How to run a weekly S&OP meeting that makes decisions
Most sales and operations planning meetings review numbers and leave the hard choices for later. Restructure the hour around named trade-offs, owners and a log.
A weekly sales and operations planning meeting usually fails quietly. Everyone arrives with a deck, the deck runs to forty slides, and the meeting ends with the same three problems on the agenda it started with. The fix is rarely a better template. It is a different purpose. The meeting exists to settle a small number of explicit trade-offs, and each one has a named owner who leaves the room committed to it. Everything else is reading, and reading can happen before the meeting.
This guide is for the people who chair or sit on that meeting: a chief operating officer, a head of supply chain, a commercial director, a finance lead. It assumes you already hold a weekly session that covers demand, supply and inventory, and that you want it to produce decisions rather than updates. It does not offer industry benchmarks for meeting length or attendance. Where a number would help you, pull your own from your systems, because the figures that matter are the ones your plants, warehouses and customers actually produce.
Before you start
Three things must be true before you change the agenda. If one is missing, the new format will fail in the same way as the old one, and the team will conclude that the idea was wrong when the real problem was the preconditions. Check them with the people who will sit in the room, not only with the person who asked for the change.
- One demand number that both the commercial and the operations teams accept for the planning horizon, even if it is imperfect. An argument about the baseline will consume the hour before any choice is put on the table.
- One named person who owns the monthly plan and can say no on the company's behalf. If nobody can, every item will be deferred to someone who is not in the room.
- A single page listing the constraints the company will not break: service levels promised to key customers, the cash limit set by the board, contractual minimums, and labour agreements. Decisions are easier when the boundaries are visible.
- Step 1
Write the decisions before the agenda
Ask the chair to list, two working days before the meeting, the choices the group must settle this week. Examples include which orders to prioritise when supply falls short, whether to build ahead of a promotion, and whether to buy material at a premium to protect a customer. Each choice becomes an agenda item. What usually goes wrong is that the list fills up with topics such as 'review third quarter' rather than choices. If an item cannot be phrased as an either-or, it is not ready for the mee
- Step 2
Put the numbers beside each trade-off
Send a short pack a day ahead that contains only the figures each decision needs: the cost and lead time of each option, the customers and products affected, and the cash tied up in the choice. Leave out the rest of the dashboard. What usually goes wrong is that the pack reproduces everything, and the room spends forty minutes searching for the one line that matters. A sensible test is whether a reader can state each option's consequence in a single sentence after reading the pack once.
- Step 3
Name one owner for each decision
Every decision gets a single person who will carry it out, and that person states the choice aloud before the group moves on. A committee can discuss a trade-off but cannot be held to one. What usually goes wrong is that the owner is recorded as 'the supply chain team' or 'commercial'. Ask who will make the telephone call to the plant manager or the key account on Monday morning. That person is the owner.
- Step 4
Take the decision inside the meeting
Ask for a yes, a no, or a dated return with a specific question that must be answered. Record the outcome in a decision log before the next item begins. At the close of the meeting, the test is simple: can someone who was not in the room say what was decided and who is doing it? What usually goes wrong is that items are parked for offline discussion, which in practice means they are decided later by whoever is most persistent in the corridor.
- Step 5
Set the escalation limits in advance
Agree in writing which decisions the group may take alone, which need the chief executive or the finance lead, and which go to the board. Express the thresholds in money, units or days, not in adjectives such as 'material' or 'significant'. What usually goes wrong is that everything above a line of management is escalated by default, and the meeting becomes a queue for approval. The limits should be reviewed once a year and printed on the same page as the constraints.
- Step 6
Move status into a pre-read
Ask each function to add one line of status to a shared document read before the meeting. Ban presenting status aloud unless it changes a decision on the agenda. What usually goes wrong is resistance from managers who fear being unprepared or being seen as idle. Reassure them that the pre-read is where their work is visible, and that the meeting will spend its time on the choices their work has made necessary.
- Step 7
Open each meeting with the log
Begin every session by reviewing the previous week's decisions: what was done, what slipped, and why. A decision that turned out wrong should be recorded as a decision that turned out wrong, with the reason and the change it prompted. What usually goes wrong is that the log is kept but never read, and the same question returns each month under a new heading. Reading the log is the cheapest way to stop the meeting going round in circles.
Where it goes wrong
The most common failure is drift. The people who hold the data are often the people who run the meeting, and they are comfortable presenting it. Within a few weeks the agenda is back to reporting, with a decision tacked on at the end if there is time. The remedy is structural rather than a matter of willpower. Make the chair responsible for the decision list, and make the pre-read the place where information lives. When a presenter starts reading out a chart, the chair should ask which decision the chart supports. If there is no answer, the chart goes back into the pre-read.
The second failure is authority without ownership. A meeting can reach a sensible trade-off and then see it dissolve because the person who must act was not in the room, or was in the room but had no power to commit resources. In our experience, this happens most often with decisions that cross a boundary: a sales commitment that needs plant capacity, or a stock build that needs cash approval. The fix is to name the owner of each cross-boundary choice before the meeting, and to invite the approver for the specific items that need them rather than for the whole session. Attendance should follow the decision list, not the org chart.
The third failure is the wish list. Some groups learn to write decisions but fill the list with aspirations such as 'improve forecast accuracy' or 'reduce expedited freight'. These are goals, and goals belong to the annual plan. A weekly decision is narrower: whether to expedite this shipment, whether to accept this order at the requested date, whether to move this production run forward. If a goal keeps appearing on the weekly list, ask what decision it is waiting for, and put that decision on the list instead.
- How long should the meeting run?
An hour is enough if the pre-read is doing its job. If the session regularly runs past ninety minutes, most items are probably reporting rather than deciding. A sensible test is to time how long the room spends on items that end in a recorded decision, and compare it with the time spent on updates. Shorten the updates first.
- Who should chair the meeting?
Choose someone with authority over both the sales and the operations sides, or a delegate who holds that authority in writing. The chair does not need to be the person who knows the most detail. The chair needs to be willing to stop a discussion that has not yet produced a choice, and to ask who will act on it.
- What if we cannot agree on a single demand number?
Pick one, record the assumptions behind it, and set a date to revisit it. An argument over the baseline is itself a decision that has been postponed, so treat it as an item on the list with an owner and a deadline. Agreeing to disagree for a month is usually better than debating the figure every week.
- Should finance attend every week?
Finance should attend whenever a decision commits cash, changes inventory valuation or affects a credit limit. Questions about tax treatment, lender covenants or contract terms should be taken to specialist advisers, and the meeting should record that the question has been referred. The group should not settle those matters on the spot.
- What do we do when a decision turns out wrong?
Record it in the log with the original reasoning, the outcome and the change that followed. The purpose of the review is to improve the next decision, not to assign blame. Groups that record failures openly tend to take clearer choices, because the cost of being wrong is visible and bounded.
This week, do three things. Cancel the agenda for the next meeting and replace it with a list of no more than five choices, each written as an either-or. Put a named owner against every item, including the person who will make the first call. Start the decision log tomorrow morning, even if it has one entry, and make it the first thing on the next agenda. At the end of that meeting, ask whether anyone who was not in the room could say what changed. If the answer is no, the next week's list should be shorter and sharper, not longer.

