How I Automated an Entire Service Business With AI

Key Concepts

• Opportunity Arbitrage: Exploiting the gap between new technology (AI) and market pricing, where the cost of delivery drops significantly while the perceived value remains high.
• AI-as-a-Service: Shifting from selling software tools to selling the outcome of a service, making the business more defensible against software updates.
• Claude Code/Desktop: An AI-powered development environment used to automate workflows, file management, and browser interactions.
• Skill.md Files: Structured documentation files that define the AI’s objectives, constraints, and methodologies for specific tasks.
• Vercel: A cloud platform used for hosting and deploying the AI-generated websites to make them accessible via public URLs.
• Wop: A payment processing platform used to generate checkout links and invoices for client services.

---

1. The Business Model: "Services are the New Software"

The core argument, supported by Sequoia Capital, is that selling software is a race to the bottom against AI updates. Conversely, selling a service—where AI is the engine—allows for higher profit margins and scalability. By automating the fulfillment of a service (e.g., website building), an individual can deliver professional results in minutes rather than weeks, creating a "disproportionate amount of value."

2. Step-by-Step Automation Framework

The presenter outlines a fully automated pipeline using Claude and various integrations:

• Phase 1: Prospecting:
  • Define a target niche (e.g., local restaurants).
  • Create a skill.md file to define the "ideal client" (e.g., businesses with poor or non-existent websites).
  • Use a scheduled task to scrape the web for leads and populate a spreadsheet with contact info (Instagram, phone, email).
• Phase 2: Service Delivery (Website Generation):
  • Create a skill.md for website design that mandates the use of real menus, branding, and photos.
  • Pro Tip: Use Canva "Brand Guidelines" templates to generate a PDF, which is then fed to the AI to ensure professional design aesthetics.
  • Deploy the site to Vercel to generate a live, shareable URL.
• Phase 3: Automated Outreach:
  • Use the Claude Chrome extension to interact with the browser.
  • Automate the sending of personalized Instagram DMs containing the link to the pre-built website.
  • Include a follow-up message with a calendar booking link to schedule a sales call.

3. Technical Implementation & Tools

• Claude Code: Used to manage the project, execute tasks, and write code.
• Automation Scheduling: Tasks are staggered (7:00 AM for prospecting, 8:00 AM for building, 9:00 AM for outreach) to allow the AI time to process without manual intervention.
• Quality Control: The presenter emphasizes that AI is prone to "hallucinations" or shortcuts (e.g., only listing five menu items). Users must refine skill.md files to enforce strict adherence to real-world data.

4. Sales and Scaling

• Closing: Once a client is interested, use Wop to generate a payment link for a one-time fee (suggested range: $500–$3,000).
• Upselling: The website is the "foot in the door." Once the relationship is established, offer recurring services like SEO, Google Reviews management, or online ordering integration.
• Referrals: Satisfied local business owners are high-value sources for word-of-mouth referrals, which can grow the business organically.

5. Notable Quotes

• "Opportunity arbitrage is when a new technology changes the way work gets done, but the market hasn't fully priced in that change yet."
• "The number one skill in the AI world is taste. AI just negates the three years it takes to physically learn how to use these old website building tools."
• "If you only sell software, you're essentially just competing with the next AI model update. But if you just sell the outcome as a service, every AI update just makes your business better."

6. Synthesis and Conclusion

The strategy presented is a blueprint for leveraging AI to perform high-value, labor-intensive tasks at scale. By automating the entire lifecycle—from lead generation to website deployment and outreach—the entrepreneur minimizes manual effort and maximizes the number of potential clients served. The ultimate takeaway is that while AI handles the technical execution, the human element—taste, quality control, and relationship building—remains the critical differentiator for long-term success.