Episode 32: Wang Yu of Tantan on Scaling China’s Top Dating App

We interviewed Wang Yu (王宇), the co-founder and CEO of Tantan (探探), China’s leading dating app. Tantan is social app that help young people in China connect with one another. It has a slide-left slide-right interface. Only when two users both slide right on each other can they start a conversation. The company was founded in 2014 and has helped users make over 10 billion matches to date. In 2018, Tantan was acquired by Momo (陌陌) for $735 million. Momo is a top location-based social networking platform in China that help people meet strangers around them. It is also one of the leading live streaming platforms in China. It is a public company on the NASDAQ and its current market cap is around $6.8 billion.

Wang Yu was born in Beijing and grew up in Sweden. He holds two master’s degrees, one on computer science and one in industrial economics. In 2007, he moved back to China and started his first business P1, a fashion community, before founding Tantan in 2014.

During this episode, Yu discussed how the failure of his first startup P1 proved crucial to the success of Tantan, why flawless execution is more important than flawless product in China, whether any social apps in China will be able to challenge WeChat, and the advantages and disadvantages of being an overseas Chinese returnee entrepreneur.

Join our listeners’ community via WeChat/Slack at 996.ggvc.com/community.

TRANSCRIPT

ZARA ZHANG: On the show today we have Wang Yu, the co-founder and CEO of Tantan, China’s leading dating app. Tantan is a social app that helps young people in China connect with one another. It has a slide left/slide right interface. Only when two users both slide right on each other can they start a conversation. The company was founded in 2014 and has helped users make over 10 billion matches to date. In 2018, Tantan was acquired by Momo for around $735 million. For those who are unfamiliar, Momo is the top location-based social networking platform in China that helps people meet strangers around them. It’s also one of the leading live streaming platforms in China. It’s a public company on the NASDAQ and its current market cap is around $6.8 billion.

HANS TUNG: Wang Yu was born in Beijing and actually grew up in Sweden. He holds two master’s degrees, one on computer science and the second one in industrial economics. In 2007, he moved back to China and started his first business, P1, a fashion community which actually I remember, before founding Tantan in 2014. Welcome to the show, Yu.

WANG YU: Thank you. Hi everyone, this is Yu. It’s great to be here.

ZARA ZHANG: You were born in China but moved to Sweden at age seven. Could you tell us about the move, like why you moved there and what was it like for you?

WANG YU: We moved because my father was doing research and he was invited to Sweden to be a guest researcher. He moved there in 1986. After two years there, we decided to move there with the whole family. We moved in ‘88. And it was a very big difference. Back in those days, the world that I knew, it was still in the early days of open up and development of China. So all our furniture didn’t belong to us; it belonged to the people. That’s the kind of surroundings I was used to. When we moved to Sweden, the first 10 years or so we lived in two different immigrant areas which is probably quite different than the Sweden people could imagine. We had one Swede in the whole class. It was kind of like United Nations. I got to get in touch with all sorts of different cultures, so that was very interesting. And in secondary school, because I was interested in math and so on, I went into the city to go to a school with a special math and science class. And that was more like the elite part of Sweden. The difference between the immigrant area called Rinkeby and Östermalm, which is the most central part of Stockholm, I would say that’s almost as different as between China and Sweden. So I got very big clashes of how to look at the world which I think has served me a lot in life.

HANS TUNG: Good. How so?

WANG YU: You’re forced basically to look at the world with different views. And it’s often drastically different. Each one has their own reasonings which make sense from their own perspectives and narratives. And you get to, early on, get acquainted with the whole notion of narrative because there is a whole history behind why you think a certain way. So you stop looking at things in terms of black and white. You kind of have to.

ZARA ZHANG: I remember Liu Zhen said the same thing on our show when she talked about her experience of going to high school in the US.

HANS TUNG: Right. In 2007 you decided to move back to China. I lived in China, worked in China between 2006 and 2013. What prompted you to move back to China back then?

WANG YU: Right between college and starting to work, I worked for a while at Sony Ericsson. I started to have a lot of business ideas. Because I was doing some VoIP work in Sony Ericsson and my manager there wanted me to join his own company to do VoIP for solving roaming between countries. And I came up with this idea of how to reduce the costs significantly through VoIP. Basically, it’s a solution where before you go on the airplane, you redirect your number to our server locally. Say you travel from Stockholm to Beijing, then you transfer it to our Stockholm number. And when you arrive in Beijing, you buy a new SIM card and then you send us an SMS telling us what your new number is. And all calls to your Swedish number will get routed to your Chinese number. So you pay two local fees. And when you want to call out, if you call a Swedish number, you just call from your normal Swedish phone. If you call a Chinese number or any foreign number, you call our local server in Stockholm and then you pay two local fees again. So instead of paying somewhere between RMB 6 and RMB 12 per minute, it becomes less than 1. And no one to this date I think has a solution like this, although roaming is much less of a problem because cost has been reduced significantly. But back in those days, this would be very new and would solve quite a big problem. So we filed for a patent and we wanted to do this in a company. The manager went as far as to offer me a CTO position with a 20% stake in the company. And for me it was like, okay, so it wasn’t harder than this. The whole solution took me a week or so to invent.

Something snapped and I understood that for any business, what you need to do, especially if you’re more or less savvy in terms of tech, you find the pain points. People complain about stuff all the time, so if you find anything people complain about, that’s a pain point. And you solve it with technology and you find a way to charge money. There’s a business. Business ideas were coming up every day in every conversation, so it felt like I was bound to start a company. And I wanted to go back to China because, even though I was quite comfortable with the Swedish culture – I have a lot of friends who are Swedes – I still felt more Chinese in terms of values and so on. And I like China more. So I wanted to go back to China, wanted to start a company. And I think of the Chinese people in Stockholm – there’s not that many Chinese people in Stockholm – I think me and Sophia were the only ones I knew of that both wanted to start a company and wanted to go back to China. So we naturally started to look for business ideas together and that’s why we did it.

