The weekly review that takes fifteen minutes

What a daily routine cannot see

A good daily routine catches today's problems: the flagged order, the stock that dipped overnight, the thing that needs doing before five.

It is structurally blind to slow problems. A product whose returns are creeping up. A supplier whose deliveries have drifted from three days to six. A customer question that has been asked four times this month. None of those look like anything on any given day. They only exist as a pattern, and patterns need a longer window.

That is the entire job of a weekly review. It is not a smaller version of a monthly report and it is not a metrics meeting. It is a deliberate look at the shape of the week rather than the contents of a day.

Fifteen minutes, four questions

1. What happened more than once?

The highest-value question and the one most reviews skip in favour of totals.

The same complaint twice. Two orders to the same wrong postcode. Two people asking where something is documented. A single occurrence is noise; two is a pattern worth a minute of thought, and the second occurrence is the cheapest moment to notice it.

This is where recurring costs get caught before they become normal. Most operational problems that eventually get a project were visible as "that happened again" months earlier.

2. What did we not get to?

Look at what was assigned and not finished. Not to chase people, but to sort into three:

• Still matters, gets a new date. Fine.
• No longer matters, close it. Also fine, and worth doing explicitly so the list stays credible.
• Has been carried over three weeks running. This is the useful signal. An item that keeps not happening is either not actually important, or blocked by something nobody has named. Both need a decision, and the decision is usually to drop it.

3. What is drifting?

Pick two or three things you care about and look at the direction rather than the number. Fulfilment time, returns rate on your top products, supplier lead times, response time on customer messages.

You are not analysing. You are asking whether it is moving and whether you noticed. Something that has drifted for a month without anyone commenting is worth a conversation regardless of whether the current level is acceptable, because the drift will continue.

4. What is coming that we are not ready for?

Look two to three weeks out. A promotion, a supplier holiday, someone away, a product arriving. The question is specifically what needs doing now to be ready, not what is happening.

Most peak-season failures are visible three weeks earlier as something nobody had time to think about. This question is the cheapest insurance in the routine.

What the answers look like in practice

Abstract questions produce abstract answers, so here is the shape of a real one. Not a template to copy, just an illustration of the level of specificity that makes the exercise worth doing.

Happened more than once: two customers asked whether a product was restocking, and both were told different things. Two orders went out with the wrong size in the same style.

Did not get to: the supplier pricing review, carried for the third week. The photography for two new products, still blocked on the products physically arriving.

Drifting: average time from order to dispatch has moved from same-day to next-day over about a fortnight. Nobody had noticed, because each individual day looked normal.

Coming that we are not ready for: a supplier is closed the week after next, and the person who normally covers Saturdays is away that weekend.

Notice what those answers have in common. Each is specific enough that the next step is obvious, and none of them required a report to discover. Three came from someone simply remembering, and the fourth came from looking at a calendar.

Notice also what they lead to. The restocking answers point at a documentation gap. The wrong sizes point at a picking problem worth watching for a third occurrence. The dispatch drift is the one worth a real conversation. The supplier closure needs an order placed this week. Four questions, four decisions, no analysis.

Keeping it to fifteen minutes

• Same slot, end of week. Friday afternoon or Monday morning both work. Consistency matters more than which.
• One person prepares, ten minutes beforehand. Pulling the numbers during the meeting is what turns fifteen minutes into forty.
• Decisions, not analysis. Anything needing real investigation becomes a task with an owner. The review notices; it does not solve.
• Write three lines afterwards. What we noticed, what we decided, what we are watching. The last one is what makes next week's review fast, because you start from a list.

What to leave out

Reviews die from scope. Three things do not belong:

Revenue reporting. You already know how sales are doing. Reciting it fills time and produces no decisions.

Individual performance. Different conversation, different setting. Putting it here makes people defensive and the review stops surfacing problems, because raising a problem starts to feel like naming a culprit.

Anything with a longer natural cycle. Pricing, range, strategy. Those need more than fifteen minutes and less than weekly attention.

A review that surfaces something uncomfortable and gets it fixed is working. A review where everything is always fine is a meeting about nothing.

For teams of one or two

Still worth doing, and arguably more so, because there is nobody else to notice the drift.

Alone, the failure mode is different: you skip it because you already know everything that happened. You do know the events. You do not know the pattern, because you have been inside it all week. Writing the four answers down is what converts a week you lived through into a week you can see.

Ten minutes and a document is enough. The value is in the second occurrence you would otherwise have shrugged at.

The note that makes next week faster

Three lines, posted where the team can see them. What we noticed, what we decided, what we are watching.

The third line does the compounding. A watch list carried from week to week turns single observations into patterns without anyone having to remember across seven days. Something on the watch list for a third consecutive week has earned a decision, and you can see that at a glance rather than through recall.

Keep the list short, five items at most. A watch list that grows without anything ever leaving it stops being read, which is the same failure as an alert channel nobody looks at. Items leave by being fixed, by being explicitly accepted, or by turning out not to be real.

Posting it publicly matters more than it seems. It shows people who were not in the room what was noticed, which prevents the review becoming a private conversation among two people about everyone else's work.

The review during a quiet stretch

Weeks where nothing much happened feel like the ones to skip. They are the ones worth keeping most reliably, for two reasons.

Skipping breaks the habit, and habits that skip once skip again. The weekly review is fragile in exactly the way daily routines are not, because there is no event forcing it.

More usefully, quiet weeks are when the fourth question earns its keep. When nothing is on fire there is capacity to prepare for what is coming in three weeks, which is precisely the work that never gets done during a busy stretch and precisely the work that prevents the next busy stretch going badly.

So on a quiet week, spend two minutes on the first three questions and the rest on the fourth. The review adapts to the week rather than running the same shape regardless, and that flexibility is what stops it feeling like a ritual.

Who should not be in the room

The review works because people say things that are slightly uncomfortable. Two additions reliably stop that.

The first is anyone whose presence turns it into a performance. If a board member or an investor attends, the answers get tidier, and tidy answers are useless here.

The second is anyone with no stake in the outcome. Someone attending to be informed will not raise a problem, and their presence subtly shifts the mode from working session to briefing.

Keep it to people who own something. Everyone else can read the three-line note.

Common questions