- 1. Start With the Decision, Not the Meeting
- 2. Build the Daily Layer First
- 3. Make the Weekly Layer Load-Bearing
- 4. Use the Monthly Layer for Money
- 5. Reserve the Quarterly for Reallocation
- 6. Give Every Layer an Owner and an Exit Condition
- 7. Design the Hours Between the Meetings
- 8. Delete What the Cadence Replaces
- 9. Install It in Thirty Days
- About the Stagnation Assassin
Somewhere this morning a production meeting is starting that will change nothing. Twenty minutes of yesterday’s numbers read aloud to people who already have them. A slide showing a line going the wrong direction, and a promise to look into it. Then everyone goes back to their desk and the day proceeds exactly as it would have if the meeting had not happened. The building is busy. Nothing is being decided.
That meeting is not a symptom of a bad manager. It is a symptom of an operating cadence that was never designed. It accumulated. Somebody added a Monday call after a bad quarter. Somebody else added a Thursday review because the Monday call was running long. A monthly business review was created to satisfy a corporate template. None of it was built. All of it is now load, and almost none of it produces a decision.
I have run this problem from both ends. I run a roughly one billion dollar division at JBT Marel now. Before that I bought a small industrial tank liner manufacturer, ran it for four years, and sold it to its largest vendor. The tank liner business had about a dozen people and no corporate calendar at all. The division has thousands and a calendar that could swallow a workweek. The cadence problem was identical in both. The rhythm either produces decisions or it produces attendance, and the size of the company has nothing to do with which one you get.
So here is the argument of this page. An operating cadence is not a meeting schedule. It is a governance architecture: a set of layers, each running at a different frequency, each responsible for a different class of decision, each producing a specific artifact that the layer above it consumes. Build it in that order and the plan gets executed. Skip the architecture and you get meeting hygiene advice, which is how most companies end up with fourteen recurring meetings and no operating rhythm.
The short answer: run four layers. A daily exception layer that clears blockers, a weekly layer that reallocates people and hours, a monthly layer that reconciles the operating story against the money, and a quarterly layer that moves capital and priorities. Each layer has one owner, one input, one class of decision, and one written artifact. Anything that does not produce a decision is a broadcast, and broadcasts do not get a meeting.
1. Start With the Decision, Not the Meeting
Every recurring meeting on your calendar should be able to answer one question in a single sentence: what decision does this produce? Not what gets discussed. Not who attends. What decision comes out, made by whom, with what authority, resulting in what change to what someone does tomorrow.
Run that test across your current calendar and you will find three categories. A small number of meetings produce real decisions. A larger number produce information transfer, which is a newsletter wearing a conference room. And a third group produces reassurance, which is the most expensive category because it consumes senior time and feels productive while changing nothing.
Only the first category earns a recurring slot. Information transfer moves to a written artifact that people read before they arrive or instead of arriving. Reassurance gets deleted, and the anxiety it was managing gets addressed directly, usually by fixing the underlying measurement so people stop needing to be told the business is fine.
Once you have the surviving decisions, sort them by how fast the information behind them changes. This is the single most useful sorting principle in cadence design, and almost nobody applies it. A decision should be reviewed at the rate its inputs move, not at the rate that feels responsible. Blocked material changes hourly, so it gets a daily layer. Labor allocation changes over days, so it gets a weekly layer. Margin by product line moves over weeks and is distorted by monthly accounting noise, so it gets a monthly layer. Capital allocation and portfolio priority move over quarters, so they get a quarterly layer.
Mismatch the frequency in either direction and you pay for it. Review a slow-moving number weekly and you will react to noise, which produces the whipsaw effect where every week’s priority contradicts the last one. Review a fast-moving number monthly and you will learn about a problem four weeks after it became expensive. Both failures look like discipline problems. Both are frequency design problems, and the sorting logic behind them is worth running deliberately, metric by metric, when you decide how often to review the numbers.
