--- title: "OPC Is Not a \"Company That Hasn't Grown Up Yet,\" but a Business Entity That \"Deliberately Doesn't Grow\"" author: "Unique Research" sourcePublication: "Unique Research Substack" originalPublishedAt: "2026-05-26T13:43:18+00:00" canonical: "https://ffcap.cn/en/research/src-20260526-01html" source: "https://uniqueresearch.substack.com/p/src-20260526-01html" language: "en" --- # OPC Is Not a "Company That Hasn't Grown Up Yet," but a Business Entity That "Deliberately Doesn't Grow" _Original · Unique Research · 2026-05-26_ _Editor's note: The first-person report and its judgments belong to the original Chinese author. This English rendition retains the opening, all nine Q&A sections, and the closing. Figures, percentages and demographic claims are source or author observations, not independently audited findings. "OPC" refers to one-person company. "小龙虾" (lobster) is a colloquial reference to Claude AI. Company, platform and personal names are preserved as stated in the source. The 吴晓波频道 (Wu Xiaobo Channel) interview reference is retained._ Unique Research OPC Is Not a Retreat, It's an Upgrade A one-person company is not "a company that hasn't grown up yet," but a business entity that "deliberately doesn't grow." " OPC is not "a company that hasn't grown up yet," but a business entity that "deliberately doesn't grow." Small companies think every day about expanding teams, raising funding and scaling up; OPC thinks about maintaining profit, maintaining autonomy and staying lean. Recently I agreed to an interview with the Wu Xiaobo Channel and was hit with nine consecutive questions about OPC (one-person company). After recording and listening back, I found what I said was quite pointed. Let me organize my thoughts and share them with everyone. 01 · Is OPC a New Species? To be honest, OPC is not a new species. Freelancers, independent consultants and personal studios have all existed for a long time. But this round is genuinely different—AI has pushed back the boundary of what one person can produce. Previously, one person would be saturated serving 3 clients; now, with AI, serving 30 is not impossible. Its core difference from a small company is: OPC is not "a company that hasn't grown up yet," but "a business entity that deliberately doesn't grow." Small companies think every day about expanding teams, raising funding and scaling up; OPC thinks about maintaining profit, maintaining autonomy and staying lean. I have seen OPCs with annual revenue over a million who firmly refuse to hire—not because they can't afford it, but because they see no need. Then what about super-individuals like micro-merchants, streamers and independent creators—does using AI make them OPCs? The key is whether you have a system. If you still trade personal time for money, doing one gig to earn one fee, and AI just makes you do it faster, then you are still a super-individual. But if you use AI to build a delivery system that can run automatically—with products, SOPs and repeat purchases—then that is OPC. Compared with the 2015 "mass entrepreneurship and innovation"? That round was capital-driven; this round is tool-driven. The logic then was: make a PPT, raise funding, burn money, tell stories; now it's: start at zero cost, run it as a side hustle first, go full-time once there is profit. This is a survival model, not a narrative model. 02 · Who Is Doing OPC? The ratio I see is roughly 70/30—first-time entrepreneurs account for 70%. But frankly, those who do well tend to be the 30% serial entrepreneurs, or at least veterans with ten years of workplace experience. Many young people treat OPC as a retreat, but OPC requires far more comprehensive ability than having a job—you have to be sales, product, delivery and finance all at once. People who have never experienced a business closed loop can easily end up "taking work but failing to deliver." Common traits in the profile: mostly aged 30–40; deep accumulation in a vertical field; prioritize freedom over getting rich. In terms of personality, self-drive and the ability to withstand loneliness are two hidden filters. OPC has no colleagues, no KPIs, no one pushing you—I have seen too many people collapse in three months—not because of ability, but because without an external rhythm, they crumble on their own. 03 · Is AI Overhyped, or Underutilized? This round of OPC fever is almost synchronized with the "lobster" (Claude) craze, and this is not a coincidence—it is a causal relationship. The explosion of lobster essentially turned AI from a chat toy into a deployable employee. The underlying logic of the OPC boom is "one person + AI employees = a micro-company." Without lobster's ease of use and low cost, AI would still be stuck in the labs of large tech teams, out of reach for ordinary people. So AI is a necessary condition for OPC, but not a sufficient condition. Without AI, the cost of many OPC business models simply cannot be covered; but with only AI and no customer insight or business closed loop, it is also useless. My judgment at this stage is: overall overhyped, but underutilized in specific links. The overhyped part: many people think that buying a few AI tool subscriptions and building a workflow means they can start a company. But AI won't find clients for you, won't build trust for you, won't collect final payments for you. The essence of business has not changed; AI solves efficiency, not existence. The underutilized part: OPCs that have already run through an MVP are not continuing to invest in AI-ified operations. For example, customer service automation, content matrix distribution, knowledge base accumulation—these can save 50% of time, but most people are still stuck at manually replying on WeChat. Right now most OPCs' use of AI stays at "writing copy" and "making PPTs"—this is tool-level use, not workflow-level reconstruction. 