ARTICLE · 11 SEPTEMBER 2026
Timeboxing: how to stop work from expanding to fill the time
A to-do list says what to do but never when to stop. Timeboxing starts from the other end: fix the time first, and when it runs out, stop regardless of the result. Where the idea comes from, how it differs from time blocking, and how to apply it to meetings and to work you do alone.
01
Work expands to fill the time available
In 1955 the British historian Cyril Northcote Parkinson published a short satirical essay in The Economist. Its first sentence: "Work expands so as to fill the time available for its completion." His example was the Colonial Office, whose staff kept growing while the number of colonies it administered shrank. Less to do, more people, and the extra people generated work for one another.
The observation applies just as well to a single working day. Give a report three days and it takes three days; give it one and it takes one. The result is rarely three times better. The extra time is mostly absorbed by deliberating, polishing and postponing. Timeboxing turns the law around: if less time means less work, decide the time first.
02
What timeboxing is, and what it is not
The word timebox settled into use in software development in the early 1990s. Instead of building until a feature is finished, fix the period first and fit in what you can. A Scrum sprint is the classic timebox: when the two weeks end, the sprint ends, finished or not.
Applied to personal work the definition is: set a ceiling on the time a task gets, in advance, and stop when it is reached. All three parts are required. In advance, a ceiling, and an actual stop. Drop any one and it is no longer timeboxing.
It helps to separate it from its neighbours.
| Method | What you fix | What ends it | What it lacks |
|---|---|---|---|
| To-do list | What | The task being done | No stopping point; one item can eat a day |
| Time blocking | When | The next block starting | Nothing stops work from swelling inside the block |
| Timeboxing | How long | The time running out | Requires practice at stopping |
| Pomodoro | Length of a focus block | Every 25 minutes | Fixed length regardless of task size |
Time blocking is writing "report, 2 to 4 pm" in the calendar: it fixes a start time. Timeboxing is deciding "two hours on the report draft": it fixes a ceiling. They can be used together, and work best that way. Pomodoro is a special case of timeboxing with the box fixed at 25 minutes and the aim being sustained attention rather than completion; the differences are in the Pomodoro Technique article.
03
In meetings: a box around every agenda item
Meetings are where Parkinson's law is easiest to see. An hour-long meeting takes an hour whether it has two agenda items or six. Timing each item changes that noticeably.
- Write a time next to each item
In the invitation: "1. Schedule check (5 min) 2. Design decision (15 min) 3. Other (5 min)". The total must be less than the meeting length. If it is exactly equal, it has already overrun.
- Make the countdown visible to everyone
The chair watching the clock and the whole room seeing the remaining time are different things. In a screen-shared meeting, share the entire screen and float a 10 minute timer above it as a PiP window. Sharing only one window leaves the PiP window out of the stream.
- When time is up, move on rather than forcing a decision
No decision after 15 minutes means information or a decision-maker is missing. More time will not fix that. Note what is missing and move to the next item.
- End early when the agenda is done
If the items finish early, the meeting finishes early. Filling the remaining time is how Parkinson's law returns.
04
Working alone: the box size changes
On your own, timeboxing solves two different problems: not being able to start, and not being able to finish. Same tool, different box sizes.
- 15 minutes — the box for startingGive the task you have been avoiding exactly 15 minutes. The aim is not to finish it but to find out what it actually involves. When the 15 minutes end you may stop. Usually you keep going, but that is a bonus, not the goal.
- 45 minutes to an hour — the box for a draftReports, proposals and slide decks get a first draft in this box. No polishing. An ugly draft at the end of the box is success. No draft means the research was not done, and that is a different box's job.
- 2 hours — the box for a deadlineFor work that genuinely has to be finished, set two hours with the last 20 minutes reserved for checking. Sitting a practice exam at full length is this type. This box is not interrupted.
Whatever the size, one rule holds. When the box ends, write one line about where things stand: "draft 70%, conclusion not written." That line is the starting point for the next box, and after a few of them you can see how optimistic your estimates are.
05
Three reasons timeboxing fails
- The box is too big"The report, this afternoon" is not a box. With a four-hour ceiling, Parkinson's law operates comfortably inside it. Stay under an hour to begin with.
- Extending becomes a habitAllow "ten more minutes" once and the box no longer exists. If more time is genuinely needed, open a new box; do not stretch the current one. The distinction looks trivial and the whole method hangs on it.
- Filling the time becomes the goalA 45 minute box with no draft at the end is a failed box. Decide what has to exist when the box ends, not just how long it lasts.
All three happen more often when the remaining time is not visible. You speed up because you can see the end approaching, and you stop because you can see it has arrived. A timer left in a background tab and forgotten is not timeboxing; it is just working.
06
How to start today
- Pick the one task you have been putting off longest
Do not load several into one box.
- Set 15 minutes
Open the 15 minute timer and press PiP so it floats above your work. Starting is the goal, so keep it short.
- Stop when it ends and write one line
Even if you want to continue, stop first. Note where you got to and what comes next.
- If you continue, open a new box
This time 45 minutes. A new box, not a longer one.
After a few days a pattern in your estimates appears. Most people overestimate small tasks and underestimate big ones. Knowing your own bias is the second thing timeboxing gives you.