HANS TUNG: Right. I remember P1 back then. It was a very cool fashion website. It was the first website in China that had a black background color. And it was cool, but I never reached out to contact you because I thought, well, maybe it’s too early for its time. There are too many diaosi users in China, very grassroots or mass market. The people who can appreciate P1 will be very niche and only in Beijing and Shanghai. What were your lessons from doing P1? Why did you do this particular concept and this user interface? And what did you learn after doing it?

WANG YU: The reason we did P1 as we did it, the major reason was that Sophia had been away from China for four years at the time and I had hardly ever lived in China. We thought that if we would just flat out compete with the grassroots entrepreneurs in China, we would just lose. We would have to come up with something different. And back in those days, and to a high extent, still, a lot of the business ideas were borrowed from the US. Our stuff, we were coming from Sweden, and we were pretty sure no one had seen anything like this in China. So even if they saw our product they wouldn’t understand it.

ZARA ZHANG: Could you describe the product a little bit?

WANG YU: It was like a high-end fashion community. And it was invitation-only. That’s why it was kept very high-end. We had a lot of street style photographers who would scout for people who were really stylish in and outside of high-end shopping malls – they were just exploding all over China at the time – and then they would ask if the person wanted to get a street style picture taken. And we were so picky that people really wanted to. A lot of people dressed up just to get snapped by us. Once we took the picture, we took their contact info and invited them, and only if they sign up to P1 would they be able to get their picture. So conversion rates were super high. It was like 80%. And the photographers were paid for each qualified user they actually generated. It was not based on the working time or the photos. It was based on the amount of users we got. So user acquisition cost was around RMB 15. But RMB 15 for that quality of user was actually pretty cheap.

HANS TUNG: It’s very good.

WANG YU: So we got millions of users. To give an example of the quality of the users, at one point Lamborghini did an event, an offline event, and they wanted some fashion people, so they asked us to invite 30 users. And we didn’t think too much of it so we sent out emails to random users and we got 30 users to come. At the event, one of them bought a car. That’s the quality. Because think about this, people who went to Guomao and Xinguang Tiandi in their 20s were really stylish at that time. They have a lot of purchase power. So even though we didn’t have so much traffic, all the luxury brands were clients with us. We did generate significant CPM numbers. It was over RMB 1,000 CPM. But at max I think we had like 20,000 DAU, so it wasn’t a big business.

We took three main lessons from the whole endeavor. We really tried. We tried for seven, eight years with all sorts of different pivots in P1 to try to get it to work, but we just couldn’t. The three main lessons were, one, it was too niche like you said. We thought that even though this is a niche market, but China has so many people so that is still a big market. But if you do China internet, you want to get the mass. That’s the value of Chinese internet. So we neglected that. That’s one. The second part is probably more important and that’s the pain point. Not very strong pain point. Fashion socializing is something you can live without. If they don’t have a problem, it doesn’t matter how good the solution is. And the third part is the product was too heavy. Since we didn’t have a killer feature, we tried to compensate that by adding more features. And that’s not a good way to try to succeed with an internet startup.

When we did Tantan, we wanted to not repeat these mistakes. We wanted to go for a super strong pain point. We spent quite a lot of time trying to define that. Our positioning is solving singles have a lack of channels to meet people. That’s our positioning. It’s neither for hook-ups or for getting married because those are actually sub-portions of the market, quite small portions to be frank. Even though for guys there is a quite higher level of acceptance for hook-ups, for girls still in China and Asia, it’s not as pronounced. We’re talking about 5-10% of the female market which heavily limits how big you can be. If you define it as meeting people if you are single, that’s something everyone has a need for. By doing that, you open up the whole marketspace a lot, the whole TAM. So that’s how we defined the pain point. And it’s obviously a big mass market. The third part is that the product is very simple. But the product, even though being simple, it fits the Asian and especially the Chinese culture very well, I think much better than it does for the Western culture. That’s because Chinese people don’t have a flirting culture, so they can’t really solve the problem offline, for most of the people in China.

ZARA ZHANG: They feel awkward.

WANG YU: Yeah. It’s indecent to approach someone you don’t know. It’s still like that in the culture even though dating is very accepted. Having a lot of girlfriends, boyfriends before you get married, that’s totally fine. But getting in touch, that first step, is very awkward. That’s why there are hardly any parties in college, for example, and most people don’t go to clubs and bars. And if you meet someone on the subway or in the mall or so on, you don’t talk to them. You do that in the West, but you don’t do that in China. So how do you actually meet someone? That’s one issue. The second issue is that most Chinese young people are not in their hometown. They’re in a different city. That’s quite unique. They’re either in a different city to study or to work. They’re much more lonely because they don’t have their family and friends there and no one to introduce them to people. It’s usually just go to work, go home, sleep alone, every day. It’s super lonely. So the need is much stronger as well.

The Tantan product, even though being very simple, it fits this problem quite well because you never need to flirt. You never need to take the first step because when you slide someone left or right, that person doesn’t know. They only know if you like each other, and then both take the first step at the same time. For a guy, you never expose yourself. You never get rejected. There’s no face issue. And if you do get matched, talking to the girl is not indecent. You’re very confident. You know that she’s interested. For the girl, she doesn’t get harassed. She doesn’t get bothered, spammed by a lot of guys. And the girl also has a super high success rate. According to our numbers, guys like 60% of the girls they see. Girls like 6% of the guys. There is a ten times difference. So for the girls, it’s almost like an assured match if they like someone. They feel like the emperor flipping the hearts. And the third thing for girls is that they can be proactive. Socially, the females, even in the West, they’re waiting for the guys to approach them. They can be proactive. It’s a completely even playing field. They can choose who they like. And that’s something they usually can’t do. The people approaching them in real life are often not the ones they like. In Tantan, the ones they like has still a 60% chance of matching with them, so the user experience is very good. So even though our product is very simple – it’s basically just flip some cards and chat – it solves this major barrier of initiating the first contact which I think is a much bigger problem in China than in the West.

