A ready to run agenda, board layout and KPI set for team leaders, and how to escalate problems in hours rather than at month end.

It’s the second week of the month, and the operations pack has just landed in your inbox. Forty odd slides. Somewhere around slide twenty-three there’s a chart showing scrap on one of your lines. It’s been creeping up since the third of last month. Not a spike. Just a slow climb, from about 2% to nearly 6%, week after week.

So you ask the obvious question. When did we know?

And the honest answer, if you dig for it, is that the operators knew on the third. The team leader knew by the end of that week. The quality engineer had a hunch it was a material batch issue by the following Tuesday. The information was in the building the whole time. It just took five weeks to travel about two hundred metres, from the line to the people who could make a decision about it.

That’s the thing I’d want you to sit with for a second. Nobody hid it. It simply had no route to travel quickly, so it waited for the one meeting that was guaranteed to happen. Month end.

Now think about how a good automotive line handles a problem. An operator pulls the Andon cord, and the team leader is there within seconds. If they can’t sort it inside the cycle, it goes up again, to a group leader inside about a minute, and to an engineer or manager within a few minutes more. Toyota’s plant in Georgetown, Kentucky has reported thousands of Andon pulls in a typical day. Escalation measured in seconds.

Most mid to large UK manufacturers sit somewhere between those two worlds. You’ve probably got a morning meeting of some kind. But if I’m honest, in a lot of the factories I walk into, it’s a seated meeting in an office, it runs well past half an hour, and the problems it discusses are already a day or a week old.

We wrote about this years ago on our old Managing Business Performance page, back when we called them operations rooms. A lot of what we said then still holds. Red and green only. A handful of measures, not dozens. Problems get a countermeasure, not a deferral. But I’d change two things now. Thirty minutes is too long for the meeting at the team board. And one meeting isn’t enough. You need a short chain of them, linked, so a problem raised at half six can reach someone with authority before nine.

I first learned to run a meeting standing at a board as a young engineer in the early nineties, and I’ve been on both sides of it since: running the board, and being the manager the problems got escalated to. So this post is the practical version. Why improvements fade. How three short tiers move a problem up in hours. What the board should look like. The KPI set your team leaders can use from Monday. And how to stop the meeting itself from fading, which, to be fair, it will try to do.

Why your improvements fade

You’ve probably seen this happen. A kaizen week goes brilliantly. The team’s buzzing, the numbers look great at the report out, and somebody takes a photo of the new cell layout for the newsletter. Six months later, you walk past and it’s quietly drifted back towards how it used to be.

It isn’t just you. In 2018, Matthias Holweg, Bradley Staats and David Upton published research in Harvard Business Review that followed 204 lean improvement projects at a large bank across 14 countries over four years. Around one in five projects delivered nothing at all. Of the ones that did, a big share slid back within a year, and by the two-year mark fewer than half were still holding their gains.

The more interesting finding, I think, was about leadership. Projects with strong support from head office were far less likely to backslide: 79% were still above their starting point after a year, against 61% of those without that support. The researchers were pretty blunt about the reason. A continuous improvement culture doesn’t just emerge from the workforce on its own. Leaders have to keep engaging with it, as a habit, long after the launch.

They also named four ways it goes wrong, and you’ll probably recognise at least two. Initiative fatigue, where leaders move on to the next thing. Tool overzeal, where a stalled effort gets “relaunched” with a new methodology. Chasing short term metrics that don’t really match what the business is trying to do. And dogmatic rollouts that leave the people doing the work with nothing meaningful to contribute.

Michael Ballé at the Lean Enterprise Institute makes a point I’ve always liked. Starting 5S isn’t hard. Keeping it is, because keeping it means changing the way people work every single day. That’s a different job entirely.

I saw this play out with an aerospace composites business in Lancashire, around 450 people. They ran a really good changeover week on a press that fed their main curing line, and got a 90-minute changeover down to 35. Proper work, properly done. Four months later it was averaging about 70 again. Nobody had decided to go back. The changeover cart got borrowed. A new setter started who hadn’t been on the event. One busy week, the prestaging didn’t happen, and the next week nobody noticed.

What fixed it wasn’t another event. It was putting changeover time on the team’s daily board, with a red line at 40 minutes, and making it the first thing the team leader looked at every morning. Within about six weeks it was back under 40, and it stayed there.

So here’s how I’d put it. The improvement event wins the gain. The daily meeting is what keeps it.

Three tiers, one morning

The idea behind tiered daily management is simple. Instead of one big meeting, you run a short chain of them, each one standing at its own board, each feeding the next, all inside the first couple of hours of the shift.

Here’s a version that works for a lot of mid to large sites on a 06:00 day shift. Shift the times to suit yours, but keep the gaps.

Tier 1, the team, 06:30, 15 minutes. The team leader and their crew, five to fifteen people, standing at the team board in the work area. Yesterday’s performance, today’s risks, and actions the team owns.

Tier 2, the area or value stream, 07:15, 15 minutes. The area manager, the tier 1 team leaders, and someone each from maintenance, quality, planning and materials. Only reds and escalations. A green team gets a nod, not a recap.

