ManageTime

Article

Cutting a big piece of work into plan-sized pieces: how small is small enough

Break a big task at the points where something becomes finished and checkable, not at even intervals. A piece is the right size when you could complete it in one sitting and someone other than you could tell it is done. Cut past that and the pieces cost more to track than they save. This article covers where to make the cut, what the evidence on granular subgoals found, the point at which smaller pieces start working against you, and how the answer shifts with the way you work.

How small should you break a task down?

A line on the plan reading rewrite the onboarding guide. Three weeks on it is still there, untouched, because nothing in those four words tells you what to do at ten on a Tuesday. Cutting it up is the standard advice and it is right. What the advice leaves out is where to cut, which is the only part that is hard.

Cut where something becomes a thing that exists. An outline you could show someone. The section on permissions, written. The three screenshots taken. Each of those has a state you can check from the outside, and that is what makes it usable in a plan. Work on the guide has no such state, so it absorbs any amount of time and looks identical at five o'clock.

The second test is whether a piece fits one sitting. Something you can finish before the next interruption arrives gets started, because starting it does not require you to find a clear afternoon first. A piece that needs two uninterrupted afternoons is not a piece. It is the original task wearing a smaller label, and it will sit exactly as long.

The cut that fails most often is the one made by the clock. Four two-hour blocks on the guide is not four pieces; it is one piece with a timer attached, and at the end of the fourth block you still cannot say what is finished. Blocks protect the time once the pieces exist. They do not find them.

Those two tests pull against each other.

A finished, checkable piece tends to run larger than a one-sitting piece, so most work has a size where the two just about meet. For a few weeks of work people mostly land between half a day and two days per piece, though no study fixes that, so treat it as a place to start rather than a rule. The test that holds is whether you can say in one sentence what the piece produces.

Two of this site's methods already gesture at the problem. Eat That Frog says that when a task feels too big, your frog is its first real step rather than the whole thing. The Rapid Planning Method asks you to write the weekly and daily results that build toward a larger outcome, which is the same move at a wider scale.

Neither says how far down to go. The evidence on that is thinner than the advice, and more interesting.

What does cutting the work up actually buy you?

About 8%, in the one place it has been measured at scale. Rai and colleagues ran a preregistered field experiment with 9,108 volunteers at a national crisis counseling organization, published in the Journal of Applied Psychology in 2023. The volunteers held an annual goal of 200 hours. Reframing that into granular subgoals, either four hours every week or eight hours every fortnight, raised hours volunteered by 8% over twelve weeks.

The more useful result is which cut lasted. The experiment set the two granular versions against each other, and the looser one won on durability: eight hours every fortnight held its benefit longer than four hours every week, even though the weekly version is the finer cut. Making the pieces smaller made the schedule tighter, and the tighter schedule was easier to fall out of.

Granularity and flexibility are the same choice.

That is the finding worth carrying into your own work, because it answers a question the usual advice does not ask. Every cut you make buys you a piece you can start and hands you a piece you can miss.

A week where you were travelling, or ill, or buried in something else breaks a plan made of weekly commitments and merely dents one made of fortnightly ones. So when two cuts both look workable, the coarser one is the safer bet for anything that has to survive a bad fortnight.

Be careful how far you carry the number. These were volunteers at a crisis line working toward an hours quota they had signed up for, not employees with a report due, and the study counted hours rather than finished work. A vignette pilot of 900 participants behind the field experiment pointed at self-efficacy, goal commitment and less procrastination as the mechanism, which is plausible for office work and untested there.

What transfers is the shape: splitting helps, and splitting too finely gives some of it back. The 8% does not transfer.

Can you break a task down too far?

Yes, and it fails in a way the advice never warns you about. Amir and Ariely tested what they called discrete progress markers across four experiments, published in the Journal of Experimental Psychology: Learning, Memory, and Cognition in 2008.

The markers helped when people could not tell how far from the end they were, cutting that uncertainty and lifting both performance and how much the task appealed to them. When the distance to the goal was already clear, the same markers did the reverse. People eased off, and performance and appetite for the task both fell.

Resting on laurels is the paper's own phrase for it, and it describes something you can watch happen at a desk. A checklist reading nine items, four ticked, hands you a reason to stop at four. Nothing about the work changed. What changed is that a marker is now standing between you and the end of the job, close enough to feel like progress on its own.

