What to Code

Thousands of books teach you how to code.
AI made most of them obsolete overnight.

How to code is easy now: Use AI to learn what you need exactly when you need it.
And more and more micro-SaaS vibecoders make it even without ever having figured out how to code.

This book focuses on the far more profound question:
What to code?

Because the smartest engineer in a San Francisco- or London-based startup can still waste 12 months and $250,000 building the wrong thing.

Perfect code does not create demand. Beautiful architecture does not make users care.

If you build the wrong thing perfectly, nobody cares. You lose time. You lose money. You end up frustrated, broke, and wondering why the market ignored something so technically impressive.

But if you build a rough solution to a real, painful problem -- and you know what to build through first-principles reasoning -- you can strike gold.

A scrappy micro-SaaS with real demand beats a polished dead-end product every time. A small product that solves an expensive problem beats a big vision nobody needs.

$1,000 MRR beats applause.
$10,000 MRR beats cleverness.

What to Code is the indie hacker’s guide to finding profitable micro-SaaS ideas in the vibe coding era.

It is for AI-native builders and founders who no longer need help writing code. They need help choosing what is actually worth building.

In an age of vibe coding, cheap prototypes, and infinite AI-generated software, judgment is the new moat.

This book teaches you how to find profitable micro-SaaS opportunities, spot real market pain, validate demand before you waste months building, avoid seductive but worthless ideas, and focus on products people will actually use, buy, and recommend.

It will help you think more clearly, choose more sharply, and build with a real chance of traction.

Because in the AI era, the winners will not be the people who can code everything. They will be the people who know what to code next.