startups lately, especially from the US. One thing he’s noticed: “Overseas teams are often great at research and have clear product goals. But they’re slower to iterate. Chinese teams move fast and adapt quickly—but sometimes skip critical validation.” His team adapts to both. With overseas clients, they emphasize clear planning and detailed feedback loops. With Chinese clients, they move fast, solve problems on the fly, and build in flexibility. 3 Things to Clarify Before You Build in China For overseas founders looking to bring a smart hardware idea to China, Carson offers three critical questions to answer first: 1 What’s the core feature you can’t compromise on? Defer the rest to later versions. 2 What’s your real cost ceiling for production? Factor in testing, packaging, logistics—not just BOM. 3 Can your idea be built with existing supply chain technology? If it needs custom processes or unproven suppliers, timeline and budget will both stretch. A Different Kind of Shenzhen Supply Chain Team Carson doesn’t call his team a factory. Or a design house. Or a trading company. They’re just a group of engineers who love helping good hardware ideas become real. “If you know what you want to build, we can help you figure out how to actually make it step by step, and faster than you think.”
This Engineer Left DJI to Help Hardware Startups Survive the 0–1 Phase
Robin LuoIf your hardware idea is still on paper or halfway through a rough prototype, Carson Lin has a message for you: “The most expensive problems in hardware aren’t in R&D or manufacturing. They’re in all the wrong assumptions made in between.” Carson spent 7 years at Foxconn, Huawei, and DJI, working across the full NPI pipeline—from product definition and EVT/DVT builds to production ramp and post-launch support. Now based in Shenzhen and Dongguan, he runs a small technical supply chain team that specializes in the hardest part of the process: turning ideas into real, testable, buildable products. Why Thinking About Mass Production Too Early Can Kill Your Hardware Product Many founders think about production too early and validation too late. According to Carson, that’s where most hardware projects start to go sideways. “We tell every client: if your EVT or DVT stage isn’t painful, you’re probably doing it wrong.” He’s not exaggerating. Carson’s team spends most of their time catching problems that should’ve been discovered months earlier: Designs that skip DFM (Design for Manufacturing) Products with vague or incomplete reliability standards Early builds with no structured change log or review checkpoints And when those problems only show up during tooling or mass production, the costs multiply. What Successful Teams Do Differently From the projects that went smoothly, Carson sees five key patterns: 1 They study how users actually use the product. Not just what users want, but how they behave. “Use scenarios change everything,” he says. “People abuse hardware in ways you never expect.” 2 They treat EVT/DVT as failure-finding missions. The goal isn’t just to build something—it’s to break it. Especially in reliability testing, Carson’s team pushes hard to simulate edge cases early. 3 They document everything. Every design change. Every supplier tweak. Every BOM update. Without this, “a single missed update can create weeks of chaos.” 4 They pause to review before every stage transition. No known critical issue should cross from DVT to PVT. No exceptions. 5 They build trust between R&D and NPI. “Engineering and manufacturing speak different languages. NPI is the translator. If that bridge breaks, the whole thing breaks.” Why Carson Built His Own CNC and Pilot Line One of the most surprising things about Carson’s team is that they’ve invested in their own CNC shop and pilot-scale test line. “It wasn’t a strategy, it was survival.” Early clients were mostly startups. They needed fast, iterative prototyping. But traditional factories wouldn’t cooperate without firm volume forecasts. After too many delays, Carson decided to build in-house. Now, they can turn around small-batch mechanical samples in 1–2 days, with full control over scheduling, process tuning, and IP confidentiality. “We only do one product in the workshop at a time,” he says. “That’s how we protect client privacy and flexibility.” They Don’t Build Their Own Hardware And That’s the Point Interestingly, Carson’s team has no plans to launch their own products. “We want founders to trust that we’re here to help—not compete. That’s why we don’t do our own hardware. Instead, we’re building some general-purpose testing modules that can lower setup costs for others.” What Overseas Teams Often Miss Carson’s been working more and more with international startups lately, especially from the US. One thing he’s noticed: “Overseas teams are often great at research and have clear product goals. But they’re slower to iterate. Chinese teams move fast and adapt quickly—but sometimes skip critical validation.” His team adapts to both. With overseas clients, they emphasize clear planning and detailed feedback loops. With Chinese clients, they move fast, solve problems on the fly, and build in flexibility. 3 Things to Clarify Before You Build in China For overseas founders looking to bring a smart hardware idea to China, Carson offers three critical questions to answer first: 1 What’s the core feature you can’t compromise on? Defer the rest to later versions. 2 What’s your real cost ceiling for production? Factor in testing, packaging, logistics—not just BOM. 3 Can your idea be built with existing supply chain technology? If it needs custom processes or unproven suppliers, timeline and budget will both stretch. A Different Kind of Shenzhen Supply Chain Team Carson doesn’t call his team a factory. Or a design house. Or a trading company. They’re just a group of engineers who love helping good hardware ideas become real. “If you know what you want to build, we can help you figure out how to actually make it step by step, and faster than you think.”
