Decisions.

The framework

1. 01

   Count the decisions, not the hours — your diary hides the real load.

   Your calendar shows meetings; it does not show that Tuesday contained nineteen judgement calls, eleven of which someone two levels below you could have made without you. Spend one ordinary day keeping a tally: every time you choose, approve, arbitrate, price or sign off, mark it with the hour. Most operators are surprised twice — by the total, and by how much of it landed before eleven o'clock on matters worth almost nothing. A diary is built out of equal hours, as though each one held the same quality of thinking; they do not, and the call that mattered most is the one left to pay for that fiction.

2. 02

   Spend your clearest hours on what cannot be undone.

   Run the door test from Day 64 across everything waiting on you, and let the answer sort your hours rather than your process — the one-way doors belong in the stretch when your head is genuinely clear, which for most people is the first two hours of the working day, before the inbox has set the agenda. Protecting those hours is the cheapest quality control available on the calls that cannot be walked back, and almost nobody spends them that way. The trap is the warm-up fallacy: clearing twenty small approvals to build momentum, which spends your best cognition on your cheapest calls and leaves the hire, the contract or the price change for six o'clock.

3. 03

   Three times in the same shape is a policy you have not written.

   The third time a decision arrives in roughly the same shape, you are not deciding anything; you are re-running a policy nobody has written down, and paying full price for it again. Write the rule and let it run: discounts below this threshold are approved automatically, spend under this figure needs no sign-off, requests matching this pattern get this standard answer. Give every rule a review date so it stays a decision rather than hardening into a fossil. Every case will feel special enough to deserve your personal judgement; that flattery is how senior people spend a Wednesday on approvals and file it under leadership.

4. 04

   Build a floor for the bad hours — a stopping rule, not willpower.

   You will still be asked to decide when there is nothing left, so design what happens then instead of trusting yourself in the moment. Name the stopping rule in advance: after a set hour, or a set number of significant calls, anything irreversible is deferred to the morning — and you tell the person waiting rather than going quiet on them. Set defaults for the recurring cases so the tired answer is at least the considered one. Most of what arrives at seven o'clock feels urgent because it arrived, not because it matters tonight; before you push through, ask what actually breaks if this is settled at nine tomorrow. Deferring is itself a decision, and at seven o'clock it is usually the better one.