Time management has a hidden assumption: that an hour is an hour. Block the calendar, protect the slot, and the work gets done. But anyone who has tried to write a hard report at 4pm after three meetings knows the assumption is false. The slot was there. The capacity wasn't.
Energy is the variable time management ignores. Your ability to do focused, demanding work rises and falls across the day in patterns that are partly predictable and largely ignorable — until you start measuring them. By the end of this guide you'll be able to map your own energy curve, match the right kind of work to the right part of the day, and stop scheduling your hardest tasks into your weakest hours.
TL;DR
- An hour is not a unit of output. Your effective capacity varies 2–3x across a normal day. Treating all hours as equal is the single biggest scheduling error.
- Energy has a shape. Most people have one or two peak windows (often mid-morning) and a reliable afternoon trough. The shape is stable enough to plan around.
- Match task type to energy state, not to whatever slot is free. Demanding work goes in peaks; shallow work fills troughs. Reversing this wastes your best hours.
- Measure before you optimize. A one-week energy log beats any generic "morning person" advice, because the advice is about the average human and you are not the average human.
- Protect the peak, don't fill it. A defended 90-minute peak outperforms a calendar packed wall-to-wall with lower-grade hours.
Why time management quietly fails
The classic productivity stack — to-do list, calendar, time-blocking — manages when work happens and whether it has a slot. It says nothing about capacity. So you end up doing what feels orderly: filling every open slot, often handing your freshest hours to whatever arrived first (email, a standup, someone else's meeting) and pushing your real work into the gaps that are left.
The result is a predictable mismatch. Consider a typical knowledge worker's day, scored on a rough 1–10 capacity scale:
| Time | Typical energy | What usually gets scheduled | Mismatch? |
|---|---|---|---|
| 9:00–11:00 | 8–9 (peak) | Email, standup, "catching up" | Yes — peak spent on shallow work |
| 11:00–13:00 | 6–7 (good) | Meetings | Partial |
| 13:00–15:00 | 3–4 (trough) | The hard creative task you avoided | Yes — hardest work in the weakest hours |
| 15:00–17:00 | 5–6 (recovery) | More meetings, admin | Roughly fine |
The two "Yes" rows are where most days lose. The peak gets spent on work that any hour could have absorbed, and the trough gets handed the one task that needed a peak. You weren't lazy and you didn't run out of time. You ran out of the right kind of time, and your calendar never told you because it doesn't track capacity.
The shape of your day
Energy isn't random noise; it follows two overlapping rhythms. The first is the circadian pattern — a roughly day-long cycle that, for most people, produces a late-morning high and an early-afternoon dip (the "post-lunch dip" is largely circadian, not the sandwich). The second is the ultradian pattern — shorter cycles, commonly cited at around 90 minutes, after which focus degrades and you need a genuine break to reset.
Two caveats the analyst owes you. First, the exact numbers vary by person; chronotype shifts the whole curve earlier or later, sometimes by hours. Second, the "90-minute" figure is a useful planning heuristic, not a law you should set a timer to. Treat both as starting hypotheses to test against your own log, not facts about you.
Framework — the three energy states - Peak: full focus available. Reserve for deep work — anything cognitively demanding, creative, or high-stakes. Usually 1–2 windows a day, 60–120 minutes each. - Plateau: functional but not sharp. Good for medium work — meetings, planning, structured tasks with clear steps. - Trough: depleted. Best for shallow work — email, filing, routine admin, anything tolerant of low focus.
The goal of energy management is simply to stop fighting this shape and start scheduling along it.
Step 1 — Measure your curve
Generic advice ("wake at 5am and write") optimizes for the average human, who does not exist. Your own data takes a week to collect and beats every productivity guru's routine, because it's about you.
The instrument is deliberately crude — precision here is false comfort:
The one-week energy log - Every ~90 minutes you're awake, jot two things: the hour and your energy 1–10. - Add one word of context: meeting, deep work, after-lunch, post-coffee, post-walk. - Do nothing else for a week. No optimizing yet — you're collecting, not fixing. - At week's end, average each time slot across the days and sketch the line.
A capture tool you already keep open makes this nearly free: a single running note with one timestamped line per check-in (this is exactly the kind of low-friction logging [time-blocking](what-is-time-blocking.md) assumes you can do, applied to energy instead of tasks). After five days a shape emerges — usually clearer than you expected, and often not the shape you assumed. Plenty of self-described "night owls" discover a sharp, wasted 10am peak.
