I have sat through hundreds of tier meetings that produced nothing but shared exhaustion. The pattern is always the same: sixteen people in a room, a projector showing yesterday's output, a supervisor reading aloud numbers everyone already knows. An hour evaporates, nobody owns an action, and the same deviation appears on the board next morning. That is status theatre — a ritual of visibility without consequence.
The failure is structural, not attitudinal. When a meeting has no defined escalation criteria, no pre-prepared data and no decision authority in the room, it degenerates into reporting. People defend, explain and narrate. The moment a discussion needs a resource decision or a containment call, it is deferred to a follow-up, which is where accountability quietly dies.
Tiered meetings work when each tier answers one question: can we run today, and if not, who is fixing what by when? Anything beyond that belongs in a problem-solving forum, a project review or a one-to-one. Keeping the tiers ruthlessly narrow is what separates an escalation system from a daily briefing circus.
The architecture: three tiers, three clocks
A workable structure is Tier 1 at cell or line level, Tier 2 at value stream or plant level, Tier 3 at site leadership level. Each tier meets on its own cadence. Tier 1 runs in ten to fifteen minutes, standing, at the board nearest the line. Tier 2 takes twenty minutes maximum, once daily, with the area managers. Tier 3 meets weekly for thirty minutes and handles only what cannot be resolved below.
The critical design rule is sequencing: each tier starts after the one below it, close enough in time that an escalated item arrives hot. If Tier 1 runs at 06:30 and Tier 2 at 14:00, the escalation loses urgency and gets narrated rather than acted upon. I favour Tier 1 at shift start, Tier 2 ninety minutes later, Tier 3 same-day mid-morning on the chosen day. Freshness of information is half the value.
Attendance must be fixed and role-based, not name-based. The line lead, a quality technician and the maintenance planner own Tier 1. Area managers, planning and the quality engineer own Tier 2. Plant manager, quality director and engineering lead own Tier 3. Deputies are named in advance so the meeting never waits for a person. A tier meeting cancelled because someone is on holiday tells you the system was decorative all along.
The three-tier meeting stack
- Tier 3 — site leadershipWeekly, 30 minutes, red queue and systemic patterns only
- Tier 2 — value stream / plantDaily, 20 minutes, escalation queue as its spine
- Tier 1 — cell or lineEvery shift, 10–15 minutes, standing at the board
Escalation criteria: decided in advance, not in the room
The single biggest lever is writing the escalation triggers down before anyone argues about them. Ambiguity at the board is where an hour gets lost. An item escalates from Tier 1 to Tier 2 when the cell cannot resolve it with its own resources within the shift, when it recurs three consecutive days despite action, when it involves a customer-affecting characteristic, or when it needs a cross-functional resource such as a toolmaker, a lab test or an engineer's time.
From Tier 2 to Tier 3, the triggers shift in altitude: risk of missing a customer shipment, containment requiring spend or line stoppage, a supplier quality escape needing commercial escalation, or any repeat failure that survived two Tier 2 action cycles. Note the wording — these are conditions, not judgements. When triggers are conditions, the line lead escalates without asking permission, and the meeting stops being a negotiation about whether something is serious enough.
Recurrence deserves special treatment because it is the most commonly abused criterion. A deviation that appears, gets patched, reappears and gets patched again is not three problems; it is one problem with cosmetic fixes. My rule from hard experience: the same failure mode appearing after a declared closure escalates automatically one tier up, with the original owner presenting what was done and why it did not hold. This is uncomfortable, and that is precisely the point.

Data preparation: the board is built before the meeting
Nothing kills a tier meeting faster than someone opening a spreadsheet during it. Data preparation is a discipline of the previous hour, not a live activity. The line lead or a designated data owner updates the board beforehand: yesterday's plan versus actual with reason codes, open quality alerts, safety observations, and the action list with owners and due dates. If the data is not on the board at meeting start, the item is not discussed — that rule forces preparation better than any lecture.
Reason codes are where most plants go soft. Machine down is not a reason code; unwinder drive fault, maintenance response delayed is. Build a fixed, short code list per line — changeover overrun, material shortage, unplanned maintenance, first-pass quality loss, staffing gap — and enforce single codes per interval, not a default other that quietly becomes the most used category. Review the code distribution weekly at Tier 2; a dominant code tells you the real constraint, which is usually not what the meetings spend time on.
