Our badminton group uses a WhatsApp poll to figure out who’s playing each day. Someone posts it, people vote yes or no or maybe, and by evening we know if there’s a game.

There was no assigned person for this. It just needed to happen — sometime before it got too late for people to plan. Some days it went up early. Some days it went up late. Some days someone else posted it. Some days nobody did, and there’d be a scramble in the group chat.

I was one of the people who posted it. Not every day, but often enough that it occupied mental space. The task itself is nothing — two minutes, a few taps. What’s less nothing is the low-level awareness of it sitting in the background. The vague sense of: has someone posted it yet? Should I? Is it too early? And then later: did someone get to it, or did it get missed again?

I automated it recently using Claude Code. A message goes out every morning at a fixed time. By evening, a second message posts the vote tally and whether we’ve hit enough players to play.

The first few days I kept checking to see if it had gone out. Then that stopped too. It just happens.

It wasn’t the two minutes I got back. It was the mental slot that freed up — that small, recurring background task that’s now just gone.

There’s a version of productivity improvement that’s about doing things faster. A better system. A shortcut. A template. These save time, sure.

There’s a different version that’s about removing things from the queue entirely. Not doing them faster — not doing them at all. That’s a different category of gain. These free up attention.

The badminton poll falls into the second category. I didn’t optimise the two minutes. I moved it off my plate completely. The mental slot it occupied is now genuinely free — not reassigned, not managed better, just gone.

The clearest way I know to think about this: every task you do has two costs. The execution cost — the time it takes to actually do it. And the running cost — the attention it consumes while it’s waiting to be done. Recurring tasks with fixed deadlines have high running costs relative to their execution costs. They’re cheap to do and expensive to carry.