From Tantan to Qianshou: A "Shamelessly Persistent" Entrepreneur's Confession | Buming Entrepreneurship Camp

Ten years ago, Wealth magazine interviewed Ryan Holiday, author of Ego Is the Enemy, asking how he viewed the fact that someone with as big an ego as Steve Jobs still managed to succeed. Holiday replied that by the time Jobs returned to Apple, he was no longer as arrogant as he'd been in his youth. He had learned to reflect on his failures rather than continue being driven by ego.

A decade later, "ego" found new resonance at the fifth cohort of BlueRun Ventures' Buming Entrepreneurship Camp. The speaker was Wang Yu, founder of Qianshou Technology. A serial entrepreneur, Wang founded Tantan — a stranger-social app — in 2014. Tantan once ranked second globally in social app downloads and fifth in worldwide dating app revenue. In 2018, the Tantan team was fully acquired by Momo Technology. In 2021, Wang embarked on his third venture, founding Qianshou Technology to focus on marriage-oriented social networking.

When hiring, Wang particularly values people with smaller egos who genuinely enjoy solving problems. He believes that "as AI grows more powerful, professional skills will become less relevant while character becomes more important. When you set aside your ego, the pursuit itself becomes boundless."

In Wang's view, entrepreneurship has three ultimate goals: first, getting things done; second, doing them well; third, advancing human civilization.

As a resident mentor at Buming Entrepreneurship Camp, Wang shared his experiences and reflections from his student days through his entrepreneurial journey. After the session, participants kept returning to three words to describe it: authentic, substantive, grounded.

The following is Wang Yu's talk at Buming Entrepreneurship Camp:

---

I was born in Beijing and moved to Sweden at age seven, where I stayed for over 20 years. Before university, my studies in Sweden were remarkably easy — I skipped roughly 80% of high school classes and could basically flip through the textbook before exams and score near-perfect marks. But university blindsided me; the difficulty gap was enormous. The bachelor's and master's program totaled 180 credits over four and a half years, and in my first three years I earned only 5 credits. I basically failed everything else.

I also played games almost every day — Quake 1, for three or four years straight, 12-plus hours daily, reaching roughly top-ten worldwide. By my fourth year, I realized I couldn't continue like this. I deleted the games, dismantled my computer, and started studying seriously. At first it was excruciating — reading one page took ten minutes, I couldn't sustain it, and had to rest every half hour.

The turning point came when the school assigned me to a "motivation class," basically for the worst-performing students. The course taught something I found crucial: "Think about whether you can turn the 'I have to' in your head into 'I want to.'"

I thought about it — did I see studying as pressure, or did I genuinely want to do it? I decided I wanted to. After that, everything flowed.

This felt like a major breakthrough. I'd been relatively arrogant before — always arriving late to class, sitting in the back row, reading novels while half-listening to lectures. Later I became genuinely humble, sitting front row center, asking questions about anything I didn't understand. My classmates gave me a nickname: "the guy who asks 1,000 questions."

I earned over 110 credits in my fourth year alone, and another 100-plus in my fifth, completing two master's degrees in five years. I basically went from the worst student in the entire university to one of the best.

Smart people often struggle to face their own poor performance. Once they discover they're doing badly, they reject themselves. But who you are and how you're performing can be separated. Even if I'm doing terribly right now, I still believe I can make it better.

My graduation project was at Sony Ericsson, where I worked for half a year. One manager left to start a company doing roaming services. Back then roaming was extremely expensive — 6 yuan per minute to receive calls, 12 yuan to make them. I proposed a solution that would reduce both to roughly 0.2 yuan. This manager filed for a patent and tried to recruit me as CTO with 20% equity.

Though I didn't join that startup, something suddenly clicked about entrepreneurship. Looking around, business opportunities seemed everywhere.

My thinking model was simple: pain point + technical solution + monetization = a business model. Anything people complain about — if you use technology to solve it better than before, then charge for it — that's a business opportunity. It still holds today; opportunities are everywhere.

I've since simplified this further: the essence of business is obvious value differentiation. Put plainly: I give you value, you give me money.

