Most weekly operations reviews are a reporting meeting wearing the clothes of a decision meeting.
Everyone arrives with numbers. Operations walks through on-time delivery, quality through scrap, purchasing through late POs. The numbers are accurate, the slides are honest, and at the end the group has confirmed what it already believed on the way in. Nobody asked whether what the room assigned last week actually got done.
The meeting fails because everyone prepared. Preparing means reporting the things you already know about. What costs a manufacturer weeks is the set of things nobody knows about yet, and those never make it onto a slide, because there is no slide for an absence.
I have run this meeting as a software manufacturer, and the version I would run now is not the one I used to run.
What the review has to surface
Commitments made to you that have gone past their date. Not tasks. Commitments, meaning a specific person said they would do a specific thing by a specific time and the time has passed. The ones that matter most in a plant are the delivery commitments with a customer's name and a date on them, and the internal ones they depend on. The engineering change that has to land before the line can run it. The qualification a supplier owes you. Almost no review has this line, because the information is scattered across the Outlook mailboxes and Teams threads of the people who are owed.
A PMO director at an industrial coatings manufacturer sent a contract out for legal review with a promised date on it. She found out it was late when the contractor waiting on the document asked her about it, three days after the date. Nobody did anything wrong. There was no mechanism that noticed a date passing.
Decisions made more than two weeks ago with no observable movement. Every decision your leadership team made last quarter is finished, deliberately reversed, visibly progressing, or quietly dead. The quietly dead ones are what cost you, because the organization is still behaving as though they were progressing.
That is decision decay, where the call stays official and stops moving, so everyone would still tell you it is happening and nothing is.
For each one, ask what has changed since the room decided. Then say its fate out loud. Either it stands, and its owner names the next step and a date, or the room reopens it and writes down the new decision and the reason.
Work that changed hands since the last review. Handoffs are where things die, and they are almost never on an agenda. This is the item that pays best in a manufacturer, because the expensive handoffs run between departments that do not share a system. Design to engineering. Engineering to operations. Operations to finance. I have never seen a company where the finance crossing was the cheap one. Each crossing is a place where what was decided arrives thinner than it left. Naming the handoff out loud is a thirty-second item that catches a category of failure worth weeks.
The one number per function that would move first if something were going wrong. Not the dashboard. One leading number each, chosen in advance, that the function owner would notice before a customer does.
What got decided in this room, written down before anyone leaves, with an owner and a date on each. If your review generates decisions and does not record them where the next review can check them, you are producing decision amnesia every week: a month later the same question comes back, and nobody can say what was settled, what was ruled out, or why.
What to cut to make room
Reading out numbers everyone can read. Circulate the pack beforehand and open the meeting at the first question. If people do not read the pack, the pack is too long, and that is the thing to fix.
Status as narrative. "We are making good progress on the supplier transition" is not status, it is a feeling about status. Ask what specifically finished since the last review and what specifically is next, and let the answer be short.
Green items. A function that is genuinely fine gets fifteen seconds. Thirty if the week was strange. Giving every function equal airtime is fairness applied to the wrong thing, and it is why the item that needed twenty minutes got four.
And problems being solved in the room that involve three of the nine people present. Name it, assign it, take it offline. Everyone knows this rule. Almost nobody enforces it, because enforcing it feels rude in the moment and obviously correct forty minutes later.
The three that only show up when someone asks
Ask each one the same way every week, or people start optimizing their answers.
What did you expect to have by now that you do not have? The gap between the plan someone had in their head three weeks ago and where they actually are is real information, and it almost never gets reported, because by the time the plan slips the person has quietly revised the plan. Asking for the delta surfaces the revision.
What are you waiting on from someone else? The single most underused question in operating reviews. It catches the cross-functional blocker that neither side is escalating, the asking side because it feels like complaining about a colleague, the owing side because they never registered that anyone was blocked. Ask it of every function, every week, and say nothing when the answer is nothing. Left alone, those blockers become the conflicts between functions that a survey of nearly 8,000 managers, published in Harvard Business Review in 2015, found were "handled badly two times out of three": resolved after a significant delay, resolved quickly but poorly, or "simply left to fester."
What did we decide that you are no longer confident in? The hardest of the three. A decision the owner has privately lost faith in holds its official status and gets none of the owner's real effort, which makes it the most expensive object in the company. I have never run this one well. I have asked it and watched people answer the question they wished I had asked instead, and I still do not know whether that is a facilitation problem I could fix or whether the question is simply too expensive to answer honestly in front of eight colleagues.
The honest limit
Most of that list depends on somebody reconstructing the answer by hand before the meeting. The commitments past their date have to be gathered out of Outlook and Teams, a mailbox at a time. The decisions with no movement have to be remembered by somebody who was in the room a month ago, unless the last review wrote them down where this one can check. The handoffs have to be recalled. That reconstruction is hours of work, it falls on whoever owns the review, usually the PMO or program director rather than the COO, and it is the first thing to go in a hard week.
What it usually costs is rework, the decision made a second time because nobody could say what was settled the first time. When we asked operating leaders what a recent slipped commitment actually cost them, more named wasted time or rework than named money, a deal, or a damaged relationship.
This is the execution gap showing up in your calendar. The call got made, and the only instrument you have for finding out whether it carried is a room full of people trying to remember. Larry follows a week at the center of a company, and the two people who carry it, in The Operator's Case, the third Executive Intelligence Paper.
Run the checklist. Your ERP knows the order and its ship date. It does not know that engineering promised the drawing change by Thursday, and that promise belongs in a record the whole leadership team can see.
A weekly operations review is a manufacturer paying its most expensive people to remember out loud, once a week, what none of its systems thought to keep.