The Lean Startup, read twice
Waste is building something nobody wants, efficiently. The book that gave my folder of failed side projects a vocabulary — and the school idea a plan.
I read The Lean Startup twice this month — once fast, nodding along, and once slowly, with a notebook, because somewhere in the first pass I realized Eric Ries was describing every failed side project I've ever built. I have a folder of them: polished, feature-complete, and used by nobody. The book finally gave me the vocabulary for why.
The sentence that stung
Ries's definition of waste: building something nobody wants, efficiently. That's the one. I've spent months being efficient — clean architecture, tidy code, careful releases — on products where the fundamental question, does anyone need this?, was never asked because asking it risked the answer.
What I'm actually taking from it
- Validated learning as the unit of progress. Not features shipped, not lines written — assumptions tested. A week that kills a bad idea is a productive week.
- The MVP is an experiment, not a small product. The minimum thing that generates real learning. For the school-management idea I keep sketching, that's not software at all — it's sitting with five school administrators and watching how the paper actually flows.
- Vanity metrics are a drug. Signups, page views, downloads — numbers that go up and mean nothing. Cohorts and retention or it didn't happen.
- Pivot is a technical term — a change of strategy that keeps the learning. Knowing when to pivot requires having stated, in advance, what result would prove you wrong. Most of us never write that down, and so we're never wrong, and so we never learn.
The question is never "can we build it?" — of course we can build it. The question is "should we?" — and code can't answer that.
If I ever take the school idea seriously — and it's getting harder to ignore — this book just wrote the first three months of the plan. None of it involves a framework.