You know the task. It has been on your list for days, maybe weeks. You are not lazy — you have done harder things than this before. So why does this one thing keep sliding to tomorrow?
The task decomposition below is not a motivational trick. It uses AI to help you name the specific kind of friction that is stopping you, then generates a next action small enough that “I don’t have time” and “I don’t know where to start” stop being true. It will not fix everything that looks like procrastination — some of that is a signal to get help, not a smaller task — but for ordinary avoided tasks, this usually works in under five minutes.
Procrastination is friction, not a character flaw
The largest meta-analysis on the topic, covering 691 correlations across decades of research, found procrastination is driven far more consistently by task aversiveness, low self-efficacy, and impulsiveness than by traits like laziness or low conscientiousness alone (Steel, 2007, Psychological Bulletin). Self-efficacy — your belief that you are capable of doing the task at all — is itself a well-studied lever: people with low self-efficacy for a specific task avoid it more, independent of how capable they actually are (Bandura, 1997, Self-Efficacy: The Exercise of Control). In plain terms: the task itself is unpleasant, unclear, or you doubt you can do it well, and your brain routes around the discomfort toward something easier. That is a solvable engineering problem, not a moral failing.
A model generates text; it does not know why you specifically are avoiding this task. What it can do well is help you interrogate the avoidance systematically instead of just repeating “I should really do that” to yourself for the tenth time. You still have to answer honestly.
If the same task (or every task) has been stuck for months, if getting out of bed itself is the hard part, or if this pattern comes with persistent low mood, exhaustion, or hopelessness, this is not a friction problem a five-minute exercise will fix. See the “when this is not friction” section below before you try to decompose your way out of it.
The six-branch friction diagnostic
Most avoided tasks are stuck on one (sometimes two) of six specific frictions. Naming the actual one matters, because the fix is different for each.
| Friction | What it feels like | Wrong fix | Right fix |
|---|---|---|---|
| Ambiguity | ”I don’t know what ‘done’ looks like or where to start.” | More willpower | Define the smallest concrete first step |
| Emotion | ”I feel dread, embarrassment, or anxiety just thinking about it.” | Push through it | Reduce the emotional stakes of the first step |
| Energy | ”I could do this if I weren’t exhausted.” | Caffeine and pressure | Match the task to a real energy window |
| Environment | ”Every time I sit down something interrupts me.” | Try harder to focus | Remove the specific interruption |
| Skill | ”I’m not confident I know how to do this part.” | Avoid it until forced | Do the smallest learnable slice first |
| Consequence | ”There’s no real cost to delaying, so I don’t.” | Guilt | Create a visible, near-term stake |
Use the model to help you triage which branch actually applies — it is a good pattern-matcher against your own description once you write it down honestly.
Here is a task I've been avoiding: [describe the task and how long
you've put it off].
Using these six categories — ambiguity, emotion, energy, environment,
skill, consequence — ask me up to three short diagnostic questions to
figure out which friction (or frictions) is actually stopping me. Do
not suggest solutions yet. Just ask questions.
Illustrative output shape:
1. If you pictured actually starting this task right now, what's the
first concrete thing your hands would do — and can you name it, or
does it stay vague?
2. Is there a specific feeling (dread, boredom, embarrassment) attached
to this task, or does it feel neutral but just never gets picked?
3. Is there anything genuinely uncertain about how to do a part of
this, or do you already know how — you're just not doing it?
Answer honestly, in a sentence or two each. The pattern in your own answers usually points to one dominant friction — most people find a specific line in their own answer that names it once they see the question.
Naming the friction, then the five-minute start
Once you have a working diagnosis, ask the model to convert it into a task so small it removes the excuse to delay further:
My avoided task is: [task].
The main friction is: [ambiguity / emotion / energy / environment /
skill / consequence — pick from your diagnosis above].
Give me a next action that takes 5 minutes or less, requires no
decision-making once I start, and directly reduces that specific
friction. Then give me one clear stop point — a moment where I'm
allowed to put it down having made real progress, not just "keep
going until I burn out."
