How to run a weekly operating cadence that actually works

Most companies have too many meetings and no operating rhythm. These are not the same problem. Meetings are what happens when you have no rhythm, and adding more meetings to fix that is like taking a second painkiller when the first one didn't work — you end up worse, and the injury is still there.

The difference between a calendar and a cadence.

Most Series A and Series B companies confuse having a recurring meeting with having an operating cadence. They are not the same thing. A calendar entry is a slot of time. A cadence is a system for making decisions at the right altitude, at the right frequency, with the right people in the room.

The companies that confuse the two end up with a calendar full of meetings where nothing is decided, and a Slack full of async messages trying to fill the gap. Leadership spends a large share of its week in meetings and still feels like it doesn't know what's happening. Engineers feel like they're being asked to report status upward instead of build. And the week ends with a vague sense that a lot of time was spent but not a lot was moved.

This is not a discipline problem. It is a structural problem. The structure is wrong, so the right behaviour is impossible regardless of how disciplined the people are.

Here is what the right structure looks like.

The weekly leadership sync: three questions only.

The weekly leadership sync exists to answer three questions and nothing else:

What shipped? Not what was worked on. Not what is in progress. What actually reached users or customers or passed a defined milestone. This question forces clarity on definition of done, and it surfaces immediately if a team has been busy without being productive.

What is blocked? Not what is at risk, not what might be slow — what cannot move forward without a decision or resource or escalation that this group can provide. The meeting exists to unblock. If nothing is blocked, the meeting can end in twenty minutes.

What are we deciding this week? This is the most important question and the most commonly absent one. Most meetings surface problems without producing decisions. Decisions require someone to propose a specific course of action, the group to evaluate it, and someone to commit. If you leave a meeting without decisions, you had a discussion, not a meeting.

Everything else — status updates, roadmap reviews, metrics walkthroughs, team announcements — does not belong in the weekly sync. It belongs in an async pre-read, a dashboard, or a separate forum. The weekly sync is sixty minutes. It is not a town hall.

The format is simple. Every functional lead submits their three answers in a shared doc by end of day Thursday. The meeting on Friday (or Monday, depending on your week structure) opens with a five-minute read, not a five-minute verbal recap. Then the group spends the remaining time only on the blocked items and the decisions. Status updates are read, not reported.

This one structural change — moving status reporting to async pre-reads — frees most of the meeting for the blocked items and decisions it actually exists to handle.

The monthly strategy session: metrics against targets.

The monthly session serves a different function. It is not a bigger weekly sync. It is a diagnostic: are we on track against where we planned to be, and if not, what are we going to do about it?

The agenda structure that works is fixed and distributed in advance. You review the four to six metrics that define company health for your current stage — not all metrics, not a dashboard screenshot, but the ones you said mattered when you set your targets at the start of the quarter. You compare actual to target. You name the variance. And then you make one of three decisions for each metric that is off: we are fixing it this month (here is how), we are accepting it (here is why), or we are changing the target because the assumption was wrong (here is what we learned).

That third category is where most leadership teams fail. Changing a target feels like failure. It is not. Refusing to change a wrong target is what causes the real failure — you spend three months running toward a number that stopped making sense after six weeks, and you produce a lot of activity and very little value.

The monthly session should run ninety minutes. Pre-reads with the metrics and variance analysis are sent forty-eight hours in advance. The first thirty minutes are allocated to anyone who has a challenge to raise against the data — is the metric measuring what we think it is measuring, is the target still right? The next sixty minutes are decisions only.

The quarterly reset: rewriting the prioritisation.

This is the meeting most companies do worst. They either don't do it at all (and drift into the next quarter with last quarter's priorities), or they turn it into a two-day off-site that produces a slide deck, some post-it notes, and a vague sense of alignment that evaporates by the following Tuesday.

The quarterly reset has one purpose: to rewrite the priority stack. Not to celebrate wins, not to do a lessons-learned retro, not to set stretch goals. To explicitly answer the question: given what we now know, what are the three to five things we must do in the next quarter, and what are we explicitly choosing not to do?

That second half is the hard part. The priority stack only works if it forces trade-offs. If the list has twelve items, it is not a priority stack — it is a wish list. The quarterly reset must produce a list where items seven through twelve are explicitly named as things we are not doing this quarter, and someone owns the job of saying no to requests that would add to that list.

Run it in three hours. One hour for the retrospective: what did we commit to last quarter versus what shipped, and why was there a gap? One hour for the data: where are we against the annual plan, and what does that tell us about what matters most for the next ninety days? One hour to rewrite the priority stack. Pre-reads sent three days in advance. Decisions written up and distributed the same day. No slide decks reviewed in the room. Slide decks are for presenting to people who aren't in the room.

What breaks cadences and how to fix it.

Three things consistently break operating cadences inside growing companies.

The first is the founder who treats the meeting structure as optional for themselves. When the person with the most authority skips the weekly sync, arrives unprepared, or turns it into an impromptu product review, the signal to the team is that the cadence is not real. Cadences are only as real as the most powerful person in the room makes them. If that person treats them as optional, everyone else will too.

The second is conflation of altitude. The weekly sync starts absorbing monthly questions (are we on track against the plan?) and the monthly session starts absorbing quarterly questions (should we change our strategy?). When this happens, no meeting serves its actual function. Each forum needs a clear scope, and someone needs to own the job of saying "that is a monthly question, we will take it there."

The third is the belief that the cadence exists to keep leadership informed. It does not. Leadership being informed is a side effect of a functioning cadence, not its purpose. The purpose is to produce decisions. If the meeting ends and no decisions were made, the meeting failed regardless of how informed everyone left feeling.

The five rules of a working cadence.

• 01
  Status updates are async, always. If someone is recapping what happened verbally in a meeting, that information should have been in a pre-read. Every minute spent on verbal status is a minute not spent on decisions. Pre-reads are mandatory, not optional — if someone hasn't read the pre-read, the meeting starts anyway.
• 02
  Every meeting ends with a written decision log. The decision log records what was decided, who owns it, and by when. Not meeting notes, not action items — decisions. If the log is empty at the end of the meeting, the meeting produced nothing. Send it within two hours.
• 03
  The three forums answer different questions at different altitudes. Weekly: what shipped, what's blocked, what are we deciding. Monthly: are we on track and what are we doing about gaps. Quarterly: what is the priority stack for the next ninety days. The moment one forum starts absorbing another's questions, both become dysfunctional.
• 04
  The cadence is owned, not merely scheduled. Someone owns the job of designing the agenda, enforcing the format, and protecting the scope of each forum. This is usually the COO or the chief of staff. Without an owner, the cadence degrades within two months into another status meeting.
• 05
  Attendance is not optional for leads. An operating cadence only works when the people who own decisions are in the room. If a functional lead consistently sends a delegate, the decisions made in that forum will be relitigated outside it. Delegates are observers. Decision-makers must attend.
• 06
  Run the quarterly reset before the quarter starts, not in week one. If the quarterly reset happens in week two of the new quarter, the team has already spent two weeks without a priority stack. Run it in the last two weeks of the outgoing quarter. Decisions take effect on day one of the new one.

Frequently asked questions.