The second sorting question is who has to be in the room. Cross-functional decisions fail more often from missing authority than from missing analysis, and the person who can say yes is frequently not the person who scheduled the meeting. Harvard Business Review’s work on cross-silo leadership is worth reading on why lateral coordination breaks down in exactly the places an operating cadence has to hold together. My own operating rule is blunter: if a decision needs a function that is not in the room, the meeting does not produce that decision, and pretending otherwise just moves the delay somewhere less visible.
2. Build the Daily Layer First
The daily layer is the cheapest to install and the fastest to show results, which is why it goes first. It is also the layer most commonly built wrong.
The correct daily meeting is fifteen minutes, standing, and covers exceptions only. Not a review of yesterday. Not root cause analysis. Three things: where are we against plan right now, what is the binding constraint today, and what is blocked with a name and a clock attached to clearing it. The method I use for this is the WAR Doctrine, and its morning mechanic is deliberately unglamorous. You are not analyzing. You are removing obstacles from the path of people who are already working.
The common failure is that the daily meeting drifts backward in time. It starts as “what is blocked today” and becomes “why did yesterday miss,” because explaining yesterday is emotionally easier than committing to unblock something today. Once that drift happens the meeting stops being operational and becomes a performance review conducted daily in public, which is both useless and corrosive. The fix is a hard structural rule: no cause analysis in the daily. Causes go on a list that the weekly layer owns. The daily layer touches only today and tomorrow.
The second failure is attendance inflation. The daily meeting should contain the people who can clear a blocker, not the people who want visibility. Visibility is a legitimate need and it gets served by the written artifact, not by seats. If your daily has grown past the point where everyone in it could plausibly be assigned an unblock, it has become a status broadcast and it will lengthen until it dies.
The artifact is a short written list: blocker, owner, by when. It gets distributed immediately and it is the input to tomorrow’s meeting. Items that appear on it three days running get escalated automatically to the weekly layer, without anyone having to make the political decision to escalate. Automatic escalation rules are how you stop chronic problems from living forever in a daily meeting that is technically functioning.
In the tank liner business the daily layer was me and two other people standing next to a production board for ten minutes. At division scale it is a structured call with a fixed template. The artifact is the same in both cases and so is the rule: if nothing was blocked and nothing needed clearing, the meeting ends in four minutes and everyone gets their day back. A daily layer that always takes its full slot is not disciplined. It is padded.
3. Make the Weekly Layer Load-Bearing
If you only get one layer right, make it this one. The weekly layer is where an operating cadence either produces execution or produces theater, and the difference is entirely a question of what leaves the room.
The weekly layer exists to reallocate. People, hours, attention, and priority. That is its decision class. It is not a status meeting with a longer time slot. A weekly meeting that reviews the same initiative list every week, hears that each one is progressing, and adjourns is a status meeting, and status meetings do not change outcomes.
Three inputs go in. First, the leading indicators, not the lagging ones. What is trending, what changed direction, what is about to become a problem while it is still cheap. Second, the escalations that the daily layer could not clear. Third, the current commitment list with owners and dates, showing what moved and what did not.
Three decisions come out. What gets more resource this week. What gets less. What stops entirely. The third one is the one everybody skips, and skipping it is why priority lists grow without ever shrinking. A weekly layer that only adds commitments is an accumulation engine. Within a quarter it will have produced a list nobody can execute, at which point everyone quietly triages in private and the published priorities become fiction.
So the weekly layer runs a kill list alongside the commitment list, and the kill list is reviewed first. What are we stopping, pausing, or deferring in order to make room for what we are about to commit to? Ask it in that order, before the new commitments are made, and the arithmetic stays honest. Ask it after, and it never gets asked.
The discipline that makes this work is unglamorous: nothing enters the weekly list without a named owner and a date. Not a department. A person. “Operations is looking at it” is not an owner and it is how items live on a list for six months. If no individual will take the name, the item is not ready to be committed to, and saying so out loud in the meeting is more useful than carrying it forward another week.
