08 August 2026
When: the timing science schedulers ignore, ours included
Daniel Pink's When argues that the same task done at the wrong hour is a different task. No auto-scheduler on the market reads your chronotype. Here is what the book claims, and what an honest workaround looks like.
The argument, in one paragraph
When (2018) collects the timing research most productivity writing skips: performance moves through the day in a pattern most people share: a sharp morning peak, an afternoon trough that is measurably worse than it feels, and a late recovery, shifted earlier or later by whether you are a lark or an owl. The claim is not that afternoons are for naps; it is that timing is treated as an afterthought while behaving like a science, and that matching the kind of work to the hour is worth more than most people's entire productivity system.
The method: match work to the hour
Analytic work (anything needing vigilance and a closed door) belongs at your peak. Administrative work belongs in the trough, where its low demands survive your low state. Work that benefits from loose association (naming things, connecting ideas) often goes better in the recovery, when inhibition is down. Find your own pattern by noting alertness for a week; owls run the sequence roughly in reverse.
Where the book disagrees with automatic scheduling
Diarized does not know what kind of work a task is, and does not know your chronotype. It places by priority into the earliest gap that fits, so a task needing full concentration lands at nine because nine was free, and a lark and an owl with identical task lists get identical calendars. Pink would say we have the order right and the timing arbitrary. He is correct, and neither Motion nor Reclaim nor FlowSavvy reads a chronotype either. This is a gap in the category, not a gap we are spinning. The honest workaround today: shape your working hours so the gaps we can reach are the hours you are good in, and put an event over your trough if you never want hard work booked there.
If we ever build timing-aware placement, it starts from this book, but it is not built, and a settings page for it does not exist. The roadmap says what is real; this post will be updated if that changes.