Breaking down OpenClaw's Wealth Trap: Everyone is "Farming Lobsters" — How Do Others Make Money?

Author: Frank, PANews

In recent times, the hottest topic in the tech and startup circles isn’t a major company’s new model release, but the nationwide “lobster farming” craze.
On one hand, the “lobster farming” boom has driven growth in related industries, with large model companies and cloud server providers making huge profits. On the other hand, how much real benefit Openclaw can actually bring to users remains a mystery. Although social media is filled with myth-like stories, a closer look reveals most are virtual stories designed to attract traffic.
Is lobster farming really profitable? If so, who is actually making the money?
PANews has compiled data from TrustMRR, public cases on social media, project official websites, and cross-verified reports from multiple sources. To distinguish “verified real income” from online myths, we have excluded many rumors based solely on unverified claims or hearsay.
According to TrustMRR’s classification page for OpenClaw, there are 153 recorded projects in this ecosystem, with total income of approximately $358,600 USD over the past 30 days. Analyzing the top 30 samples, their combined income accounts for 97.3% of the total.
If we break down these projects and their underlying monetization logic by “industry value chain,” a stark truth emerges: the first to make money aren’t those using lobsters as products, but those who help others farm lobsters, teach others how to farm lobsters, or rely on hype to promote MEME tokens.
But this isn’t the most genuine answer we’re after. How exactly are those truly using Openclaw making money? To answer this, PANews has summarized five monetization strategies behind OpenClaw.

First: Selling “Shovels” and services — quick cash from exploiting “cognitive gaps”
The most discussed and financially impressive products in OpenClaw are often not specific applications but tools and one-click hosting services.
OpenClaw functions more like an infrastructure layer rather than a ready-to-use consumer product, creating high barriers for non-technical users. Once complexity exists, services will emerge.
Among the roughly $350,000 USD in the TrustMRR sample income over 30 days, “hosting deployment” and “one-click cloud hosting” projects alone contributed about $120,100 USD, accounting for 34.5%.

A typical example is QuickClaw, which packages underlying capabilities into a mobile app priced at $3.99/week or $49.99/year, generating about $8,782 USD in the past 30 days.
In Chinese communities, this logic manifests more simply: “lobster farming” services on second-hand platforms like Xianyu.
According to media reports, recently, “OpenClaw deployment services” on Xianyu and Xiaohongshu have exploded. Remote installation costs range from 100-300 RMB, while on-site setup costs 400-1000 RMB. During certain periods, daily transaction volume for related services increased by 150% compared to the previous quarter.
This logic essentially exploits “information and perception gaps.” Users are willing to pay to save 30 minutes of hassle, but this is a “window period” business. As official one-click deployment tools mature, the red-hot profit from simple deployment will quickly fade.

Second: Packaging AI expert personas — when “storytelling” becomes the most expensive product
Moving up the chain, another more valuable layer in the OpenClaw ecosystem appears: not just deployment, but training the agent.
In the top 30 TrustMRR samples, projects related to templates, skill packs, and configurations contribute 26.4% of income.
One of the most credible and complete business cases at this level is FelixCraft.
In early 2026, creator Nat Eliason launched an experiment. He named his OpenClaw robot “Felix,” invested $1,000 as startup capital, and let it build its own business. Within a week, Felix generated about $3,500 USD via Stripe.
Additionally, the crypto community issued related MEME tokens for this agent, forwarding 60% of daily transaction fees, allowing Felix to earn tokens worth up to $100,000 USD in a week.

As a case worth deep analysis, Felix has several features: Nat Eliason granted the AI high permissions, allowing it to autonomously post on Twitter and interact in communities. Before launch, Eliason spent significant effort building the agent’s framework, including memory modules, security settings, and workflows.
Eliason admits in interviews that the profit was an accident. Essentially, Felix’s main revenue comes from packaging its training process and results as a product for sale. The MEME token gains are largely driven by the story and hype it creates.
Notably, the top-earning project in TrustMRR’s OpenClaw category, Claw Mart (a marketplace for agent skills), was created by Felix and has earned $71,300 USD so far. The story of Felix autonomously creating projects and automating work is a powerful endorsement of this product.
Felix’s success reveals a high-level monetization path: giving agents continuous identity. When OpenClaw is branded with a specific name (Felix), a sellable guide, reusable skill packs, and a compelling “AI entrepreneurship” story, it transforms into a highly viral personal brand. The core obstacle isn’t AI itself but Eliason’s strong agent training skills and marketing ideas.

