TL;DR. An hour at 20 percent capacity is not an hour, which is the entire argument. Log your energy on a 1 to 10 scale roughly hourly for a week, tag what you were doing, and change nothing while you measure. Most people find one strong peak, a weaker second one, and a reliable trough. Then match task difficulty to level instead of to the calendar: hardest thinking in the peak, admin and review in the trough. "Energy management" is a consulting framework, not a validated scientific construct, and if you are sleeping badly or genuinely overloaded, no scheduling trick fixes the input.
Two hours on a Tuesday. Same length, same desk, same task waiting.
At 9am I can hold the whole problem in my head and write something that survives the next morning. At 3pm I can open the same file, read the same lines four times, and produce something I will delete.
The calendar thinks those two hours are identical. Everyone knows they are not, and almost nobody plans as though that were true.
That is the whole idea. The rest is implementation.
Where the idea comes from, honestly
The phrase most people are echoing comes from Tony Schwartz and Jim Loehr, who argued in The Power of Full Engagement and then in a widely circulated Harvard Business Review piece that time is a fixed resource while energy is a renewable one, and that organisations should train the second the way athletes do. Their model splits energy into physical, emotional, mental and spiritual dimensions.
It is a good frame. It is also a consulting framework born in corporate performance training, and should be labelled that way.
There is no instrument that measures personal energy. Sleep can be measured, heart rate variability can be measured. "Energy" is a self-report with a number on it.
What holds up is narrower and still enough to act on: alertness and cognitive performance vary across the day, sleep loss degrades both, and genuine breaks restore more than switching screens. The framework is a decent container for those findings, not itself a finding.
The one-week protocol
The point is to stop arguing about population averages and get your own. It costs about ten seconds, six or seven times a day.
- Rate your energy 1 to 10, roughly hourly. First instinct. Deliberating gives you a more thoughtful number and a less consistent one, and consistency is what makes the week comparable.
- Tag what you were doing. Sleep, coffee, meals, exercise, meetings, deep work, breaks, screens, other people, stress. These usually explain more of the variance than any internal rhythm does.
- Change nothing else. Not your bedtime, not your caffeine, not your schedule. Optimise during the measuring week and you will measure the optimisation.
- Log at least five days including one non-work day. Weekday patterns are shaped heavily by your calendar, and that is worth seeing separately.
- Read for shape, not precision. You want your best hours and your reliable low, not a curve you can publish.
- Then move exactly one thing. One task, into one better hour. Leave the rest alone so you can tell whether it worked.
Step six is the one people skip, and skipping it is why energy tracking so often ends as a week of interesting data and no change.
Reading the week
The common shape is two peaks and a trough: a strong morning peak, a lower late-afternoon or evening one, and a dip after lunch. Common is not universal. Plenty of people get one peak. Some get theirs at 9pm.
Two things worth checking.
How wide is the peak, really? People say "I am a morning person" and mean four hours. The log often shows one strong hour and a decent one either side. If your peak is 90 minutes, planning three hard tasks into the morning was never going to work.
Is the trough about biology or about lunch? Look at the tags. If every low point sits an hour after a large meal or straight after a meeting block, that is a cause you can act on, which beats any rhythm. Same test for the weekend day: a jagged weekday next to a smooth Saturday tells you about your calendar rather than about you.
Match difficulty to level, not to the calendar
The practice: stop asking what time it is and start asking what this hour can hold.
| Level | What that hour can hold |
|---|---|
| 8 to 10 | The one thing with no template: the hard problem, the argument that must hold, the decision you keep deferring |
| 6 to 7 | Drafting from an outline, code you have written before, meetings that need you |
| 4 to 5 | Inbox, admin, expenses, review, planning tomorrow, calls where you mostly listen |
| 1 to 3 | Stop. Walk, nap, or accept this hour is not for output |
Two rules make the table work.
Never spend a peak on something a trough could do. Answering email at 9am to "clear the decks" is the most expensive habit in this whole area. You are spending your best hour on work your worst hour is perfectly capable of.
Never put a hard task in a trough to prove something. It does not build discipline. It builds an association between that task and feeling stupid, and then you avoid it.
Apply the first rule to the first hour of your day and you have most of the benefit.
What goes in the trough
The trough is not wasted, it is a different tool. Give it batching work: inbox triage, invoices, expenses, small refactors, tidying the task list, planning tomorrow, listening-shaped calls.
