Why I stopped optimizing everything

There was a year when I optimized everything. My morning routine, my inbox, my keyboard shortcuts, the title tags of pages nobody visited. Every workflow got squeezed, every process got a spreadsheet. It felt like being extremely good at my job. Looking back, it was elaborate procrastination with charts.

The local maximum problem

Optimization, by definition, makes you better at climbing the hill you're already on. It never asks whether you're on the right hill. In SEO this failure mode is everywhere: the fifth re-optimization of an ageing article, chasing a position for a keyword that stopped mattering, while the piece your readers actually keep asking for stays unwritten. I've done this. The dashboard rewards it — the line goes up — which is exactly what makes it dangerous.

The questions that actually moved my work were never optimization questions. They were direction questions: should this page exist at all? Is this channel worth a single additional hour? Optimization can't answer those. It can only make the current answer more efficient — including when the current answer is wrong.

Slack is not waste

Queueing theory has a result that should be printed above every calendar app: as a system approaches 100% utilization, waiting times don't grow gradually — they explode. A machine shop scheduled at full capacity isn't fast; it's a traffic jam with invoices. Every small surprise cascades, because there's nowhere for it to go.

I had built myself exactly that. A perfectly optimized week is a week where the first sick child, the first urgent client email, the first idea that deserves an afternoon — breaks everything downstream. And ideas were the biggest casualty. Thinking, it turns out, mostly happens in the slack. Remove the slack and you haven't become more productive; you've just made sure your work can only ever contain what you already knew.

A system with no slack isn't efficient. It's brittle — and brittleness always presents itself as efficiency, right up until the day it doesn't.

What I still optimize

I didn't swing to the other extreme; I narrowed the scope. Three things still get the full engineering treatment:

• The few numbers actually attached to income — everything else is dashboard theatre.
• Anything I repeat weekly. A process that runs fifty times a year earns its optimization; a one-off task never does.
• Recovery time — how quickly the system returns to normal after a bad week. This one predicts more than any productivity metric I've tracked.

Everything else is now allowed to be a little inefficient, a little manual, a little slow. That's not negligence — it's a design decision, the same one that keeps a calm business calm. The goal was never a perfect machine. It was a machine I could still want to run in ten years.