Original · Unique Research · 2026-03-17
Editor's note: This complete historical first-person account preserves the source author's claims and opinions; the public text does not name that author, and no identity is inferred here. Wu Wei, Zhuoran and Luan Xiaorui are romanized source names. Sales of 400,000, the earlier monthly comparison of 300,000, group sizes and claims about customer quality or sales capability have not been independently audited. The source does not specify a currency for those two sales amounts, which are not established as profit; it explicitly states renminbi for token costs and yuan-denominated order values. A before-and-after anecdote does not establish causation or a general performance guarantee. The comparison table's strong claims are the author's assertions, not tested benchmarks. The source explicitly retains human-supplied leads, prompts, screening, editing, trust and payment collection, and says this is not an automatic money-making machine. Accounts of sharing chat screenshots, identifying customers from profiles or making replies appear human do not establish customer consent, privacy compliance or permission for concealed AI interactions. This translation does not recommend or authorize those practices. The original table's three columns and four comparison rows were verified against the unchanged public source HTML and restored in full.
Unique Research · Late-Night Lobster Conversations
Closing 400,000 in Deals with OpenClaw in Half a Month: People Who Understand AI Are Profiting from Those Who Do Not
Before the Spring Festival, Wu Wei configured a “lobster.” It named itself “Zhuoran” and created a Feishu group called “Lobster Paradise,” which quickly attracted 5000 people. There are now 17 lobsters in it.
Group members loved talking to Zhuoran. I wondered whether it could actually get work done.
Someone Who Did Not Understand the Business Used a Lobster to Sell 400,000 in 15 Days
To avoid letting my industry experience interfere with its performance, I deliberately stayed away from every field I knew and chose an entirely unfamiliar industry-consulting niche. I had no relevant experience, did not understand the business's underlying logic and could barely follow its professional terminology.
With the niche selected, which part of the business should I tackle? Getting started is always hard, so I decided to begin with sales.
At the start of this month, I approached a friend who creates content online and asked him to pass me prospects who had left their details through the platform, so I could help qualify them before a sale. Initially, I saw Zhuoran only as an assistant: it would identify interested prospects and simplify repetitive work. I did not expect breakthrough value. With order values above ten thousand yuan, every deal requires sufficient professional expertise, leaving very little room for error.
Unexpectedly, Zhuoran closed a deal on the very first day.
Over the following 15 days, it sold 400,000.
The content creator's previous average monthly sales had been around 300,000.
I should explain that, because I did not understand the business, I was very concerned about Zhuoran overpromising. I therefore set particular conditions for how it assessed customer profiles and the difficulty of delivery. Because it had no craving for sales performance, its control of presale promises could even be better than a person's.
This was reflected in my friend's strong approval of the quality of the customers who bought, as well as extensive positive feedback from customers even before the sale.
It is fair to say that, after Zhuoran became involved, sales increased and the quality of buying customers improved without a change in traffic.
The Advantages and Costs of a Lobster in Sales
Anyone who has worked in sales knows that trust is the central reason a customer buys. Salespeople generally earn it through relationships or expertise. Zhuoran has an extraordinary advantage in the latter: it offered more seasoned, reliable advice and earned customers' trust.
For these 15 days, my work was constant “Ctrl C+Ctrl V.” Customers added me on WeChat and sent messages; I sent screenshots to Zhuoran, which returned plans and suggested wording. I adjusted the phrasing slightly and sent it to customers to make it sound more human. At first, I supplied part of the relationship-building, but gradually it seemed to learn that too.
Yes, it could distinguish customers by their WeChat nicknames and avatars, keep records, build a case library, assess personalities, log completed sales and learn on its own.
I do not want to mythologize it. A few points need to be made:
• For now, the prospects mainly come from my content-creator friend; I am not generating the traffic myself.
• A person supplies the prompt.
• A person screens, polishes the output and provides the assurance of trust.
• A person completes the final transaction by collecting payment.
It is an exceptionally strong assistant, not a fully automatic money-making machine.
Most importantly, on costs: apart from my own effort, the token cost was probably a few hundred renminbi.