HANS TUNG: There have been many dating websites and then apps in China as well as outside of China. When you were doing Tantan, what were the features that you were thinking of that’s different from everything that’s out there? And what are the features that are distinctively Chinese that a Western app wouldn’t be able to satisfy?

WANG YU: I don’t think it’s so much about what features you have as opposed to how you operate the whole service and what kind of values you have while doing that. When we did Tantan, there were 100 products doing more or less the same product. And they all failed. And they failed not because we were doing a good job and out-competing them. They failed because they were doing a bad job. And even though I don’t think to this day that it’s really that difficult, it becomes very difficult if you don’t have the proper experience. And we only gained the experience by basically failing with P1 for seven, eight years. We basically made all the mistakes we could make and we succeeded by not repeating them. For a new startup, for new entrepreneurs who haven’t made those mistakes, I think you would need to have a lot of luck to avoid them.

For us specifically in this kind of service, first off you need to decide is everyone going to see each other only once or are you going to repeat the people you see. Since you don’t have too many users in the beginning, most startups choose to repeat. That’s a bad idea because if you like someone, you don’t need to like them again. You’re just waiting for them to like you back. If you see them again, it’s just a waste of your time. If you don’t like someone, you really don’t want to see them. What you want to optimize for is time efficiency, the amount of time you spend versus getting someone you can meet offline that likes you that’s a proper person. That’s what you want. So you only see each person once. The second is that you want to see fresh people. You don’t want to see people who were active three months ago. That’s also a big issue for a new service since you don’t have too many users. In many of the other services, you see people who were active three months ago and then you feel like it’s a ghost town. We decided early on you can only see people who are active maximum within 48 hours. They have to be active within 48 hours. With those two restrictions, then you can understand what you’re doing. I don’t think hardly any of the competitors we have realize this. In Tantan, each person slides 200 cards on average per day.

HANS TUNG: 200? That’s a lot.

WANG YU: 200. Since day one it hasn’t really changed. And if each person slides 200 cards and you only see each person once and you want to see fresh users, what does it mean for a new service? It means that locally, within the city, you have to grow by 200 guys and 200 girls a day for this to be sustainable. But at the same time, if you do grow by 200 guys, 200 girls a day, no one will hit the bottom of the stack. Because most users aren’t active every day. They are active, on average, maybe every second day. So 200 in terms of new growth is enough per city. And if they don’t hit the bottom of the stack, they don’t care if there’s one card left or a million cards left. You’re essentially operating as if you were dominating the market already in terms of user experience. So you can very early on give users that user experience. And since we had this offline street style acquisition experience, we could get good-looking guys and girls offline. We just skipped the whole street style part. We just used male and female models, male models to attract good-looking girls and female models to attract good-looking guys.

HANS TUNG: That’s the concept of yunying or operations.

WANG YU: Yes. If you get that part right, you get the city to live. I don’t think any of the competitors, of the 100 competitors, managed to do that. We first did that in Beijing and then in Chengdu. It didn’t work in Shanghai. That’s very interesting. Shanghai people, they’re too snobby for this offline thing to work. So we got Beijing, and Chengdu was awesome. That was the early experience. In Beijing, for example, you get 10 matches as a guy. You talk to all those 10 girls, maybe 5 will answer you. In Chengdu if you get 10 matches, if you don’t say anything, 5 of the girls will talk to you. That’s how big of a difference there is between cities. Once you get Beijing and Chengdu to work, we had this problem, how do we handle the rest of the cities? If we had to, offline, get each city to work, that’s a massive investment. It doesn’t feel like it’s going to scale that well. So something we tried that was very interesting was that we changed the maximum distance from 100 km to 100-plus km. 100-plus means infinite.

HANS TUNG: That’s a pretty long distance.

WANG YU: That’s the single biggest optimization we have done in the history of Tantan. Just that single change made everyone active. In a small city, before, you could maybe flip five cards and then you’re out. Now you can flip those five cards and people in the other surrounding cities. And if those were not enough, you could see people in Beijing and Chengdu who were super happy users. Even if you were not so happy, as long as you interacted with happy users, people tended to not care so much about location, much less than we thought. So immediately, we got the whole China to be as active more or less, in terms of retention and so on, as Beijing and Chengdu. We then went to just do online promotions from that point on.

ZARA ZHANG: The atmosphere of the whole company was uplifted.

WANG YU: Yes.

HANS TUNG: So what was the maximum distance? You said 100-plus km.

WANG YU: 100-plus is like infinite. It’s actually, I think, 2,000 km. Half China.

HANS TUNG: 2,000? Wow. That’s like 1,200 miles. That’s big.

WANG YU: Yeah. You still see the people closest to you first, so if there is a lot of people close by, you don’t notice that limit. It’s only when there’s not people around that you notice the limit.

HANS TUNG: But how do you date when you’re dealing with someone who lives far away from you?

WANG YU: Early on, there were plenty of users who took a plane to meet their dates. It’s insane. The first day, since you can only see each person once, we had one of our own female employees who ran downstairs. We were like, “What’s up?” She bought a new SIM card to register a new account because she mis-swiped a guy. She wanted to swipe that guy yes. And within the first month, we had the first married couple who wrote us a thank-you email. That’s how efficient it is. It’s quite amazing. That’s one. And the other part with China is that, for example, if you look at Western dating apps, they usually connect to Facebook. By doing that, they get a lot for free because Facebook profiles are very good. First of all, at least in the West, it’s almost always of the real person. And it’s of the real person in a live situation. It’s not trying to look good. It’s quite authentic. It actually works better than the profile pictures you upload for a dating app. That’s one. The second is they get their contacts for free. They get their Facebook friends for free, so they know who their friends are. They also get their profile information for free. The only thing you need to do, basically, is write a description of yourself. Everything else you get for free. Also, you get a much higher conversion rate from downloading the app to signing up because it’s a one button press.