04 · Is "Multi-Lobster Collaboration" Common? Prevalence does not exceed 10%. It sounds sexy, but the reality is stark. Let me break it down for everyone: Process complexity: high. True multi-agent collaboration is not opening five ChatGPT windows and chatting separately—it requires clear input-output interfaces, exception handling, quality verification and state synchronization. Essentially, you are building a micro-ERP. If any link in the middle has the wrong format, or the client temporarily changes requirements, the entire chain breaks. Technical capability requirement: medium-high. Low-code platforms like Dify and Coze have lowered the barrier now, but you still need to understand concepts like API calls, prompt engineering and RAG knowledge bases. People with purely business backgrounds get stuck here; people with technical backgrounds often get stuck on business understanding. People who understand both business and agent building are as rare as phoenix feathers and unicorn horns. Cost: severely underestimated. Looking at subscription fees alone, it's just a few tens of yuan a month, but the hidden costs are extremely high. Building a stable multi-agent workflow often takes 40 to 60 hours. Many people only count the $20 API fee and don't count their own 40 hours of time cost—valued at an OPC entrepreneur's time, that's equivalent to spending several thousand yuan. Effect deployment: currently only two types of people have succeeded. One is product managers with a technical background; the other is people with extremely standardized business, such as those who only do "resume editing" as a single SKU. The simpler the business, the more effective multi-lobster collaboration is; the more complex the business, the more irreplaceable human intervention is. My recommendation: for 90% of OPCs, rather than raising multiple lobsters, first raise one lobster to the point where it can stand on its own. Most people haven't even used one AI assistant well, so talking about multi-agent collaboration is a castle in the air. 05 · High Homogenization, Why? It is not a limitation of AI's capability, but a limitation of human imagination. Three causes: low-hanging fruit effect, information cocoon, avoiding verticalization. Should you break out? It depends on what you want. If you want to raise funding, build a brand and build barriers, homogenization is a dead end; but if you just want to earn 50,000 to 100,000 a month and live freely, you can entirely carve out a niche audience in a red-ocean track and live very well. OPC's first goal is profit, not differentiation. There are only two paths to breakout: cut downward—drive general capabilities into extremely narrow industry gaps; cut backward—don't do tools, do operations, helping people who bought tools but don't know how to use them with deployment. 06 · Common Dilemmas and the Divide Between Success and Failure Top three: no clients, no product, no rhythm. Customer acquisition dilemma: many people have skills but can't sell. When at a company, clients were found by the boss; now you have to find them yourself, and you suddenly discover that "strong professional ability" and "able to get orders" are two different things. Productization dilemma: being able to do the work and being able to sell a product are two different things. Many people take on custom projects until they die—they look busy, but their hourly rate works out to less than having a job. Without productization there is no leverage; without leverage, OPC is just another form of employment. Rhythm dilemma: the most common problem for a person working alone is infinite procrastination without an external deadline, or simultaneously opening ten fronts without going deep on any. At a company job there are people pushing you, meetings and quarterly goals; OPC relies entirely on yourself, and many people fall into "false busyness"—scrolling through news for a day, tuning tools for a day, taking courses for a day, but producing no commercial results. Where is the root cause? OPC is not a downgrade, it is an upgrade. Many people think that going solo from a company is "anyway I'm better than the boss," but in reality you lose the organization's leverage—brand, channels, collaboration, financial support. You must rebuild leverage with systems and tools, and most people haven't thought this through clearly, thinking it's just a change of workplace. 