vs. Lean Startup

LDD owes a direct debt to Lean Startup. The learning loop descends from Build-Measure-Learn. "Validated learning as the unit of progress" is Eric Ries's language. If you've practiced Lean Startup, LDD will feel familiar.

But the world that produced Lean Startup and the world LDD operates in have a fundamental difference: the economics of building are inverted.

The constraint shifted

Lean Startup was written when building was expensive. An MVP took weeks or months. You built the minimum because building more was wasteful. The whole method was organized around a scarce resource: developer time.

LDD operates in a world where an agent can build a feature in an afternoon and rebuild it from scratch the next morning. Building is nearly free. What's still expensive is learning: getting in front of real users, interpreting evidence, making good decisions about what to do next.

Same lineage. Different constraint. And that different constraint changes almost everything about how you work.

The slice is not an MVP

This is the sharpest difference. An MVP is the smallest product you can ship. It exists because you can't afford to build more until you know if the idea works.

A slice is the smallest exposure you can validate. You might have already built the whole feature. But you put it behind a feature flag and turn on one slice for real users. You validate one belief, then the next.

"Build big, show little" has no equivalent in Lean Startup. Ries would have said "don't build big," because building was the bottleneck. That's no longer true. Building, rebuilding, and refactoring now takes minutes to hours. The code is cheap. The learning is what you keep.

LDD is the process, not a phase

Lean Startup is often practiced as a discovery phase you graduate from. You run experiments, find product-market fit, then switch to "real" development with Scrum or whatever the team uses. The learning stops when the building starts.

LDD doesn't have a graduation moment. The learning loop is how you build, period. You keep running loops on mature features, not just new ideas. There is no point where you stop asking "did this actually matter?"

It's agent-native

Lean Startup assumed humans do everything. LDD is designed around the reality that agents are execution partners. The spec isn't just a document for humans to read. It's a context layer that enables agents to build well. That changes what specs look like, how you write them, and the skills you need.

Not startup-exclusive

Lean Startup was framed around founders searching for product-market fit. LDD applies to solo builders, feature teams in established companies, anyone building with AI. The learning loop works whether you're finding PMF or shipping the fourteenth iteration of an existing feature.