Every Company Has a Kerosene Burner

Hungry Marines shuffle in line under the cover of night, clutching paper trays and awaiting confirmation that their efforts today have earned them a hot meal.

The heat of the day is replaced by cold drafts running through a quiet desert backdrop. The hike to their location wasn’t particularly long — a 10-mile ruck out — but the training of the day had taken its toll. A brief respite to lift morale seemed in order.

They wait quietly, expectantly, weighed down by their equipment.

But word never comes.

For reasons the Marines couldn’t know from where they stood, isolated in the field, there would be no hot meals — another night of Meals-Ready-to-Eat. Field rations.

Abruptly, coldly, their hope evaporated like the sweat on their uniforms, leaving only the salt behind.

The evening is seasoned with disappointment.

The value and comfort of a meal are easy to underestimate—until we’re forced to go without. That discomfort isn’t just hunger, nor is it about doing without as a form of discipline; it is the point where a small oversight, invisible in daily routines, becomes impossible to ignore. In a complex system, this is how hidden gaps surface: not as abstract risks, but as very real, tangible problems that eventually demand attention. The overlooked meal is a small version of the larger blind spots explored here—those unclaimed corners of responsibility that only reveal themselves when something essential stops working.

In field training, often far from any established cities or towns, Marines may periodically be served hot meals from a mess kit. The essential component of the system is the kerosene burner. It is also equipment almost no one is truly responsible for — often not locally supported, not meaningfully maintained, not the kind of thing anyone thinks about until it stops working, or when it’s time to inventory stock.

The Marines who use it know it well. They know how it works, what basic maintenance involves, and when it is broken. What they did not have — and couldn’t see from where they stood — was what it actually takes to replace one.

Replacing one is genuinely complicated. It doesn’t work like ordering a part from a catalog. It runs through a specialized process, with its own rules, at a depot that is states away, which even more experienced Marines may struggle to fully understand. So, what happens is what always happens with something unfamiliar: people avoid it. Requests get deferred, denied, or left untouched — not because anyone is negligent, but because the knowledge lives somewhere else, with someone else, and reaching it is harder than hoping the problem holds off a little longer. The Marines in the field are speaking a different language than the system that could deliver a solution.

It isn’t that nobody could learn the process. It’s that the process is esoteric enough, and held narrowly enough by those who do understand it, that the people who depend on it most steer clear of it — if they are able to see it at all. The knowledge sits with a few specialists who aren’t always reachable, and everyone else treats the whole thing as someone else’s problem — right up until it isn’t.

Nobody owns the whole picture.
The Marines held their half.
I held mine.

The gap between those halves wasn’t anyone’s fault — it was just invisible from both sides. It stayed invisible until people’s stomachs started growling in the field.

The failure was already there the whole time. It just didn’t register as a failure until it surfaced as hunger, far from where it had actually broken — and by then it was too late to fix quickly.

I didn’t run a company. My background is military logistics, not business. But I spent years inside large, complex organizations watching the same structural failure repeat, and I’ve come to believe this is not a logistics problem or a military problem. It’s a structural one, and it shows up anywhere people have to rely on each other to keep something running.

So far, this has been ubiquitous.

We rely on others not just because we want to, but because we must. No one can hold the whole picture of a complex operation in their head, so everyone chooses a lane and holds their pieces. That’s not a weakness — it’s the only way anything complicated functions at all.

But choosing a lane has a cost, and it’s a cost nobody puts on the invoice. When you choose a lane, you give up a full appraisal of the complexity it takes to keep the engine running. You see your half. You trust that someone sees the other half. And in the seam between those halves, exposure collects — quietly, unowned, invisible — exactly the way it did with that burner.

If you run a small company, you have kerosene burners. Plenty of them. The system only one person half-understands. The process that’s too tangled or too specialized for anyone to want to touch, so it quietly becomes nobody’s job. The dependency that lives in one person’s head, or with one outside specialist you can’t always reach. None of it looks like a problem right now. That’s the whole point. It won’t look like a problem until stomachs start growling.

And the reason these gaps persist isn’t laziness. It’s that the things most likely to expose you are often the things people most avoid looking at — because they’re complicated, because they’re unfamiliar, because the knowledge is held narrowly by someone who isn’t in the room. Avoidance feels reasonable in the moment. It’s almost always cheaper to assume the burner will keep working than to go figure out what happens when it doesn’t.

There’s a second version of this that’s often more insidious than plain avoidance. Sometimes someone does step in and try to fix the thing, and leaves it half done. The hard part gets deferred, the edges get left loose, but a record now exists that it was addressed. And that’s the trap: a half-finished fix looks like a solved problem. It reads as handled, so people stop checking it. The exposure is still live, but now it’s wearing the appearance of resolution, which means it gets noticed even later than the thing nobody touched at all. At least the untouched burner is honestly broken. To anyone looking at the paperwork, the problem is in motion. To the Marines in the field, nothing has changed. But now the person who might have started a fresh request assumes it’s already being handled, and the person who abandoned the first one assumes someone else picked it up. The burner is just as broken as before — but the system now incorrectly believes it is being fixed.

You can’t fix this by being more careful. You can’t assign every burner an owner and call it solved — because the moment your operation is complex enough to require specialization, it’s complex enough that no one sees the whole picture, and the seams just reappear somewhere else. The exposure isn’t a bug you can patch out. It’s the price of the structure working at all. It’s built-in, permanent, and it moves.

So I’m not going to tell you I can close the gap. I don’t think anyone can. The decentralization that makes your company function is the same decentralization that hides the risk, and you don’t get one without the other. What you can do is see it earlier. You can shorten the gap between when something is actually broken and when you find out. That’s the entire game: not eliminating the delay between a problem and your awareness of it, but closing it. Catching the burner before the field goes hungry instead of after.

In practice, that’s less about a system than a habit of attention. It’s noticing where the hesitations and the I-don’t-knows cluster, because that’s usually where a seam is. It’s being suspicious of the processes that rest entirely on one person, and more suspicious still of the problems that were recently “handled” — because a temporary patch treated as a permanent fix is the half-finished burner all over again. Mostly, it’s a willingness to walk the system end-to-end every so often and ask how it actually works, rather than how it’s supposed to, and to let someone from outside the usual lane ask the questions the insiders have stopped thinking to ask.

None of that closes every seam. Nothing does. But it greatly improves your chances of meeting a failure in its earliest, most manageable form — while it’s still a quiet gap and not yet a line of hungry Marines.

That’s not an optimistic outlook. But it’s the true one.

In my experience, the people who run things well aren’t the ones who believe they’ve eliminated their blind spots. They’re the ones who know the blind spots are structural, permanent, and moving — and who build the habit of looking for them a little sooner than hunger does.