What’s Different About China’s Tech Scene and How Are They Innovating?
Interview with Shunwei VP Meng Xing
We sit down with Shunwei VP/EIR Meng Xing to explore a broad range of socioeconomic, institutional, and financial factors that shape the Chinese tech environment. For example, why are the Chinese typically fast followers and excel at 1–100 over 0–1? What are primary founder incentives, and why is there a lack of talent in middle management? Why are there so many unicorns in seemingly niche markets, and how come there are so many humongous, diversified platform plays such as the BAT (Baidu, Alibaba, Tencent)?
Additionally, we examine frontier technologies such as AI, AR / VR, and provide examples of the Chinese successfully productizing these technologies via innovative business models that we haven’t yet seen in the US.
Meng Xing is VP/EIR at Shunwei Capital covering frontier-tech, AI, robotics, AR/VR, and more. In the past ten years prior, he also founded two AI start-ups that were acquired by Amazon and a listed Chinese company. In between, he worked as an investment banker at J.P. Morgan Hong Kong and casino giant Caesars Entertainment, on top of getting an MBA from the Sloan School of Management at MIT.
Listen to the podcast on SoundCloud (here)
Audio transcribed by Shaolong Lin, edited by Yvonne Yan
Foundational Differences between China and the U.S.
Adam: Meng Xing, thanks for joining us. Before we dive in, I know you’ve founded companies and invested in start-ups both in the U.S. and China. Can you tell us a bit more about your experiences and touch on some of the major differences between both markets?
Meng Xing: I started my first company while I was getting my MBA at MIT, and as you are probably familiar, MIT is a school that very much encourages entrepreneurship. In a way that the entire ecosystem is well built up from startup competition to mentoring systems, to entrepreneurs in residence at school, the angel investment environment around the school, the Tech Stars, all the start-up incubators around the vicinity, and ultimately up to the VCs and even later stage PE and so forth. The ecosystem is well built up. On top of that, the ecosystem sort of drives on the same value, whereas you are looking for a massive destructive technology or business model that could enable a company to expand into a supersize company.
In China, today we also have all of that. For most of those incubators, the startup programs, the competitions, they happened in the past three or four years. However, I think most of them are driven by different values. In schools, the organizers of the competitions have different value propositions than the people who compete, and also have a different value proposition from the administrative branch that oversees the schools. Also you get those incubators: a lot of incubators here are rental companies in a way, and they also get their profits from the government by listing all projects, building those incubators and getting government grants so that they can get the land for a cheaper price. Essentially, they’re a real estate company, they don’t really care about the success of the company that got incubated in that sense. On surface, we have those players around, but I think they are strung by different forces and so forth, so being able to navigate through this is key and oftentimes a burden on the founder. Having their eyes on the ball, being sharp all the time, and knowing what they really need and want from different forces are really important. I think that is one difference.
The second point I would talk about is that there is a huge difference between the overall social economic environment in China and the U.S., and because of that, there are a lot of new opportunities in China for start-ups that do not really exist in the U.S. For example, there is a Chinese company that’s very popular, called Kuaishou (快手). Essentially, they are a short video streaming platform on mobile. You can think of it as a short version of YouTube or Vine, plus a streaming platform. Because live streaming is not as popular as it is in the U.S., there’s probably no good comparison to it. Kuaishou today runs about 50 million DAU and with more than 150 million MAU, which is massive. They are targeting, at least initially, only the population of the rural, tier 4, 5 cities and even below. You never see a similar product like Vine for the countryside, the factory workers, but there is that very popular product in China that is only targeting for this niche market. In China, because there are so many people out there, there aren’t really any niche markets.
You can look at another company called Omni Prime (买单侠). They are a payday loan company essentially for the factory workers from Foxconn and so forth. The company supports workers when their savings aren’t enough to purchase expensive goods such as cellphones. The factory workers from Foxconn alone comprise 1 million people, and adding on top of that all the smaller factories — that is a huge population with their own social dynamics. Doing a start-up for those niche populations alone is building significant value. And because their income level and social environment are vastly different from people live in the 1st tier cities, you have to build a startup that caters to their specific needs and not to build a uniform product that essentially goes everywhere. There are numerous unicorns within these niche, fragmented areas in China.
The third difference is that, in the U.S., startup founders are by nature or by culture very fond of being the first to invent something and take that idea, execute it and build a large company. It just so happens that the economic returns for doing it this way justifies them to be the first mover. In China, being the first mover, at least in the past, has not always been so favorable in return, because the market is so huge, and the execution capability in China is so good that it’s very difficult to differentiate yourself just by being the first mover. You will get a lot of copy cats coming after you, so people focus more on the execution in protecting their edge and being able to grow sustainably.