In China, obviously you can’t do that. And since the Western apps, half their success is because of this connection to Facebook, if you want to get this to work in China, you have to find a way to emulate that success. That’s also something that I don’t think any competitor really got that right. From the first day, we manually moderated all user profiles. We removed all profiles who did not have a proper profile picture, which is like 40% because that’s how China is. And you obviously can’t use WeChat because WeChat is even worse than 40%. Every user when you sign up, you’re actually invisible. No one can see you. You can swipe the other users, but no one can see you. But our moderation is quite efficient, so within a few seconds someone will have looked at your profile. If you pass then people will start seeing you. It’s seamless. You don’t need to know. If you didn’t pass moderation, you’ll get the prompt saying you can’t use this, and you have to change your profile picture. We did that from day one.

HANS TUNG: So your system automatically pings users. As they become more active, they need to contribute more to their own profile.

WANG YU: Yes. On Tantan, no one has ever been seen that has not manually passed moderation. That’s a very big sacrifice. So far I think we have probably killed off like 80 million accounts.

HANS TUNG: Because it’s not compliant. It doesn’t show enough.

WANG YU: Yeah.

ZARA ZHANG: And how do you have enough manpower to do that manual moderation?

WANG YU: We have hundreds of people doing that by now.

HANS TUNG: And this makes sense for you because, like you said, most users end up looking at 200 photos a day, so you want to make sure that these are high-quality 200 photos.

WANG YU: Yes. You don’t want to see a water bottle. If you do that, your excitement goes down a lot. So we had to do that. Then we had to develop a profile information system for users to fill in their profiles, otherwise they’re just a profile picture. They need to feel more alive. So we have profile tags. And I think it’s like 85% of the users fill in those tags, over 10 tags each. It’s also quite a big success in terms of that product. And then, since they sign up with their phone numbers, we use their phone address book instead of the Facebook contacts. What we do there is that we block all your phone contacts. And that’s why people give us their contacts. Almost everyone gives their address book because they don’t want to see their coworkers or their boss, or them to see them. But we push people who are secondary degree contacts to them. Friends of friends are actually okay. That’s how we use the address book.

ZARA ZHANG: It’s opposite to a lot of social apps where you sign up and they automatically ping your whole contact list.

WANG YU: Yes. That’s horrible for a dating app. That’s what I mean. You can call it features but I think it’s more like operational rationale.

HANS TUNG: Yes. It’s definitely how you operate, how you yunying and you carefully think through the usage cases and make sure the user experience is optimized.

ZARA ZHANG: For your moderation, do you use automation like machine learning?

WANG YU: We’re looking at that now but so far it’s all been manual. That’s the thing with China. It’s quite cheap. The funny thing is that we looked at some external AI solutions for doing this which everyone else uses, and it’s actually more expensive than using moderators.

HANS TUNG: Well, like you said, in the West, Facebook is so popular so you can automate this. You don’t need to yunying (operate) it. In China you can’t, so you have to be able to do that.

ZARA ZHANG: How do you prevent people from just moving the conversation to WeChat once they meet on Tantan?

WANG YU: Our rationale there has been we want to maximize user experience, so we don’t want to prevent them. That’s a natural need. Would we prevent them from exchanging phone numbers? That’s silly. And WeChat is probably more convenient than phone numbers at this point. We’re a dating app. The problem we solve, like I said, is for singles to meet people. We’re not solving the whole keep in touch with your friends part because that would be just directly competing with WeChat which is stupid.

ZARA ZHANG: Yeah. And WeChat is not good for meeting new people.

WANG YU: No. That’s intentional. Their biggest issue is that it’s getting diluted, and dating is a heavy diluter of friends management. So they don’t want that.

HANS TUNG: Right. It’s very weak ties.

WANG YU: Yeah. So we’re not competing. They don’t want to compete with us. We don’t want to compete with them. Another thing is that even if they exchange WeChat, the conversations tend to be quite short-lived. Because it’s not a friend you met. This is a potential girlfriend or boyfriend, so either it works out or it doesn’t. If it works out then that’s one person that you got. If it doesn’t work out, you never talk to them again. So it doesn’t really matter if that thing happens on Tantan or WeChat because it’s not a long-term relationship.

ZARA ZHANG: Could you discuss the monetization strategy for Tantan? And would you monetize the same way if it was a Western product?

WANG YU: In the short term, I think our revenue strategy is very similar to how we would do it in the West. Right now, it’s subscription-based. It’s value-added services. You pay a monthly fee for different sorts of premium features. But it’s a freemium product. You can use it very well as a free user, but if you buy a subscription it’s much better. So that’s what we’re doing and it’s looking quite good actually. I can’t go into numbers but it’s looking quite good. But one of the things with China versus the West, I think, is that the median ARPUs tend to be lower, but you can definitely get the average ARPU to be higher.

ZARA ZHANG: Yeah. It’s a lot of outliers.

WANG YU: It’s mostly about the top 1%.

HANS TUNG: Yeah. The whales.

WANG YU: If you find revenue models where there is no limit on how much you can spend– because the subscription, obviously, if you want to spend more, there’s a ceiling. But if you find revenue models such as in games or live streaming where you can spend as much as you want then there is people spending tens of millions a month. Basically, those are 1% generating 90% of the revenues. You significantly increase the average ARPU. Those are things we are going to look at going forward, but there is no need for us to be too aggressive about it yet because we’re still mostly growth-oriented. It’s mostly about growth still.

