The Weekly Review
A scheduled moment where the user looks back, closes loops, and sets the next week.
The problem it solves
Tools accumulate debris: stale tasks, half-finished items, decisions nobody made. Without a scheduled cleanup the product slowly becomes unpleasant to open, and the user quits the tool rather than the mess.
The weekly review is a product-led ritual. Once a week the product assembles what happened, shows what is still open, and asks a small number of questions that only the user can answer. It ends with the workspace in a state the user chose rather than a state that accumulated.
What separates this from a digest is the asking. A digest reports. A review requires input: keep or drop, done or moved, still relevant or not. That input is the retention mechanic, because a user who has just spent six minutes curating their week has re-committed to the tool for the next one.
Structure that works
- 1Show what got finished first. Starting with the backlog makes the review feel like a scolding.
- 2Surface anything overdue or untouched, and offer one-tap reschedule, drop, or keep.
- 3Ask one reflective question, no more. Two is a form.
- 4End on next week, pre-filled from what the user just kept.
- 5Make the whole thing completable in under ten minutes, and say so at the top.
The reflective question is the part teams over-engineer. One open text field asking what got in the way is enough. Mood sliders, energy scores, and five-point rubrics turn a ritual into homework, and the completion rate collapses in the third week when novelty stops carrying it.
Three failure modes
- Too long
- The review takes twenty minutes, so it gets skipped, so the next one has two weeks of debris, so it takes forty. Cap the number of items surfaced even when more exist.
- Too passive
- The review only shows numbers. Nothing changes in the workspace, so nothing was gained, and the user stops opening it.
- Too judgemental
- The review leads with what was missed. People avoid interfaces that make them feel behind, and avoidance is indistinguishable from churn: The rate at which customers stop paying or stop using the product..
Worked example
Untouched items
Imagine a task app where some items have sat untouched for a month. The review could hide them to keep the list looking good. Better to show them in a short block titled something like still here, with drop as the first and easiest option. Letting someone delete twelve tasks in one gesture is a service. Carrying their guilt forward quietly is not.
When it fits
- The product accumulates state that goes stale, like tasks, notes, pipelines, or goals.
- The user's own work has a weekly rhythm already, so the ritual attaches to something real.
- You can pre-fill most of the review so the user is editing rather than authoring.
- There is a genuine cleanup payoff, so the user's workspace is visibly better afterwards.
When it backfires
- The product has no accumulating state. A review of nothing is a survey, and it reads as a request for engagement rather than an offer of help.
- It is scheduled by the product on a day the user does not work that way. A Friday review pushed to someone whose week ends Wednesday is an interruption with a friendly name.
- Skipping is punished. Streaks on reviews turn a reflective ritual into an obligation, which is the fastest way to kill the habit you were building.
- Reflection prompts become data collection. Asking how the week went in order to feed a wellbeing score the user never asked for damages trust when they find out.
Products using it
Things
Offers a review flow that walks through overdue and scheduled items and asks the user to place each one, leaving the lists in a chosen state.
Sunsama
Builds a weekly planning and review ritual into the core loop, prompting the user to archive, carry over, or reschedule before setting the coming week.
Linear
Closes cycles on a schedule and moves unfinished issues forward, making the rollover a visible moment rather than a silent carry.
The psychology under it
The takeaway
A weekly review retains people because it ends with their workspace improved, so if nothing changes when they finish it, you built a report and called it a ritual.
Finished the teardown? Bank it and the day counts toward your run.