Tier 3, the site, 08:15, 15 to 20 minutes. The operations director or site manager and the heads of function. Site trends, resource clashes, and the decisions that need authority or money.

The gaps matter. Twenty to thirty minutes between tiers gives leaders time to walk from board to board and means escalations arrive fresh rather than as second hand summaries. If your site’s smaller, two tiers is fine: team and site. The mechanics are the same.

What makes it a system rather than three separate meetings is the escalation. So write the rules down and pin them next to every board. Something like this:

1. The team fixes it. If the team can sort it with its own people and tools, it goes on the board as an action with an owner and a date, and it’s closed within 24 hours.
2. It goes up the same morning. If the fix needs another function, affects safety, has happened three times in a fortnight, or has kept a measure red for two days running, it goes on the escalation lane at tier 1 and gets raised at tier 2 that morning. Safety concerns don’t wait for any meeting at all.
3. An answer comes back the same shift. Every escalation that’s accepted gets a named owner and a response time. The team gets at least an interim answer before the shift ends: a fix, a dated plan, or an honest reason why not.
4. Nothing ages quietly. No escalation lives longer than 48 hours without an owner and a date. If it does, it goes straight to the next tier up.

Think of the whole thing as a clock. A month end review moves problems in weeks. A weekly meeting moves them in days. A working tier system moves them in hours. Andon moves them in seconds. You don’t need to be Toyota tomorrow. Moving from weeks to hours is a huge step on its own.

The physical detail that holds all this together is what I’d call the escalation lane. A strip on every board with four columns: what the problem is, when it was raised, who owns it, and when it was answered. If an escalation only exists in somebody’s notebook, it’s already in trouble, because nobody can see it getting older.

I remember a team leader at a chilled food plant in Lincolnshire, a site of about 600 people. When we went back through her shift log, she’d written up the same sticking conveyor guard eleven times in three months. Eleven. Every entry was polite and accurate. None of them had an owner. The first week it went on an escalation lane with a date next to it, maintenance had it fixed in two days. She wasn’t cross about it, which honestly surprised me. She just said, “I thought somebody was reading those.”

The board: what goes where

The board is the meeting. If it isn’t on the board, it isn’t in the meeting. That sounds strict, but it’s the single rule that stops a 15-minute stand up turning into a 45-minute chat about whatever’s on people’s minds.

Here’s a layout that works for most team boards, reading left to right.

Header strip. Team name, date and shift. Who’s in, who’s off, and how the gaps are covered. A simple skills cover grid, showing who’s trained on which station, earns its place here more often than you’d think.

SQCDP letters. One panel each for safety, quality, cost, delivery and people. Each has a letter tracker, which is the big letter divided into one box per day of the month, coloured red or green. Under each letter, a small trend chart with the target drawn on it, so you can see whether today’s red is a blip or the third in a week.

Problems and countermeasures. Every red gets a line: what happened, the containment for today, one action, an owner and a due date. Old fashioned, I know. It works.

The escalation lane. As above. What, raised when, owner, answered when.

Standards we’re protecting. This is the strip most boards are missing, and it’s the one that ties the meeting to your lean work. List the two or three improvements this team has made recently, with the measure that proves each one is holding. Changeover time on the press. 5S audit score in the cell. First off approval time. If one slips, it goes red like everything else.

Today. The plan, the biggest risk, and the standard agenda printed at the bottom so anybody can run the meeting.

A few rules from our old operations room days still stand up. Use red and green only. Amber is where problems go to hide. Keep to between three and six measures the team can actually influence. And get yesterday’s numbers on the board before the meeting starts, not during it. If the data turns up at lunchtime, your 06:30 meeting is really a meeting about the day before yesterday.

What about digital boards? If you’ve got machine monitoring feeding live OEE and downtime to a screen, brilliant. It saves somebody a 05:45 spreadsheet ritual. But I’d still start with a whiteboard and marker pens, and I’d keep the conversation standing at the board either way. The screen is the data. The meeting is people deciding what to do about it.

The KPI set and the 15-minute agenda

People often ask me for “the” KPI list. There isn’t one, to be honest. But here’s a starting set for a team leader that I’ve seen work in machining, assembly, food and process plants, with one main measure per letter and a supporting one where it helps.

Safety. Accidents and near misses reported yesterday. Yes, reported. A team that reports near misses is healthier than one that never seems to have any. Supporting: open safety actions.

Quality. Right first-time percentage, or scrap and rework count. Supporting: internal and customer complaints traced back to this team.

Cost. Scrap value or labour hours against standard. Supporting: overtime hours.

Delivery. Plan attainment, meaning what we said we’d make yesterday against what we made. Supporting: OEE or the biggest downtime cause on the constraint machine.

People. Attendance and skills cover. Supporting: improvement ideas raised and implemented this month.

Then, for tier 2, the area manager sees the same letters rolled up by team, plus the escalation lane for the whole area. At tier 3, the site leadership sees safety, on time in full to customers, cost of poor quality, output against plan at the constraint, and one more number that I’d argue matters as much as any of them: how many escalations are open, and how old the oldest one is.