Illustrative output shape (friction: ambiguity, task: “write performance review self-assessment”):
Next action (5 min): Open a blank doc and write down the three projects
you worked on this quarter as a bare list — no descriptions, just names.
That's it.
Stop point: Once the three project names are listed, stop. You now have
the actual skeleton of the document instead of a blank page, which was
the real blocker.
Notice what this is not: it is not “just do the whole task in one sitting.” The friction was ambiguity — not knowing where to start — so the fix is a concrete, bounded first move. If your friction were energy, the same task might get a different next action: “Do this at 9am tomorrow when you’re fresh, not tonight.” If it were environment, it might be: “Put your phone in another room for these 5 minutes, nothing else.”
This is a direct application of what researchers call implementation intentions — pre-deciding the specific “if situation X, then I do Y” plan rather than relying on in-the-moment willpower, shown across decades of studies to reliably increase follow-through (Gollwitzer, 1999, American Psychologist). The task itself does not need to feel less daunting; it needs a small enough entry point that starting no longer requires a decision.
The stop point matters as much as the start
A next action without a stop point tends to either not get started (it looks like the whole task in disguise) or spiral once you’re in motion (five minutes becomes three hours, and next time the task feels even more daunting because it swallowed your evening). Define both before you start:
- Start: a single, concrete, 5-minute-or-less action.
- Stop: a specific, visible marker of progress after which you are allowed to stop without guilt — not “keep going if it’s flowing,” which erodes the boundary you just built.
If the five minutes go well and you want to keep going, that is a bonus, not the plan. Stopping on schedule, having made visible progress, is what makes the technique repeatable tomorrow instead of a one-off burst.
When this is not friction
Task decomposition helps with ordinary avoidance: a task that is unpleasant, unclear, or poorly timed, on top of a baseline where you can generally get things done. It does not treat depression, ADHD, anxiety, burnout, or unsafe working conditions, and applying “just find the 5-minute version” to someone whose difficulty comes from one of those is not just unhelpful — it can reinforce the idea that the problem is a personal failing to engineer around rather than something that deserves support.
Depression specifically often looks like procrastination from the outside — the World Health Organization’s fact sheet on depressive disorder lists poor concentration, hopelessness, disrupted sleep, and persistent low energy among its core symptoms, present most of the day, nearly every day, for at least two weeks (WHO, “Depressive disorder,” 2025). That is a materially different pattern from “I keep avoiding this one report.”
Signs this is bigger than friction:
- The pattern covers nearly everything, not one or two specific tasks.
- It has lasted months, not weeks.
- It comes with exhaustion, low mood, or difficulty concentrating that shows up outside of “avoided tasks” too.
- Sleep, eating, or basic daily functioning are also affected.
- You have tried decomposition-style fixes repeatedly and nothing sticks.
If any of that describes your situation, the next step is a conversation with a clinician, a workplace accommodation conversation with your employer, a tutor or learning-support contact if this is school-related, or a coach — not another prompt. The World Health Organization’s guidelines on mental health at work explicitly recommend manager training and return-to-work support precisely because workplaces are expected to accommodate conditions like this, not just expect individuals to engineer around them alone (WHO/ILO, 2022). See when AI is not therapy for how to tell the difference between “I need structure” and “I need a person.”
What AI cannot do here
The model cannot tell whether your avoidance is ordinary friction or something that needs support — it only sees the words you give it, and it will happily generate a five-minute-task breakdown for a problem that actually needs a doctor, because breaking tasks into steps is what it was asked to do. It also cannot hold you accountable to the stop point or the next action; a habit tracker, a friend, or your own calendar entry does that.
Try it today
Pick one specific task you have avoided for at least three days. Run the diagnostic prompt, answer honestly, then generate the five-minute start and stop point. Use the task friction diagnostic to write it down — the act of writing the friction name and the stop point on paper (not just in the chat) makes both easier to actually follow. If you are building a longer learning or work habit around this, see the 30-day AI learning plan and protecting your attention while using AI for what comes after the first five minutes.