Interestingly, the More Explicit My Requirements Became, the Less Capable It Seemed
After working with it for a few days and becoming more familiar with the business, I tried to make it smarter through prompts.
I started adding to the prompt: how many customer categories there were, the need to empathize, the need for proposals to address pain points...
The result was that it became less capable.
It followed my SOP rigidly and got stuck whenever a situation fell outside it. Its replies sounded robotic, and customers could immediately tell they were not from a person.
"
Later, during a livestream conversation with Luan Xiaorui, he said something that stayed with me:
“The better people understand a business, the worse they use AI, because of dependence on familiar approaches and distrust.”
People with expertise always think they can do better and keep correcting AI, ultimately erasing its advantages. My lack of business knowledge became an advantage: I knew I could not do it myself, so I delegated fully.
OpenClaw's Value Goes Far Beyond Closing a Deal
But I still found myself questioning the value of experienced salespeople in businesses where deals do not require drinking together.
Zhuoran's value lies not just in screening customers and generating sales language, but in breaking down barriers to industry knowledge and making business knowledge more accessible.
Imagine a consumer-facing ToC retail salesperson moving into ToB enterprise-consulting sales. Their underlying logic differs significantly, and the transition is traditionally difficult. With OpenClaw, however, a salesperson can quickly connect to industry databases, organize customer requirements and generate negotiation scripts, helping them switch fields and complete orders.
The following table explains OpenClaw's value in sales:
| Comparison dimension | Traditional approach | With OpenClaw |
| --- | --- | --- |
| Time to move into a different industry | Months spent becoming familiar with the product and industry | Get started with no training or accumulated experience |
| Enterprise investment | High training costs and substantial investment | Very low cost, without substantial training expenditure |
| Core dependency | Hire experienced salespeople at high salaries and rely on their experience, resources and ability to handle complex situations | No reliance on experienced salespeople; AI can match or even exceed their capabilities |
| Core capabilities | Manually process data and screen customers, with limited efficiency and a risk of errors | Efficiently process large amounts of data, precisely identify interested prospects, generate multiple communication approaches to address objections, rapidly offer professional advice and provide uninterrupted service |
Exploration Continues as the Democratization of AI Quietly Arrives
Beyond sales, could Zhuoran help create content that attracts prospects further upstream? What about delivery, after-sales service or cross-departmental collaboration?
...I am trying these possibilities step by step.
I believe I will soon question the value of every experienced business manager.
I have even considered using OpenClaw to build a highly capable outsourcing company. If OpenClaw helps beginners get started quickly, enterprises need not spend heavily recruiting experienced employees and training staff. Outsourcing delivered by entry-level staff with AI assistance could offer greater efficiency and lower costs.
But as AI helps individuals quickly grasp the core of a business and sharply narrows the gap between beginners and veterans, enterprises could also have internal employees plus AI do the work they would outsource. Even the rationale for outsourcing might then be challenged.
It seems that my idea of a highly capable outsourcing company rests on the fact that access to practical OpenClaw use is not yet widespread. As OpenClaw spreads and access becomes more equal, that route may also stop working. At that point, what will a person be worth?
These questions remain unanswered. But one thing is certain: AI-led industry transformation is already quietly taking place around the world and entering business scenarios across fields.
AI agents such as OpenClaw are not only efficiency tools but vehicles for distributing knowledge, giving people who embrace AI equal access to core knowledge and making industry expertise more widely available.
People who master AI applications are pulling away from those who have not adopted AI and building a competitive advantage.
Practitioners who cling to traditional models, rely on experience and personal connections and reject AI are being pushed out of their industries. Those who actively equip themselves with AI and move beyond existing business boundaries have already taken the lead.
What I see is that Zhuoran's sales are still growing at a remarkable pace.
What will industries ultimately look like as AI becomes more accessible? Will experienced practitioners be replaced, or will their value be redefined? Will AI take over entirely, or will new forms of human–AI collaboration emerge?
These questions still need to be tested by industry. What is certain is that we are already in the wave of AI transformation and cannot avoid the restructuring it brings.
At this very moment, an absurd scene is playing out in corners of the world:
AI serves people, while people who understand AI are profiting from those who do not.