What's value differentiation? Like Yao Ming in a room — obviously tall. If you make your value obviously differentiated, marketing isn't hard; you simply state your differentiation. Even if you don't, users will say it for you. If it's not obvious, you haven't done well enough.

For entrepreneurship itself, value operates on three levels:

Level one: getting things done. Before founding Tantan, we spent seven years building P1, a fashion community, stumbling through countless pitfalls. When we started in 2007, we faced near-bankruptcy roughly every three months. There weren't many VC firms in China then, so we went to Sweden to fundraise, raising 20 to 30 rounds on amounts as small as 100,000 yuan. By the time we built Tantan, we felt we'd finally caught a break and were determined to make it work.

Level two: doing things well. The shift from "done" to "well" completely changes how you think. Doing things well means thinking for the user, obsessing daily over how to deliver value.

Level three: value for human civilization. Moving all of human civilization forward. Few companies operate at this level — most of what Elon Musk does qualifies, and DeepSeek does too.

If you orient toward level three in your thinking and problem-solving, the capability gains far exceed the lower levels. At level one alone, you're fixated on growth, user acquisition, making money. But these don't necessarily deliver user value.

Truly few people aim for human civilization-level value, pursuing obvious value differentiation. Of course, such people may also fail. Munger once noted that a core method of success is finding "weak competition." If you choose to do something sufficiently valuable, you may find surprisingly few competitors, giving yourself more time to genuinely do it well.

I believe character is the critical factor in doing and making things well. When I hire now, regardless of role, I basically look for four character traits.

First, they feel satisfaction from doing things well. Not from what they personally gain, but from the thing itself being well done. Such people will push extraordinarily hard because they enjoy it.

Second, small ego — the "thing" matters, the "I" doesn't. This attitude is genuinely important. People with big egos do things to prove "I'm awesome." If anything threatens their sense of self, they'll refuse to do it, or do something actually wrong but find ways to argue it's right.

Having ego or not has nothing to do with whether someone will argue with you passionately. You can have personal convictions, your own value judgments — as long as you care about the "thing," not the "I." When someone sets aside ego, the "thing" they pursue can become infinite. People who care too much about "I" rarely accomplish much.

Third, they enjoy solving problems, habitually mulling things over. People who like solving problems are generally smart, but I've met extremely smart people who don't enjoy problem-solving — they basically never get things done.

Fourth, "dogged persistence." Whatever happens, smooth or rough, they keep pushing forward with basically unchanged effort.

Anyone with these four traits, regardless of prior experience, can do things well. You tell them "this thing is awesome" and they charge ahead. Because they want to do it well, they're dogged, they put things above themselves, and they keep thinking. When problems arise, you tell them "this part has issues," and they think: "Fantastic! I can make this even better."

Especially as AI grows more powerful, professional skills become less relevant while character becomes more important. The most crucial character trait is that dogged persistence — getting things done.

All three of my ventures relate to social connection and human emotional bonding. A critical point in building social platforms is how to enable precise matching between people.

The film A Beautiful Mind presents a classic game theory example: in a bar with five men and five women, all men are simultaneously attracted to the same woman. The question follows: if everyone pursues her, likely no one succeeds, and they may even alienate the other women. How can everyone be satisfied?

Game theory founder Nash pointed out that if everyone maximizes only their own self-interest — say, all chasing the most attractive person — it leads to suboptimal outcomes for both the group and individuals. The equilibrium requires considering both self-interest and overall coordination. In other words, you can't greedily fixate on the "best" option; you must slightly step back, seeking better solutions for the entire system.

But this problem isn't so severe in reality. Dating and social connection are fundamentally "one radish, one hole." Everyone eventually finds someone they love. The platform's challenge is whether it can deliver genuine personalization. We want to solve "dating" so thoroughly that it becomes a "problem" no longer repeatedly discussed, like transportation or food delivery.

For the vast majority of people, in the sea of humanity there exists at least one person, often multiple, such that if you could meet and deeply know each other, you'd be strongly attracted. The difficulty isn't "whether they exist" but "meeting" and "knowing."