A nap belongs here too if your situation allows one, and the length matters, because the wrong one leaves you worse off than no nap. The power nap calculator works out the windows.
The trough is also where choices get expensive rather than just slow. If your late afternoon is where you default to the familiar option and buy things you would have talked yourself out of at 10am, that pattern has a name and a fix that is not willpower: see decision fatigue.
What does not belong here is caffeine used to overwrite the dip. It blocks the tiredness signal without clearing what is behind it, then charges you in the evening for the hour it lent you. That arithmetic is in the ultradian rhythm post, which also explains why the popular 90 minute cycle claim is shakier than the internet presents it.
How this fits with chronotype and 90 minute cycles
Three ideas keep getting merged into one.
Chronotype is the position of your whole curve: whether your peak sits early or late. Relatively stable, largely not a choice, useful mainly as a starting hypothesis. The chronotype quiz gives you the label in a few minutes. A week of your own logging beats the label, and if the two disagree, believe the log.
The 90 minute cycle is the shape within a block, and its waking-hours version is a contested hypothesis rather than settled science. The template that comes with it, a long focus block and a real recovery break, is worth running regardless.
Energy management is the assignment problem: given a curve you did not choose, which work goes where.
They stack rather than compete.
Where the calendar wins anyway
The honest constraint: your peak may not be yours.
Mine is the first two hours after waking, and those belong to a small daughter and a client standup. No tracker solves that. The evenings I build in are consistently a 6 out of 10.
So the version that works when you do not own your day:
- Protect one hour, not the whole peak, and put your hardest recurring task in it, even if that hour is only mediocre in absolute terms.
- Defend the trough downward instead of upward. You may not be able to claim a peak, but you can usually stop agreeing to hard conversations at 3pm.
- Where you have no schedule at all, use a ceiling instead: a limit on how long something gets is more realistic than a place for it, which is the distinction in timeboxing vs time blocking.
The tracker
The energy level tracker is what I built for the logging week. Precisely what it does:
A check-in is a 1 to 10 rating with labels from Exhausted up to Peak Energy, twelve optional activity tags covering sleep, caffeine, meals, exercise, meetings, deep work, breaks, screens, stress and naps, and an optional note. The Today tab draws your day as a timeline from 5 AM to 11 PM with the check-in log underneath, and you can page back through previous days. The Patterns tab averages the last seven days by hour, then names your three highest hours and your three lowest, plus the average energy recorded alongside each tag. The Insights tab turns that into plain scheduling suggestions and lets you copy a text summary.
Three honest notes. It stamps the hour you are in, so you cannot backfill 11am at 4pm: log in the moment or lose the row. The "cognitive budget" panel is a heuristic I invented, where each draining activity you tag costs a flat 15 points out of 100: a nudge, not a measurement. Everything lives in your browser, so nothing is uploaded and nothing survives clearing site data. Entries older than 90 days are dropped.
Where the hours themselves went is a different exercise: the time audit guide. The audit tells you where the time went, the energy log what it was worth.
The limit worth stating plainly
If your line is flat and low all week, this article cannot help you, and pretending otherwise is how productivity writing does harm.
A flat low usually means the input is broken: short sleep, too many commitments, or something that has nothing to do with scheduling. Fix sleep first because it moves every other number, and the sleep debt calculator will at least size the hole. Subtraction is the second lever and the one nobody wants: fewer commitments, not better placed ones. If the word that keeps arriving is depleted rather than tired, the burnout assessment is a more honest starting point.
And if you are persistently exhausted in a way a normal week does not explain, that is a question for a doctor, not a tracker. I build habit software. This is not medical advice.
What actually changed for me
Not much, and the small part that changed has held for months.
I stopped opening email first. That is it. The first block after the house goes quiet goes to the hardest thing on the list, and the admin that used to eat it happens in the last twenty minutes when I am useless anyway.
One task, one hour, moved once. The rest of the week is as chaotic as before, which is the version of this that survives a real life.
Last updated: November 2026.
Written by Eusebiu, the solo founder building Loggd in public. I write code for clients during the day and build this in the evenings around a small daughter, so my building hours are permanently a 6 out of 10 and I plan accordingly. I share what the aggregate data keeps showing on Threads.
Track the pattern over weeks, not days. Loggd shows your consistency as a forgiving contribution grid, so a low week stays a lighter square instead of a reset to zero. Start free.