07 · "Holding a Hammer Looking for a Nail"—What Are the Characteristics of Successful and Failed OPCs? This is the biggest bubble in the current OPC field, bar none. Specific manifestation: first learn an AI tool, then everywhere look for "what business can I do with this." The result is making "what AI can do" rather than "what the market needs." I have seen too many OPC entrepreneurs whose first sentence is "I'm good at using AI to do XXX," rather than "I solved a pain point for a specific group of people." This is a fatal thinking-order error. What does a successful OPC look like? The nail is clear—first there are specific clients and specific pain points, then use AI to amplify. The monetization loop is short, with repeat purchases or referrals, and humans control key nodes. What does a failed OPC look like? Buy tools first then find scenarios, over-rely on AI leading to unstable delivery quality, no pricing power, business model hidden in "possibility." How much does success or failure relate to AI usage? AI is an amplifier, not a starting point. 70% of OPC success or failure depends on business basics—customer acquisition, product, delivery—and 30% depends on how well AI is used. AI can take a 70-point person to 90 points, but it is very difficult to pull a 30-point person to the passing line. 08 · What Is the Ecosystem Missing? The tool layer, information layer and platform layer are relatively complete. But the credit layer, finance layer, collaboration layer and security layer are severely lacking. Credit layer: there is no endorsement mechanism for person-to-person commercial trust, and the cost of breach is low. Between companies you can sign contracts, use corporate accounts and check credit reports; between OPCs it often relies on WeChat promises, and when there is a dispute there is no arbitrator. Finance layer: personal payment collection, tax planning and social insurance payment are still institutionally treated as individual businesses or flexible employment, with no intermediate layer adapted to the new form. Many OPCs have annual income of 500,000 to 1 million, but pay social insurance at the minimum tier, have no housing fund, dare not get sick and dare not have any interruption in cash flow. This is not entrepreneurs being melodramatic; it is an institutional gap. Collaboration layer: when OPCs temporarily team up for projects, there are no standardized cooperation contracts, profit-sharing mechanisms or intellectual property ownership arrangements. Every cooperation requires renegotiation, and the trust cost is extremely high. Security layer: there is no commercial insurance, professional liability insurance or income-interruption insurance adapted to OPC. Having a job comes with work injury insurance and unemployment benefits; OPC has nothing—once you get sick or something happens at home, cash flow immediately breaks. The most realistic dilemma: OPC is a "neither fish nor fowl" in the current legal framework. It is not a company, not an employee, and the registration process and tax treatment of individual businesses are severely lagging behind the actual business form. Many people either risk using personal payment collection or are forced to register a company increasing management costs, with no better option in between. 09 · Five Pieces of Advice, No Nonsense First, sell first then build, don't build first then sell. Even if only one client pays a deposit, it is ten times better than building a product in isolation. The fastest way to validate demand is to "sell something you haven't built yet"—if the client is willing to pay, then you start working. Second, choose one place where clients can search for you and go deep. Official Account, Xiaohongshu, Jike, Zhihu, Video Account—choose one, don't do all of them. OPC has no team to build out a content matrix, so you must achieve "when they search the keyword they see you" on one platform. Third, treat AI as an intern, not a partner. An intern can help you write first drafts, organize materials and reply to customer service, but ultimately clients pay for your judgment and professionalism. Maintain human control over key delivery links—this is your quality baseline and pricing basis. Fourth, you must receive your first money within three months. If there is no income after three months, it means either your skills are not needed by the market or your customer acquisition method is wrong. Don't wait until "you're ready to start"—market feedback is the best preparation. Fifth, save enough living expenses for six months, then cut off your retreat. The most dangerous state for OPC is "half-hearted"—working a job while running a side hustle, reluctant to give up either. Either honestly work a job and save enough capital, or go all-in full-time and force yourself to run through it. The worst is being stuck in the middle—neither having the stability of a job nor the ruthlessness of entrepreneurship. These are my views. If you want to discuss any of them in depth, just find me directly. --- Original publication: https://uniqueresearch.substack.com/p/src-20260526-01html On-site reading page: https://ffcap.cn/en/research/src-20260526-01html