HANS TUNG: If you look at the profiles of the most active users on Tantan, are they all just young users or are they different profiles, different kinds of users?

WANG YU: We used to be very much about very young people, but it’s been expanding. It’s more expanding towards all sorts of people, but it’s more of a focus around young people.

HANS TUNG: How young are we talking about?

WANG YU: I think the average age used to be 22. Now it’s 25.

HANS TUNG: Interesting. It used to be people in college. Now it’s people who have been out of college and working for a couple of years.

WANG YU: Yeah. We’re seeing a lot of surge of users who are 35-plus and so on as well.

HANS TUNG: Good. And for those who are 35 or even older, what has attracted them to come?

WANG YU: The thing is, like I said, there are really no alternatives, neither online or offline. So if you want to meet someone, if you’re a single or a divorced person or something, that’s a gangxu. That’s a super pain point. You have to solve it. So as long as they know about us, they’re going to try us out.

ZARA ZHANG: How do you think dating culture has changed and do you think Tantan contributed to that change?

WANG YU: I definitely hope we have contributed to the change, but I think the revolutionary service was really QQ. They opened up this whole web friend notion. It’s like QQ, Facebook, WeChat in the early days were all about dating. And they were smart about it, so they quickly switched over to friends management which is much larger. For us, since WeChat already owns that market, it doesn’t really make sense. That’s how they did it. But early on for QQ, it was all about strangers. No one had any friends. People my age that I know of who grew up in China, half of them met their partners through QQ. That’s how big I think it is.

So I think the whole online romance thing was really revolutionized by QQ. It has been widely accepted among the population for years. That’s nothing we really made any difference to. I think what we hopefully contributed to was that we made it easier in recent years. Because in the recent years, there has not been too many good solutions. I wouldn’t say we are that awesome either, but we’re definitely better than most of the competitors. Even though I think we have big room for improvement, we’re still better. That’s why we have been able to generate this significant amount of matches, over 10 billion matches to date. And each match is for two people, so it’s like 20 billion match experiences. We don’t know how many people got together because of this, but even a very conservative estimate is tens of millions. And there is like 300 million singles or people who could be counted in that category, so tens of millions is quite significant.

HANS TUNG: Sure. As with any dating app, you need to have effective measures that prevent frauds and harassment and so forth, especially for the protection of the women users. What are the things that you have done that you’re proud of that have worked well? And what are the other new measures you may want to add over time to make the user experience even better and safer?

WANG YU: I can talk in broad strokes about what we have done. I can’t go into details because it’s secret. It’s like a hacker/anti-hacker game, ever-continuous. The spammers and so on, they have a huge supply chain. There are people who are specialized in terms of creating accounts. There are people who are specialized in terms of getting matches. There are people who are specialized in terms of getting those matches to add their other accounts on WeChat. And then there’s another group of people who chats on WeChat to get their business done.

ZARA ZHANG: It’s like an assembly line.

WANG YU: Yeah, it’s like an assembly line, and all four parts are operated by different people. It’s quite professionally operated. I think we even know the offices of where they work. By now I think there’s only one or two players left that still are able to do anything because the tech barrier of being able to do anything is so big now. I think most people just can’t do it. So it’s all about this hacker/anti-hacker game where you prevent them from being able to do each of these four steps. And you have to do it in a way where, even if they know what you do, they can’t do anything about it. It cannot be based on strategies where if they don’t know what we’re doing then it works. If they know then they can just circumvent it. For example, if they build a certain type of profiles and we find a certain keyword and we ban them based on that keyword, then it’s very easy for them to A/B test, and they just remove that keyword. By doing this, what we did is that we educated them. We made them better. We made them create profiles in the future that look more like a real person, which is really bad. You don’t want to iterate this. If you iterate this, in the end you can’t catch them at all. It’s kind of like Enigma. You can never use the information you get from Enigma. You have to kill them with other means because the whole thing is that you can’t let them know that you know.

It’s quite complicated, but in the end you have to use methods where even if they know what you do, there’s just no way around it. It’s a hard block. For example, they buy bunches of phone numbers, like 10,000 sequential phone numbers. So if you see certain patterns in the phone numbers, you ban all 10,000 and it becomes much more expensive to buy those numbers. Stuff like that. And we have probably implemented thousands of different rules and different AI formulas for how to catch them. It’s actually quite difficult to operate a spam account in Tantan. I think we have probably done this better than any other player in the social sphere and I include all the friends management services out there. We talked to the people managing anti-spam on WeChat and they were scared of how high we set the bar. 80% of the people we kill off are not even defined as spammers on WeChat. The result is that you don’t actually see that many spammers on Tantan, at least compared to most other services.

ZARA ZHANG: In 2018, we all witnessed this so-called fengkou or peak for social apps in China with a lot of companies launching new chat apps or social networks and investors pouring a lot of money into these products. But it’s commonly agreed that it’s still very difficult to challenge the dominant position that WeChat has in all Chinese people’s social lives. Do you think startups in the field should seek to challenge at all or should they just work with WeChat?

WANG YU: I think social is a pretty broad concept. This fengkou we’re referring to, I think part of it, where we’re talking about socializing with strangers, that erupted because of our acquisition I think. People suddenly saw, okay, you can actually succeed with a social app and there are actually quite good exit opportunities. That’s, I think, more than half to the fengkou. And then the other part, which is people doing the friends management part trying to challenge WeChat, that I think is much more difficult. I wouldn’t say it’s impossible. I think there’s definitely opportunities out there because to date there has not been a proper implementation of something like Instagram or Snapchat in China. People failed for years with doing Facebook in China until WeChat got it right. You had Kaixin. You had Weibo. People were debating whether Facebook with the whole friends feed is something that works at all in China. And obviously it works. It’s not about if the concept works or not. It’s just about if you are able to operate it correctly, like I said with Tantan’s case as well. Social is quite difficult. You have to really get into the minds of users and then you have to implement a simple solution. So I think you need a lot of experience. Zhang Xiaolong had a lot of experience before he did WeChat.

