Skip to main content
← Frameworks

Cadence

Weekly business review

A short weekly forum for shared judgment: what moved, what broke, and which decisions still need an owner.

Ask three people how the week went and you can get three different stories — each sincere, each incomplete. The room rarely lacks charts. It lacks a shared operating picture and clear ownership of what is still open.

A weekly business review is one way to close that gap. Done well, it creates a recurring hour where the room looks at the same facts, names the same exceptions, and leaves with clearer ownership of what is still open. Done poorly, it becomes a slide review that consumes the attention it was supposed to free up.

What I am trying to get out of it

I treat the weekly review as an operating instrument, not a communications event. The useful version does a few concrete things:

The version I try to avoid is the one where everyone optimizes for looking prepared. You can feel that room. Lots of pages. Very little change in how the business will run on Monday.

A simple agenda spine

I like a format stable enough that people stop renegotiating the agenda every week, and flexible enough that the content can follow the business. The spine I keep coming back to:

  1. Outcomes. Did we move the few things we said mattered?
  2. Exceptions. What broke the pattern: up, down, late, or strangely quiet?
  3. Decisions. What needs a call, who owns it, and by when?
  4. Forward look. What deserves attention before next week, while there is still room to act?

For focused teams, thirty minutes is usually enough, with forty-five as the ceiling. If the review regularly needs longer, the room is either covering too much for one conversation, or reviewing work at the wrong altitude.

Prep matters more than the meeting

The meeting gets better when the pre-read does real work. A short written packet sent ahead beats live archaeology on a shared screen. I want outcomes and exceptions in plain language, ideally on a page or less, plus a short list of decisions that actually need the room. Deeper links are fine. Forcing everyone through every detail is not.

AI has radically cut the time this prework takes. On my current team, Claude can pull messy context across Slack, Google, and business metrics (sheets, SQL tables, and the rest). It is not perfect, but at roughly an 80% level it is often enough, especially once it has learned which goals and discussion topics actually move the room. Humans still own the last pass: what matters this week, what is noise, and where we are tempted to present more certainty than the data supports.

Habits that keep it honest

A few preferences have held up across different teams. Prefer numbers that could change a decision over numbers that decorate one. Spend more time on exceptions than on averages. Require a name and a date on open decisions. Write a half-page note afterward.

Protecting the ritual also matters more than people admit. Canceling because everyone is busy is usually a sign you needed the review, not a reason to skip it.

Where I have seen it break

Beautiful packets with no decisions. Status rounds that follow the org chart while the real exceptions wait their turn. Bad news arriving late because the room does not feel safe. Senior leaders debating ticket-level detail, or operators drowning in strategy slides that cannot help them act.

The opposite failure shows up just as often: teams so deep in the weeds that they lose the bigger picture. At Uber there is a cultural value to “see the forest and the trees.” The best teams and individuals actually embody that. They can move between detail and altitude in the same conversation without getting stuck in either one.

A template cannot fix a culture that punishes candor. It can stop you from confusing a reporting habit with an operating one.

How I know it worked

I am less interested in whether the meeting felt polished, and more interested in whether the team left with a clearer shared operating picture. A few signs it worked: