By Garbo Tian

Spring 2026 marked a tipping point for the “One-Person Company” (OPC), which rapidly became a dominant buzzword in China’s startup discourse.

Advances in artificial intelligence have made it increasingly plausible that a single individual can execute tasks that once required full teams. Across Beijing, Shanghai, Shenzhen, and Hangzhou, local governments have doubled down on this narrative, rolling out subsidies such as free workspaces and housing, compute vouchers, and startup grants. Even OpenAI CEO Sam Altman’s remark about a “one-person unicorn” has been repeatedly cited as a shorthand for this emerging entrepreneurial imagination.

Yet the distance between narrative and reality remains wide.

In March, a reporter from South Reviews visited Shanghai and entered two early-stage “one-person company” communities and offline networks. The goal was to examine how this model functions in practice. What emerged was a quieter, more grounded reality: while AI can generate pitch decks, assist consulting work, and build simple websites, it has not yet replaced core decision-making labor. The real burden remains structural and human—what should I do, what do users actually want, and how can this be sustained?

This does not mean the “one-person company” is fiction. Rather, it signals a structural transition. As young professionals exit or are displaced from large corporations, state-owned enterprises, or private firms, entrepreneurship is no longer a simple substitution for employment or income. It requires a fundamentally different mode of work.

Once the institutional scaffolding of employment—stability, benefits, and organizational buffering—is removed, individuals are left directly exposed to market volatility and behavioral uncertainty. The transition is neither smooth nor linear. It brings disorientation and strain, but also autonomy and, in some cases, a new kind of enclosure built on self-discipline rather than hierarchy.

What follows is a portrait of individuals attempting to reconstruct work in a volatile environment, outside the boundaries of traditional firms.

1. Becoming a One-Person Company

Karen has not held a formal job for two years.

More precisely, she is no longer bound by an employment contract, daily supervision, or a fixed salary schedule.

She left her previous company after more than a decade in the workforce, not as a calculated entrepreneurial move, but as a consequence of prolonged workplace conflict that left her physically and mentally exhausted.

Initially, she assumed she would simply find another job in 2024. That did not happen. Most of her applications went unanswered. She then attempted co-founding a startup, but fundraising proved difficult.

“If you don’t already have revenue and a strong technical background, it is very hard to raise capital,” she said.

The investment climate had already shifted. The era when a pitch deck alone could unlock venture funding was fading. Investors were increasingly focused on return metrics. Capital had become more selective, more conservative, and less willing to subsidize early-stage uncertainty.

Karen recalibrated her expectations. Instead of immediately seeking funding, hiring, or office space, she began looking for low-capital ways to work. She encountered the concept of the “one-person company” in a foreign book. The question she could not shake was whether OPCs could exist beyond content creation and self-media work.

From June 2024 onward, this question became persistent. She began organizing gatherings, searching online for OPC case studies, and hosting weekly meetings. After engaging with more than 2,000 people, she built an offline community called SoloNest, which gradually became her new work.

2. From Manager to Solopreneur

Renault is a regular presence in that community.

Six years ago, he was a mid-level manager at a major internet platform, earning a high salary. Three years ago, he ran a startup with more than 60 employees. Today, he describes himself as a “one-person project manager,” having outsourced all his staff.

Looking back, launching a startup in 2020 carried significant risk. But at 28, he felt compelled to leave.

He sensed stagnation. His work had become repetitive for a full year, with no meaningful technological or product iteration.

“In big tech, there’s a lifecycle. The dividend window for new technologies is about three years,” he said. That period allowed him to rely on accumulated momentum. At the same time, he had secured a promising new project with confirmed investment, which gave him the confidence to leave a stable salary and start a company.

After two years in which his income dropped to roughly one-tenth of its previous level, performance eventually improved. But new pressures emerged.

A major client stopped cooperating after one year and began building its own team, wiping out more than 60% of revenue. Around the same time, a partner refused to pay outstanding balances, leaving a RMB 2 million contract partially unpaid and pushing the company to the edge of collapse.