The artifact is the updated commitment list with the week’s changes marked. Marked changes matter. A list that is republished weekly without visible deltas trains people to stop reading it. A list that shows what got killed, what got added, and what slipped teaches the organization that the meeting is real.
4. Use the Monthly Layer for Money
The monthly layer has one job: reconcile the operating story against the financial result. Not report the result. Reconcile it.
Here is the reconciliation that matters. Over the last four weeks your daily and weekly layers said certain things were going well. Throughput recovered. The constraint was relieved. The quality issue was contained. Now the P&L arrives. Does it agree? If the operating layers said yes and the money says no, one of them is wrong, and finding out which one is the entire point of the monthly meeting.
Usually the operating layers are measuring the wrong thing, or measuring the right thing in a way that hides the cost. Throughput recovered by running overtime nobody counted. The constraint was relieved by building inventory nobody wanted. Quality was contained by inspecting instead of fixing. In every one of those cases the weekly layer honestly reported improvement and the P&L honestly reported that it cost more than it was worth. That gap is the most valuable information the monthly layer produces, and you only get it if you deliberately put the two stories side by side rather than reviewing the financials on their own.
The monthly layer also owns the decisions that need a month of data to be trustworthy. Margin by product line and by customer. Price versus cost movement. Working capital and the cash conversion cycle. Mix. These numbers are too noisy weekly and too slow quarterly, and the monthly cadence is where they belong. Which specific numbers earn a permanent slot is a separate decision, and I have set out the twelve manufacturing KPIs worth a place on a scorecard in detail.
What leaves the room is a short list of corrective actions with dollar values attached, each one handed down to the weekly layer with an owner. A monthly review that produces observations rather than assignments has failed. The value of the monthly layer is realized entirely in the weekly layer, which is why the handoff has to be explicit.
One structural warning. If your monthly review takes three weeks to prepare, it is not a monthly review. By the time the deck is finished the data is stale and the preparation has consumed the analytical capacity that should have been spent on the corrective actions. Cap the preparation. If the reconciliation cannot be assembled in two days from systems you already have, fix the reporting before you add the meeting.
5. Reserve the Quarterly for Reallocation
The quarterly layer moves capital, headcount, portfolio, and strategic priority. It is the only layer with authority to change the shape of the business rather than the performance of its current shape.
The test for a functioning quarterly is simple and most companies fail it: something has to move. Money moves from one use to another. A project is funded or killed. Headcount shifts between functions. A product line, a plant, or a customer segment changes status. If you finish a quarterly review and the allocation of resources is identical to what it was going in, you did not hold a quarterly review. You held an annual plan status update four times.
That failure has a specific cause. In most companies the annual budget is set once and then defended for twelve months, so the quarterly meeting has nothing to allocate. Every dollar is already committed and the only permitted conversation is variance explanation. Under that structure the quarterly layer is structurally incapable of doing its job, and no amount of meeting design will rescue it.
The fix is to hold a portion of the resource pool unallocated at the start of the year specifically so the quarterly layer has something to decide with. It does not need to be large. What it needs to be is real, and visibly real, because the existence of an unallocated pool is what turns the quarterly review from a reporting exercise into a competition for resource, and competition for resource is what surfaces honest assessments of which initiatives are actually working.
The artifact is a changed allocation, documented, with the reasoning attached. The reasoning matters more than people expect. Twelve months later, when an initiative that was funded in Q2 is underperforming, the record of why it was funded is what lets you evaluate the decision rather than just the outcome. Companies that never write this down cannot tell the difference between bad decisions and bad luck, which means they cannot improve their decision-making at all.
| Layer | Frequency | Decision class | Who is in the room | Artifact |
| Daily | 15 minutes, every working day | Clear today’s blockers | Anyone who can remove an obstacle | Blocker list with owner and clock |
| Weekly | 60 to 90 minutes | Reallocate people, hours, priority. Start, stop, continue | Function leads with commitment authority | Commitment list with marked deltas and a kill list |
| Monthly | Half a day | Reconcile operating story against the money. Corrective action | Leadership plus finance | Corrective actions with dollar values and owners |
| Quarterly | One to two days | Move capital, headcount, portfolio, strategic priority | Decision makers with allocation authority | A changed allocation, with reasoning on the record |
6. Give Every Layer an Owner and an Exit Condition
Two governance rules keep a cadence from rotting, and both of them are about giving someone the standing to remove things.
