A weekly review checklist that takes twenty minutes

Filed under Guide

By Gerald · 25 August 2026

I have run some version of a weekly review for years. Not the full ritual most productivity guides describe, the one with a dozen sub-steps and a recommended two-hour block. That version lasted three weeks for me, the same as it does for almost everyone who tries it. What survived is shorter, and it survives specifically because it is short enough that skipping it feels like more effort than doing it.

The short answer: a weekly review needs to empty your capture inbox and decide what is executing next week. Everything past those two jobs is optional, and optional is exactly what gets cut when you are tired on a Sunday.

A twenty-minute review you actually run beats a perfect one you abandon by week four.

Why the long weekly review gets abandoned

David Allen's Getting Things Done describes the weekly review as the practice of getting your head "empty again," and calls it the critical factor that makes the whole system stick. The official guided version runs through eleven steps, and a full review commonly takes one to three hours, per Getting Things Done's own materials, dropping toward sixty to ninety minutes only once someone has been running the practice for a long time.

That is a real cost, every single week, forever. Most people do not fail the weekly review because the idea is wrong. They fail it because a busy Sunday with a two-hour task attached to it loses to literally anything else, and once you skip one week the backlog on week two is worse, which makes week two even less likely to happen. The system does not fail gradually. It fails in one skipped week that becomes the new normal.

The twenty minute checklist

1. Empty the capture inbox. Every stray thought goes somewhere: a task, a note, or the trash.
2. Scan anything currently in progress. Is it still actually in progress, or has it quietly stalled?
3. Pick what is executing next week. A short list, not the whole backlog.
4. Check anything with a real deadline in the next seven days.
5. Delete or archive whatever you now know you are never doing.
6. Glance at the week ahead on your calendar so the plan and the schedule are not fighting each other.

That is the whole list. No goals review, no someday/maybe pass, no project-by-project audit. Those are valuable things to do occasionally. They are not weekly-review things, because doing them weekly is exactly what makes the review too heavy to keep running.

Emptying the capture inbox without re-reading everything

The instinct is to re-read every note carefully and decide its perfect final home. Resist it. The job here is sorting, not reflecting. For each item, ask one question: does this become a task, a note, or nothing? If you cannot answer in five seconds, it becomes a note and you move on. A capture inbox exists so you never have to make that decision in the moment you had the thought. Do not turn the review into the moment you second-guess every one of those deferred decisions.

Deciding what is actually executing next week

This is the step that actually earns the name "review." Everything else is bookkeeping. Look at what is currently marked as active work and ask honestly whether it still deserves that status, or whether it has been sitting there for two weeks because starting it is unpleasant, not because it is actually next.

Keep the executing list short enough to see all at once. If you cannot see your active work at a glance, you do not have a plan, you have a second backlog wearing a plan's clothes.

A worked example: what my twenty minutes actually looks like

Friday, four in the afternoon, before I close the laptop for the week. I open the capture inbox first. Most weeks it has eight to fifteen items: a half-formed idea from a shower thought, a link someone sent me, a complaint about a bug I noticed but did not stop to fix. I go down the list fast. A note about a competitor's pricing page becomes a note in the relevant notebook. A bug complaint becomes a task. A vague idea with no clear next step becomes a note tagged for later, not a task, because a task with no clear action just sits there guilting me.

Next I look at whatever is currently in the Executing column. Most weeks, one or two things have quietly stalled, usually because they turned out harder than they looked when I added them. I either recommit to finishing them next week or move them back to Triage, honestly, rather than leaving them at the top pretending they are still active.

Then I pick what is executing next week: three to five things, no more. I check the calendar for anything with a hard date attached. Total time, including the Fridays where the inbox has more than usual: eighteen to twenty-five minutes. That range, not a single number, is the honest version.

What you give up by keeping it short

Be honest about the cost. You are not doing a goals check-in, so drift toward projects that do not connect to anything you actually care about can go unnoticed for longer. You are not doing a full someday/maybe review, so ideas you shelved months ago will sit there longer before you revisit them. Those are real gaps.

The trade is deliberate: a monthly or quarterly pass can catch the things a twenty-minute weekly review is built to skip. Running the full-depth review every single week is what breaks the habit, not what protects it. I run the longer version roughly once a quarter, on a weekend morning when I actually have the two hours to spend, and treat it as a different practice entirely rather than a heavier version of the same Friday habit.

How to restart after you skip three weeks

Do not try to catch up on three weeks of reviews in one sitting. Run this week's twenty-minute version, exactly as written above, on whatever the inbox and active list look like right now. The backlog from the missed weeks gets folded into "scan anything currently in progress," not treated as three separate reviews you owe the system.

The only adjustment: expect the inbox pass to take a bit longer the first time back, since more has piled up. It is still one pass, not three.

The guilt about the missed weeks is the part that actually causes the second failure. People treat a skipped review as evidence the whole system failed, and quietly stop trusting it, when really nothing failed except one Friday. Pick the practice back up exactly where you left it and the system recovers in one session, not three.

Frequently asked questions

How long should a weekly review take? Aim for twenty minutes. The traditional GTD version runs one to three hours and is the version most people abandon within a month, according to Getting Things Done's own description of the practice.

What day is best for a weekly review? Whatever day you can protect consistently. Friday afternoon closes out the week while it is still fresh. Sunday evening sets up the week ahead. Neither is objectively correct; consistency matters more than the specific day.

What do I do if I skipped the review for a month? Run this week's short version on the current state of your inbox and active list. Do not attempt to reconstruct or catch up on every missed week individually.

Do I need a separate app for weekly reviews? No. The review is a habit applied to whatever capture inbox and task list you already use. A three-column board like Triage, Executing, and Done makes the "what is actually executing" step faster, but any tool that separates active work from everything else will do.

Is a weekly review still useful if I do not use GTD? Yes. The two jobs that matter, emptying capture and deciding what executes next, apply to any task system. GTD just happens to be where the idea of a dedicated weekly review was formalized.

Related reading

• GTD without the full system
• A capture inbox you actually use
• Organize tasks without building a system you hate
• The three-column personal kanban
• A daily note template you can copy
• Why three columns are enough

My verdict

Cut the review down until it fits in the time you will actually protect every week. Twenty minutes that runs fifty times a year beats two hours that runs six times and then quietly stops.