How do you stop working when the work isn’t finished?

Sources checked 17 October 2026. The routines, notes and workplace scenarios below are illustrative applications of the research.

It is 6:12. The proposal is almost finished, except for the price. You could decide that now. But deciding it means checking the supplier’s estimate, and the estimate raises a question you should probably send tonight.

An hour later, the proposal is still open. You have answered two other emails and remembered something for tomorrow’s meeting. There was no obvious moment to stop because every completed task exposed another task beside it.

For a founder, freelancer or anyone with an open-ended job, an empty list is a difficult condition to attach to an evening off. You can always improve the draft, approach another customer or clear another message. The question is what needs to be settled before you leave, and what needs a reliable way back into your attention tomorrow.

What the unfinished work is asking you to remember

An incomplete task contains more than its title. There is the last decision you made, the source you were checking, the question you could not answer, and the worry that you will forget all three. Closing the document leaves those details somewhere. Often, you have left yourself to remember them.

In experiments published in 2011, E. J. Masicampo and Roy Baumeister found that unfinished goals could intrude during unrelated tasks and interfere with performance. Making specific plans for those goals removed the interference effects in their studies. That is evidence for a useful role of planning under those conditions, rather than a guarantee that writing a list will clear anyone’s mind. [1]

“Finish proposal” is a reminder that a problem exists. “Open the pricing sheet, compare the two supplier quotes, then fill in section four” preserves a way to work on it. The second note asks less of tomorrow’s memory.

The note does not have to explain the whole project. Write enough that someone with your knowledge, arriving tomorrow without tonight’s train of thought, could begin. Usually that means the current state, the first physical action, the relevant file or missing input, and a realistic time to return.

For a designer, it might be: “Client approved the layout. Export the mobile version from frame seven. The replacement photograph is in the shared folder. Start after the 10:00 call.” For a researcher: “The chart uses the old denominator. Check table two in the source paper before editing the caption.” These notes preserve different details because the work is different.

The evidence gets more specific than “switch off”

Brandon Smit studied 103 employees pursuing 1,127 goals. Incomplete goals, particularly important ones, were harder to detach from later in the day. An experimental planning intervention helped employees who were typically more prone to difficulty detaching. The reported benefit was conditional; it was not equally established for every worker. The paper appeared online in 2015 and in a 2016 journal issue. [2]

Sophie Leroy and Theresa Glomb examined a related problem in four studies. Anticipating time pressure when returning to interrupted work made it harder to focus on the intervening task. A brief plan for resuming the original work reduced that effect. Their research concerns task interruptions, so applying it to the end of a workday is a practical extension, not the exact situation they tested. [3]

These distinctions matter because productivity advice can acquire more certainty each time it is retold. A laboratory effect becomes a universal account of the brain. A useful routine becomes a promise that you will stop worrying. The studies offer a reason to try a concrete return plan and see whether it helps you. They leave room for workload, responsibility and the conditions of your job to matter.

Give the end of the day some working time

Cal Newport has described a shutdown routine that reviews his tasks, calendar and weekly plan before he closes the computer. His account is a personal practice, including his report of feeling less troubled by work in the evening. The transferable part is the review: he checks what he has committed to before calling the day finished. [4]

You could reserve the final ten minutes of a work period for a shorter version. Ten is a convenient allowance, not an experimentally established dose. If you intend to leave at six, the review belongs before six. Adding it after the time you promised to stop quietly lengthens the day.

First, check whether anything requires action before you next return. Look at the actual deadline and consequence. A supplier waiting for confirmation before an overnight dispatch deserves different treatment from a non-urgent request that arrived five minutes ago. Deal with the first, arrange coverage, or tell the affected person what has changed.

Then leave a return note for the work most likely to follow you home. You do not need to rewrite the entire task list. Capture the new commitments in the place you already use, inspect tomorrow’s calendar, and give the important unfinished work a plausible next slot.

Finally, stop opening new material. Reviewing an inbox can turn into processing an inbox, and processing it can create another hour of replies. The review needs a stopping condition of its own: urgent responsibilities handled or handed over, new commitments recorded, and the next work period possible to begin.

If the review repeatedly uncovers more work than tomorrow can hold, change the plan while the problem is visible. Move a deadline, reduce the scope or ask which commitment takes priority. Assigning every task to tomorrow creates a list you have little reason to trust.

Some unfinished work belongs to someone else

Waiting is an awkward state. You cannot finish the proposal until the supplier sends a price, but you can keep checking for it all evening.

