Written by: Ada, Deep Tide TechFlow

At the end of 2024, Israeli developer Maor Shlomo retired from the reserves, opened his laptop, and began working on a project. With no funding, no team, and no Slack channel, six months later, Wix acquired his company Base44 for 80 million USD in cash. By that time, the product had 250,000 users and a monthly profit of 189,000 USD. In the three months leading up to the acquisition, he hadn’t written a single line of frontend code.

Dutchman Pieter Levels is even more dramatic. One person, no employees, using the most primitive PHP and jQuery, while simultaneously running three products: Nomad List, Remote OK, and Photo AI. His total revenue in 2022 reached 2.7 million USD. He has never had a job, never raised funds, and has lived a digital nomad life across more than 40 countries and over 150 cities.

These stories are too good, good enough to create an illusion that one person plus one computer can build skyscrapers from nothing.

One-person company incubators are flourishing in Shenzhen, Shanghai, Suzhou, and Hangzhou. Nanshan Model Camp is 100,000 square meters with 700 companies applying; Lingang Zero Boundary Cube has free co-working spaces, with 300 spots in the first phase snatched up and the second phase of 8000 square meters on the way. Dan Shao, who is building a one-person company community in Chengdu, filled all his events to capacity in less than a month of operation.

The trend is indeed explosive. According to the "China OPC Development Trend Report (2025-2030)," by June 2025, the number of one-person limited liability companies nationwide has surpassed 16 million, with the number of new OPC registrations in the first half of 2025 reaching 2.86 million, a year-on-year surge of 47%, accounting for 23.8% of all newly registered companies.

But what lies beneath this trend?

We spoke with three individuals who are currently walking this path: an observer operating a one-person company community for nearly two years, holding more than 2,500 real samples; a post-00s entrepreneur who has built two companies bridging the US and China in Silicon Valley; and an independent developer who transitioned from a primary market FA to making AI Agent products.

The stories they told are somewhat different from the narrative seen on social media.

The Underlying Logic of Success

Dai Wenqian operates a one-person company community called SoloNest in Shanghai, and the motivation is quite simple. In June 2024, she just left the internet education industry and wanted to know what one person could actually achieve. After searching through books and videos without finding answers, she decided to conduct field research on her own: organizing meetups, observing samples, and interviewing.

In nearly two years, she accumulated over 2,500 samples, of which 20% successfully ran a business loop and achieved results. She also wrote a book based on her observations of these samples: "One-Person Company."

A joint study by Qichacha and Xiaobang Report shows that the three-year survival rate of companies established in China in 2021 has dropped to 71%, with nearly a quarter going bust within the first three years. Clearly, SoloNest's 20% success rate has outperformed the larger market.

But Dai Wenqian is more concerned about the 80%.

"Those who didn’t succeed fall into two categories. One is those who tried but didn't manage to make it work due to disrupted traffic or unsustainable models. The other group never even started." She analyzed.

The number of people who haven't started is much larger than one might think.

"Because they are not feeling enough pain, there is an escape route. Their brains trick them into thinking they want to start, but essentially they are just afraid of missing out. Most of them are worried about their current jobs but feel that starting an OPC might be a solution, yet fear-driven impulses might be a poor starting point," Dai Wenqian said.

Leon, who has experience as a primary market FA, has encountered even more of such individuals. He attended offline events for one-person companies and found that many people lack direction, participating in various events and buying courses everywhere. "No one can help you figure out how to make money. The only right path is to do it, to step into pitfalls, and to suffer losses," Leo said.

Among those who have started, who has survived?

The answer is quite counterintuitive. Dai Wenqian noticed a stable commonality among the successful samples: those who ran a business loop have almost never continued doing the work of their original industry.

They do not choose their path based on "what I am good at," but start from "where there is unmet demand."

Dai Wenqian gives her own example. Previously at Ximalaya as a brand manager, she had never done offline events or built a community. However, she possesses curiosity about people, aesthetic sense in product creation, and structured communication skills—these foundational abilities are not limited by industry. Once she found the market demand points, scenarios, and industry pain points, she could transfer those foundational abilities over.

There is also a guy in the SoloNest community who makes tennis bags, and he was not originally in the bag business. Due to his love for tennis, he discovered unmet needs in the current market offerings and prepared 100,000 yuan in startup funds to create an original tennis bag. After two years, he now consistently sells over 300 bags a month. This corresponds precisely with the methodology of Pieter Levels. In 2014, Levels set himself a challenge: to create 12 products in 12 months and put them on the market to see which ones received responses. Nomad List was the seventh product and the only one that succeeded.