-
How to Work Shenzhen’s Supply Chain: A Field Guide for Overseas Founders
Robin LuoCharley Shen has spent the last 15 years inside the supply chains that power global hardware, starting from components, moving through sourcing, and now helping overseas founders scale full-stack hardware products from Shenzhen.
How to Work Shenzhen’s Supply Chain: A Field Guide for Overseas Founders
Robin LuoCharley Shen has spent the last 15 years inside the supply chains that power global hardware, starting from components, moving through sourcing, and now helping overseas founders scale full-stack hardware products from Shenzhen.
-
Before the Company Exists: Samuel and the Work That Comes First
Robin LuoSince last year, Samuel has been moving between the US and Shenzhen while building early hardware projects. This story follows what shapes hardware ideas before a company exists, cost, iteration, and the decision to show up in person.
Before the Company Exists: Samuel and the Work That Comes First
Robin LuoSince last year, Samuel has been moving between the US and Shenzhen while building early hardware projects. This story follows what shapes hardware ideas before a company exists, cost, iteration, and the decision to show up in person.
China Business Reality
View all-
Three Things That Blindside Foreigners Running a Business in China
Robin LuoMost foreigners who run into trouble in China hit one of three problems: a labor dispute they can't resolve, a visa transfer that gets stuck, or a tax bill nobody warned them about. This is what each of those looks like from the inside.
Three Things That Blindside Foreigners Running a Business in China
Robin LuoMost foreigners who run into trouble in China hit one of three problems: a labor dispute they can't resolve, a visa transfer that gets stuck, or a tax bill nobody warned them about. This is what each of those looks like from the inside.
-
Inside a VC’s Mind: AI, Hardware, and What Actually Gets Funded
Robin LuoAfter years working across Africa and the Middle East, Michelle returned to China and now moves between Shanghai, Beijing, Shenzhen, and Silicon Valley. In early-stage investing, she pays close attention to founder clarity, team conviction, and whether a product is built for short-term buzz or long-term traction. 1. When an AI project first lands on your desk, what are the fastest reasons you decide not to spend more time on it? To be honest, I rarely pass on a project purely out of instinct. Most of the time, I prefer to speak with the founders or the founding team before making any decision. Many AI startups today have pitch decks that are quite abstract. They’re filled with new concepts and technical jargon, and sometimes investors don’t fully understand what the team is actually building just from the slides. Misunderstandings often come from storytelling rather than from the product itself. So I usually schedule a 45-minute to one-hour conversation with the founders to understand clearly what they’re building and how they think about the product. After that conversation, the key factor for me becomes the founding team, especially the founder’s motivation. I care deeply about why you’re in this business. Why did you decide to build this product? Entrepreneurship is abstract and extremely difficult. It requires conviction and resilience. If someone is building a company because their Stanford roommates raised money from a16z, or because they want fame, reputation, or to become a KOL, that’s when I would pass. There are simply too many hardships in entrepreneurship. If the motivation is superficial, it won’t sustain the company. 2. At the early stage, how would you rank the problem being solved, the founding team, and the technology or solution and why? At the early stage, I rank the founding team first. The reason is simple: products can pivot, technology can evolve, but the founding team determines whether the company survives and adapts. The problem being solved is important, of course. But even a strong problem won’t matter if the team cannot execute or lacks conviction. Technology or the solution is necessary—especially in AI. But at an early stage, technology alone is not enough without the right team and mindset behind it. Entrepreneurship is abstract and uncertain. What I look for is whether the team truly understands what they’re doing and whether they’re prepared for a long-term challenge. 3. What is the single strongest signal that makes you want to keep digging into a project? The strongest signal is genuine conviction from the founder. I want to understand why you’re here and why you care about this problem. I want to see whether you’ve thought deeply about the product and whether your motivation is internal rather than external. If I feel clarity and a strong sense of purpose from the founder, I’m much more willing to continue exploring the opportunity. 