Build · Lesson 2 of 2

Ship small, learn fast

Make it small, show it to someone, learn, repeat.

5 min read · Free — no account needed to read

Execution is a loop: make the smallest useful version, put it in front of someone, learn, adjust. Each loop teaches you something no plan could.

Things will go wrong. Resilience is not pretending they didn't; it is asking "what did this teach me?" and running the next loop.

With AI you can run loops much faster — a prototype in an afternoon instead of a month. That makes the judgement about which loop to run next more important, not less.

See it happen

You have been building a study tool for your classmates for two months and nobody outside has seen it.

The read that arrives on its own

It is not ready. Showing something half-finished would give the wrong impression.

The same situation, with this idea

You put the one screen that does work in front of two classmates. Both immediately try to use it in a way you never designed for, and one of the three features you were about to spend a month on turns out to be unnecessary.

What changed

Two months of work had been judged by one person against an imagined user. Twenty minutes with two real ones replaced the imagining. The embarrassment was small and it arrived early, which is much the cheapest time for it to arrive.

The takeaway

Make it small, show it to someone, learn, repeat.

Easy to take the wrong way

Reading "ship small" as "ship sloppy". Small means narrow — one thing that genuinely works — not a wide thing that half works. A broken demo teaches you about the bug, not about the idea.

Try this

Pick one idea and define the smallest version someone could actually use this week.