Picture a Tuesday afternoon on the shop floor. A shift supervisor is walking the line, stopping at every station, checking every reading twice, signing off on jobs that honestly did not need signing off at all. The operators know the drill by now. They wait. They do not fix small problems themselves because that is not really their job, is it. It is his.
If that scene feels familiar, you are not alone. Most manufacturing leaders did not deliberately set out to micromanage their teams, it crept in gradually, one safety scare or one scrapped batch at a time, until checking everything started to feel like the only responsible option available to them. The trouble is that constant checking quietly erodes the very thing you need most, which is people who think for themselves under pressure. In a sector already stretched thin by skills shortages and an ageing workforce, that erosion is a cost you genuinely cannot afford right now.
This post is about building a different kind of control. Not looser standards. Not stepping back and hoping for the best. A framework built on two things that sound almost too simple: clear expectations and honest after-action reviews, done without blame. Get those right and something interesting happens. Your team starts owning the result, not just the task in front of them. You will still walk the floor. You just will not need to sign off on everything you see there.
Why manufacturers reach for control first, and what it actually costs you
Let me be fair to the instinct, because it is not an unreasonable one. Manufacturing runs on precision, and a missed torque spec, a contaminated batch, or a guard left off a machine can carry consequences that are genuinely serious rather than merely inconvenient. When the cost of a single error is that high, checking everything starts to feel like due diligence rather than distrust, and I understand why supervisors default to it. I did the same thing myself, earlier in my career, well before I properly understood what that habit was quietly costing me.
Here is the thing nobody quite says out loud, though. UK manufacturing is genuinely short on hands at the moment, with tens of thousands of unfilled vacancies across the sector at any given time, and roughly a fifth of the current workforce aged 55 or over and heading towards retirement within the next decade. Add rising costs into that mix, with a substantial share of manufacturers reporting increases of up to twenty percent over the past year, and you begin to understand why margins for error feel considerably thinner than they used to. When you are already short of experienced people, and every mistake now costs more than it did two years ago, the temptation to hover over the people you do have becomes stronger still. You cannot afford a mistake, so you check, and then you check again, which is a fairly rational response to a genuinely difficult set of conditions. It simply does not achieve what it feels like it should achieve.
The problem is that constant checking does not actually protect you from mistakes long term. It just moves the risk somewhere else. People stop flagging small issues because they assume someone above them will catch it anyway. Decisions queue up waiting for a signature that adds nothing except a delay. New starters, already hard to find and expensive to train, learn quickly that initiative gets corrected rather than praised, so they quietly stop offering it. Then they leave, and you are back to square one, recruiting into a market where good candidates already have other options and know it. It is probably no coincidence that recent industry research found fewer than half of manufacturers now describe themselves as genuinely agile, the lowest figure in five years. Rigid, closely supervised teams do not adapt quickly. They wait for permission.
I remember spending a morning on the floor of a components plant in the West Midlands a few years back, watching a line leader physically walk over and approve a changeover that the operator had done a hundred times before without a hitch. Competent bloke, clearly capable, just waiting for a nod that added nothing except a few minutes of dead time and a quiet message that his judgement was not quite trusted yet. Multiply that across a shift, a week, a year, and you start to see where the fatigue comes from. Not from the work itself. From being managed like you cannot be relied on.
None of this means loosening your standards. It means moving the control from your physical presence at someone’s shoulder to something more reliable: clear expectations up front, and a proper look back afterwards. That is the shift the rest of this post is about.
Start with expectations so clear nobody has to guess
Most accountability problems do not actually start at the point of failure. They start much earlier, at the point where somebody was told to keep on top of quality or make sure output stays up and left to work out what that actually means in practice.
Vague instructions are the enemy here, I think, more consistently than laziness or carelessness ever are. If a target is not observable, meaning it is not something you could point at and confirm definitively happened or did not happen, then you have not actually set an expectation. You have set a hope instead, and hopes are notoriously difficult to hold anyone accountable to fairly.
So get specific. Instead of keep quality high, try something like scrap rate stays under two percent per shift, and any reject gets logged with a reason code before the next changeover. Instead of watch the schedule, try every delay over fifteen minutes gets flagged to the team leader within the hour, not at the end of the shift when it is too late to do anything useful about it. These are not complicated asks. They are just precise enough that everyone knows what good looks like and, just as importantly, what falls short of it.
It also helps to be explicit about where the boundaries of someone’s own decision making actually sit. What can an operator decide alone, without asking anyone first. What needs a quick check with the team leader. What genuinely has to go up the chain before anything moves. You do not need a fancy framework for this, though plenty of businesses use something like a simple breakdown of who is responsible for a task, who needs to be consulted, and who just needs to be kept informed. What matters is writing it down somewhere everyone can see it, not keeping it in your head as an unspoken rule that people only discover after they have already broken it.
This matters even more in manufacturing because of shift patterns and handovers. A message that lives only in one supervisor’s head is not much use to the night shift, or to the agency worker filling in on a Friday who has never worked your line before. I have seen sites transform their handover quality just by moving from a quick verbal chat at the door to a short-written log with three or four fixed questions. What ran well. What did not. What needs watching next shift. It takes two minutes to fill in and it means expectations travel with the work itself, not just with whoever happened to be standing there when something was decided.
One more thing worth saying honestly. Clear expectations are not a document you write once and forget about. Targets shift, lines change, new kit gets installed, people move roles. Revisit them properly every few months, out loud, with the team present, not just as a memo nobody actually reads.
Swap the blame game for after action reviews
This is probably the part of the framework that feels most uncomfortable at first, and also the part that does the most work once it beds in.
