The Efficiency Trap: How We Optimised Ourselves Into Fragility
For the best part of thirty years, efficiency was the virtue nobody questioned. Strip out cost. Remove the buffer. Consolidate the suppliers. Run the asset harder. Hold less stock. Every layer of slack was treated as waste to be hunted down and eliminated, and the people who led it were rewarded. It felt like obvious good management, because for a long time it was.
Then a virus, a ship wedged across a canal and wars in Europe and the Middle East reminded us what we had actually built. Not lean, high-performing machines. Systems with no give in them at all. And a system with no give does not bend when it is hit. It breaks.
What we actually built
The playbook was everywhere and it was consistent:
Just-in-time inventory, so nothing sat on a shelf earning nothing
Single sourcing, because one big supplier is cheaper than three smaller ones
Assets run at ninety-plus percent utilisation, because idle capacity is money left on the table
Long, globe-spanning supply chains, because somewhere further away could always make it for less.
Each decision was rational on its own terms. Together they produced organisations optimised to a knife edge, tuned for a world that was stable, predictable and endlessly cheap to move goods around. The moment that world wobbled, the same choices that had looked so smart became the reason the shelves were empty, the line stopped and the order never came.
The trade nobody wrote down
Here is the uncomfortable truth underneath all of it. Efficiency and resilience pull against each other, and we spent decades pretending they did not.
Every buffer you remove is a risk you take on.
Every supplier you consolidate is a single point of failure you create.
Every point of spare capacity you strip out is a shock you can no longer absorb.
We only ever counted one side of that ledger. The cost of the buffer was visible, immediate and easy to put in a business case. The value of the buffer was invisible right up until the day you desperately needed it and did not have it. So we optimised away the thing we could see and quietly accumulated a liability we could not, and we called the result progress.
Why we did it anyway
We need to be fair about how we got here, because it wasn't through stupidity. It was short term reward.
A buffer is invisible when it works. Nobody praises the inventory that stopped a stockout that therefore never happened, or the second supplier that was never needed this quarter. But the cost of that buffer shows up on every report, every month, in plain sight. So the person who cuts it looks decisive and gets promoted, and the person who defends it looks cautious and gets overruled.
Resilience pays off rarely but enormously. Efficiency pays off constantly and modestly.
It's not really surprising that more often than not, organisations will choose the constant, visible win over the rare, invisible one almost every time, right up until the rare event arrives and takes the building down.
Optimised for a world that no longer exists
The deeper problem is that the environment which justified all of this has gone. The hyper-optimised model was the correct answer to a specific set of conditions:
Stable geopolitics
Cheap and reliable supply chain and logistics
Predictable demand
Globalisation running only in one direction
Reading that list back, its quite stark how the world has changed in just a few years. Not one of those conditions holds the way it did. We are now in a world of trade friction, weaponised supply chains, energy volatility, extreme weather and demand that lurches without warning.
The war in Europe was not the end of it.
Conflict in the Middle East has repeatedly impacted two of the arteries the world depends on. Attacks around the Red Sea have pushed container ships off the Suez route and the long way round Africa, adding weeks and cost to once routine journeys. And every flare-up puts more than a question mark over the Strait of Hormuz, the narrow channel through which roughly a fifth of the world's oil and much of its liquefied natural gas has to pass, where every closure ripples through global energy prices within days.
Next... well who knows, but heightened tensions between China and Taiwan hang over the most concentrated choke point in the modern economy, the advanced semiconductors that almost everything now depends on, alongside some of the busiest shipping lanes on earth.
Running a buffer-free operation in that environment is not efficiency. It is standing in an open field during a lightning storm because it hasn't rained in years.
The same trap, at national scale
If you want to see where this logic ends up, look at the national infrastructure we all depend on. The same instinct that stripped the buffers out of businesses stripped them out of the systems that keep the taps running and the country warm, and for the same reason. Storage costs money to build and maintain, and for a long time it looked like money we could save.
Water is the clearest example. Only one new potable supply reservoir has been completed in England since the water industry was privatised in 1989, Carsington in 1992, and even that was really an inheritance from the old era. It was first planned back in the 1960s and had been under construction since 1979, opening late only because a dam collapse in 1984 set it back by the best part of a decade. Take that one delayed project out, and the picture is starker still: not a single new potable supply reservoir has been commissioned in England in the more than thirty-five years since privatisation, even as the population has grown by more than ten million people.
More homes, more demand, and not one new store of water conceived to serve them. And we did not simply stop building. Existing supply reservoirs have been sold off or decommissioned rather than maintained, with Clydach in South Wales among the recent losses, and a GMB union investigation counting around thirty-five reservoirs sold since 2017 against just two built. The National Infrastructure Commission now warns of a potential shortfall of up to four billion litres a day by the 2050s, and puts the risk of a severe drought hitting much of England within thirty years at around one in four. The government's answer, nine new reservoirs, is welcome, but the first will not come online until the mid 2030s at the earliest. We spent three decades letting the buffer shrink, and now the key question is, are we rebuilding the buffer, or are we building to maintain the same level of risk we have today?
