A trick aimed at the wrong problem reads as a personal failure

Work out which problem you actually have

These questions are not a quiz and there is no score. They are the fastest way to find out which section of the usual advice is addressed to you, because the four states feel similar from inside and are not similar at all.

• Do you know exactly what needs doing and find yourself unable to begin? That is a starting problem. It is the most common reason a planner full of correct entries sits untouched.
• Do you begin fine and lose the thread partway, surfacing somewhere else entirely? That is an attention problem, and it is the one most of the standard focus advice was written for.
• Do you intend to leave at a particular time and discover the time has gone, or budget an hour for something that takes four? That is time not registering, and no amount of planning fixes it because the planning was not the thing that failed.
• Do you lose the thing you were holding in mind the moment anything interrupts? That is working memory, and it is the one where almost every available fix is external rather than internal.
• Do you have the opposite trouble, unable to stop once you have begun, with the cost arriving later? That is its own pattern and the strategies for starting make it worse rather than better.

If the problem is starting

The advice here is always to break the task down, which is correct and is not a strategy. It is a description of the outcome you want. The strategy would have to be how to do the breaking down from a state where holding the whole task in mind is precisely what has stopped working, and that step is the one the lists skip.

What makes a task startable is a first physical action small enough that agreeing to it costs nothing. Not the first step conceptually, the first thing your hands do. Open the document is startable. Write the report is not, and neither is plan the report, which is the same refusal wearing a smaller coat.

The second half of this is removing the decision from the moment of starting. A stall is expensive partly because you are deciding and beginning at once, and a decision made the night before, by somebody else, or by a tool, is one less thing to carry into a bad morning.

• Reduce the unit until it is a physical action rather than an intention. If the step still needs thinking about, it is too large.
• Decide the next action at a time when deciding is cheap, which is any time other than the moment you are stuck in front of it.
• Start with the version that does not count. Open the file and change nothing. Beginning badly is reliably easier than beginning, and the two usually turn out to be the same event.
• Shorten the distance between deciding and doing. Every step in between, finding the file, logging in, clearing the desk, is a place the decision gets lost.
• Hand the breaking down outside your own head when you cannot do it. That is the one genuinely hard job in this list and the only one a tool meaningfully takes over.

If the problem is holding the thread

This is the territory the timers and the distraction blockers were built for, and in this territory they work. A session with an external end to it holds a shape that your own sense of the work does not, and a blocked site is one fewer place to surface in.

The thing worth knowing is that a fixed interval cuts in both directions. If you have been circling a task for an hour and have finally locked on, an alarm arriving at that moment asks you to stop at the exact point where stopping is most expensive, and the cost of getting back is not a matter of sitting down again. People resolve that by ignoring the timer, and a timer you have ignored twice is decoration.

The version that holds up is an interval that comes from the work rather than from the technique, which means the piece of work has to be small enough to have an end you can see. Then the end of the step and the end of the session are one event instead of two competing ones.

If the problem is time not registering

Almost every strategy offered for this is a planning strategy, and planning is downstream of the actual difficulty. A plan is a set of claims about how long things take, and if your estimates are wrong the plan is wrong in exactly the same proportion. Better planning applied to the same estimates produces a tidier version of the same miss.

What helps is making elapsed time visible rather than countable. A number you have to look up and subtract is a calculation, and a calculation during a task is one more thing competing for the attention the task needed. Something continuously visible, a running clock in view, a timer on a lock screen, a sound at intervals, reports the passage of time without being consulted.

The second half is recording how long things actually took, for a fortnight, with no attempt to improve anything. It is dull, it is the only thing that fixes an estimate, and the gap it reveals is usually larger and more consistent than people expect. The time blindness guide on this site covers the mechanism in detail.

If the problem is holding things in your head

Nearly every fix for this is external, and that is not a workaround. Working memory is the one place where the advice to try harder is most obviously wrong, because effort is not the input and a thing you are holding does not become more held by being held more firmly.

The principle is that anything you are keeping in mind should be somewhere else. One place, not several: two competing lists is worse than no list, because now you also have to remember which one is current. The capture has to be fast enough to survive a bad day, which usually means one gesture rather than a tidy system.

• Write it where you are, not where it belongs. Filing it correctly later is a separate job and a far cheaper one than remembering it.
• Keep one inbox. The value of a capture system is entirely in there being exactly one place to look.
• Put the physical objects in the path you already walk rather than in the place they are supposed to live.
• Convert anything recurring into a standing arrangement, so it stops being something you have to recall at the right moment.
• Close the loop out loud. A commitment made silently to yourself is the easiest kind to lose.

If the problem is stopping

This belongs on the list because it is routinely treated as the absence of a problem. A long stretch of intense output does not look like something to manage, and from outside it looks like the thing everyone was asking for.

The cost is paid forward rather than at the time, and it is paid in meals, sleep, movement and other people, none of which register as a decision while the work is going well. The bill arrives afterwards as a flat period that is then read as laziness, and that reading is what produces a harder system, which consumes the capacity that was left.

The strategies that help here are the opposite of the strategies for starting. An external interruption you have agreed to in advance, a stopping point attached to an event rather than to a feeling, somebody else expecting you somewhere. The hyperfocus and burnout guides on this site are the long version, and reading them together is the point: they are one process described from its two ends.

What a strategy list cannot do

This page is about systems and it is not medical advice. Nothing here diagnoses anything, nothing here treats anything, and a focus app is not a treatment. Those are not disclaimers added at the end, they are a description of the limit.

There is a specific failure worth naming, because it is the most common way advice like this does harm. If a strategy aimed at the right problem still does nothing, repeatedly, that is information about the problem rather than about you. The response that works is a conversation with a clinician, not a longer list and not a stricter version of the same list.

Medication, therapy, coaching and assessment all sit outside what a page can offer, and they are frequently the thing that makes any of the above usable in the first place. A system is a multiplier on capacity rather than a source of it.

Where Task Assistant sits, stated narrowly

Task Assistant addresses one of the four problems above, which is worth being plain about rather than claiming the set. You say what you are stuck on and it returns small steps with time estimates attached, and the timer runs for the length of a step rather than for a standard block, so the end of the work and the end of the session are the same moment.

That is the starting problem, plus a little of the attention one. It does not supply motivation, it does not improve your estimates, it does not remember things for you, and it will not stop you working through dinner. A companion sits on screen while a session runs, which is body doubling rather than a scoring mechanism, and there are no streaks to break.

The task breaker is the same decomposition in a browser with nothing to install, which is the quickest way to find out whether the step you could not find was the whole of the obstacle. If the answer is no, the honest conclusion is that your problem was one of the other three, and the sections above point at what actually addresses those.

Try it now

• task breaker

Questions