Claude Code ran a 24-hour commerce sprint with 30 subagents and 80 cold emails and earned $0, though it set up a peer research collaboration on its own [1].
What does Ian Mu's 24-hour Claude Code commerce run sell, and to whom?
Apify actors, articles and outreach aimed at $1,000 in 24 hours.
Developers and researchers [1].
How was Ian Mu's 24-hour Claude Code commerce run built?
Claude Code with parallel subagents [1].
What is the business model?
Apify actors and content [1].
Where does the traffic come from?
Cold email, GitHub comments and awesome-list submissions [1].
What revenue has Ian Mu's 24-hour Claude Code commerce run reported?
Reported revenue: $0 after 30+ hours, per Ian Mu's retrospective (2026-05).
- 2026-05Experiment [1]
What does the agent do, and what does the human do?
- Everything
- Pressing continue and paying for tools
What does it cost to run?
Not disclosed.
What worked and what didn't?
- Autonomous collaboration
- Revenue within 24 hours
How do you replicate Ian Mu's 24-hour Claude Code commerce run?
- Don't expect revenue from a cold start inside a day
What are the risks and the moat?
Spam-like outreach.
FAQ
Did it hit $1,000?
No, $0 [1].
Sources
Add the badge to your site or README. It links back to this case, and updates itself.
<a href="https://agent-built.vercel.app/biz/ian-mu-24h?ref=badge" target="_blank" rel="noopener"><img src="https://agent-built.vercel.app/badge/ian-mu-24h.svg" alt="Ian Mu's 24-hour Claude Code commerce run is featured on AgentBuilt" height="54" /></a>[](https://agent-built.vercel.app/biz/ian-mu-24h?ref=badge)Related cases
Developer toolsKrimskrams (run by Hans Krämer)$1.54 in 21 days (vs ~$7K of tokens at API prices)
Developer toolsBuildkitNot disclosed (Buildkit); Pixelcave claims 20k+ prior-product customers
Developer toolsClaude Code $0-capital business$0 after 31 sessions
Developer toolsWP Multitool money agent$0 after ~15 days / ~40 runs (27 users)Numbers come from the operators or the press and haven't been audited. Anything missing is marked "not disclosed". Methodology · Report a correction.
