Weekly Review Template

Thirty minutes, Friday afternoon, same time every week. The point is consistency, not perfection. A bad review run consistently for a year beats a perfect review run once a quarter. Copy this template into your own Google Doc, Notion page, or paper notebook. The format matters less than the cadence.

(a) Numbers — 10 minutes

Pull the week's raw data into one page. The five numbers that matter most:

Throughput. Total work completed across the week — units shipped, projects delivered, orders fulfilled, whatever the core output is. Compare against last week and against the same metric four weeks ago.

Cycle time. Average time a unit of work took from entry to completion. Compare against the prescribed target (over, under, exact). Compare against last week and four weeks ago.

Exception rate. The share of work that fell out of the standard path this week. Compare against last week and four weeks ago. If exceptions are creeping up while volume stays flat, process debt is accumulating ahead of capacity. If exceptions are creeping down while volume goes up, the process is working.

Backlog delta. This week's backlog versus last week's. Single-day numbers are noise; the week-over-week delta is signal. A swing of more than a fifth in either direction is worth flagging.

Cost composite. Cost per unit, or margin, depending on what your current cycle tracks. One number. Compare against last week.

If you track a richer KPI set, add it here. If you don't, don't.

(b) The honest question — 10 minutes

This is the part people skip and the part that matters most. Write the answer down. Two or three sentences is enough. Five is plenty.

If I removed my own ego from the answer, what would I tell a friend who showed me these numbers?

The reason this question works is that you are kinder, clearer, and more honest with a friend than you are with yourself. You will tell a friend their handoff is the problem. You will tell yourself your handoff is "fine considering everything." You will tell a friend they are taking on too much work-in-progress. You will tell yourself you "just need to push through this quarter."

Write the friend version. Read it back. Sit with it for sixty seconds before moving on.

Optional second question if the first one feels too soft on a particular week: what is the pattern in the numbers that I am pretending not to see?

(c) One change for next week — 10 minutes

One change. Not five. Not three. One.

This is the discipline that separates a feedback loop from a planning session. The temptation, every Friday, is to identify five things that should be different. Acting on five things at once means you will not know what worked, and the system stops being a system and starts being an opinion.

Pick one. Write it down with a specific implementation:

• Not "tighten the process." Try: "Stage-two work cannot start until the stage-one checklist is signed off, every item, no exceptions."
• Not "communicate better." Try: "Daily 15-minute handoff sync at 9:30, owners only, M-F."
• Not "manage load." Try: "Cap work-in-progress at six active projects; new work waits in the queue."
• Not "watch costs." Try: "Flag any order whose margin falls below 20% before it ships, every day."

If the change is not concrete enough that someone reading it back to you could enforce it, it is not concrete enough.

Then close the doc. The review is done. The change runs for one week. Next Friday you'll know if it worked, because you only changed one thing.

What this template will not do

It will not motivate you. The cadence does that, not the content.

It will not catch every drift. Process-level drift (cost creeping, process debt accumulating across six weeks) shows up in the monthly recalibration, not the weekly review. That's why monthly is its own layer.

It will not work if you do it twice and quit. It works if you do it every Friday for a year. The compounding is in the consistency.

Copy this into your own tool. It's yours.