The first is that every layer has exactly one owner, and that owner has the authority to cancel the meeting. Not reschedule. Cancel. A daily meeting with no blockers should be cancelled that morning. A monthly review where the reconciliation shows no gap should be twenty minutes instead of four hours. When nobody has cancellation authority, every slot gets filled to its allotted length regardless of whether there is content, because leaving a scheduled meeting unheld feels like a failure rather than the efficiency it actually is.
The second rule is that every recurring agenda item has an exit condition stated when it is added. What has to be true for this to stop being reviewed? A quality issue gets reviewed weekly until first-pass yield holds above target for four consecutive weeks. A customer recovery gets reviewed until the account is back to baseline volume for two months. Write the condition down at the moment the item is added, when everyone is still thinking clearly about it, because nobody will ever propose removing it later.
Items without exit conditions are how cadence becomes theater. They accumulate silently. Every one of them was urgent when it was added and none of them is ever consciously retired, so three years later the weekly meeting opens with eleven standing items, nine of which are historical artifacts that get a nod and a “still fine” from someone who has stopped looking at the underlying data. The meeting is now ninety minutes long, has four decisions in it, and everyone has learned to check email through the first hour.
Audit the standing items once a quarter against their exit conditions. Anything that met its condition comes off. Anything that has been reviewed for two quarters without meeting its condition is not being managed by the cadence, and it needs either a different intervention or an honest admission that it is not going to be fixed at the current level of resource.
7. Design the Hours Between the Meetings
A weekly cadence occupies perhaps ninety minutes out of one hundred and sixty-eight hours. Whatever else it is, an operating cadence is mostly not meetings, and the companies that get execution right have designed the gaps as carefully as the gatherings.
Three things have to exist in the gaps.
Escalation rules, written down, that tell a person exactly when they are permitted to interrupt the cadence rather than wait for the next slot. Without them, one of two failures occurs. Either nothing escalates, and problems sit politely in a queue until Thursday while they compound, or everything escalates, and the cadence is drowned by interruption and stops meaning anything. The rule should name thresholds: this magnitude of miss, this class of customer impact, this dollar value. Thresholds turn escalation from a judgment about courage into a matter of procedure.
Decision rights, so that the great majority of decisions never need the cadence at all. The purpose of the meeting structure is not to make decisions centrally. It is to make sure the decisions that genuinely need multiple functions get made on a rhythm, and everything else gets made by whoever is closest to the work. If your weekly meeting is full of decisions that one person could have made alone, you do not have a cadence problem, you have a delegation problem wearing a cadence costume.
A written record, because a decision that was not written down was not made. This sounds like bureaucracy and it is the opposite. Two people leave a meeting with different understandings of what was decided, both act accordingly, and three weeks later the conflict surfaces as an execution failure that gets blamed on middle management. A two-line written record of what was decided, by whom, and what happens next, published within the hour, eliminates an entire category of organizational friction for almost no cost.
A decision that was not written down was not made. Two people leave the room with different understandings, both act accordingly, and three weeks later it surfaces as an execution failure and gets blamed on middle management.
8. Delete What the Cadence Replaces
This is the step that gets skipped and it is the step that determines whether the whole exercise helps or hurts.
When you install four clean layers on top of an existing calendar, you have not built an operating cadence. You have added four meetings to a calendar that was already too full, and the organization’s honest response will be to attend the new ones at low effort while continuing to run the business through the old informal channels that actually work. Within two months the new cadence is a compliance ritual and the old dysfunction is intact underneath it.