An after-action review is a short, structured conversation that takes place immediately after a job, a shift, or an incident, organised around four honest questions. What was supposed to happen. What actually happened. Why did a gap emerge between the two. What will we do differently next time as a result. It is a method that originated in military training and was subsequently adopted extensively within lean manufacturing, Toyota included, because it converts everyday experience into genuine learning instead of allowing it to simply evaporate at the end of a shift.
Compare that with how most root cause meetings actually go in practice. Someone made an error, so everyone gathers in a small room, and even with the best intentions in the world, the conversation tends to circle around who did it rather than what allowed it to happen in the first place. People get defensive. Honest explanations start turning into excuses. And the next time something goes slightly wrong, the person involved is that little bit more likely to quietly fix it themselves and say nothing, rather than risk sitting through another one of those meetings.
That is genuinely dangerous in a manufacturing environment, maybe more than in most workplaces. Near misses, small quality slips, odd noises from a machine that probably mean nothing, these are exactly the things you want reported early, while they are still cheap and safe to deal with. A culture that punishes honesty does not stop mistakes from happening. It just stops you hearing about them until they get too big to hide.
The fix is not removing accountability. It is separating accountability from blame, which are not actually the same thing even though we tend to treat them like they are. Accountability means owning what happened and what you will do about it next. Blame is about assigning fault to a person and leaving it there. You can have deep accountability with almost no blame at all, and honestly, that combination tends to produce far better results than the blame heavy version ever does, because people keep telling you the truth.
Practically, this can be as simple as a ten-minute huddle at the end of a shift or a job run. Three or four questions, everyone allowed to speak, no names attached to problems, just facts and fixes. I worked alongside a food production site where this felt awkward for about the first fortnight. People were clearly waiting for someone to get told off. Nobody did. By week three, operators were flagging their own near misses before a supervisor even noticed there was anything to flag, because they had learned the review was about the process, not about catching them out personally. That shift, from hiding problems to surfacing them voluntarily, is honestly the clearest sign the framework is starting to work.
Keep these reviews short, keep them regular, and resist the urge to turn them into a lecture with extra steps. The moment they start feeling like a telling off in disguise, people stop being honest in them, and you lose the entire point of running them at all.
Give people room to move inside clear guardrails
Once expectations are clear and the review habit is properly bedded in, you have actually earned the right to loosen your grip a bit. This is the step a lot of manufacturers skip, because it feels like the riskiest one.
Autonomy on a factory floor does not mean anything goes. It means defining the guardrails properly so people can move freely inside them. Your SOPs already do some of this work, they fix the how for safety critical steps, and that is right, some things genuinely should not be left to individual judgement on any given day. But most SOPs stop there. They rarely say when to escalate, when a small deviation is fine to handle locally, or what decisions comfortably sit within someone’s own role without asking. Fill in that gap and you give people a proper map, not just a fence around them.
Start small if trust has been thin on the ground until now. Let a team leader decide on minor changeover sequencing without asking first. Let an operator adjust a process parameter within an agreed range rather than calling someone over every single time a reading drifts slightly off centre. Watch what happens through your after-action reviews rather than through standing over their shoulder all shift. If it goes well, which it usually does, widen the scope a little further next time round. Trust built this way is earned in both directions, and it tends to actually stick.
There is a psychological safety piece underneath all of this that is easy to skip past but really is not optional. People will only use the freedom you give them if they genuinely believe raising a problem early will not backfire on them personally. This links straight back to the review culture from the last section. If someone brings a near miss to an after-action review and gets thanked for the honesty, that behaviour repeats itself. If they get quietly punished for it, even subtly, even just a slight change in tone, that behaviour stops overnight, and you are right back to people hiding things until they cannot be hidden any longer.
I think the leaders who get this right tend to share one habit in particular. When something goes wrong and gets flagged early, their first question is never who did this. It is always something closer to what did we miss, or what would have made this easier to catch sooner. That single change in the first question you ask, more than any policy document you could write, sets the tone for whether your team owns problems openly or buries them quietly instead.
None of this happens overnight, to be honest. You are not flipping a switch from controlled to autonomous in one meeting. You are widening the guardrails gradually, checking the reviews stay honest, and learning to trust the process you have built rather than your own instinct to check everything yourself, one small step at a time.
Where to start this week
Here is the shape of the whole thing, stripped right back. Set expectations specific enough that nobody on your team has to guess what good looks like. Review the work honestly and regularly, separating accountability from blame so people keep telling you the truth even when it is inconvenient. Then, and only then, widen the space people have to make their own calls, watching through those reviews rather than through constant supervision.
None of this is about loosening your standards, whatever it might look like from the outside. If anything, it raises them, because a team that genuinely owns its results tends to catch more, fix more, and hide less than one that is simply waiting to be told what to do next. It does feel slower to build than just doing it yourself or hovering until a job is done right in front of you. I will not pretend otherwise, because it is. But the payoff, fewer surprises, fewer good people walking out the door towards a competitor twenty miles up the road, less of your week spent firefighting things that should never have reached your desk, is worth the slower start.
If you want to take one thing away from this and actually try it this week, make it the after-action review. Pick one shift, one job, one changeover. Run the four questions honestly, no names, no blame, just facts and fixes on the table. See what your team tells you once they believe it is genuinely safe to be honest with you. It usually says more about where your accountability gaps really are than any formal audit ever will.
This is exactly the kind of practical, day to day skill we build out properly in the High Performing Teams workshop, if you want a fuller toolkit rather than just a starting point. But even without that, the four questions above will get you moving in the right direction. Your team already knows more about where things go wrong than you probably give them credit for. Give them a framework that lets them actually tell you and watch what they do with the responsibility once they know it will not be used against them.
