How to break large tasks into smaller, actionable steps

There is a task on your list right now that has been there for three weeks.

You have meant to do it every day. You have meant to do it every weekend. You have not done it.

The honest reason, almost certainly, is not that you are lazy. It is that the task is too big to start. Somewhere inside it is a real next action — but you cannot see it from the outside of the task, and you cannot see it from the list either. So you scroll past, again, and reassure yourself that you will get to it tomorrow.

This article is about how to take a task that is too big and find the small thing that will actually let you start. It draws on the research and practice of executive-function coaching, productivity writing, and the small, unglamorous tradition of just breaking the work down. None of it is novel. All of it works.

Why your brain refuses to start

Three things tend to be true about tasks that get stuck on a list for weeks.

The first is that working memory is small. Research summarised by the productivity coach Melissa Gratias suggests that working memory can hold about three to five items at once.¹ If a task contains seven distinct steps, your brain cannot hold all of them at the same time. You look at the task, feel the shape of it being too big to fit, and bounce.

The second is that the form of a goal matters more than its content. Goal-setting research (Locke 1968; Latham) has consistently found that specific goals produce more progress than general ones. The same study tradition gives us the now-cliché advice to make goals SMART. The reason that advice persists is because it works: vague tasks generate avoidance; specific tasks generate action.²

The third is the Zeigarnik effect. Unfinished tasks hold a privileged place in memory — your brain keeps surfacing them as background noise. The Zeigarnik effect is supposed to help you get back to the task. In practice, with a list of fifty unfinished items, it ends up generating a persistent low hum of "you should be doing something" without telling you which thing.³

Combine those three: the task is too big for working memory, too vague to act on, and quietly nagging at you in the background. No wonder you keep scrolling past it.

The fix is not motivation. The fix is making the task small enough and specific enough that starting it does not feel like starting it.

The two-minute first step

The simplest possible decomposition technique is: what is the first thing I could do in under two minutes?

This is not the same as "what is the smallest possible part of this task?" That question is too abstract. It produces answers like "open the document," which sounds small but is not actually a true first step.

The two-minute question is more concrete. What is the literal next physical action? If the task is write a chapter, the two-minute step is usually not "open the document." It is something earlier than that: clear off the desk, or find the notes file from last week, or put the kettle on and find the chair I write best in. These look ridiculous when you say them out loud. They are also, almost always, the actual reason the task has not started.

The executive-function coach who runs Beyond BookSmart told a story about a podcast intro she had been avoiding for days. She thought the friction was the recording. The actual friction was that her desk was too cluttered to set up the mic. Once she identified that — about two minutes of clearing — the rest took five minutes.⁴

Two-minute first steps almost always exist. The reason they are hard to spot is that they look beneath the task, not inside it.

A simple decomposition framework: STM

A useful, easy-to-remember structure: Steps, Sequence, Time, Mapping.

Steps. Write down every single thing the task contains. Aim for granular. Write the report is not a step; open the notes file from last Tuesday is.
Sequence. Put the steps in order. Some of this is obvious. Some of it will reveal that what you thought was the first step is actually third.
Time. Estimate how long each step will take. You will be wrong, in a specific direction (most people underestimate by about 50%). Estimate anyway, and adjust over time.
Mapping. Put the steps into your calendar, or into your week, in actual time slots. Not all of them on the same day. Different steps on different days.

This is, structurally, the same framework that David Allen's Getting Things Done method uses at the project level — capture, clarify, organize, reflect, engage — but compressed into a single sitting for a single task.⁵ You do not need to convert your whole life to a methodology. You just need a quiet 15 minutes to break this one task down.

STM is most useful when the task has been on your list for more than a week. For smaller tasks, the two-minute first step is enough. For tasks that have been on your list for more than a month, STM plus a calendar slot is usually what gets them moving.

How big should a "small step" be?

This is the question most articles about task decomposition skip.

A good small step has three properties:

It is clearly finishable within the time you have for it (15 minutes, 30 minutes, an hour).
It produces a visible artefact — a paragraph written, a section outlined, a meeting scheduled, a document opened and not closed.
It does not require any new decision to start. The decision was made when you wrote it down.

The last one matters most. The most common reason a "small step" still doesn't get started is that it secretly contains a decision. Outline the report sounds small, but if you haven't decided what shape the report takes, you cannot start. Draft a one-page outline using the structure from last quarter's report removes the decision and makes the step starter-ready.

If a step still feels heavy after you have written it down, look for the decision hiding inside it. Move that decision into its own step. The task will start moving again.

Common traps when breaking tasks down

Three things tend to go wrong.

The first is over-granularising until the breakdown itself becomes the work. If you have written 47 sub-steps for a single task, you have not decomposed the task — you have avoided it by inventing a planning ritual. Aim for between three and eight steps for most tasks. Add detail only when a step refuses to start.

The second is treating breakdown as one-time. Most large tasks do not yield to a single decomposition session. The first session gets you started. The second session, after you have made some progress, refines what's left. Plan to revisit the breakdown after the first concrete step is done.

The third is using breakdown as procrastination cover. There is a particular pattern where you "plan" the task three weeks in a row and never do it. The clue is that the plan stays roughly the same each time. If your decomposition has produced the same first step three weeks running, the problem is not the breakdown. The problem is the calendar slot. Put the first step on the calendar this week. Stop re-planning.

A short note on Oasa's Sprouts

In Oasa, smaller steps inside a Seed are called Sprouts. The vocabulary is doing a small piece of psychological work: a Sprout is something that emerges from a Seed, not a thing assigned to it. It implies growth happening, not work imposed.

A user can Tend (focus on) a single Sprout at a time. The Seed has not disappeared; it is just temporarily zoomed in. Completed Sprouts become small visible additions to the Garden, alongside completed Seeds.

This is one specific implementation of the principles in this article. There are other apps with similar mechanisms — subtasks, checklists, project-level breakdowns. The mechanic matters less than the discipline of using it.

If you would like to see how the language fits together, the home page walks through it in a single screen.

Key takeaways

A task that has been on your list for three weeks is usually too big to start, not too unimportant to start.
Three reasons your brain bounces off it: working memory is small, vague goals produce avoidance, and the Zeigarnik effect generates low-grade nagging without specifying the next action.
The two-minute first step is usually beneath the task, not inside it.
STM (Steps, Sequence, Time, Mapping) is enough for most stuck tasks. Use it once. Then put the first step on the calendar.
A good small step is clearly finishable, produces a visible artefact, and contains no hidden decisions.
Beware the three traps: over-granularising, treating breakdown as one-time, and using breakdown as procrastination cover.