Step 2 — Match work to state
Once you can see the curve, scheduling becomes a sorting problem: rank your work by cognitive demand, rank your hours by energy, and line them up. High demand to peaks, low demand to troughs.
| Work type | Cognitive demand | Schedule into |
|---|---|---|
| Writing, design, hard analysis, strategy | High | Peak |
| Meetings, reviews, planning, coding routine features | Medium | Plateau |
| Email, scheduling, expenses, filing | Low | Trough |
Worked example. Maya logs her week and finds a clean peak at 9:30–11:00 and a deep trough at 14:00–15:30. Before, she ran email at 9am ("clearing the decks") and tried to write the strategy doc at 2pm. After, she flips it: the strategy doc moves to 9:30 with notifications off, and email is deliberately demoted to the 2pm trough — a slot where its low demands are a feature, not a waste. Same eight hours, same task list. The hard work now happens when she's sharp and the easy work absorbs the slump. Nothing about her time changed. Her output did.
Step 3 — Protect the peak
A peak is only worth identifying if you defend it. An unprotected peak gets colonized by other people's meetings and your own notifications, and a fragmented peak is barely a peak at all — attention has a switching cost, and every interruption taxes the minutes after it, not just the minutes during it (see [attention residue](what-is-attention-residue.md) for why a "quick" interruption costs far more than its clock time).
Protecting the peak — a checklist - [ ] Block your top energy window on the shared calendar as busy. Don't label it "free for focus." - [ ] Notifications off — not silent, off — for the duration. - [ ] One task in the window. If a second one's pulling at you, it goes on a list for later, not into the peak. - [ ] Default-decline meetings that land in it; offer your plateau hours instead. - [ ] End on a real break (walk, not a feed-scroll), so the next cycle starts from recovery.
The aim is one defended peak a day, not a calendar crammed wall-to-wall. A single protected 90-minute window of genuine deep work routinely beats a fully booked day of fragmented, mid-grade hours — because deep work compounds within an unbroken window and resets at every interruption — so a long, clean block buys disproportionately more than its minutes suggest.
Step 4 — Manage the inputs, not just the schedule
Your curve isn't fixed scenery; a few inputs move it, and managing them is part of managing energy:
| Lever | Effect on the curve | Practical rule |
|---|---|---|
| Sleep | Sets the ceiling on every peak | Non-negotiable; no schedule survives a sleep deficit |
| Caffeine timing | Shifts a peak earlier; borrows from later | Use it before a peak, not to paper over a trough |
| Movement | Lifts a trough by a point or two | A 10-minute walk beats pushing through the dip |
| Food | Heavy meals deepen the afternoon dip | Lighter lunch on days with an afternoon peak |
These are dials, not miracles. The point isn't to engineer an eight-hour superhuman peak — it's to stop accidentally sabotaging the peaks you already have.
Common mistakes
- Optimizing before measuring. Adopting someone else's 5am routine without logging your own curve. You're tuning to the average human's data, not yours.
- Treating the trough as a personal failing. The afternoon dip is a rhythm, not weakness. Fighting it with willpower is less effective than scheduling around it — feed it shallow work and let it pass.
- Filling the peak instead of protecting it. A peak booked solid with meetings is a peak deleted. Empty calendar space in your best window is the most productive thing on the schedule.
- Mistaking busy for spent. Energy management isn't about doing more hours of work; it's about doing the right work in the right hours. A well-matched six-hour day beats a mismatched ten.
- Ignoring recovery. Skipping breaks to "stay in flow" borrows from the next cycle at interest. The break is what makes the next peak possible.
Summary + next step
Time management asks when and whether. Energy management adds the question that actually predicts output: with what capacity? The method is small — log your curve for a week, sort your work by demand, match high to peak and low to trough, and defend one window a day. No new app, no 5am alarm, just scheduling along the shape you already have instead of against it.
The next move is to find your curve. Keep a single running note open and drop a timestamped energy score into it every 90 minutes for a week — the same lightweight logging habit behind [the weekly review](the-weekly-review.md), pointed at your energy instead of your tasks. After five days you'll have something no generic routine can give you: a map of your own best hours, and a reason to stop wasting them.