There is a fair debate to be had about whether Britain has enough total rainfall. But that debate rather misses the point. Rain you cannot capture and hold is not much use to you in August during a drought. The more places you have to store water, the better your odds of getting through a dry spell, whatever falls in an average year. Every reservoir sold off or never built is simply one less place to hold the wet months against the dry ones. Storage is the buffer, and we let it wither.
Another part of our national infrastructure, Gas, tells the same tale even more starkly. When the Rough facility closed in 2017, we removed around seventy per cent of the country's gas storage capacity in a single decision, because the site was ageing and no longer looked economic to maintain. As a result the UK can now hold only a sliver of its annual gas demand, on the order of two per cent, against twenty-five per cent or more in the large European storage nations. Measured in days, that is roughly a week or so of cushion here, against something closer to three months in Germany, France and the Netherlands. We swapped a strategic reserve for a just-in-time model that runs on daily imports, and it works beautifully right up until global prices spike or the weather turns, at which point we discover we are holding almost nothing in hand.
None of these decisions were mad in isolation. Each one saved real money and could be defended on the day it was taken. But strung together over decades they are the efficiency trap playing out on a national scale: a country that quietly optimised away its own shock absorbers, and only noticed once the shocks arrived.
This is not an argument against efficiency
Jim Collins put it best in Great by Choice, his study of the companies that thrive in chaos:
"The only mistakes you can learn from are the ones you survive."
To be clear, the answer is not to swing to the other extreme and pad everything with waste. Slack for its own sake really is just cost. The point is sharper than that: optimising blindly for a single variable is what gets you hurt, and for a generation that single variable was efficiency.
Resilience is not the opposite of good management. It is insurance, and like all insurance it looks like an unnecessary expense right up to the moment it's the only thing standing between you and a very bad year. Collins found this in the data, not just in theory. The companies that outperformed through turbulence practised what he called "productive paranoia", preparing ahead of time and holding deliberately, almost irrationally large margins of safety. Far from being the loosely run ones, they carried bigger cash reserves and buffers than their rivals around eighty per cent of the time. Buffer was not their weakness, it was their discipline.
The skill is not choosing efficiency or resilience. It is knowing precisely where you can afford to be lean and precisely where you cannot.
What buying back slack looks like
In practice, resilience is a series of deliberate choices, not a vague instinct to hold more of everything. A handful of tangible moves make most of the difference:
Dual-source the things that would stop you.
Map the handful of suppliers, materials or components whose failure actually halts production, and qualify a second source for those alone. Stay lean and single-sourced on everything you could replace within a week, so you are paying for resilience only where it genuinely counts.
Hold buffer where it buys time, not everywhere.
Instead of carrying more of everything, carry deliberate days of cover on the critical few inputs and lines. It stops being "cash trapped on a shelf" and becomes "the week of breathing room that keeps us trading when a supplier goes down."
Write down your single points of failure.
Run the simple exercise most businesses never do: for each critical process, name the one supplier, system, site or person that, if it vanished tomorrow, stops you. You cannot manage a fragility you have never actually listed.
Leave headroom on the constraint.
On the machine, line or team sitting on your critical path, run at a sustainable utilisation rather than flat out. The last few points of "efficiency" there are exactly what turns a single breakdown or absence into a missed order.
Design for switchability.
Set up contracts, carriers, routes and integrations so you can change supplier or plan in days, not a six-month project. Pre-qualify the backup, keep the integration ready, and treat that option as cheap insurance rather than wasted effort.
Stress-test the bad day, then rehearse it.
Walk through concrete scenarios, a key supplier fails, a site goes dark, demand doubles overnight, and agree the response in advance. The aim is a plan that already exists when the shock lands, not one improvised in the middle of it.
None of that is anti-efficiency. It is efficiency with its eyes open.
The shift leaders actually have to make
There is a line from James Clear, in Atomic Habits, that sits underneath all of this:
"You do not rise to the level of your goals. You fall to the level of your systems."
A business does not become fragile because anybody chose fragility. It becomes fragile because its targets, its incentives and its reporting were all quietly tuned to reward a single thing. You fall to the level of what you measure.
The hardest part is not operational, it is how we measure and reward. As long as the only number on the wall is utilisation or unit cost, people will keep optimising the buffers away, because that is what we are asking them to do.
Resilience has to become something we measure, name and put on the risk register alongside every other material exposure, with an honest price attached to the failure it is there to prevent. The question in the boardroom cannot only be "what is this buffer costing us?" It has to sit next to "what does it cost us the day it is not there?"
Developing a solid understanding of the the cost of each, and a lot of decisions that looked obvious start to look reckless.
The edge for the next decade
The most efficient system and the most resilient system are almost never the same system. For thirty years we optimised relentlessly for the first and assumed the second would take care of itself. The last few years, sadly, made it very clear that it won't.
The businesses that thrive from here will be the ones that stop treating slack as a failure of discipline and start treating it as a deliberate, priced, strategic choice.
In a stable world, the leanest operator wins. In the world we actually have, the one that can take a hit and keep running does. Building that is not a retreat from good management... in this environment, it's what might just make a difference to the survival of your organisation.
