The typical inventory meeting fails before it starts: purchasing brings one spreadsheet, sales brings another, finance brings a valuation report, and the first forty minutes go to reconciling numbers. An evidence-backed inventory meeting starts from a single ranked list, shows the same three facts for every flagged SKU, and ends with named owners. It is shorter, and things actually happen afterwards.
Key takeaways
- One list, distributed before the meeting, replaces the reconciliation debate. If people arrive with their own numbers, the meeting has already lost.
- Every flagged SKU carries the same three facts: days of cover, velocity trend and last movement date. Evidence travels with the line, not with the person defending it.
- Overrides are allowed and logged. The log is where you discover wrong lead times, stale pack sizes and account-specific deals nobody wrote down.
- Each action gets one owner and one date. "We'll look into it" is not an action.
- Finance joins for clear-down and margin decisions only, not for every reorder line.
Why inventory meetings drift
Charts start debates. A turnover graph invites interpretation; three people read it three ways and the meeting becomes a discussion about the chart. Evidence ends debates. "Cover is 9 days, velocity is up 18 percent over four weeks, lead time is 21 days" leaves one sensible question: how many units, and by when.
The second cause of drift is scope. Meetings that try to cover expedite decisions, weekly reorders and next quarter's supplier strategy in one sitting run long and resolve nothing. Split the horizons: a daily stand-up for the 14-day list, a weekly block for reorders, a monthly review for structure. This article covers the weekly meeting, the one most teams get wrong.
Before the meeting: one ranked list
Distribute the list the evening before. It should contain, for every flagged SKU:
- Risk tier (low, medium, high, critical) and the score behind it
- Recommended action: reorder, stop-buy, transfer, discount or escalate, with a quantity where relevant
- Three evidence facts: days of cover, velocity trend (4 weeks versus 12), last movement date
- Money at stake: revenue at risk for stockouts, cash trapped for overstock
- Proposed owner
Cap the list to what the team can execute in a week. Receiving has a capacity; so does the buyer's afternoon. A 120-line list guarantees that line 80 is never read. Thirty to fifty lines is a working range for a four-buyer team.
Why three facts, not ten
Three facts fit on one line and can be checked in seconds. Ten facts become a report and the meeting turns into reading. Anyone who wants more can ask, and the answer should be available on demand, but the default view stays lean.
During the meeting: a fixed agenda
| Block | Time | Content | Output |
|---|---|---|---|
| Last week's actions | 5 min | Completion rate on the previous list; anything still open | Carry-over items with new dates |
| Critical and high tier | 20 min | Top lines by money at stake; approve, adjust or override each | Approved quantities, owners |
| Stop-buy and clear-down | 15 min | Overstock lines; finance present for budget calls | Freeze list, clearance budget |
| Overrides | 10 min | Each override stated in one sentence: why, and what happens instead | Override log entries |
| Patterns | 5 min | Anything that reappeared three weeks running | Items for the monthly review |
Two rules keep it honest. First, no new SKUs join the floor unless they displace a line of equal risk; otherwise the list grows until nobody reads it. Second, time-box overrides. The buyer or sales lead states why they disagree with the recommendation and what they will do instead, in one sentence, and the disagreement is recorded rather than argued.
The override log is the real asset
An override is not a failure of the list. It is information the list did not have. Reviewed monthly, the log reveals systematic gaps:
- Lead times in the system that no longer match receipt history, so every reorder for that supplier looks too late
- Pack sizes or minimum order quantities that make the recommended quantity impossible to place
- Account-specific commitments (a contracted delivery, a promotion) that explain a demand spike the data reads as a trend
- Master data errors: a SKU replaced by a successor, still carried as active
Fix the root cause and the override disappears. Ignore the log and the same argument repeats every week, with the same person defending the same exception from memory.
After the meeting: actions leave the room
Approved reorders become purchase order drafts in the ERP. Transfers become transfer requests. Escalations become tasks in the CRM or a message to the account owner. Each carries its owner and date from the meeting notes. The following week opens by checking completion rate. If the rate is consistently below 70 percent, the list was too long; shorten it rather than adding a second meeting.
Finance participation
Invite finance for the clear-down block and for margin flags. They approve the clearance budget and hear the carrying-cost argument directly. They do not need to sit through forty reorder lines, and asking them to do so is how finance stops attending.
How Flowra handles this
Flowra produces the ranked list, refreshed nightly or on demand, with the evidence attached to every line: the risk tier and its weighted components, the stockout or overstock signal, days of cover, the 4-week versus 12-week velocity trend, last movement date and the money at stake with its basis stated. The recommended action is a draft; Flowra is read-only by default and a purchase order is only created after a human approves it, with the connector, the source and the user's role all having to allow it.
Overrides live in the same place as the recommendation. When a buyer corrects a quantity or refuses a proposal, the correction is recorded next to the original evidence, which is the override log this article describes. The list can be delivered where the team already works: a Monday digest by email, daily alert rules to Slack or Teams for critical lines, and the assistant available in-channel to ask "why this quantity" during the meeting, answering only from the data it holds. The ask-anywhere section shows what that looks like in Slack, and the wholesale use case shows a purchasing team running exactly this cadence.
- Fact
- Lyon: 8 days of cover, stockout signal critical. Central: 96 days of cover on the same SKU, overstock tier high.
- Forecast
- Lyon velocity steady at 6 units/day; transfer restores 38 days of cover. Central remains above its 45-day target after the move.
- Recommendation
- Transfer 180 units. Alternative: reorder 200 units from the supplier at 14 days lead time and accept a 6-day gap.
- Hypotheses
- Inter-site transport within 3 days; Lyon demand is not an unlogged promotion. Confidence is medium because central stock is reconstructed from flow and the last physical count is 5 months old.
- Next step
- Approve, adjust the quantity, or ask why. Nothing changes in the ERP until you do.
Frequently asked questions
How long should a weekly inventory meeting take?
Under an hour with a capped list. A fixed agenda of five blocks (last week's actions, critical lines, stop-buy and clear-down, overrides, patterns) fits in 55 minutes for a four-buyer team reviewing 30 to 50 flagged SKUs.
What evidence should accompany each SKU in an inventory review?
Three facts on one line: days of cover, velocity trend over four weeks versus twelve, and last movement date. Add the risk tier, the recommended action with a quantity and the money at stake. More detail should be available on request, not by default.
Should buyers be allowed to override the recommended action?
Yes, and every override should be logged in one sentence: why, and what happens instead. Reviewed monthly, the log exposes wrong lead times, pack sizes and account commitments the data did not know about, so the list improves.
See it on your own data. Upload a CSV or connect Odoo, WooCommerce or SQL — first stock-risk report in minutes, no credit card.
Get my free inventory report →