Eric Simons: How WebContainer fueled Bolt's 40M ARR breakout

Bolt’s apparent overnight success (0 → ~$40M ARR in months) rested on years spent building WebContainer, a browser-based OS that makes dev environments fast, local, and cheap—without that, Bolt’s current scale and responsiveness would be impossible.

By executing dev environments locally via WebAssembly instead of cloud VMs, Bolt avoids slow spin-ups, capacity limits, and abuse-prone free tiers; this architecture gives it speed, reliability, and virtually infinite scalability on top of users’ devices.

StackBlitz scaled Bolt to tens of millions in ARR with ~15–20 people by keeping burn low, hiring multi-skilled, ego-free “Spartans,” and running a daily all-hands Zoom to minimize communication loss and enable rapid, front-to-back decision-making.

Because LLMs can now reliably generate production-grade code, people who excel at defining problems, specifying products, and judging quality (PMs, designers, entrepreneurs) can directly create working software, reducing reliance on developers for much UI and CRUD work.

The best Bolt users treat it like a developer teammate: they provide clear scope, constraints, and context—often akin to a well-written Jira or Linear ticket—rather than vague one-liners, which dramatically improves output quality and iteration speed.