The Idea Maze

TL;DR: Good startup ideas aren’t single points — they’re multi-year plans that anticipate many possible paths through a maze. Good founders know which turns lead to treasure and which lead to death. The “ideas don’t matter, execution is everything” orthodoxy is wrong, and Dixon’s essay is the cleanest rebuttal of it.

What it means

Dixon argues against the “ideas don’t matter, execution is everything” orthodoxy that’s dominated startup discourse since 2010. Ideas do matter — just not the one-line pitch. What matters is the founder’s ability to navigate the maze of decisions that follows. A bad founder runs into the “photosharing maze” without knowing the history, the players, or the technologies that might move the walls. A good founder has spent years building the map.

The reason this matters is that the maze is mostly invisible to people who haven’t been in it. From the outside, every photo-sharing startup looks like every other photo-sharing startup. From inside the maze, the differences are enormous and they explain why one company became a unicorn and another died in 18 months despite shipping the same features.

The argument

Build the maze three ways. Direct experience (often at work, in the relevant industry, before you start the company). Studying history — almost every “new” idea has been tried before, and the post-mortems are sitting on the internet for free. Reasoning by analogy to similar businesses: “If you are building a peer-economy company, look at what Airbnb did right.” Put yourself in interesting mazes and give yourself time to figure them out before you start trying to win one.

Your real competition is wasted time, not other startups. Other startups in your maze are usually a distraction — they won’t take the same path you take, and their presence mostly validates the opportunity rather than threatening it. The real threat is spending years going down the wrong path. Other founders going down the wrong path with you are not your competitors; they’re your future case studies.

Stealth mode is a bad idea. The benefits of learning about the maze far outweigh the risks of idea theft. Almost all good ideas have been tried before — understanding what previous attempts did right and wrong is essential, and you can’t learn that if you’re hiding from the people who tried them. The number of founders who think their idea is uniquely vulnerable to theft and isn’t is essentially 100%.