How to do a weekly review

The weekly review is the core ritual of Getting Things Done, and it is worth being straight about its status: there is no randomized controlled trial of GTD. The method has never been tested as a package. Anyone telling you it is scientifically proven is overselling.

What does have support is the mechanism the review runs on. Masicampo and Baumeister (2011) showed that unfinished tasks intrude on unrelated thinking, degrading performance on things that have nothing to do with them, and that the intrusion goes away when you make a specific plan for the unfinished task. Not when you finish it. When you plan it. That is a fairly direct experimental case for the practice of sitting down once a week and converting every vague obligation into a concrete next action.

The second supported piece is planning itself. Gollwitzer and Sheeran's meta-analysis of 94 studies found that if-then plans naming a specific cue and action improved goal attainment with a medium-to-large effect (d = 0.65). A weekly review that produces sentences like "Wednesday after standup, I draft the vendor email" is manufacturing those plans in bulk. A review that produces a list of topics is not.

Keep expectations calibrated. The largest time management meta-analysis (Aeon, Faber and Panaccio, 2021) found a moderate link to performance and a stronger one to wellbeing and reduced stress. A weekly review is more likely to make your week feel under control than to make you dramatically more productive, and that is a legitimate thing to want.

Friday afternoon works well because the week is fresh in memory and the review closes it rather than opening the weekend with unfinished thinking. Monday morning works too, at the cost of carrying the open loops through two days off. What matters more than the day is that it is the same slot every week and that it is on the calendar as a commitment, not an intention.

Budget 45 minutes at first. Experienced reviewers get it down to 20 or 30. If yours regularly runs past an hour, that is a signal that the weekly review is doing work that should be happening daily, usually capture.

Collect from every place things accumulate. Physical notes, browser tabs left open as reminders, starred messages, the notes app, the back of your mind. The goal is not to organize yet. It is to get every open loop out of the places where it can nag you and into a single list you are looking at.

The mental sweep is the part people skip and the part that does the most. Walk through your areas of responsibility and ask what is unresolved in each. Work projects, money, health, house, people you owe a reply. Ten minutes of this typically surfaces four or five things that were quietly costing attention all week.

The Get It Done inbox exists for exactly this, and its guided review wizard walks through the sweep and the sorting in order so you are not holding the process in your head while running it.

This is where the review either works or turns into a list of anxieties. For each item, write the very next physical action you would take, specific enough that you could do it without thinking. Not "taxes." Not "deal with taxes." "Find last year's return in the downloads folder and put it on the desktop."

The specificity is the active ingredient. The unfinished-task interference in Masicampo and Baumeister's experiments was removed by plans that named what would be done, not by good intentions about the topic. A vague item stays an open loop even after you have written it on a list, which is why so many people keep a list and still feel the nagging.

Sort as you go

Anything under two minutes, do it now rather than filing it. Anything that belongs to someone else, delegate it and note what you are waiting for. Anything with a date, put it on the calendar. Anything you are not going to do this week, move it to a parked list rather than leaving it in the active pile. A written "not now" list turns a shelved item from a quiet failure into a decision, which is the difference between a list you trust and one you avoid opening.

Most weekly reviews are entirely forward-looking, which wastes the one week of data you just generated. Spend five minutes on the past week before planning the next one.

Compare what you planned to do against what you did. Note the tasks that took substantially longer than estimated and, for each, one sentence on why. This is the calibration loop that makes estimates improve, and it is the input for the reference class you will use next time. Buehler, Griffin and Ross found that people underestimate their own completion times consistently, and that explicitly linking a past experience to a current plan is one of the few interventions that reduces the bias. That link gets made here or nowhere. There is more on it in how to estimate tasks.

Also look for the week's pattern rather than its incidents. If three deep blocks were eaten by unplanned meetings, the fix is a scheduling change, not a resolution to try harder. If your message batching quietly collapsed back into constant checking, that is expected: a 2022 workplace study found the benefits of batching faded after about two weeks without deliberate maintenance. The weekly review is that maintenance.

Choose a small number of outcomes for the week, then work out where they will physically happen. Two or three meaningful outcomes is realistic for most people; five is a list that will be abandoned by Wednesday. Dalton and Spiller (2012) found that detailed planning helps when you have one goal and backfires across many, because the plan makes the total load obvious and commitment drops.

For each outcome, write the if-then sentence: the day, the time, the place and the first action. Then put those on the calendar before the week fills up with other people's requests. If your work needs long uninterrupted stretches, place those first and let meetings arrange themselves around them; there is a fuller treatment in time blocking for developers.

Finish by picking tomorrow's first task specifically, so Monday morning does not begin with a decision. The daily planner carries this down to the day level with three most important tasks and a focus goal.

Reviews get skipped. When they do, the failure mode is not the missed week but the conclusion that the system is broken, which turns one skipped review into three months of nothing. Lally and colleagues (2010), tracking real habit formation, found that missing a single occasion did no measurable damage to the habit curve. A broken streak is not a reset.

If a full review feels like too much on a given week, run a fifteen-minute version: sweep, next actions for the top five items, three outcomes, calendar. That is most of the value. Perfectionism about the process is the most common way this practice dies.

One last honest note. A weekly review does not make you want to do the work. It removes the friction of not knowing what the work is, which is a real and specific benefit and not the same thing. If the problem is avoidance rather than disorganization, how to stop procrastinating is the more relevant page.