The key is not just to choose the right path but also to quickly validate enough assumptions.

Dai Wenqian breaks this process down into three critical milestones. The first is whether you dare to create something and throw it onto the market; unfortunately, many people lack even the mindset of validation and merely think without action. The second milestone is whether you can sell it once someone shows interest; there is a gap between "someone thinks it is good" and "people are willing to pay continuously." The third milestone is whether you can free yourself from delivery.

The first two milestones filter out those who are inactive or stuck.

The third is the real hard battle.

The 1.2 Million Trap

Those who pass the first two hurdles will find that while they are alive, there is a ceiling above their heads, firmly welded in place.

Dai Wenqian provides a precise figure: relying solely on self-delivery, the annual revenue ceiling for one person is about 1 to 1.2 million yuan.

"No matter how hard you work, selling time has its limits," she said.

This is the most realistic dilemma for one-person companies in China. On social media, they discuss Maor Shlomo's 80 million USD exit and Pieter Levels’ 2.7 million USD annual income, but those are stories of Silicon Valley SaaS and globalized digital products. In the Chinese context, one-person companies are more common in the C-end, tertiary industry, and experience economy, where the delivery chain is heavy and people are interlinked.

Barry, who operates businesses across the U.S. and China, sees a more fundamental divide. American kids think of entrepreneurship in terms of B2B SaaS and AI Agents. Chinese kids are more focused on tangible industries such as pets, elderly care, and food. This isn't about intelligence; it's about the differences in industrial structure and willingness to pay between the two countries. American companies have a strong awareness of payment, so a small SaaS tool can run successfully, while China's B2B ecosystem is completely different.

So how can we break the 1.2 million ceiling?

The most intuitive path is automation, using AI to remove oneself from the delivery chain.

But this path is much more difficult than what the narrative suggests.

There is a typical case in the SoloNest community. Jason's business involves job support, helping internet operations interns and graduates with resume revisions, mock interviews, and job coaching. He started purely selling time, taking on a dozen clients a month.

Many peers get stuck at this stage. Jason's approach was to find a group of peers who had difficulty with sustainable customer acquisition, train them, and direct clients to them. They share profits per deal, with no employment relationship. Dai Wenqian calls this "multiple one-person companies collaborating, not a single multi-person company." He later expanded into B2B services, transforming from purely C-end to C plus B.

Now, Jason is working on the third step: using the consulting data from the past two years to create a knowledge base and develop a semi-automated delivery product. But he has only completed 60% in two months.

Why so slow? Dai Wenqian provided a mathematical model: "Assume your delivery chain has five key nodes, and each node can only achieve 80% efficiency through automation, can we say the entire chain's automation qualification degree is 80%? It is not; it would likely be 0.8x0.8x0.8x0.8x0.8—only 33%. The longer the chain, the more difficult it is to automate. It's not additive; it's multiplicative."

This is why lobster seems like it can be automated suddenly, but when you actually build it up, you realize that if any single step is not done well, the output is garbage. And the prerequisite for using AI well is that your manual work is already well done; otherwise, you cannot see where the problems lie.

Leon is the one among the three interviewees who has the strongest technical foundation. He is currently creating AI Agent products without writing a single line of code, delegating all development to AI. The penetration of AI in his workflow is nearly 100%.

But he is very restrained in his judgment about automation: "To assess whether a task can be assigned to AI, look at three points: whether the cost is low, whether the risk is high, and whether the effect is good. High-net-worth services cannot rely on AI; the working style of AI assumes you allow it to make mistakes, and it optimizes its strategy through errors. But in services to high-net-worth individuals, mistakes are not allowed; if you make a mistake during one communication, the entire business is ruined."

Some business processes simply cannot be replaced by AI.

Dai Wenqian also acknowledges that her AI penetration rate is only 30%. Because her core delivery involves offline human-to-human interaction, this part cannot be automated. What she can do is partial automation, including content acquisition, knowledge base accumulation, etc., but she cannot completely remove herself from the business.

She works more than 14 hours a day. Creating content to attract new users, chatting with people for screening, maintaining partners, designing products, breaking down samples, etc., and there are also two fixed offline events every weekend.

"Many one-person company founders won't share this kind of thing online. No one looks. Everyone likes to see the glamorous scenes: drinking coffee here, going to exhibitions there, earning millions a year, the leading lady. But the reality is that entrepreneurship involves a lot of dirty work, tedious tasks, constantly repeating and iterating," she said.

A One-Person Company is Not the Final Destination

Automation is one way, but not the only way.