4. Could you briefly introduce yourself and what you focus on as a VC today? I’m still a relatively novice VC investor. I graduated from college in 2022, and over the past three and a half years, I worked in Africa and the Middle East. During that time, I invested on behalf of a global company, working within the strategy, venture, and investment arm of a corporate venture capital setup. I returned to China at the end of last year and joined a venture capital firm around mid-November, so it’s been less than six months in this role. Today, I invest in AI software and consumer electronics / AI-powered hardware, mostly in China or with Asian founders building for global markets. That’s also why I split my time frequently between Shanghai, Beijing, and Shenzhen. Many software companies are in Beijing, while consumer electronics and hardware companies are concentrated in Shenzhen. I also spend time between China and the United States, especially Silicon Valley. There are founders who grew up there, studied there, or worked there, and now want to combine their understanding of North American or European markets with Asia’s efficient supply chain to build something global. 5. What kinds of consumer AI–related projects are you seeing most often and which ones actually excite investors? It’s a very interesting time for VCs. We’re seeing a wave of impressive founders coming from companies like Google, Apple, Amazon, ByteDance, and Tesla, deciding to start their own companies. There’s a wide range of software and hardware projects, some of them are things I personally never imagined people would build. On the software side, one major category is AI social apps. Founders are reimagining how AI might reshape interpersonal interaction, job hunting, dating, making friends, and communication in general, once AI becomes a variable inside social behavior. But social apps are also extremely difficult to build. We also see many productivity tools, especially targeting what people call OPC—one-person companies. AI can dramatically increase the efficiency of an individual coder or artist, so productivity products are a major theme. On the hardware side, it feels like AI could reshape almost everything in daily life from laptops to personal assistants. Founders are imagining AI rebuilding many of the tools we interact with every day. What excites investors isn’t AI as a label. It’s whether the product direction makes sense and translates into real usage, especially when founders can connect the technology to a strong consumer experience. 6. Which consumer AI categories do founders most overestimate? Investors currently have strong appetite for almost anything AI-related, it often comes down to individual taste. If you pitch the right person, you can find funding. Personally, I’m more cautious about products that sit too close to what large internet platforms can already do. In software especially, ByteDance can release something overnight that’s ten times better than what a two-person team is building. Some founders are also building features very close to what Gemini already includes in student or bundled plans. If startups try to charge $10–15 per month for something similar, they risk being squeezed by large platforms offering one-stop solutions. Founders increasingly understand they need to know the capabilities and boundaries of tech giants and avoid building too close to their core. Hardware is different. Hardware behaves more like consumer goods. AI agents may disrupt SaaS, but hardware is closer to consumer electronics and consumer branding. In many cases, it’s a consumer brand first, and AI-powered hardware second. For consumer products, opportunities continue across generations. Each generation grows up with different devices and ways of entertaining, learning, and communicating. The opportunity remains; the question is which startups can integrate AI software effectively into hardware and supply chain realities. 7. What are the biggest differences when evaluating AI hardware versus AI software at an early stage? The moat functions very differently. There are also different types of AI hardware. For example, in industries like 3D printing, being an early mover can create meaningful advantage. But if you’re building something like an AI camera or AI recording device, you’re entering a space with a very mature supply chain in Shenzhen. In that case, hardware itself is not the moat, the moat lies in the AI software or agent layer built on top. With AI software and agents, the issue is often the lack of moat. Open-source solutions appear quickly on GitHub, and similar tools can be replicated rapidly. So for consumer software and agents, you need a highly niche and accurate entry point, true product market fit and a strong ability to tell your story. When code moats weaken, trust becomes the moat. Why should users trust your team and your product? That’s where brand and storytelling matter more than people expect. We’ve seen examples like Manus. Many products try to imitate what it does, but it built advantage through strong consumer experience, a clear early positioning, and effective brand storytelling as a first mover. 8. What kinds of founding teams do you personally trust more? I trust teams with clear ownership and complementary strengths, teams that can genuinely connect technology to product. In AI, technical ability matters, but product instinct matters just as much. The best founders identify pain points users may not even articulate clearly, and then express those needs through the product. I also look at whether founders understand how competitive and fast-moving this market is and whether they’re building for the long term, not just chasing a short-term trend. 