Most daily production meetings run twenty five minutes and produce almost nothing. Everyone goes around the room, says roughly how their area did yesterday, someone mentions a machine that’s been acting up, and the meeting ends with vague agreement to keep an eye on things. It’s not that the people in the room don’t care. It’s that the meeting is built on memory and impression instead of numbers, and a meeting built on impression can only produce more impressions.
A tier board meeting fixes the structure, not the people. Same time, same short agenda, real numbers on the wall before anyone starts talking, and it ends with a list of specific actions with an owner and a date, not a status update everyone already sort of knew.
Why the impression-based version fails
Ask someone how their shift went and they’ll answer as best they can, from what they remember. What they remember is shaped by whatever felt significant in the moment, a loud breakdown, a customer call, a bad morning, not necessarily by what actually cost the most capacity. A twenty minute breakdown that happened in front of the supervisor gets reported. A changeover that ran fifteen minutes over standard four times that week, with nobody standing there timing it, doesn’t, even though it may have cost more in total.
That’s the same failure mode as a clipboard OEE number, covered in Your OEE Dropped Because You Started Measuring Everything, applied to a meeting instead of a metric. A meeting run on memory ranks problems by how loud they were, not by how much they cost, and a plant that manages by loudness ends up fixing the wrong things in the right order of urgency.
The 10 minute structure
A tier board meeting works because it’s short enough that nobody has time to wander, and structured enough that the same five minutes of small talk can’t eat the whole slot.
Headline numbers, on the wall before anyone talks
The meeting should never open with “how did yesterday go.” That question invites an opinion. It should open with the numbers already visible, OEE against the plan, output against the target, and any open safety or quality item, so the room starts from the same shared facts instead of negotiating what “went well” even means. A wallboard or a shared screen showing this before the meeting starts turns the first two minutes into confirmation, not discovery, which is exactly what makes the rest of the meeting move fast.
The biggest loss, ranked by dollars, not by noise
This is the segment that actually earns the meeting its keep. Pull the top one or two items off a downtime Pareto ranked by cost, the same discipline covered in how much downtime actually costs, and spend the time on those, specifically, not a general tour of everything that happened. If the biggest loss is a changeover running long on one specific product pairing four times a week, that’s the conversation, not a general complaint about changeovers being slow.
The discipline here is resisting the pull toward whatever machine had the dramatic stop yesterday, if it’s not actually the biggest cost in the window. A supervisor who watched a breakdown happen wants to talk about the breakdown. The meeting’s job is to check the ranked number instead of the memory, and talk about whatever’s actually costing the most, even when it’s quieter.
Open items, checked before anything new gets added
Every action from the previous meeting gets a status: done, in progress, or blocked, in that order, fast. This is the segment most meetings skip entirely, and skipping it is what turns a tier board into a talking exercise instead of a management tool. An action that gets raised, discussed, and never followed up on trains the room to stop taking the meeting seriously, because nothing that gets said there actually changes anything. Checking status first, before any new topic gets introduced, is what keeps a backlog of promises from piling up unaddressed.
New actions, with an owner and a date, or it doesn’t count
A meeting that ends with “we should look into that” hasn’t produced an action, it’s produced a feeling. The last two minutes exist specifically to convert whatever came up in the loss discussion into something with a name and a day attached. “Maintenance will inspect the fixture on CNC 3 by Thursday” is an action. “We’ll keep an eye on it” is not, and letting that pass as if it were is how a tier board slowly drifts back into being a status meeting with extra steps.
Why this only works with real numbers behind it
None of this structure matters if the numbers on the wall aren’t trustworthy. A tier board built on a hand-updated whiteboard that’s a day behind, or a spreadsheet someone forgot to refresh, degrades back into the same impression-driven meeting it was supposed to replace, just with a nicer looking template. The numbers need to reflect what actually happened on the floor the day before, pulled from the machines themselves rather than reconstructed from memory at 7am, which is the entire argument for measuring downtime and output at the source instead of on a form. A shift comparison view that’s already built for this rhythm, and a downtime Pareto that updates automatically overnight, is what makes the two minute headline segment take two minutes instead of ten.
What derails a tier board after the first month
The structure above is easy to run once, on the day someone’s watching. The failure mode that actually kills a tier board shows up in week three or four, once the novelty wears off, and it’s rarely the structure itself falling apart, it’s small erosions that each look harmless.
The most common one is letting the meeting run long “just this once” because yesterday was unusual, a bad breakdown, a customer escalation, something that felt worth more than two minutes. That’s a reasonable instinct in the moment, and it’s also exactly how a ten minute meeting becomes a twenty five minute one again within a few weeks, because every day has a reason that feels like an exception. The fix isn’t refusing to ever discuss something important, it’s moving that discussion to a separate, scheduled conversation with the people who actually need to be in it, instead of letting it eat the daily meeting that everyone else is depending on staying short.
The second erosion is letting the loss segment drift back toward whoever’s loudest, a few minutes at a time, once the person running the meeting gets comfortable and stops checking the ranked list before the meeting starts. That’s worth guarding against specifically, because it’s invisible while it’s happening. A meeting that’s slid back to running on impression looks identical to one still running on data, right up until someone compares what got discussed against what the Pareto actually said that week.
The third is skipping the open-items check when nothing dramatic happened the day before, treating it as optional when the list looks short. That’s precisely when it matters most, because a short list of open items is the easiest one to let slide unnoticed, and a habit of skipping it “when there’s not much to report” is how a backlog of unresolved actions builds up instead of all at once.
Quick recap
- A tier board meeting fails when it runs on memory instead of numbers, because memory ranks problems by how loud they were, not by what they cost
- Ten minutes, four segments, same order every day: headline numbers, biggest loss by dollars, open items, new actions
- Numbers go on the wall before anyone talks, so the meeting starts from shared facts instead of impressions
- Spend the loss segment on the ranked Pareto, not whatever machine had the dramatic stop yesterday
- Check every open item’s status before introducing anything new, or the backlog piles up unnoticed
- Every new action needs an owner and a date, “we’ll keep an eye on it” is not an action