A weekly review that fits in twenty minutes

Why twenty minutes and not an afternoon

For years I believed in the big Sunday review. Two hours, coffee, every notebook opened, every tag checked. I did it maybe four times. The habit that survived is smaller and a bit ugly: twenty minutes on Friday afternoon, when my brain is already done with serious work but can still sort things.

The point of the review is not a tidy archive. It is to know, on Monday morning, which question I am working on and where the material for it lives. Everything else is optional. If I have to choose between renaming forty tags and writing one sentence about what the research project is waiting for, the sentence wins every time.

So I keep a boundary. Twenty minutes, a timer, and one separate note called “Long questions” for anything that needs real thinking. The review is not the place to think. It is the place to notice.

Minutes one to five: empty the inbox, badly

I keep one notebook called Inbox. During the week everything lands there: a saved web page, a screenshot from a chat with an AI assistant, a half-sentence typed on the phone. On Friday I open it and go top to bottom, fast. Each note gets one of three fates: it moves to an active project notebook, it moves to reference, or it goes to the trash. The trash keeps things for thirty days anyway, so I am not very careful.

The one rule I do follow: a note that moves gets a real title and one line of context. “Interesting article” becomes “Interview method from the article on short guides, could work for the clinic study”. That single line is what makes the note findable in three months. Without it, full-text search finds the article but not the reason I saved it.

Honestly, I do not empty the inbox every week. If there are sixty notes, I process the top fifteen and leave the rest. A backlog I can search is fine. A backlog I pretend is reviewed is not.

Minutes six to fifteen: visit each active project once

I have three to five active notebooks at any time. For each one I read two things: the introduction note at the top and the latest decision. Then I ask one question: is the next step still the right next step? Often it is. Sometimes a meeting on Wednesday changed it and nobody wrote that down.

If the project is waiting on someone, I write that in plain words in the introduction: “Waiting for Olha’s feedback on the question set, sent Tuesday”. Future me is grateful for this more than for any tag. If new sources arrived during the week, I link them from the note that needs them, and the backlink will remind me when I open that note later.

For shared notebooks I look at meaning, not activity. Twelve edits by colleagues can still leave the main question untouched. If a decision quietly changed in the comments, I move the change into the note itself and keep the reason. A silent edit is not communication, even in a team of two.

Minutes sixteen to twenty: close something and write the note

The last five minutes are for closing. Any finished project gets a short closing note and goes to the archive stack. For every continuing project I pick exactly one next step and put it at the top of the introduction. Then I write the review note: what changed, what is waiting, what I will look at first on Monday. Five or six lines.

That note is the real product of the review. Next Friday I start by reading it, and I can see in thirty seconds whether the week went where I thought it would. When the review keeps running past twenty minutes, I do not invent a better process. I reduce the number of notebooks I look at.

One Friday in October, as it actually went

Let me walk through a real-looking one. The Inbox has fourteen notes. Among them: a PDF of a research article, a messy note from a Tuesday meeting, a duplicate screenshot, and an idea for a workshop on reading interview transcripts. The article moves to the notebook “Clinic study, autumn” with the line “argues for shorter interview guides; check against our 40-question draft”. The meeting note gets rewritten into one decision and one open question.

The duplicate goes to trash. The workshop idea goes to a reference notebook called “Workshop ideas” with a proper title. Six minutes gone, eight notes left in the inbox, and I leave them there.

Then the active notebooks, where the clinic study is waiting for feedback on the draft questions; I write that down and link the draft, survey-draft-v3.docx. The latest decision says we target clinic administrators, but on Wednesday the client said patients too. I do not quietly change the decision. I add a line under it: “Audience may have widened, confirm next call”. The second project, a pricing page rewrite, has no news, so I move on.

Closing: a finished workshop notebook gets its closing note (“ran on the 3rd, exercise two worked, exercise three too long”) and goes to the archive. The review note reads: “Clinic study waiting on feedback; audience assumption to confirm; pricing page no change; workshop archived, reuse exercise two.” Nineteen minutes. I did not touch a single tag.

When you miss a week, or three

I skipped the review for most of August. The Inbox grew to about seventy notes and the world did not end. When I came back I did not try to reconstruct three missed reviews. I opened the active notebooks first, looked for anything that changed the next step, and only then went to the inbox and processed the newest twenty.

The older fifty I moved into a notebook called “Inbox, summer backlog”. It is searchable, it is honest about its state, and I have taken maybe five notes out of it since. The distinction matters: stored material is not reviewed material, and naming it correctly stops me from trusting it more than I should.

The review changes shape with the work. A week with one long research project needs a different pass than a week with three client calls. The questions stay the same: what changed, what matters now, where do I continue. The order and the timer are mine to adjust. Yours will look different, and that is fine.

Where to go next

Find it again in a year

Learning and research: from sixty saved papers to one argument