People describe their jobs in hours. Hours are a poor measure of what work costs, and the mismatch explains a familiar confusion: the week that looked light and left you flattened, and the week that looked brutal and was fine.
What makes work expensive is mostly not the quantity.
What actually costs
Transitions. Switching between tasks is more expensive than doing either. A day with twelve context changes and a day with two can contain identical hours and completely different cost. This is the single most underweighted factor in how anybody plans a week.
Unpredictability. Work that arrives without warning is more expensive than the same work scheduled, because it forces a transition and prevents any of the preparation that makes a task cheap to enter.
Held-open decisions. A decision you have not made consumes capacity continuously until it is made. Six open decisions is a background process running all day. This is why a light week with an unresolved question in it can be exhausting.
Emotional labour. Managing your own reaction, managing someone else's, holding back an opinion, being in a room with an unresolved conflict. Draws on the same account and appears nowhere on a calendar.
Ambiguity about what matters. Five tasks is manageable; five tasks with no stated priority requires you to supply the ordering, and ordering is expensive.
Working memory demands. Work requiring several things held simultaneously — reconciling, cross-referencing, coordinating — is far costlier than work of the same difficulty done sequentially.
Interruption during deep work. Not the duration of the interruption. Attention does not snap back, and there is a tail during which you are working at a fraction of capacity while feeling like you are working.
Sensory load. Open-plan noise, overhead lighting, a busy visual field. Genuinely costly for some people, and generally treated as a preference rather than a cost.
What costs less than it looks
Long hours on one thing. Genuinely demanding, and often cheaper per hour than a fragmented day, because the transition is paid once.
Difficult work you find interesting. Still tiring, meaningfully cheaper to enter and sustain.
Physically demanding but cognitively simple work. Tiring in a different account.
Volume of routine tasks. A large number of similar small things, batched, is not particularly expensive. A small number of dissimilar things, scattered, is.
What follows from this
Count transitions when you plan. A day with five scattered meetings does not have three hours of usable gaps in it. It has none, for anything demanding. Planning deep work around scattered meetings is the most common planning error there is, and it produces a reliable sense of having failed at a day that was never possible.
Batch aggressively. Similar things together, so you pay one transition. This is most of what makes a well-run week feel manageable.
Close decisions rather than carrying them. An imperfect decision made today is cheaper than a perfect one made next week, because of what the carrying costs in between. Make the small ones by default.
Ask for priorities, not just tasks. "Which of these matters most" is a short question that removes an expensive judgement from your plate.
Ask for meetings to be clustered. One of the easiest adjustments to obtain and one of the most effective. It protects blocks rather than fragments.
Protect one uninterrupted block rather than trying to optimise the whole day.
Notice the emotional load. If the week was expensive and the calendar was light, look at what you were managing rather than what you were doing. It is usually there.
Why this matters beyond productivity
Because "I am exhausted and my week was not even busy" is a statement people use as evidence against themselves. It gets read as weakness, or as evidence that something is medically wrong when the explanation is structural.
Both readings are wrong often enough that the distinction is worth making. A week of twelve context switches, three open decisions and two difficult relationships is an expensive week, and being flattened by it is the correct response rather than a symptom.
Which also means the reverse is worth taking seriously: if your weeks are genuinely light by all of these measures — few transitions, nothing unresolved, no sensory load — and you are still exhausted, that is a different observation and it is worth raising clinically.
What to write down
If you want to know what costs you, two weeks of one line a day:
Date — roughly how expensive today felt — number of context switches — anything unresolved hanging over.
Most people find it is transitions and open decisions rather than hours, and that the pattern is not what they assumed. That is useful for planning, and it is also useful in an appointment, because "my exhaustion tracks the number of times I have to switch tasks, not the hours I work" is a much more specific observation than "work is stressful."
The part that is not yours to fix
Some of this is structural and not adjustable by you. A role built on constant interruption will remain expensive regardless of how you batch.
Worth being honest about which category you are in, because there is a failure mode where someone spends two years optimising their own systems to survive an arrangement that needed changing. Adjustments help. They do not convert one kind of job into another.