So the cut earns its place when you cannot see the end, and costs you when you can.

A second line of evidence says you were already running this calculation. Binder and colleagues, in Cognitive Science in 2025, had people choose between subgoals while building virtual block towers, then tested which computational models predicted those choices.

People favoured subgoals that balanced progress toward the finished tower against the effort of working the subgoal out. They told apart subgoals that looked equally ambitious and were not equally cheap to solve. Their choices may also have tracked the cost of what would be left after the next piece, though the authors offer that as a possibility their data fit rather than a finding.

Block towers in a lab are a long way from a quarter of work, and none of this was measured on office tasks. What the studies agree on is still worth having: the pieces carry a price. Every extra cut adds something to write down, somewhere to tick it off, and a judgement later about whether it counts as done.

So once the pieces are small enough that maintaining them is the work, the cutting has gone past the point where it helps. Timeboxing is the usual tell. A piece that one box disposes of did not need its own line on the plan.

There is a test for this. Look at the last three pieces you ticked and ask what changed outside your own tracker when each one was done. If the honest answer for two of them is nothing, the cut is finer than the work, and the ticks are measuring your bookkeeping.

How does your working style change the answer?

None of the three studies sorted people by how they plan, so what follows is this site's reading rather than a result. The cut divides the four working styles on one question: which failure you are more likely to produce, a piece too big to start or a plan too fine to keep up.

If you want the whole structure visible before you begin, the Architect's pattern, your risk is the second one. The breakdown will be thorough, nested, and finer than the work needs, and it will take on upkeep of its own. Set a limit before you start cutting: one level of pieces rather than two, and no piece that exists only to be ticked.

Rai's durability result is aimed at you. The looser cut is the one still standing after the fortnight your plan went sideways.

Burst workers, nearest the Sprinter, have the opposite problem and it hides better. A big piece suits you on the days the burst arrives and leaves you nothing to do on the days it does not. Cut the first piece small and leave the rest coarse. What you need is a way in on a flat Tuesday, not a complete map.

If you think in outcomes, closer to the Visionary, you will cut by subject rather than by state: three pieces named after the three parts of the topic, each able to absorb a week. Rename them by what exists once they are finished. A piece called permissions tells you nothing. The permissions section drafted tells you when to stop.

A week that never repeats, the Improviser's normal, makes a fixed schedule of pieces a fiction by Wednesday. Amir and Ariely's warning is the one to watch here. A long list of small ticks offers plenty of visible progress on a day when none of it moved the thing that mattered. So keep the pieces and drop the sequence. Know which piece is next and which one is the end, then let the middle reorder itself.

Take the biggest unstarted line on your plan and write what its first piece produces, in the past tense, as something another person could verify. If that sentence takes more than a minute to write, the piece is still too big, and the line has been sitting there for that reason rather than yours. If you are not sure which style is yours, the working style assessment takes a few minutes, and the planning pillar covers where the pieces meet the daily and weekly passes.

Frequently asked questions

How small should you break a task down?

Small enough to finish in one sitting, and large enough that finishing it leaves something another person could check. For a few weeks of work most people land between half a day and two days per piece. No study fixes that figure, so treat it as a starting point. What settles it is whether you can say in one sentence what the piece produces.

How do you break a big piece of work into smaller tasks at work?

Cut at the points where something becomes finished, rather than at even intervals of time. Ask what will exist when each piece is done: an outline, a drafted section, the numbers requested. A piece named after a state can be checked from outside. One named after an activity absorbs any amount of time and looks the same at five o'clock.

Can you break a task down too far?

Yes, and the cost shows up as easing off after each piece rather than as wasted admin. Amir and Ariely found across four experiments that progress markers helped when people could not judge how far from the end they were. When the distance was already clear, the same markers reduced both performance and appetite for the task. The other cost is upkeep: once maintaining the pieces takes longer than the pieces save, the cut has gone too far.

Should each piece of a broken-down task get its own deadline?

Not every one, because dating every piece makes the plan tighter than it needs to be and easier to fall out of. In Rai and colleagues' field experiment, eight hours every fortnight held its benefit longer than four hours every week, the finer of the two cuts. Date the pieces other people are waiting on, and leave the rest in order without dates.

Sources