Break a procrastinated task into easy starting steps
Reduce ambiguity and activation cost without diagnosing motivation or trivializing real barriers
4 min setupTested with:ChatGPTReviewed: 2026-08-28
1
Add your context
Your text stays in this browser. AILesson Prompts does not send it to a model or server.
2
Your prompt
Unfilled fields remain visible as placeholders, so you can still copy and edit the prompt
Break the delayed task into steps that are easy to start while still leading to the real outcome.
Task, done definition, deadline source, current state, and importance:
[task]
Observed decision, input, emotional, physical, tool, and attempt barriers:
[barriers]
Time, energy, environment, support, minimum progress, and boundaries:
[capacity]
Do not diagnose the person, moralize, or assume laziness. Identify which supplied barriers make starting costly: ambiguity, task size, missing permission/input, risk, perfectionism, context setup, low energy, or conflict. Separate work from decisions and waiting. Return a first action that takes no more than the available start window and produces visible evidence; the next 3–7 small steps with stopping points; a minimum version; if-then responses for likely barriers; a request for help when needed; and a restart note for interruption. Preserve safety, health, and stated boundaries. Do not hide a major unresolved decision inside a tiny task.
Private by defaultPrompt assembly happens locally in your browser. Avoid placing confidential information into any AI service unless your organization allows it.
From input to outcome
A worked example
See how concrete context turns this recipe into a usable result
Actual input
Delayed task and outcome
Submit a two-page conference proposal by 30 September 17:00. Done means title, 250-word abstract, three learning outcomes, 80-word bio, and manager approval uploaded to the portal. Today is 20 September. I have scattered research notes but no draft. The topic matters for my professional development.
Observed barriers
I keep trying to find the perfect angle. Three possible audiences are mixed in my notes. The manager needs two working days to review. I do not know whether customer numbers may be disclosed. Prior attempts ended after reorganizing notes. No health concern stated.
Available capacity and support
I have 15 minutes now and 45 minutes tomorrow morning. Low energy now. A colleague can read one paragraph Thursday. Minimum useful progress today is something I can reopen. Do not schedule weekend work.
Example output
First 15-minute action: create a document named “Conference proposal—working draft”; paste the required five sections as headings; under Audience write the three candidates; choose one provisional audience for tomorrow and add “decision can change after abstract v1.” Visible evidence: saved skeleton plus one provisional choice. Do not reorganize all notes.
Tomorrow's 45 minutes: 1) 5 minutes—write one sentence: audience, problem, promised change. 2) 15 minutes—draft an intentionally rough 180–300-word abstract using only non-sensitive claims. Stop and save. 3) 10 minutes—derive three observable learning outcomes. 4) 5 minutes—mark every desired customer number “permission needed,” not as fact. 5) 10 minutes—write title options and a restart note. Thursday: colleague reads only abstract for clarity; then request disclosure decision and manager review no later than the date that preserves two working days.
Minimum version contains one clear audience, source-supported abstract, three outcomes, factual bio, and no unapproved numbers. If angle perfection returns, keep the provisional audience through abstract v1. If disclosure remains unresolved, omit numbers or use approved aggregate wording. Restart note: “Next: resolve comments in abstract paragraph 2; audience currently provisional; customer figures blocked.”
Why this works
1
A visible first artifact reduces ambiguity more effectively than the abstract instruction to begin.
2
Stopping points and restart notes reduce the cost of an imperfect or interrupted work session.
Check the result
Can the first action finish within the stated start window and leave visible evidence?
Do the small steps collectively reach the real definition of done?
Are missing decisions and external dependencies exposed rather than disguised as effort?
ChatGPT