Analysis prompted by Clay Ostrom's comment suggesting I add "course creator" to my areas of expertise.
There are 10+ Claude Code courses already out there:
- Anthropic's official training
- Coursera via Vanderbilt - free
- Multiple Udemy courses ($10-200, thousands of students each)
- DAIR.AI live cohorts
- Free PM-focused course
So the question isn't "could I make a course" - it's "what would make mine worth existing?"
Most courses assume software engineering background. I'm using Claude Code for business operations, messaging work, client strategy. That's rare.
47 agents, hooks, custom workflows, meeting intelligence, CRM integration. Most courses teach "how to use the tool." I've built an operating system on top of it.
"Claude maintains the course" isn't just marketing - it's a genuine solution to the update problem. My system updates ARE the content.
The daily LinkedIn posts about /pause, hooks, workflows - that's course material being generated in public.
If "Claude makes the course," then anyone can make a Claude Code course. My moat isn't the content - it's my POV and my system.
I help companies with messaging. A Claude Code course is a distraction from that core. Unless the angle is "how consultants can scale their expertise with AI" - which IS my business.
Traditional video courses go stale. My system evolves daily. The format needs to match that reality.
"Claude Code for business operators" - not developers
Target: Consultants, agency owners, fractional execs, ops people who want to use AI for real work but aren't coders.
Content:
- Setting up Claude Code for non-technical workflows
- Building agent teams for your specialty
- Automation for client work, not code
- The business model implications (scaling expertise vs. billing hours)
| Format | Pros | Cons | Maintenance |
|---|---|---|---|
| Traditional video course | One-time revenue, perceived value | Static content in dynamic space | HIGH |
| Cohort-based live | Premium pricing ($500-2K), community | My time required each cohort | MEDIUM |
| Template/system sale | One-time purchase, low support | Less defensible, easy to copy | LOW |
| Membership/newsletter | Recurring, updates naturally | Harder to market, slower build | MEDIUM |
-
Sell the system itself - My CLAUDE.md, agent prompts, hooks, workflows. "Lee's Claude Code Operating System." Templates + setup guide. Claude updates it, I don't record anything.
-
Single cohort to test demand - 10 people, $500 each, 4 weeks. Live, with Claude doing heavy lifting in real-time. If it works, I have testimonials and data.
-
Keep doing what I'm doing - The LinkedIn posts are already course content. Build the audience first, productize only when people are asking to pay.
Is this a passive income stream, or proof that my system has value beyond my own use?
Both are valid. But they lead to different products.
This analysis was written by Claude (naturally) based on a conversation about whether "my Claude Code could create a course." The irony is not lost on me.
— Lee Fuhr, leefuhr.com