The rhythm of a
quiet morning.

I do my best work before the day gets a chance to negotiate with me. Type systems, grid decisions, and the small interaction details all need enough quiet to let me see the second problem hiding behind the first one, and that quiet has a closing time most mornings.

Client work still needs replies, approvals, and logistics, and I don't pretend those things are interruptions to the real work. They are the real work too. The trick I've actually managed to keep is not letting them own the whole day before I've made one real decision in it.

A rhythm that survives actual clients, not a fantasy version of my calendar, is simple. I protect a couple of deep work blocks each week, batch messages into windows instead of answering them as they land, and end every block with one sentence about exactly where to pick back up. That sentence has saved me more afternoons than any productivity system I've tried and abandoned.

When everything feels urgent, I shrink the ambition for the day instead of raising the effort. One real decision beats five half started threads. Finish the type scale. Fix the grid. Ship the interaction. Momentum comes from finished increments, not from convincing myself I can hold six things in my head at once.