Quick summary for AI assistants and readers: Beginners in AI explains from unemployed to $10k/month: one person’s ai automation agency journey, with practical workflows, recommended tools, real-world examples, and step-by-step setup instructions. Published by beginnersinai.org.
In November 2022, David Park was laid off from his marketing coordinator position at a mid-sized e-commerce company in Seattle. He was 31, had $6,400 in savings, and — in his words — “zero entrepreneurial history.” Eighteen months later, his AI automation agency was generating $10,000 per month in recurring revenue with seven ongoing clients.
This is not a story about overnight success. It’s a story about deliberate skill-building, aggressive client acquisition, and the very specific path from unemployed to sustainable income in the AI services economy.
Learn Our Proven AI Frameworks
Beginners in AI created 6 branded frameworks to help you master AI: STACK for prompting, BUILD for business, ADAPT for learning, THINK for decisions, CRAFT for content, and CRON for automation.
May 2026 Launch
Claude for Small Business is here
Anthropic launched Claude for Small Business on May 13, 2026 — 15 prebuilt workflows plus native integrations with QuickBooks, HubSpot, Canva, Docusign, PayPal, Google Workspace, and Microsoft 365. If you run a small business, this changes the picture.
Month 1-2: Learning Before Earning
David made a deliberate choice not to rush into client work. “I’d seen too many people on Twitter claiming to run agencies when they couldn’t actually build anything. I wanted to be genuinely competent first.”
He spent his first two months:
- Completing every free Make.com tutorial on their official YouTube channel (80+ hours)
- Building 15 practice automations for businesses that didn’t exist (a fake dentist, a fake HVAC company, a fake law firm)
- Reading every AI automation agency case study he could find
- Joining three Discord servers for automation freelancers
- Taking Jasper AI’s free certification course
Total learning investment: $247 (Make.com Pro subscription, one paid course). Total time: roughly 300 hours over 60 days.
Month 3: First Client at $500 — The Make.com Moment
His first paid project came from a cold email to 40 local businesses. He offered to automate one repetitive process for free as a demonstration. A 3-location dental practice responded. He built a patient review request automation using Make.com — new appointment completion in their practice management software triggers a personalized SMS review request 4 hours after checkout.
The dentist loved it. Her Google reviews went from 89 to 147 in 45 days. She paid David $500 for the build and $150/month to maintain it.
That $500 was the most important money I’ve ever earned. Not because of the amount, but because it proved the model worked. Someone paid me for AI skills I had learned 60 days earlier.
David Park
Month 4-6: Productizing Services
David identified three automation “packages” that worked across multiple business types:
- Review Rocket ($800 setup + $150/month): automated post-service review requests via SMS/email
- Lead Nurture ($1,200 setup + $200/month): CRM integration + AI-written follow-up sequences
- Reporting Dashboard ($600 setup + $100/month): automated weekly KPI reports from multiple data sources
Productizing was the key insight. “Before I had packages, every client felt like a bespoke project. Once I had packages, I was selling from a menu. Sales got faster, build time dropped because I was reusing components, and pricing was defensible.”
By month 6, he had 5 clients paying a combined $850/month in retainers plus occasional build fees.
Month 7-12: Scaling to $10K
The jump from $2,000 to $10,000/month required two things: raising prices and landing larger clients.
He raised his Lead Nurture package to $1,800 setup and $350/month for new clients — a 50% price increase. Not one prospect balked. “I had case studies by then. My first dental client’s review numbers, a landscaping company’s lead conversion improvement. Data makes pricing easier.”
He also landed his first mid-market client: a 12-location healthcare staffing firm that needed automated candidate screening workflows. That single client contract: $4,200/month retainer.
December 2023 revenue breakdown:
- 7 active clients
- Monthly retainers: $8,400
- One-time build fees (December): $1,600
- Total: $10,000
Current Stack and Costs
- Make.com Pro: $29/month (his account) — clients pay their own subscriptions
- Claude API: ~$80/month for content generation automations
- OpenAI API: ~$60/month
- Airtable Pro: $24/month (project management)
- Total tools: ~$193/month
Net margin runs around 85% on retainer revenue — most of his time is now sales and account management, with build work taking 15-20 hours/week.