HANS TUNG: A lot of experience. WeChat didn’t come out of nowhere. He did Foxmail for a long time.

WANG YU: Yeah. I think you need entrepreneurs like that who know what they’re doing. For example, I think if Wang Xing would try something like that, he would have a pretty good chance. He knows his stuff with social. But there’s not that many entrepreneurs who really have that insight. For us, we failed for seven, eight years to learn what we learned. There’s not that many people with that experience. If you do e-commerce, you really don’t become much better at social. That’s, I think, the biggest block. You need the right kind of entrepreneurs and you need a lot of luck. That combination just hasn’t really happened yet.

HANS TUNG: Only very few can do it. Maybe you can do it again or someone like Alex Zhu or Louis from Musical.ly if they come out and do it again maybe. It’s just not many people who can do it.

WANG YU: Yeah, exactly. There’s not a lot of people. And you need a lot of luck. And WeChat, I think, is much more dominant than Facebook is as well because they have the whole IM part. And IM I think is even more important than friends feed.

HANS TUNG: Earlier you mentioned that when it comes to ARPU for Chinese users, the median ARPU is low but the mean could be high because of the whales on the high end. In terms of monetization, you also mentioned that maybe gaming or live streaming are better ways to capture the whales and get them to spend without much upper limit. So when it comes to monetization for you, why did you choose not to build that out yourself and instead decide to sell to Momo who was pretty good at those two areas? And when you were thinking about selling, besides Momo did you think of anyone else like Tencent who also has capabilities in these two areas as well?

WANG YU: The thing with Momo, we were not looking for an acquisition. We were raising our E round. And it was going exceptionally well. I think two weeks after we opened up the round, we already had term sheets and even finished the DD. So we would have been able to close the E round. But at that point, Momo approached us through our financial advisor, China Renaissance. We had never met with Tang Yan before, so we didn’t really know what to expect. It was quite an interesting idea to meet up. And we did it and we hit if off quite well.

HANS TUNG: Both of you are product guys.

WANG YU: Yeah. And, like I said, it’s quite difficult to succeed in the social sphere. You notice that a lot of the philosophies are actually quite similar. Momo has gotten a lot of traction from media and all sorts of people.

HANS TUNG: Tang Yan’s very good. I met him before. I know how smart he is. And he understands social.

WANG YU: Yeah. That was a really good feeling. And the other part is that at that point, we hadn’t really gotten started with monetization, and we didn’t really know what to expect with it. Nowadays, since we have succeeded quite well with it already, we are quite confident. But in those early days, just a year ago, we had less than 10 times what we have now. So we didn’t really know what would happen. And for us to keep growing, we would need to spend much more in terms of marketing. We would need to significantly increase our marketing budget. So the risks were definitely there. And I thought that on the stock exchanges, the market caps of companies were a little bit too high. Very big chance of a big drop. Now it happened just because of this Trump thing, but anything else could also have made it happen and it could have been even worse. If that happened and we would stand there having spent our money on marketing and not really have gotten revenues to grow, we would be in a pretty bad spot. So we definitely had that risk to assess.

The second part is that working along with Momo and Tang Yan, if you look at us today, we are running the company quite independently in terms of the product and operationally. For us, the difference of getting acquired versus raising a round is actually not that different. The feeling is quite similar. And we’re getting a lot of help from Momo. They have a much stronger, much more experienced tech staff than we have. We have a lot of different strengths in terms of how we operate our products, so by exchanging these strengths it’s quite helpful for both sides. Together, we more or less own the whole market. And the value of that versus just competing I think is much better. So that’s how we thought about it and I don’t regret it, even though we have managed to get revenues and so on to succeed nowadays. I think it was a good decision.

HANS TUNG: Alibaba invested in Momo at IPO I think. Did Ali or Tencent approach you as well?

WANG YU: Alibaba was an investor in Tantan too actually. And Tencent is someone we have spoken with for a long time. The issue there, it’s like I said, their main challenge is the dilution. It’s hard to see how we can significantly contribute strategically to Tencent somehow. We can’t really do that much. Regardless of how well we do, it’s more like a financial investment for them. They want to focus more on strategical investments. If we would cooperate with Tencent in the way we do with Momo, we would not get as much help and priority as we’re getting. The synergy would be quite different.

HANS TUNG: Yeah. The synergy with WeChat is not there. The synergy with QQ could, but that will be a very different conversation.

WANG YU: Yes. And QQ is a monster in terms of both the product and the organization.

HANS TUNG: Yeah. There’s not one single person that you can talk to like you do with Tang Yan at Momo.

WANG YU: Yeah. It’s really been a pleasure working with Tuong Nguyen and the whole management team at Momo.

ZARA ZHANG: Has your day-to-day changed after?

WANG YU: Not too much. We do try to extract as much synergies as possible, but we both think it’s better to operate Momo and Tantan as separate products.

HANS TUNG: And when it comes to live streaming and gaming operations, do you plan to leverage their know-how and do more on your own platform?

WANG YU: We’re definitely going to leverage their know-how, but we’re trying to find ways to get inspired by gaming and live streaming to create revenue models that are similar but not necessarily exactly the same. If we do something like live streaming, I want it to be a little bit more hidden, kind of like how Kuaishou does it. Because Kuaishou is probably the biggest live streaming player out there and you could also say they are like a live streaming company, but you definitely don’t feel that they’re a live streaming product.

HANS TUNG: No. It’s a lot more social.

WANG YU: Yeah. If we do something like that, that’s how I want it to feel like. And that’s a challenge but there’s no urgency either. It’s more about growth.