If we summarize that, we’re looking at more U.S. companies that, at least in the early stage, are doing the zero to one part. They start with a groundbreaking idea in a groundbreaking niche market or new industry, found a new product and bring that to the market. In China, a lot more people, you will see very few people doing that groundbreaking work. You will see much more people expand upon whatever they see that works and compete in how fast they raise capital, how fast they expand geographically, and how fast they can expand vertically or integrate more products and become a platform company rather than a product company.
And you see this a lot because in China, almost all gigantic companies are platform companies in a way or another. They want to do anything for them to stay sustainable. As a large company, they have to do everything. It’s very difficult to find likes of Tencent, Alibaba, and Baidu, where take Baidu for example, their search is obviously a firm, but they are also the largest consumer-facing discussion forum company in China as well. They are also an O2O company where we say O2O is an online to offline food delivery company. Also, they own the number one online video company in China as well. Baidu is the smaller of the three to begin with, and if you look at Tencent, it’s even huger in terms of what the company integrates into its product offerings.
Adam: Just let me cut in a little bit, and I understood that these macro factors have significant impact, but what about the founders themselves at an individual level. What are some of the differences when it comes to motivations, environments, upbringing such that might impact how these start-ups develop and scale.
Meng Xing: When you look at founders in China and the U.S., and we touched upon that a little bit earlier, all companies start with an idea, take that idea into the product and find that product-market fit. You expand the market for that company, and so on and so forth. From idea to product is the early stage of the company, and from the product to huge conglomerate is the later stage. That later stage I would say is mainly dependent on your execution, while at the early stage it’s more dependent on your idea formation, conception, and your building of a team.
At least in the U.S., the burden on the founder to take this from beginning to end is less because up to certain stage, as long as you can prove your product market fit, then you can purchase the execution capabilities via capital you raised. There are a lot of good people that you can hire, professional managers who can help you execute better and take your company into a later stage company, or from a product to a platform company. But in China, such talent isn’t easily acquired.
Number one: the supply just isn’t there because in China there isn’t that long of a history of entrepreneurship and professional management at the different stages of the start-up cycle.
Number two: there is a lot more investment money in China, and so the opportunity cost for professional managers is a lot higher because if they are good, they can start their own company and easily raise funding. From the Chinese culture, as well as from realistic economic stand point, it would be very difficult to break your “respect ceiling” or “wealth ceiling” by just being a manager. You only able to break that when you are a founder of a company, and that makes a significant difference. In China that’s the difference between whether you can buy your own apartment or not. So in most cases, the capable managers will go on and become founders on their own.
Number three: I think there’s a big difference in RMB funds and U.S. dollar funds. Because of the difference in terms of how the fund structured, as well as their exit strategies, they encourage different dynamics in the startup world. The RMB funds usually have a shorter term. For example, if we take the typical U.S. dollar fund, it might have a 10 years period for a typical term, whereas a RMB fund might only have 3 or 5 years. There is much less tolerance for long term projects here.
Also on the exits. RMB funds are usually exiting in the Shanghai Stock Exchange, one of the stock exchanges in China. Because of Chinese regulation, they require the startup companies to have a level of revenue for sure and a certain level of profits. So that condition pressures RMB backed start-ups to build a company within 3 to 5 years and have significant profit so that they can exit into the public market. That’s a lot of stress on founders and can impact their incentives and motivations when it comes to building their businesses.
So you’ll see fewer super unicorns that took RMB funding compared to money from U.S. funds. But because RMB funding is more abundant in China nowadays and there are better exit routes for those companies as well, they’re getting quite a bit of attraction. At the end of the day, as the capital market grows, as the maturity of start-up founders grows, and overall market develops… I think some of the differences will diminish. But I think some of the differences will grow and increase even, so things won’t converge anytime soon.
Frontier Technologies in China
Adam: Clearly a lot of different factors that impact the differences between both markets. But I was going ask, given your focus at Shunwei for AI, AR/VR, a lot of these frontier technologies, I’m curious about what you see there. Can you tell us a little bit about what you see in China when it comes to frontier tech?
Meng Xing: Frontier tech is getting very hot and becoming major theme for the venture capitalist over the past two years. This includes new technologies in AI such as natural language understanding, computer vision, 3D reconstruction and so forth. In terms of technology in its own, there’s not a lot of difference between China and the U.S., but the applications of these technologies are more interesting.