Visual standards matter more than software. The board should show the escalation status of every open item with a colour or symbol agreed plant-wide: green on-track, amber at-risk, red overdue or escalated. Anyone walking past should grasp the state of the area in ten seconds. Digital boards are fine — I have used them in plants with distributed leadership — but only if the update discipline matches paper. A digital board last touched at Tuesday lunchtime is worse than a hand-drawn one maintained hourly.
Running the meeting: script, clock, and the owner speaks
The Tier 1 script is short enough to memorise. Safety first, always, even if trivial. Then plan attainment with the top reason code. Then each red or amber item: the owner states what happened, what is being done, and what they need. New actions are confirmed with owner and time on the board. The line lead runs it; the area manager listens and removes obstacles — and must resist the urge to run it themselves. I have seen well-meaning managers turn a ten-minute huddle into a forty-minute interrogation.
A physical timer on the board is not gimmickry. When an item cannot be closed in two minutes, the correct move is to name the two people who will solve it elsewhere within the hour and move on. Parking is not deferral if it carries a time and a place; it is deferral only when it carries neither. The chair's core skill is distinguishing needs-a-discussion from needs-a-decision, and the latter should take seconds when the escalation criteria are clear.
Tier 2 works the same way but with the escalation queue as its spine. Area managers bring what their tiers could not close. The quality engineer brings the week's defect Pareto — not as a presentation but as a decision aid: which categories need escalation, which need containment, which need resource. Tier 3 reviews only the red queue and systemic patterns — recurring suppliers, chronic equipment, resource conflicts — and either assigns resolution authority or sets a deadline for a recommendation.
Escalation flows up; authority flows down. When both flow up, the top tier drowns.
The ten-minute Tier 1 sequence
- 01SafetyAlways first, even when trivial — it sets the order of priorities
- 02Plan vs actualTop reason code stated, not debated
- 03Red and amber itemsOwner speaks: what happened, what is done, what is needed
- 04New actionsNamed owner and time written on the board before close
- 05Close on clockAnything longer than two minutes is parked to two people, same hour
What to measure about the meetings themselves
Treat the meeting system as a process with its own indicators, because it will drift without them. Measure cycle time adherence: did each tier finish inside its clock? Track escalation volume by tier per week — a Tier 1 that never escalates is hiding problems; a Tier 1 that escalates daily is under-resourced or under-trained. Track action closure rate against due date, and more revealing, the age profile of the open action list. A list where nothing is older than a week is healthy; one with items entering their fourth week is a museum.
The metric most plants never collect is recurrence rate: of the actions closed this month, how many addressed a failure mode that had already been closed before? That number tells you whether the tier system is solving problems or rotating them. Pair it with a simple audit — monthly, someone from outside the area observes a Tier 1 and scores it against a five-point checklist: board updated before start, safety covered, owners speaking, actions captured with times, meeting ended on clock. Publish the scores. Areas improve what is visibly scored.
Watch for the classic decay modes. Slide decks creeping into Tier 2 are the first symptom of theatre returning. Chairmanship migrating to the most senior person present is the second. Actions without named individuals — quality will look into it — are the third. Audit for these quarterly and correct them on the spot. System integrity is maintained by small, fast corrections, not by annual redesigns nobody follows.
Making it stick
Sustainability comes from the tiers being genuinely useful to the people who attend them, which means leadership must respond. Every escalation that reaches Tier 3 and dies there without a decision teaches the whole plant that the system is ornamental. Responding visibly — a resource granted, a supplier challenged, a spend approved — reinforces the behaviour faster than any communication campaign I have ever run.
Be patient with the first quarter and ruthless with the clock from day one. Teams take six to eight weeks to trust that escalation is help rather than blame, and that trust is built through consistent, unemotional handling of bad news. Praise the escalation that caught a containment gap before it reached the customer; never shoot the messenger who brings a red item to Tier 2.
The day a line lead hesitates to escalate because of how it will be received, the hour a day you saved becomes the least of your losses. A tier system is a promise that raising a problem earns a response. Keep the promise on time, every time, and the meetings shrink to their proper size: short, uncomfortable where they should be, and impossible to mistake for theatre.
