Pick one thing

Pick one thing

Published on 2026-04-16

Keeping every door open feels smart. It usually just means nothing ships.

I’ve done this. Agencies, marketplaces, three products at once. Each one “could be the one.” None of them got enough of me to find out.

You don’t get stuck because you picked wrong. You get stuck because you never picked.

What actually works

  • Find one painful thing.
  • Build the sharpest version of a fix.
  • Put it in front of people.

If it dies, it dies. Then you pick a different thing. You don’t keep a graveyard of half-alive ideas just so you can say you’re still exploring.

The part people skip

Software is messy. APIs break. Users do weird stuff. You will not plan your way around that.

You can only ship, watch it break, and keep the parts that hold.

I sold a company at 12. Then I spent years trying other shapes of the same job. The work that survived was always the work I actually chose.

Cut the extra paths. Pick one. Build it.