ZARA ZHANG: I think a lot of founders in China, especially the younger founders, tend to have this attachment to what they’ve built and they’re less open to M&A ideas. What is your advice to founders? At what point should you say, I built this but it’s probably better off as part of another company?

WANG YU: I don’t know how blunt I can be. To be just completely blunt, if you do an IPO as a founder, you’re quite far from seeing the money. That’s the major difference. You get a lot of face, but not so much cash. I don’t know if you’re doing a startup for face or for money in the long run. I think most people are for the money.

HANS TUNG: Right. This way you get the best of both worlds. But the trick is finding someone like Tuong Nguyen who can be a partner. Without that, you have to deal with bureaucracy. Then you’re pretty much giving up your baby and taking the cash out and you won’t have much control of the product afterwards anymore.

WANG YU: Yes, exactly.

ZARA ZHANG: For this episode, we experimented with crowdsourcing questions from our listeners community. I’ll ask a couple of questions from our listeners. By the way, here’s a message for our listeners. If you want to contribute questions that Hans and I ask on the show, please join our listeners’ WeChat group or Slack channel at 996.ggvc.com.

Here’s a question from Jenny who is a co-founder of WalktheChat, a WeChat marketing agency. What is the crazy story of managing data in China? What kind of regulatory difficulties have you encountered with regard to data on your platform? For example, users might post revealing photos of themselves. And how do you make full use of data other than for optimizing the product?

WANG YU: Tantan, we try to be a very data-driven company. We ran P1 for many years and we did it in all sorts of wrong ways. But I think in the long run, we felt everything we saw from Facebook made so much sense. So we wanted to operate the product being data-driven. So, run A/B tests. Don’t try to be too top-down-driven. Don’t rely on genius leaders. Get anyone with bright ideas to try out their stuff. And if data-wise it’s good, you launch it. If it’s not, you drop it. And fail fast, fail early. Release early and iterate, that sort of thinking. And aside from that in terms of data, obviously, AI is a big part, so algorithms. And for us, the people you see is based on an algorithm. That’s the key part of our product. That’s obviously a big part of how we use data to try to optimize the user experience for users. But I don’t think of that as having any privacy implications because no one is looking at it. It’s just machine learning.

In terms of other parts, like I said, we manually moderate all user content. So all explicit contents are filtered out. I don’t think you would ever see anything bad on Tantan. But the thing is that since that’s been the culture for so long now, people don’t upload explicit content. I think it’s less than 0.1% of our content that actually is bad in that way. Most of the content we remove, it’s not a person. It’s just a shoe or something. That’s the biggest issue. Explicit content and stuff like that is actually not a big issue. And like I said, spammers, even if they do something, they don’t dare to do anything on Tantan because we have automatic monitoring of user activity. They tend to try to do it on WeChat. Even people saying bad words in chat and so on, it’s a super small percentage of the actual chat. I don’t know if that answers the question.

HANS TUNG: Here’s the next question from Jason Liu, a software engineer at Google. How did you validate your product market fit and get the first round of active users? And what kind of metrics did you use to help you to do product iterations initially?

WANG YU: I think I went through most of that. One part was this offline acquisition of users. One part was that we invited all users on P1 to join Tantan. That’s a unique asset no one else has because that’s basically the best-looking guys and girls you can find in China. We didn’t get too many people to sign up. It was like 20,000. In terms of scale it didn’t make a big difference, but in terms of seed users, you just can’t get anything better. That’s one part. And the rest I described in terms of the philosophy and the operational rationale.

HANS TUNG: I think a lot of people who want to do their own startup underestimate how much learning one has to have the patience to pay for. Without doing P1, you would never have made Tantan work. The idea could be exactly the same.

WANG YU: Oh, yes. We were so far out wrong in so many fields of our philosophies with P1 at different time points that it would just be laughable.

HANS TUNG: And a lot of times, experience from a big company, whether it’s even a Tencent or Ali or Google, may help somewhat, but unless you make your fair share of mistakes and learn from them, it’s not enough.

WANG YU: Exactly. One example: the first two, three years of P1, we were not product-driven nor tech-driven. We were brand-driven. All was about building this high-end brand people would believe in. And we did that quite well, but that’s not how you succeed with scaling an internet company.

HANS TUNG: But without that experience, you wouldn’t have the initial users to do Tantan.

WANG YU: Exactly. There’s no way you can succeed with Tantan with trying to do it brand-driven. There’s just no way.

HANS TUNG: It also feels like no matter what you want to do, you have to do something you personally, passionately believe in so that even if the road you took is different from other people’s route, over time you learn and adjust and everything you did before becomes useful for the next venture.

WANG YU: Yes. Definitely.

ZARA ZHANG: Finally, a question from Ini Wong, a business development professional in Hong Kong who is also Swedish-Chinese. What are the skill sets and values from Sweden that has helped you in building your startup in China? What are the advantages and disadvantages of being Swedish-Chinese?

WANG YU: In terms of first disadvantages, I knew very little about China. And the problem was that I was so bad, I didn’t understand how bad I was.

ZARA ZHANG: You didn’t know what you didn’t know.

WANG YU: That’s a common issue, I think, with most things. I’ve seen many employees or competitors, you’re not good enough to understand how bad you are. It took me seven years to realize that, okay, I don’t understand this market.

HANS TUNG: That’s very honest and fair.

WANG YU: After I realized that, I tried to compensate for that. And it only took a year to start to understand the market much better. But unless you see the problem, it’s hard to solve it. That’s one of the disadvantages. Advantages, I think, like I said, having seen many perspectives of different things, I think you have a much open mind about everything. And Sweden is a very communistic society.

HANS TUNG: Socialist. Socialist society.

WANG YU: Yeah. I think it’s even on communist level.