He described recurring mornings and late nights spent worrying about payroll for more than 60 employees.

“I’d wake up thinking: this month I need to spend 600,000 RMB,” he said. “Revenue becomes secondary. The cash outflow comes first.”

Eventually, he could no longer sustain the structure. He reorganized the company. Some departments were spun off into independent entities. Others were no longer directly employed, instead operating through project-based contracts. Employment shifted toward modular collaboration rather than fixed roles.

By converting full-time staff into external contractors, Renault preserved his own position. He became a “one-person project operator” without office overhead or payroll obligations.

Karen and Renault represent two different entry paths into the OPC model: one driven by employment breakdown and job market friction, the other by the financial pressure of scaling firms with volatile revenue structures. In both cases, the employment contract is loosening.

3. After Employment Ends

Once the employment structure is dismantled, Renault found that his work changed fundamentally.

Managing a 60-person company meant constant administrative and managerial load. Reducing the organization to a one-person core eliminated much of that friction. His focus shifted to output rather than supervision.

But this shift is not cosmetic. It rewrites the entire operating logic: hiring systems, compensation structures, evaluation mechanisms, and risk allocation.

In his live-streaming business, roles such as camera operators, stage assistants, and hosts are now outsourced. Previously, he paid a host a monthly salary of RMB 15,000, plus about RMB 3,700 in social insurance costs. Now he pays RMB 300 per hour. A six-hour session is settled immediately after completion.

“It has nothing to do with sales volume,” he said.

On the surface, the arrangement appears neutral. Performers gain flexibility and potentially higher hourly wages. Renault gains flexibility in assembling teams without fixed payroll obligations.

Yet the underlying logic is market-based discipline. If performance is insufficient, there is no next contract.

“I don’t enjoy management,” Renault said. “But I still need to ensure fairness.”

He estimates that roughly 80% of past revenue came from his own efforts, while the remaining 20% came from more than ten salespeople. In his view, that imbalance required correction. “The cost structure inside companies is often misaligned with contribution. That’s why layoffs, performance reviews, and internal systems exist.”

In this model, income is increasingly tied to output rather than employment status. Evaluation becomes externalized to the market.

Renault reports a net margin of about 45% under this structure. However, his former finance manager—doing the same job with more flexibility—now earns roughly half of what they previously made and must take additional freelance work to compensate.

The shift exposes the other side of the equation: risk has not disappeared, it has been redistributed.

4. Fragmentation and Its Limits

In the film industry, this outsourcing logic is already common. Independent producer Lola is direct about its implications.

“Risk has been transferred,” she said. “To maintain flexibility and lower costs, the uncertainty is pushed onto execution teams.”

From the employer’s perspective, fixed costs are reduced. From the contractor’s perspective, employment protections—stable income, social insurance, and legal safeguards—are weakened or removed.

“At times, I feel like a business opportunist,” she said. “If a project fits you, we work together. If not, I move on.”

In practice, however, the ecosystem is uneven. While the industry structure allows flexible recombination across genres and production cycles, access is highly dependent on networks.

When they met, Lola listed her contacts at speed: “93 screenwriters, 113 directors, 93 AI creators, 36 MCN agencies, 18 investment channels…” It took her six years to build this network.

But not everyone has such infrastructure. Without strong connections or institutional backing, many independent creators struggle to secure consistent work and may eventually return to traditional employment for stability.

The OPC model, in other words, is not evenly accessible. It depends heavily on accumulated social capital, reputation, and prior institutional exposure.

5. No Clear Winners

The shift from traditional corporations to one-person companies does not produce clear winners or losers.

What changes is the invisible boundary of risk. In traditional employment structures, organizations absorb volatility and distribute it across participants. Efficiency is partially sacrificed in exchange for stability.

Once that buffer disappears, individuals are directly exposed to market judgment. Strong performers may earn significantly more. Others lose the safety net that once stabilized their income and identity.

This raises a structural question: when cooperation becomes modular and contractual rather than hierarchical, what holds it together?

That question becomes central in the next case.

❓ Frequently Asked Questions