For example, I founded a facial recognition company a few years back while I was in the US, and back then facial recognition was not a massive market technology. It was there for some niche markets, and it was not that good. When we launched the product, we were looking at mostly working with internet companies where our customers would be able to understand where to integrate our technology and most of that is into the social networks for aggregating similar faces and photos, or dating sites and so forth. That was the trend in the U.S. for those kind of start-ups in the last few years until today.
But in China, our counterpart in China which does facial recognition companies, they started similarly, serving companies in the Internet market, but they soon found out that the internet companies are stingy, not willing to pay a lot of money for that service. Once they think your service is important, they would hire similar people to build a similar product. So the Chinese start-ups quickly turned to traditional industries, like surveillance companies. You’ll probably be surprised to learn that there is a $50 billion market cap surveillance company in China, called Haikangweishi (海康威视), and they own a huge market share — the number might be 25–30 percent of the China’s security cameras that are placed in airports, in cities and in sensitive areas. They build the hardware, they build the servers, they also build the transmission protocols and systems for graphics and do the video surveillance platform on the software level as well. What they lack is artificial intelligence technology for facial recognition, and when the entire platform is updated to the next generation of surveillance and policing, they certainly found their technology insufficient and so those facial recognition companies that we’ve talked about, they saw the opportunity and they went in and scooped up some of the opportunities that were left on the table.
And as you can probably imagine, the customers for this sector includes the government, the police department, and a lot of the agencies, the airports, the train stations and so forth, and they have a lot of budgets to spend. Furthermore, they are much more willing to spend that budget than the internet companies. We were seeing most facial recognition companies, or most valuable facial recognition companies in China, starting their products with a surveillance camera, a search engine for police departments for face searching, etc. And they are able to make $40–50 million a year in revenue and probably even more. You don’t see those types of numbers in the US for facial recognition start-ups. Usually they are still valued for their tech and talents rather than from a business stand point.
Adam: That’s a really interesting example. Taking a step back, it seems that business model innovation is happening, and there are a lot different examples of companies leveraging their technology and creating lots of really cool business models on top of it. So based on that, can you tell us a little bit about other examples you’ve seen in China that attempt to productize A.I. and other frontier tech to create many cool new innovative commercial models out of that?
Meng Xing: Another example I would share is autonomous driving. There’s a lot of questions in terms of how soon autonomous driving will arrive and whether it would actually impact our lives, whether there would be a bunch of government regulations. In reality, there are lots of other problems that technology has to solve first before this actually impacts us. I think the assumption is that we want to introduce autonomous driving into our normal driving behavior pattern in local and highway street, in our sedans, vehicles and so forth. But if you are introducing autonomous driving into a confined environment with low speed vehicles, with communication configuration with speed lights and with the road itself, that’s a much easier problem to tackle with lots less regulations to consider.
But that environment does not really exist in most parts of the world. I think there’re a lot of unique opportunities in China for this to happen. For in China there’re a lot of economic zones that are built from ground up, hosting new technology companies, mostly built by the government. Those are usually built from sand (no man’s land). They construct the road and the buildings, and that entire district would be planned out very scientifically from the ground up. In those areas, there’s usually no residential area. The residential areas and the business areas will be very much segregated. They build autonomous vehicles for transportation within those confined districts or new economic zones, and by zones we are talking about 2 x 2 miles squared. It’s not a big area. Building an autonomous vehicle for that including transportation, street cleaning, and a lot of garbage trucks is relatively possible today, and we invested in a company that does exactly that.
We see about 100 of those economic zones out there requiring such type of vehicles and technology as we speak today, and we see the number is growing as well. That’s very unique in China, to be able to apply the tech products on the road first before introducing to the mass market.
AR / VR
Adam: Why don’t we switch over to AR/VR. It’s a hot area in general globally, with plenty of ups and downs. Based on what you’ve seen over the past few years, can you tell us a little bit more?
Meng Xing: Sure. VR has been an overly-discussed topic on both sides of the world, in the U.S. and in China, as it all started when Facebook decided to acquire Oculus for $2 billion. Everybody on the other side of the world started looking at this and said, “this is the future”. A lot of companies started to move fast around the time that Oculus was founded and launched into Kickstarter and later acquired by Facebook. The first mover did raise a lot of money in China, and they became from a nobody-cares to everybody-cares to super-hype company, but unfortunately today they are back at where they began, because the time to market was too long. Despite their raising a lot of money, they spend a lot of money on development and marketing as well, and in the end the first mover almost got wiped-out; the entire sector almost got wiped out. They are hanging on, but with difficulty.
