Naming the First Thing

Before starting any task, I identify the first concrete action before opening anything.

The friction is rarely the work itself — it’s starting. An undefined task has no entry point, which means every session begins with a low-grade decision problem: where do I start? That decision is difficult because it is undefined and unconstrained. Being faced with this decision often leads to procrastination. It also often results in a suboptimal choice: whatever feels easiest rather than whatever moves the objective forward.

Naming the first thing solves this. I have been using Claude to help. I hand it a project, ask what the next step is, get a specific answer, start there. The answer is usually obvious in retrospect. The value is not having to figure it out at the moment you’re supposed to be working.