9. How would you describe the founder mindset in Shanghai, Beijing, and Shenzhen? Interestingly, I don’t see a huge difference in mindset. You do see different founder profiles shaped by the environment, more software and AI agents in Beijing, more hardware and consumer electronics in Shenzhen. But everyone recognizes the importance of AI. Founders are learning aggressively and absorbing information from the global market, not just China. They’re tracking GitHub, open-source models, and global innovation, not only ByteDance, DeepSeek, or Kimi. In that sense, the mindset is quite similar. The anxiety level is also similar. In Shenzhen, your neighbor could become your competitor overnight. In Beijing, you might discover a GitHub repo or a release from ByteDance, Gemini, or Facebook, building exactly what you had in mind. It feels like the best of times and the worst of times for founders. 10. What misunderstandings do overseas founders have when approaching Chinese VCs? Frankly, we don’t get approached by foreign founders very often, partly due to geopolitical dynamics. It’s true that Silicon Valley VCs often give more generous valuations for AI startups. I do speak with founders raising in both China and Silicon Valley. One common misunderstanding is assuming our capital is political. I don’t think that’s the right framing. We’re in an era of deglobalization, but there’s still respect for innovation and entrepreneurship that transcends politics. I see investors in Silicon Valley and Shanghai sharing similar mindsets and I see this even more strongly among founders. Founders are trying to build global products, software, hardware, agents, consumer electronics for a global audience, as day-one global teams. On a macro level, the world may be separating. On a micro level inside tech, people are bridging gaps with code and products. Products tell the founder’s vision. Personally, I’m open to conversations with founders anywhere in the world who want to build global companies. AI products should be global. 11. If a founder plans to reach out to VCs soon, what should they fix first? I would summarize it in three questions: why now, why this, and why you. “Why now”. Saying AI is a paradigm shift is not enough. You need to show your niche entry point, your product-market fit, and your ability to scale quickly with AI. Investors want to see that you’re not only thinking about the next three months. You need to explain how your team, product, and strategy will evolve over the next 12 months and the next three years. “Why this”. What pain point are you addressing, why does it matter, and why would users pay for it? Some needs aren’t obvious; strong product founders identify and articulate them through their product. And “why you”. This isn’t always asked directly, but we assess it through conversation and storytelling. It’s critical. So timing, product, and team: why now, why this, why you. These are the core questions every founder should think through.
Inside a VC’s Mind: AI, Hardware, and What Actually Gets Funded
Robin LuoAfter years working across Africa and the Middle East, Michelle returned to China and now moves between Shanghai, Beijing, Shenzhen, and Silicon Valley. In early-stage investing, she pays close attention to founder clarity, team conviction, and whether a product is built for short-term buzz or long-term traction. 1. When an AI project first lands on your desk, what are the fastest reasons you decide not to spend more time on it? To be honest, I rarely pass on a project purely out of instinct. Most of the time, I prefer to speak with the founders or the founding team before making any decision. Many AI startups today have pitch decks that are quite abstract. They’re filled with new concepts and technical jargon, and sometimes investors don’t fully understand what the team is actually building just from the slides. Misunderstandings often come from storytelling rather than from the product itself. So I usually schedule a 45-minute to one-hour conversation with the founders to understand clearly what they’re building and how they think about the product. After that conversation, the key factor for me becomes the founding team, especially the founder’s motivation. I care deeply about why you’re in this business. Why did you decide to build this product? Entrepreneurship is abstract and extremely difficult. It requires conviction and resilience. If someone is building a company because their Stanford roommates raised money from a16z, or because they want fame, reputation, or to become a KOL, that’s when I would pass. There are simply too many hardships in entrepreneurship. If the motivation is superficial, it won’t sustain the company. 2. At the early stage, how would you rank the problem being solved, the founding team, and the technology or solution and why? At the early stage, I rank the founding team first. The reason is simple: products can pivot, technology can evolve, but the founding team determines whether the company survives and adapts. The problem being solved is important, of course. But even a strong problem won’t matter if the team cannot execute or lacks conviction. Technology or the solution is necessary—especially in AI. But at an early stage, technology alone is not enough without the right team and mindset behind it. Entrepreneurship is abstract and uncertain. What I look for is whether the team truly understands what they’re doing and whether they’re prepared for a long-term challenge. 3. What is the single strongest signal that makes you want to keep digging into a project? The strongest signal is genuine conviction from the founder. I want to understand why you’re here and why you care about this problem. I want to see whether you’ve thought deeply about the product and whether your motivation is internal rather than external. If I feel clarity and a strong sense of purpose from the founder, I’m much more willing to continue exploring the opportunity. 