The only survivors in this market today we are looking at, for most part, are teams that have been backed by large players like Tencent, Xiaomi, Huawei and so forth. They are still hanging out and they are doing very well. They are developing the technology in the right path as the founder would expect to, and not from the super-hype expectation perspective. And I suspect the same thing would happen in AR as well.
We see many companies start very early, raising a lot of funding early in China. But going forward you are not looking at a time-to-market for months, you are looking at that in years. For hardware companies, starting early is not necessarily a good thing. For AR hardware companies that are independent, I would very much suggest them to look into partners from an equity standpoint or from a business standpoint with a large firm that can take them further and mitigate the risks over the next few years when there’s not significant revenue or even significant user excitement.
There’s a different story though, on the software side. In China, we usually joke about this, where we say there are so much money going to the AR sector, and investing technology, but the only type of company that makes money in AR in China is the company that builds AR contents on mobile phones that are cater to kids for education purposes. There’s a company called NeoBear, where they develop those study cards, physical study cards, and on top of those study cards are printed with dinosaur, different types of animals, and when you open your camera of the app on your phone and you scan the cards, there would be a 3D character of that dinosaur that jumps out on your phone and you would be able to interact with that. It’s a very simple technology and very simple idea as well, but they are able to sell that to mass market and even become a phenomenon in the early child care world. It became a very hot product in this world.
Adam: I think now for NeoBear, they are not only selling these cards with AR components, but also selling the brand and IP of the animal and the bear itself and have become an icon amongst the children. I believe they are even creating a cartoon based of it and other shows. This speaks to the importance of developing a business model that actually works based on real use cases and expanding beyond that.
Meng Xing: The founder of NeoBear was a producer at a CCTV English program, and he has a very well-established background in producing TV shows and building IPs in the media side of the world. If I ask you which company would you think, given the founder’s background, would be doing the best in AR… you’d probably say someone from Microsoft in computer vision or hardware from Intel that builds super advanced products out there. But no, NeoBear is an example where they take a different approach in building a unique product, serving the market at the right time. They built a product that is acceptable to the market, but also very new and eye catching because people haven’t seen that before, with not too much of the technology barriers to overcome.
Adam: That’s a very good example of leveraging technology and building a very valid use case and business around it. But when we take a step back and again look more broadly at AR software, for examples, some of the platform plays in China: Baidu, Alibaba and Tencent, they’re all working on building a platform for AR. It seems to me that certain types of use cases would be eaten by these guys, including advertising, whereby if you have your own third party app, you have to scale to actually have enough users to use that consistently over time to build up enough engagement for the advertisers and brands. That’s going to be kind of hard unless you work with the BAT type of companies or platforms.
But we’ve also worked with or seen a few other companies that work on AR software that were developing SDKs and kits and such, they are able to make money on focusing that specific verticals, for example education as you’ve mentioned, or it could be enterprise-based offerings. Could you speak to AR software more generally? What other opportunities are in that space?
Meng Xing: When we look at the AR market, beyond hardware, we want to see what would be the killer app that would make this expensive kit/glass useful. That is very hard to answer, I don’t think anybody has the answer yet, until it reveals itself. We think hard and try different things, but I think one very interesting thing for the application for AR is to do telepresence communication as well as viewing events (sports events or social life events for example) from a 3D viewpoint. You see some of these demos at the HoloLens demo, for example the “Holoportation” experience, but you don’t see a lot of commercial companies or start-ups out there being able to launch an equal demo out there. The HoloLens team obviously did a very good job, but that’s something very futuristic, more for a demo purpose rather than from a practical product standpoint.
We think being able to communicate in that way virtually as if the person on the other side of the world is sitting right in front of you, and you can stand up and look at him from the side, from the back, almost as if you were interacting with them in the real world, is the future. It’s amazing and brings a lot of rich experiences and value to owning an AR product.
So we invested a company called Owlii that does exactly that. Recently they were able to put together the real time demo where they accomplish exactly what I’ve just been telling you about. Owlii is not the first company to come up with this real-time demo, but they spent a lot of time in making sure that the faces/edges, the representation of human, and the environment is accurate and delicate enough for this product to be useful. They are doing very well and pretty advanced from a global perspective in terms of technology. You’ll see more and more companies like that from China that are not innovative from a business perspective, but also actually leading or becoming one of the leaders in the technology that they are approaching.
Adam: Thanks Meng Xing, until next time!