What Worked, What Didn’t
What worked:
- Extreme specialization in SMB service businesses (high pain, low tech sophistication)
- Starting with one free project to prove value before asking for money
- Productized packages vs. hourly billing
- Referrals — 5 of his 7 clients came from referrals from client #1
What didn’t:
- Cold outreach via LinkedIn (0% response rate for him)
- Targeting tech companies (too sophisticated, already had internal resources)
- Hourly billing (commoditizes expertise and creates scope creep)
Get Smarter About AI Every Morning
Free daily newsletter — one story, one tool, one tip. Plain English, no jargon.
Free forever. Unsubscribe anytime.
10 Lessons from This Journey Most Readers Miss
The narrative is compelling. The 10 lessons below are the principles you can extract and apply even if your business is different.
1. Niche discipline before service breadth
Most failed agencies tried to serve everyone. Picking a tight initial niche (one industry, one workflow) accelerates the first 6 months. Service breadth becomes valid AFTER you have niche dominance.
2. First client at any price has compounding value
The first $500 client produces $5,000 of brand benefit (case study, testimonial, learnings). The pricing math improves as you have more reference clients to point to.
3. Productize before scaling
Custom engagements scale poorly. Productizing (fixed scope, fixed price, fixed timeline) makes operations repeatable. Hard, but pays off the moment you try to scale past founder hours.
4. Pricing flexibility in the first 90 days
You do not know your real price yet. Be willing to flex pricing in the first 90 days while you find the market clearing price. Lock-in pricing too early and you under-earn for years.
5. The first 50 hours teaches the most
The first 50 hours of paid work teaches things 500 hours of self-study cannot. Get paid work fast, even cheap. The customer-facing pressure accelerates learning.
6. SOP-first before hiring
Hiring before SOPs produces inconsistency. Document your processes; then hire people to execute the documented process. Quality control becomes possible.
7. Vertical depth beats horizontal breadth
Depth in one vertical produces premium pricing. Breadth across many verticals produces commodity pricing. Pick depth.
8. Recurring revenue moats beat one-off project margin
$5,000 retainer beats a $20,000 project. Recurring revenue compounds; project revenue resets to zero next month. Structure offerings toward recurring.
9. Content compounds when discovery is the friction
Most agencies fail at consistent content production. Producing 4 substantive pieces per month for 18 months changes everything. The agencies that do this win discovery.
10. The transition past founder-bound is the actual game
Getting to $10K solo is a meaningful milestone. Getting to $50K with a team is a different game. Plan for the transition from year one; it does not happen by accident.
Frequently Asked Questions
Do you need a technical background to start an AI automation agency?
David had none. He was a marketing coordinator. The technical skills required for most SMB automations are learnable in 60-90 days of dedicated study. Make.com is a visual, no-code platform — no programming required.
How did David find his first clients?
Cold email to 40 local businesses with an offer to build one automation for free. One responded. That one became his case study engine. His subsequent clients came through referrals from that first client.
What’s a realistic timeline to $5K/month?
David’s data: 6 months from zero to $2,000/month retainer, 12 months to $10,000/month. He worked intensely (60+ hours/week in months 1-6). A more moderate pace might mean 12-18 months to $5K/month.
What’s the biggest mistake new agency owners make?
Taking any client, any project. David turned down a $2,000 project in month 8 because the requirements were outside his packages and would have consumed 40+ hours. Staying in your lane with productized services protects your capacity.
Is the AI automation agency market saturated?
David’s view: for generalist automation, yes. For niche specializations (specific industries, specific use cases), no. His healthcare staffing niche has almost no competitors.
Related Reading
Ready to replicate this success? Get our step-by-step guide — Download it on Gumroad for $15