4. Could you briefly introduce yourself and what you focus on as a VC today? I’m still a relatively novice VC investor. I graduated from college in 2022, and over the past three and a half years, I worked in Africa and the Middle East. During that time, I invested on behalf of a global company, working within the strategy, venture, and investment arm of a corporate venture capital setup. I returned to China at the end of last year and joined a venture capital firm around mid-November, so it’s been less than six months in this role. Today, I invest in AI software and consumer electronics / AI-powered hardware, mostly in China or with Asian founders building for global markets. That’s also why I split my time frequently between Shanghai, Beijing, and Shenzhen. Many software companies are in Beijing, while consumer electronics and hardware companies are concentrated in Shenzhen. I also spend time between China and the United States, especially Silicon Valley. There are founders who grew up there, studied there, or worked there, and now want to combine their understanding of North American or European markets with Asia’s efficient supply chain to build something global. 5. What kinds of consumer AI–related projects are you seeing most often and which ones actually excite investors? It’s a very interesting time for VCs. We’re seeing a wave of impressive founders coming from companies like Google, Apple, Amazon, ByteDance, and Tesla, deciding to start their own companies. There’s a wide range of software and hardware projects, some of them are things I personally never imagined people would build. On the software side, one major category is AI social apps. Founders are reimagining how AI might reshape interpersonal interaction, job hunting, dating, making friends, and communication in general, once AI becomes a variable inside social behavior. But social apps are also extremely difficult to build. We also see many productivity tools, especially targeting what people call OPC—one-person companies. AI can dramatically increase the efficiency of an individual coder or artist, so productivity products are a major theme. On the hardware side, it feels like AI could reshape almost everything in daily life from laptops to personal assistants. Founders are imagining AI rebuilding many of the tools we interact with every day. What excites investors isn’t AI as a label. It’s whether the product direction makes sense and translates into real usage, especially when founders can connect the technology to a strong consumer experience. 6. Which consumer AI categories do founders most overestimate? Investors currently have strong appetite for almost anything AI-related, it often comes down to individual taste. If you pitch the right person, you can find funding. Personally, I’m more cautious about products that sit too close to what large internet platforms can already do. In software especially, ByteDance can release something overnight that’s ten times better than what a two-person team is building. Some founders are also building features very close to what Gemini already includes in student or bundled plans. If startups try to charge $10–15 per month for something similar, they risk being squeezed by large platforms offering one-stop solutions. Founders increasingly understand they need to know the capabilities and boundaries of tech giants and avoid building too close to their core. Hardware is different. Hardware behaves more like consumer goods. AI agents may disrupt SaaS, but hardware is closer to consumer electronics and consumer branding. In many cases, it’s a consumer brand first, and AI-powered hardware second. For consumer products, opportunities continue across generations. Each generation grows up with different devices and ways of entertaining, learning, and communicating. The opportunity remains; the question is which startups can integrate AI software effectively into hardware and supply chain realities. 7. What are the biggest differences when evaluating AI hardware versus AI software at an early stage? The moat functions very differently. There are also different types of AI hardware. For example, in industries like 3D printing, being an early mover can create meaningful advantage. But if you’re building something like an AI camera or AI recording device, you’re entering a space with a very mature supply chain in Shenzhen. In that case, hardware itself is not the moat, the moat lies in the AI software or agent layer built on top. With AI software and agents, the issue is often the lack of moat. Open-source solutions appear quickly on GitHub, and similar tools can be replicated rapidly. So for consumer software and agents, you need a highly niche and accurate entry point, true product market fit and a strong ability to tell your story. When code moats weaken, trust becomes the moat. Why should users trust your team and your product? That’s where brand and storytelling matter more than people expect. We’ve seen examples like Manus. Many products try to imitate what it does, but it built advantage through strong consumer experience, a clear early positioning, and effective brand storytelling as a first mover. 