Third: Selling efficiency myths — using AI to work and monetizing through storytelling
Among all monetization paths, the most recognized is: replacing manual work with OpenClaw, turning saved costs into profit.
This has become a reality in content operations. Developer Oliver Henry named his agent “Larry,” responsible for his TikTok account. Larry automatically calls large models to generate images, write titles, and upload drafts. Henry only spends 60 seconds daily choosing background music and clicking publish.
Henry states that within five days, Larry’s videos surpassed 500,000 views, earning him about $588 USD (from paid app recommendations in his videos). Additionally, Larry generated $4,000 USD through MEME tokens.
Interestingly, Henry’s tweet about this story has reached 7.1 million views, similar to Felix’s case, where storytelling seems more commercially valuable than the agent itself.

Fusheng, founder of Cheetah Mobile, built a team of eight agents called “30,000,” achieving daily updates from a few articles per year, reaching a record of over 1 million reads on Bosheng’s account, attracting social attention. The viral post explaining how the agent works also drew over a million reads.

This suggests that in content creation, whether agent-generated content can go viral remains unproven. Most viral stories are about agents making money or improving work efficiency.
The biggest current topic in content creation is the “little lobster” story.

Fourth: Deep industry customization — moving beyond tool competition to earn “service premiums”
If deploying “lobsters” is about earning “entry barriers,” then packaging “lobsters” into personalized products is another level.
RoofClaw exemplifies this. TrustMRR shows it earned about $49,800 USD in the past 30 days, with total revenue reaching $1.8 million USD. It offers “personalized customization and delivery of a MacBook Air equipped with OpenClaw.”
This means their business isn’t just pre-installing a lobster but embedding it into a MacBook with tailored services, training the lobster to meet specific needs.

Such services likely target the future commercial demand for “lobsters.” Users probably don’t want just a “usable” lobster but a fully trained, customized one. Behind this demand is deep service for agents.
Simply put, we foresee many companies relying on agents in the future, but how these agents are trained or “cultivated” will become an unavoidable necessity.

Fifth: On-chain trading legends — the most tempting poisoned apple and traffic bait
On social media, the most sensational stories about OpenClaw are always about getting rich overnight.
Currently, one verifiable on-chain account is 0x8dxd on the prediction market Polymarket, a high-frequency trading bot. Many social media posts speculate that this bot relies on OpenClaw for high-frequency trading, but PANews’s analysis shows the actual controller behind this address has never published such claims. The stories claiming “OpenClaw designed an automated trading system earning $100,000/month” are just promotional articles, mostly promoting their automated trading programs.
This case is listed as a warning: as previous PANews research indicates, agents and high-frequency trading bots are not the same. People are often misled and fantasize about their mystery.

Final reflection: The person who teaches you how to make money is the real winner
After analyzing the entire ecosystem, we notice a phenomenon more worth pondering than any single case: sharing “I made X amount with OpenClaw” on social media is itself a very stable business.

When a post like “I earn 50,000/month with OpenClaw” goes viral, traffic becomes a lure. The author naturally directs viewers to paid communities, consulting, or product links.
“Showing off income” is the top of the funnel for acquiring customers, and “making money myths” are the strongest marketing material. This creates a perfect self-reinforcing cycle: selling stories of earning money — attracting traffic — monetizing traffic — then sharing “secrets” as a mentor — leveraging even more.
Essentially, this has spawned a new business chain: bottom layer is deployment and infrastructure, middle layer is skill packs and workflow replacements, top layer is industry solutions and consulting.
If you understand sales, marketing, and have traffic, OpenClaw can drastically reduce costs and amplify productivity.
Many are sharing how OpenClaw optimized workflows and enabled many conveniences, but it’s far from a secret to wealth. The “herd effect” it triggers is the core of this traffic story: when you desperately push through the crowd to the front, you find nothing there — and you are the one being waited for.
(PS: This article was created without using the “little lobster” analogy.)