Now the agenda. Same order every day, so it runs on rhythm rather than on whoever’s facilitating.

Minutes 0 to 2: safety and people. Anything that happened, any near misses, who’s missing and how we’re covering.

Minutes 2 to 6: yesterday against target. Walk the letters. The team calls each one red or green, not the team leader. That small shift matters more than it looks.

Minutes 6 to 10: the reds. For each one: what happened, how we’re containing it today, one action, one owner, one date. No root cause hunting here. That happens after the meeting with the people who are needed.

Minutes 10 to 12: yesterday’s actions and our standards. Did what we said we’d do actually get done? Are the improvements on our “protecting” strip holding?

Minutes 12 to 14: escalate. What can’t we fix ourselves? One sentence each on the lane, with the impact in units, minutes or pounds.

Minutes 14 to 15: today. The plan and the one thing most likely to go wrong.

Standing up isn’t just a lean habit, by the way. Back in 1999, Allen Bluedorn and colleagues at the University of Missouri compared more than a hundred five person groups and found sit down meetings ran 34% longer than stand up ones, with no better decisions to show for it. Standing keeps you honest about the clock.

A handful of ground rules stop the meeting going soft. Start on time, even if the boss is late. Especially if the boss is late. No chairs and no laptops. Challenge with respect, which was on our old page, and I’d keep word for word. And anything that needs more than a couple of minutes leaves the meeting with a name and a time attached.

Keeping the meeting from fading too

Here’s the slightly awkward truth. The daily meeting is the thing that stops your improvements fading, and the daily meeting will try to fade as well. It rarely dies loudly. It just gets a bit longer, then moves into the office “because it’s cold out there”, then somebody brings a laptop, then the area manager starts sending apologies. Three months later it’s a status update with chairs.

David Mann, in Creating a Lean Culture, describes a lean management system as four things working together: leader standard work, visual controls, a daily accountability process and leadership discipline. The meeting and the board are only two of them. The other two are about what leaders do, every day, to keep them honest.

So build that into the routine of every manager above tier 1.

Attend one tier 1 a day as a coach, not the chair. The area manager stands at the back, says almost nothing, and gives the team leader two minutes of feedback afterwards. Rotate round the teams.

Audit the meeting monthly. A manager from a different area watches one meeting and asks five questions. Did it start on time, standing, at the board? Did it finish inside 15 minutes? Was yesterday’s data up before it started? Does every red have an owner and a date? Did every open escalation get a status?

Rotate who facilitates. Every week or two, someone else in the team runs it. The meeting then survives holidays, sickness and promotions.

Measure the system itself. Track two numbers at tier 3: the average age of open actions, and how long escalations take to get answered. If either starts creeping up, the system’s fading, whatever the SQCDP letters say.

And please, invest in the people running the board. Research from the Chartered Management Institute found that 82% of people stepping into management roles in the UK have had no formal management or leadership training. On a factory floor, the person running your tier 1 was quite possibly your best operator eighteen months ago. Running a sharp, fair, 15-minute meeting in front of their old mates is a genuine leadership skill. It needs teaching and coaching, not just a laminated agenda.

One last thing, and it’s maybe the most important. In my experience, tier meetings almost never die because people get bored of them. They die because escalations stop coming back answered. A team leader raises something, it disappears upstairs, nothing happens, and after the third time they stop raising things. Why wouldn’t they? So if you only check one thing as a senior leader, check that. Every escalation, answered, visibly, on the board that raised it.

Hours, not weeks

So, back to that month end pack, and the scrap that climbed for five weeks before anyone with authority saw it.

With a working tier system, that story looks different. The operators flag rising scrap at tier 1 on day one. By day two it’s red for the second time, so it goes on the escalation lane. Quality picks it up at tier 2 that morning, with a response time. By the end of the shift there’s a containment in place and a suspect material batch quarantined. The month end pack still shows it, but as a small blip with a note next to it. Not five weeks of cost.

That’s really the whole case. Improvements fade because nobody checks the new standard every day. A 15-minute meeting at a board, linked to two more short meetings above it, checks it every day and moves problems up in hours. Keep the board simple, with red and green, three to six measures, a problems log, an escalation lane and a strip for the standards you’re protecting. Use a fixed agenda. Write down your escalation rules. Then make sure your managers coach the meeting, audit it, and above all answer what comes up to them.

If you’d like a hand setting this up, it’s exactly the kind of thing we do at TCMUK. Our Lean Coaching and Mentoring Programme trains your teams on site and then coaches your own people through live projects, with check ins at days 10, 30, 60, 90 and 120, so the daily management system is built by your team rather than handed to them. And if your team leaders are new to leading people, our Helping Managers to Succeed and Lead Programme gives them a mentor and practical tools for running teams day to day. Or, if you’d just like to talk through how your morning meetings work now, book a 15-minute call with Adam. No pitch. Just a conversation.

Your factory probably already knows about most of its problems. The question is how long it takes that knowledge to reach someone who can act on it.