They way I understand it, most people aren't advocating that. "fail fast" I think embraces: - finding if your idea works FAST. Stuff like Pretotyping. Yes, t

They way I understand it, most people aren't advocating that. "fail fast" I think embraces: - finding if your idea works FAST. Stuff like Pretotyping. Yes, that part suggests showing something half-baked ("B/C/D class stuff"), to see if it works - example: mocking up something the size of a Palm and seeing if you would carry it around all the time, to validate that hypothesis. Or the modern-day landing page. http://www.pretotyping.org/ http://www.pretotyping.org/ However, once you validated the first part, the MVP does not have to be half-baked... on the contrary, it has to be crafted as well as you can !!! http://differential.io/blog/misconceptions-about-lean-startup http://differential.io/blog/misconceptions-about-lean-startu... But with only the minimal set of stuff the user needs. The original iPhone didn't have cut&paste - do you think it was a failure? Most succesful early-runners do one thing and do it well, and then build on it. "Some people equate the word lean with cheap, and cheap with low quality. They are using the wrong definition of the word lean. "Lean" in Lean Startup is derived from the Toyota Production System, which is decades old. The definition of Toyota's lean is "to design out waste and overburden."" The point is, find out what the customer needs first and fast ("fail fast"), then deliver, but deliver something of quality. With code it's tricky because you're iterating and it might be tempting to keep in low-quality stuff, but even the iPhone iterated :)