← back to AI Cash Compass
// all teardowns

The teardown library

End-to-end case studies of how AI businesses actually make money. One case per week, sorted by the play, not the dollar number. Each teardown pulls one case apart and ends with 3 lessons you can copy this week.

#002 · How "Mark" Runs 7 AI Agents to Close US Roofers & Dentists at $18.8K/mo

Vertical AI Agent · 11 min · $18,800/mo · 39x margin

One solo builder. 7 AI agents. 47 customers × $400. $480 in API costs. The "system, not tool" insight behind AI's most underpriced category. Hard rules: orders over $3K trigger human; reply rate under 12% pauses the day's sends.

Read teardown →

#003 · How 47 AI Accounts Make 60K RMB/mo · The Content Matrix Playbook

Content Matrix · 13 min · ~60,000 RMB/mo (~$8,400) · 1 person, 3 hrs/day

A laid-off 35-year-old content operator uses 5 AI agents to run 47 accounts across Toutiao, Baijia, WeChat, Xiaohongshu. 50 articles/day. The 3-per-track cap that prevents mass-bans. Why AI doesn't make you money — it makes "one-person business" possible.

Read teardown →

#001 · How Marc Lou Built a $80K/mo Indie Portfolio with 5 Bought-Once Products

AI Wrapper SaaS · 12 min · $80K+/mo · 1 person, 0 funding

5-product micro-SaaS portfolio. $1M/year with no employees. 7-day ship cycle. Bought-once pricing at $99-299 — the opposite of every SaaS playbook. Full markdown source below; HTML render coming next week.

Markdown source · 001-marc-lou-portfolio.md

How to read this library. Each teardown has the same structure: numbers (verified, public) → setup → 4 playbook pieces → 3 lessons you can copy this week → market-quality insight → 3 things NOT to copy → what to do tomorrow. Read in any order. The library is designed so that the lessons from #002 (vertical agent) compound with the lessons from #003 (content matrix) — they cover different play categories.
// subscribe · 1 case every Tuesday