So the subtraction happens in the same week as the addition, not later. Go through the calendar meeting by meeting and ask which layer now owns its decision. If the Thursday operations call’s decisions now happen in the weekly layer, the Thursday call is cancelled, not shortened. If the monthly business review duplicates the monthly reconciliation, one of them dies and it should be the one built to satisfy a template rather than to produce a corrective action.
Expect resistance, and expect it to be sincere. Standing meetings accrue constituencies. People have built their own information flow around them and losing the meeting means losing access. That is a real cost and it has a real answer: the artifacts. If the daily blocker list, the weekly commitment list, and the monthly corrective actions are published and readable, most of the people fighting to keep a meeting are actually fighting to keep information, and the artifact serves them better than the seat did.
Set a target and hold it. Total recurring meeting hours after installation should be lower than before, not higher. If they are higher, you did not install a cadence. You added ceremony.
9. Install It in Thirty Days
The whole architecture goes in over four weeks. Longer than that and the effort loses momentum and gets absorbed into whatever crisis arrives in week five.
Week one is inventory and design. List every recurring meeting with its attendees, its length, and the decision it produces. Most of that list will have a blank in the decision column, and the blanks are the entire diagnosis. Sort the surviving decisions by the rate at which their inputs move. Write down the four layers, their owners, their inputs, and their artifacts on a single page. If it does not fit on one page, it is too complicated to be executed.
Week two installs the daily layer. Nothing else changes. Fifteen minutes, exceptions only, blocker list published every day. Two weeks of running it cleanly before anything is built on top of it, because the daily layer is what feeds the weekly layer’s escalation queue, and a weekly layer sitting on a broken daily layer will fill its time with the wrong material.
Week three installs the weekly layer, with the kill list reviewed first and the commitment list published with visible deltas. This is the week where the biggest behavioral shift happens and it is worth the owner personally chairing it for the first month to hold the structure.
Week four installs the monthly reconciliation and performs the subtraction. Cancel what the new layers replaced, publicly, in the same week the new structure is complete. The quarterly layer goes in at the next natural quarter boundary, and the unallocated resource pool has to be created at the next annual planning cycle, which means the full architecture is not finished for a year. That is fine. The first three layers deliver most of the execution improvement, and they deliver it inside a month.
Then measure the only thing that matters about a cadence: are decisions being made faster, and are they sticking? Count the decisions produced per layer per month. Count how many commitments from four weeks ago are still on the list unchanged. If the counts do not move, the architecture is right and the enforcement is missing, and the person who has to fix that is the one who owns the layer. Usually that is you.
One last thing worth saying, because it is the objection I hear most. People assume a structured cadence makes an organization rigid, and that fast-moving businesses should stay loose. The opposite is true and the reasoning is simple. An organization without a cadence does not become flexible, it becomes governed by whoever escalates loudest, because there is no scheduled path for a decision and the only route is interruption. A designed cadence is what makes it safe to say no to an interruption, because everyone knows exactly when the decision will be made instead.
Monday morning: open your calendar, write the decision each recurring meeting produces next to its name, and cancel every meeting where that column stays blank.
About the Stagnation Assassin
Todd Hagopian is a Fortune 500 transformation executive who has generated $3B in shareholder value across Berkshire Hathaway, Illinois Tool Works, Whirlpool, and JBT Marel, where he serves as Global VP of Product Strategy and Marketing. Known as The Stagnation Assassin, he is the author of two published books: The Unfair Advantage: Weaponizing the Hypomanic Toolbox and Stagnation Assassin: The Anti-Consultant Manifesto. His blog is published in 15+ languages and read by operators worldwide. Bring him to your stage via his speaking page or connect with him on LinkedIn.
Decision Velocity Audit. Send me your current recurring meeting list with the decision each one produces, and I will tell you which layers you are missing, which meetings are broadcasts, and where your decisions are queuing. No charge and no pitch at the end of it. Request the audit here.
See also: How Often to Review the Numbers