Give the dependency an owner and a check time. “Waiting for the supplier’s revised quote; requested today at 3:00. Check at 10:00 tomorrow. If it has not arrived, call and tell the client whether the proposal date changes.” That is enough information to act later without pretending the quote has arrived.

If another person is taking over, a note alone is insufficient. They need to receive the handover, understand the relevant deadline and accept responsibility. Putting someone’s name beside a task does not establish that they have agreed to cover it.

A freelancer can also send a short expectation-setting message: “The draft is complete. I am waiting for the delivery cost before confirming the total. I will update you by 11:00 tomorrow.” It tells the client what exists, what is missing and when to expect news. Send it only when that update time is realistic.

For an employee, the conversation may need to happen with a manager: “I can finish the customer revision today or the internal deck. Which should move to tomorrow?” The difficulty may be an unresolved priority rather than a missing personal ritual. Making the conflict explicit gives someone a decision to make.

When everyone can still reach you

A closed laptop does little if the phone continues to deliver the same work. This becomes especially awkward when the team has never agreed what an evening message means. One person is recording a thought for tomorrow. Another reads it as a request to respond now.

Make the expectation specific enough to use. Routine replies can wait until the next work period. Urgent issues need a defined channel and someone responsible for receiving them. On-call work needs an actual rota or other coverage arrangement. The person finishing a shift should know who has taken over; the person starting one should know what remains unresolved.

A founder’s habits influence this agreement. If you regularly request something at 10:00 p.m., a sentence about flexible hours may not make the request feel optional. Where practical, schedule routine messages for working hours and state the deadline when it is unclear. Then respond consistently when someone waits until that deadline.

Some workplaces give people little control over any of this. A precarious contract, an understaffed team or a manager who expects constant availability can make stopping costly. A better note can reduce the effort of restarting; it cannot supply missing staff or negotiate a boundary on your behalf. That is a reason to address the work arrangement, rather than blaming yourself when a shutdown checklist fails to produce a peaceful evening.

A thought can return after you have stopped

You may remember the proposal halfway through dinner. The existence of that thought does not require an immediate work session. Check whether it contains new information or repeats something already in the return note.

If it is genuinely new and can wait, write a brief line in the same capture place and leave it there. “Ask whether delivery includes installation” is a reminder. Opening the quotation, searching the terms and drafting a reply is work. You can choose to do it, but count it honestly when deciding whether your evenings are available to you.

Avoid turning the capture step into a second planning session. One line can preserve the question. Tomorrow’s review can decide what to do with it. If an urgent responsibility really has appeared, use the coverage arrangement you made rather than assuming every resurfacing thought qualifies as an emergency.

It also helps to have somewhere to go after the desk: a meal, a walk, a conversation, a book you were already reading. There is no requirement to make that time useful to tomorrow’s output. The work period has ended, and another part of the day can begin.

Check the note in the morning

The next morning, test whether the note is useful. Open it before reconstructing the task from the inbox. If the situation has changed, update the next action. If it has not, begin where you said you would.

After a few days, notice whether the notes are usable. “Work on pricing” may still be too vague. A paragraph explaining every detail may take longer to write than you want to spend. Keep the smallest amount of context that lets you begin without repeating last night’s decisions.

Do the same before a weekend, with the return date made explicit. A note that says “tomorrow” on Thursday can become ambiguous after a change of plans. “Sunday, after the team check-in” or “Monday, first work block” is easier to interpret when you come back.

For tonight’s proposal, you might leave this: “Scope and draft complete. Supplier quote still missing. At 9:30, check for the quote; if absent, call. Then finish the price table in section four. Client update due at 11:00.”

Save the draft beside the note. Check that the update fits tomorrow’s calendar. Close the document with section four still unfinished.

Sources

• [1] Masicampo & Baumeister, “Consider it done! Plan making can eliminate the cognitive effects of unfulfilled goals” (2011). Journal of Personality and Social Psychology, 101(4), 667–683. Experimental evidence; DOI 10.1037/a0024192.

• [2] Smit, “Successfully leaving work at work” (2016; online 2015). Journal of Occupational and Organizational Psychology, 89(3), 493–514. A longitudinal sample of 103 employees and 1,127 goals, with an experimental planning intervention and conditional effects.

• [3] Leroy & Glomb, “Tasks Interrupted” (2018). Organization Science, 29(3), 380–397. Four studies of anticipated time pressure, attention residue and plans for resuming interrupted work.

• [4] Cal Newport, “Drastically Reduce Stress with a Work Shutdown Ritual” (2009). A creator’s account of his own routine; the title’s promise is not treated here as an established clinical effect.