Decisional Delays
Decisional Delays
Between finishing one task and starting the next there is a stretch that appears on no schedule. That’s where the next move usually gets decided — which task, what first step, where the materials are — and deciding it there is expensive: the deciding eats the attention the next start needed, and whatever is already open on the desk gets first claim on it. That stretch is a decisional delay, and closing it costs a few minutes the night before: write the next task, its first action, its materials, and a fallback for when it’s blocked. Then finishing something is only finishing it. The next start is already decided.
What the gap costs, and how to see it
The visible minutes between tasks are only part of the delay. The bigger cost is the break in Flow State — the absorption, not a timer, that the gap snaps. When the next action is decided in that gap, four things happen together. Attention leaves the work that just ended. Planning consumes the effort that should have gone into the next start. Uncertainty invites whatever is already open on the desk. Re-entry into focused work takes time of its own. The decision feels small. Repeated across a day, the switching adds up to a large loss.
The repair starts by seeing the leak. For about a week, mark when a focused task ends and when the next focused task actually begins. Compare the planned schedule to the real flow. Count unplanned transition time and the focus breaks that follow those gaps. The first days of the log usually see nothing; awareness of the gap builds over the week. The goal is not a feeling of being more decisive. It is to see whether decisions between tasks are creating hidden drag.
Those two jobs belong together: plan the decisions outside execution time, and measure how much time disappears between tasks. The audit names the leak. The day-edge plan stops it.
Plan outside the block
The next action has to be specific before the block starts, and the block itself has to be protected from new choices. That is what replaces figuring it out as the work goes. Execution mode should not require constant re-planning.
Batch planning and prioritisation at the start or end of the day so mid-day planning is unnecessary. During a focus zone, block inputs so new options do not enter.
A usable plan names six things: the next task; the start condition; the stopping condition; the materials needed; the fallback if the task is blocked; and what not to do during the block. “Study biology” still contains decisions. “At 9:00, Aim the endocrine chapter for 25 minutes” — write what the material must answer before opening it — “and produce five why/how questions” removes more of the ambiguity.
When the plan fails, the failure usually has a named shape.
| Failure | Repair |
|---|---|
| Vague plan | Write the next action |
| Mid-task planning | Batch at the day edges |
| Open distraction channels | Block inputs |
| Overplanned day | Fallbacks and buffers |
| Invisible transition loss | Track a week |
An overplanned day recreates decisions the moment anything slips. Fallbacks and buffers are the signal to loosen a schedule that is too tight to survive contact.
The stretch after the last task is still there. Tonight’s six items have already spent the decision, so the next start is just a start.
Links into the knowledge base
- Decision Making — hub for how much thinking a choice is worth.
- Choice Throttling — a different problem: a tangled choice, not a between-task gap.
- Flow State — the state the gap breaks — absorption, not a timer.
- Focus Management: How to Enter & Recover Inside a Work Block — inside-the-block recover; this page is between blocks.
- Attention Management: Preserving Flow — between-block friction from the attention side.
- Procrastination: a System Problem — when the gap becomes avoidance.
- Self-Management — parent dimension.
- Self-Regulation — control of the switch.
- Marginal Gains — small leaks compound.
Open questions
Which transitions create the most delay?
Does a written next-action queue, calendar blocks, or both close the gap better?
Sources
- Mark, Gudith & Klocke 2008, CHI — 23 min 15 sec to return to an interrupted task after intervening tasks. Bound only: that study measures resume-after-others, not a self-posed “what next?” after a task ends.
- Gollwitzer & Sheeran 2006 — implementation intentions: a specific next action reduces start friction. Support for writing the next action before the gap, not a timer for the gap itself.