What platforms should truly do is understand people on users' behalf, simultaneously raising the probability of "knowing" and "meeting" to tens of percent — that would be hundreds of thousands of times more efficient. And this matching goes far beyond appearance, encompassing whether users share commonalities: similar life experiences, interests, and deeper motivational drives.

For example, someone working hard itself doesn't move me, but if her core motivation is self-actualization, I resonate. Shared overseas upbringing, specific interests — these all affect connection.

Ultimately, a relationship approaching "100 points" must satisfy at least three conditions simultaneously: you're attracted to them, they're attracted to you, and you share genuine, stable spiritual resonance. Missing any one, it barely holds. And what we're trying to do is make all three happen together as much as possible.

When we built Tantan, we were also doggedly persistent. At least 200 domestic companies made similar products, but only a handful survived. We succeeded mainly because we'd stumbled through so many pitfalls, figured out some key insights — not because we were especially smart or capable.

To achieve results, we'd try method after method. For example, how should a social app display user profile photos? We observed that users on overseas platforms like Tinder and Facebook tended toward authentic lifestyle photos, well-suited to dating scenarios. How to achieve this in China?

Our app supported WeChat account login, but WeChat profile photos were hard to directly reuse on Tantan. Because over half of WeChat users' photos weren't real people — either cartoon characters or celebrities.

So we manually reviewed all photos, requiring real-person photos for registration. This review basically lost us 40% of users. We'd tell them they could re-register with a different photo; roughly 10% would return.

Another example: during cold start, we tried setting up booths on university campuses, handing out Cornetto ice creams to attract student registrations. On day one, a single school generated 100,000 user swipes — but almost all left swipes (dislikes). Students weren't putting effort into presenting themselves well.

We internally discussed and concluded this user acquisition approach wouldn't work. We tried redirecting trendy users from our fashion community P1 to Tantan. But these people were limited in number, only bringing Tantan tens of thousands of daily active users.

Later we adjusted our approach, doing ground promotion in Sanlitun, Xidan, and Nanluoguxiang — trendy areas with lots of young people — specifically hiring attractive guys to recruit attractive women, and women to recruit guys. It worked extremely well; users found it easier to discover people they liked. During cold start, we made the user experience hit maximum satisfaction immediately.

Our character and attitude let us pull our feet out of these pitfalls and keep walking.

Q: If an app can collect very detailed information about people, could AI clone user data to chat and break the ice on users' behalf? Is this direction viable?

Wang Yu: Many AI social startups are pursuing this direction. I think it's probably not reliable.

Before considering whether AI can do things humans find difficult, first look at whether AI can do what humans already do well.

Put yourself in this scenario: can your best buddy date on your behalf? He has no idea who you like, can't represent you in conversation, and even if he chats well it means nothing for you. Even if my AI chats well with your AI, it's completely meaningless for you personally.

Don't walk around with a hammer looking for nails. You now have AI as a tool and think "how do I use it?" — that approach doesn't work. Start from problems and value: what do people need? How do you solve it? Then see what technical methods can achieve it. Starting from the tool is basically unreliable.

Q: With AI emerging now, what new opportunities do you see?

Wang Yu: Actually, opportunities exist everywhere regardless of AI. For AI applications or AI assistant-type products, the core is that it comes to you, not you going to it. Like Siri — even with strong capabilities, you never think to use it. The breakthrough point is that it comes to you, but at the right timing and in the right scenario, otherwise it becomes intrusive.

Q: When recruiting, how do you screen for suitable partners?

Wang Yu: People's character is hard to disguise. In interviews, listen to them talk about what they've done, watch what makes their eyes light up — their fundamental character becomes clear, and you'll know whether they get excited about the work itself or about others' praise.

If you want to assess someone's skills, best to have them solve an actual business problem you've struggled with on the spot. You've gone through the struggle and effort to solve it, so you know where the pain points are. If they can quickly crack it, that's impressive.

Yutong Zhang of Moonshot AI Shares Her AI Entrepreneurship "Chain of Thought" | Buming Entrepreneurship Camp

AI Native Builder: You Can Always Find a Way to Greet the World | Buming Entrepreneurship Camp

Winning the World's Admiration with Bare Hands in the AI Era | Buming Entrepreneurship Camp Now Recruiting