Dai Wenqian observed another breakthrough method: instead of replacing herself, she connects herself together.

Jason’s case embodies this logic. He doesn’t hire anyone and collaborates with other one-person companies. Each Lego piece is an independent entity, each with capabilities and customers, and when combined, they create value.

Moreover, if each one-person company can be enhanced by AI, then when they are combined, it’s like augmented Legos. Dai Wenqian believes this is the greatest source of imagination for one-person companies: "It’s like Legos; not every piece of Lego needs to be 100% automated, but each piece can be augmented by AI. Three augmented Legos combined are not 1+1+1 but 3×3×3."

Another path is to replicate one’s experience and methodology to more people. Barry has validated this model. He is the founder of two one-person companies, and from 0 to 1, he explored everything himself; after running the business loop, he stepped back to delegate it to the team, passing the baton and moving on to other businesses.

Maor Shlomo also had a similar choice. Base44 grew to 250,000 users and nearly 200,000 USD in monthly profit in six months, but he still chose to sell to Wix. His explanation was that although the growth was astonishing, the scale and volume we needed could not organically grow from one person. One person can take a product from 0 to 1, but going from 1 to 100 requires organization, resources, and distribution capabilities—things one person cannot achieve.

Three different paths: AI productization, collaborative connections, partner expansion, but the underlying logic is the same: a one-person company is not the final state. It is a springboard. Verify something with minimal cost; once it is successful, you must find a way to ensure you are no longer the bottleneck. Otherwise, you will be forever welded to that 1.2 million line.

Before the Door Closes

The 2026 data looks impressive. Shenzhen has issued an OPC entrepreneurial ecosystem action plan, aiming to build over ten 10,000 square meter OPC communities by 2027. Shanghai Pudong offers new one-person companies up to 300,000 RMB in free computing power. Suzhou attracted 300,000 college students in 2025, rapidly expanding its talent pool.

But Dai Wenqian said something sobering.

"The barriers are significantly lower now. Previously, finding money, people, and venues had high startup costs. Now you can almost validate anything at zero cost. But this is a benefit that indiscriminately impacts everyone; your easing also eases others, more players lead to more expensive traffic. This is an arms race."

Pieter Levels can earn 2.7 million USD per year alone because he started in 2014, accumulating a decade’s worth of SEO moats and community trust. Maor Shlomo of Base44 could sell his company in six months because he had already founded a data company that raised 125 million USD before, giving him a circle, judgment, and speed that didn’t start from zero.

These individuals are not the "ordinary people can also succeed" narrative of one-person companies. They are the standout points of survivorship bias.

The real world of one-person companies is among the more than 2,500 samples in the SoloNest community: 20% have continuously made money and are moving on to the next phase; 40% are stuck in various ways but still iterating and seeking breakthroughs; while the remaining 40% are still confused about finding direction. Among the 20% who are alive, most earn less than 1.2 million. They work until dawn, fill weekdays, and have no weekends.

In fact, the business of one-person companies profits from the time difference between "a niche demand being discovered but not yet occupied by organized capital." This time gap has a name: expiration date.

The expiration date depends on two things: the timing when you discover this demand and how quickly you can execute it.

Lowering the barriers will not prolong the expiration date. On the contrary, it will shorten it. Because what you can validate at zero cost, others can do too. The demand you see, others can see as well. If you manually produce an MVP and throw it out without failing for three months, tomorrow ten identical products will appear on the same user’s phone.

This is also why most people get "stuck." The essence of being stuck is not a capability issue, but that their realization speed is not keeping up with the crowded market.

Maor Shlomo and Pieter Levels are advertisements rather than samples precisely because they resolved the expiration date issue in two opposing ways. Levels extended the expiration date to ten years through first-mover advantage and compound interest, while Shlomo compressed it to six months through speed and exit.

The middle path is the most dangerous. For most one-person company founders in China, there is neither a ten-year timeframe to slowly build a flywheel nor Wix coming to cut checks. They work 14-hour days maintaining their 1.2 million ceiling, thinking if they just hold on a little longer, they will break through. But the market will not wait for you to hold on. The next peer, who can validate at zero cost, may appear at any time, flattening your little moat.

A one-person company has never been a long-term state; it is a window with an expiration date.

When the window opens, the barriers are low, tools are cheap, and the demand is clear; it looks like the best time for ordinary people. But the window will not stay open forever. It will fill up with later entrants and be crushed by more efficient tools, eventually being completely shut down by a well-funded startup or a suddenly active line of a major company.

Whether you can move away from that bottleneck position before it closes is the only true issue of this business.