8. What kinds of founding teams do you personally trust more? I trust teams with clear ownership and complementary strengths, teams that can genuinely connect technology to product. In AI, technical ability matters, but product instinct matters just as much. The best founders identify pain points users may not even articulate clearly, and then express those needs through the product. I also look at whether founders understand how competitive and fast-moving this market is and whether they’re building for the long term, not just chasing a short-term trend. 9. How would you describe the founder mindset in Shanghai, Beijing, and Shenzhen? Interestingly, I don’t see a huge difference in mindset. You do see different founder profiles shaped by the environment, more software and AI agents in Beijing, more hardware and consumer electronics in Shenzhen. But everyone recognizes the importance of AI. Founders are learning aggressively and absorbing information from the global market, not just China. They’re tracking GitHub, open-source models, and global innovation, not only ByteDance, DeepSeek, or Kimi. In that sense, the mindset is quite similar. The anxiety level is also similar. In Shenzhen, your neighbor could become your competitor overnight. In Beijing, you might discover a GitHub repo or a release from ByteDance, Gemini, or Facebook, building exactly what you had in mind. It feels like the best of times and the worst of times for founders. 10. What misunderstandings do overseas founders have when approaching Chinese VCs? Frankly, we don’t get approached by foreign founders very often, partly due to geopolitical dynamics. It’s true that Silicon Valley VCs often give more generous valuations for AI startups. I do speak with founders raising in both China and Silicon Valley. One common misunderstanding is assuming our capital is political. I don’t think that’s the right framing. We’re in an era of deglobalization, but there’s still respect for innovation and entrepreneurship that transcends politics. I see investors in Silicon Valley and Shanghai sharing similar mindsets and I see this even more strongly among founders. Founders are trying to build global products, software, hardware, agents, consumer electronics for a global audience, as day-one global teams. On a macro level, the world may be separating. On a micro level inside tech, people are bridging gaps with code and products. Products tell the founder’s vision. Personally, I’m open to conversations with founders anywhere in the world who want to build global companies. AI products should be global. 11. If a founder plans to reach out to VCs soon, what should they fix first? I would summarize it in three questions: why now, why this, and why you. “Why now”. Saying AI is a paradigm shift is not enough. You need to show your niche entry point, your product-market fit, and your ability to scale quickly with AI. Investors want to see that you’re not only thinking about the next three months. You need to explain how your team, product, and strategy will evolve over the next 12 months and the next three years. “Why this”. What pain point are you addressing, why does it matter, and why would users pay for it? Some needs aren’t obvious; strong product founders identify and articulate them through their product. And “why you”. This isn’t always asked directly, but we assess it through conversation and storytelling. It’s critical. So timing, product, and team: why now, why this, why you. These are the core questions every founder should think through.
-
How Foreign Entrepreneurs Can Protect Their Rights in China
Robin LuoShenzhen lawyer Vicky from Yingke Law Firm shares legal insights for foreign entrepreneurs in China, from contracts and IP protection to dispute resolution. Learn how to avoid common risks and operate safely in the Chinese market.
How Foreign Entrepreneurs Can Protect Their Rights in China
Robin LuoShenzhen lawyer Vicky from Yingke Law Firm shares legal insights for foreign entrepreneurs in China, from contracts and IP protection to dispute resolution. Learn how to avoid common risks and operate safely in the Chinese market.
Field notes
View all-
Why Every Overseas Hardware Team Needs a Local Partner in Shenzhen
Robin LuoFor overseas founders building hardware for the first time, this city looks like a dream: hundreds of factories, lightning-fast prototyping, and prices that seem too good to be true. But speed cuts both ways. The same pace that helps you move quickly can also wreck your timeline, your budget, or your product if no one on the ground is keeping things honest.
Why Every Overseas Hardware Team Needs a Local Partner in Shenzhen
Robin LuoFor overseas founders building hardware for the first time, this city looks like a dream: hundreds of factories, lightning-fast prototyping, and prices that seem too good to be true. But speed cuts both ways. The same pace that helps you move quickly can also wreck your timeline, your budget, or your product if no one on the ground is keeping things honest.
-
Thinking of Building Hardware in Shenzhen? Start with These 5 Realities
Luo RobinShenzhen is where hardware gets real. But for overseas founders, sourcing teams, the city’s speed and density can backfire if you land without a strategy.
Thinking of Building Hardware in Shenzhen? Start with These 5 Realities
Luo RobinShenzhen is where hardware gets real. But for overseas founders, sourcing teams, the city’s speed and density can backfire if you land without a strategy.