HANS TUNG: Okay. We’ll see. Maybe that’s a controversial statement.

WANG YU: It’s a culture issue. Everyone is very equal. And that, I think, keeps employees happy. It’s easier to interact with people when you don’t have this bossy attitude. I think that’s one of the things that we have had going for us. We have pretty happy employees generally.

HANS TUNG: You mentioned that you don’t believe in having a genius boss. Everybody can have an idea and use data to test it. If it tested as good, you implement it.

WANG YU: Yeah. If you look at Facebook, most of the stuff that actually made it work were not from Zuckerberg. It’s from just random employees trying out something. The genius of Zuckerberg is trying to build a platform that encourages this and makes it very easy. Back in the days where they were doing just PHP and websites, they could A/B test single lines of code. I don’t think anyone has ever been able to do that except for them.

ZARA ZHANG: And I also think because of your experience living in Western society, you know how dating could be different and how the ways people meet each other could be different. If you only grow up in China, all you see is the Chinese way of interacting.

WANG YU: Oh yes. And also we have gotten a lot of inspiration from Western companies, and I think it’s quite hard to understand them without having my experience from Sweden.

HANS TUNG: I want to ask one more question before Zara goes into the quick-fire session which is, like you said, not many founders have experience living and working overseas in addition to making something work in China. Yet a lot of Chinese models can be exported to other emerging markets around the world. The next new 1 billion users will come from outside of the US and China. What do you think of companies like ByteDance trying to expand in foreign markets? And for you personally, do you feel that either inside Tantan or Momo or something new on your own later, do you want to do something that’s more global in the next 10, 20 years?

WANG YU: The answer to that is just a simple yes, definitely. Once in a time in P1, we tried to define our core values. They didn’t help at all but for what it’s worth, one of them was a global perspective. I think that’s one of the things that makes us different from other companies. We definitely have a global outlook of everything we do, but we have been focused on China so far. We definitely think we have a good shot, both expanding into different areas in social in China and also within dating and social abroad. That’s something we have already started doing some work in, but it’s not the highest priority yet.

In terms of ByteDance, I think that’s a super impressive company. They’re probably the one in China who has implemented this whole Facebook type of operation the best, this whole data-driven, A/B testing, opened up, everyone can try different sorts of things culture and processes the best. They are very aggressive, maybe too aggressive. If there’s anything that would be to their disadvantage, it might be their aggressiveness. But that’s also their strength. We will see how it goes, but it’s definitely a very exciting company and one of a kind I think in the world. There is no one that I’ve seen in the West who are close to their level of aggressiveness.

HANS TUNG: That’s fair. We do see a lot of founders from Latin America, from Southeast Asia, from India beginning not just wanting to learn from US, but increasingly learn from what founders in China have done as well. It’s just super exciting for us to be in the middle of everything and help founders learn from each other. We think that building a global community, global product is a much more interesting way to live.

WANG YU: Yeah, definitely.

ZARA ZHANG: So next we’ll jump into a round of quick-fire questions so you can just say the first thing that comes to your mind. Who is the entrepreneur you admire the most and why?

WANG YU: I would say Steve Jobs. It’s just because he has been able to have so clear visions of what he’s doing and have been nailing them so well and have been able to get his whole team to raise the whole standard to his level and get them to actually ship the stuff on time. I think that also came from his misfortunes that he learned from. He would never have been able to do it if everything just went well.

ZARA ZHANG: What’s a book you’ve read recently that you would recommend?

WANG YU: I have not read too many books recently.

HANS TUNG: Do not have time.

WANG YU: No. San Ti, The Three-Body Problem. I was not too impressed by the first book. And then I was even less impressed by the beginning of the second book. But from the second part of the second book until the whole of the third book, amazing in terms of just the sci-fi part. Wow. I think the people interactions could be better, but in terms of just the sci-fi part, awesome.

ZARA ZHANG: Lastly, what do you do for fun?

WANG YU: I just came from boxing training. I’ve been doing a lot of boxing. Not hitting other people. Mostly trainers with pads. I need to lose weight, so this is super good for that. Aside from that, I don’t know. Watch some movies. Play some games. I’ve basically bought every game you can buy, but I have not played anything really since college days. I used to be an elite Quake player. I played Quake for like 12 hours a day for four years. I was top 10 in the world at a point in Quake 1. But then that took too much time so I quit and haven’t really picked up games too much.

ZARA ZHANG: What do you see yourself doing in 10 years?

WANG YU: 10 years? I have a pretty strong entrepreneurial spirit so I’m probably going to be in some kind of entrepreneurial project.

HANS TUNG: I’m sure you will. But this time will be more global.

WANG YU: Yeah, probably.

ZARA ZHANG: Thank you for your time and we really enjoyed this conversation.

WANG YU: Thank you. It was really fun.

HANS TUNG: Thanks for listening to this episode of 996.

ZARA ZHANG: GGV Capital is a multi-stage venture capital firm based in Silicon Valley, Shanghai and Beijing. We have been partnering with leading technology entrepreneurs for the past 18 years from Seed to pre-IPO. With $6.2 billion in capital under management across 13 funds, GGV invests in consumer new retail, social, digital internet, enterprise cloud and frontier tech.

GGV has invested in over 290 companies with more than 45 companies valued at over $1 billion. Portfolio companies include Airbnb, Alibaba, Ctrip, Didi Chuxing, DOMO, HashiCorp, Hellobike, Houzz, Keep, Slack, Square, Toutiao, Wish, Xiaohongshu, YY and others. Find out more at ggvc.com.

We also highly recommend joining our listeners’ WeChat group and Slack channel where we regularly share insights, events and job opportunities related to tech in China. Join these groups at 996.ggvc.com/community.

HANS TUNG: If you have any feedback on this podcast or would like to recommend a guest, please email us at 996@ggvc.com.