The Buffer Is the First Thing Cut

The leanest supply chain in the world is the one most likely to break.

That sounds like a cheap paradox, so let me try to knock it down. If it survives the beating, it tells you something about why smart, well-run companies keep walking into the same wall.

First, the honest case for lean, because it is a good one. For most of the twentieth century, factories sat on mountains of parts. Warehouses full of inputs, aisles of finished goods, capital frozen in stuff nobody had bought yet. Toyota looked at that and saw waste. Its answer, refined over decades, was to hold almost nothing: parts arrive as the line needs them, the next stage pulls from the last, inventory shrinks toward zero. Just-in-time. It was not a gimmick. It cut costs, exposed defects faster, freed up cash, and let a company turn on a dime when demand shifted. The idea was so obviously right that the whole manufacturing world copied it, and prices came down for all of us because of it.

So the buffer looks like pure fat. A shelf of spare parts earns nothing. A second supplier costs money to keep warm. Three months of chips in a drawer is cash doing nothing but waiting. In good times, every one of those is a line item a clever manager can cut to make this quarter look sharper than the last.

And they can cut it for years with no punishment at all.

That is the trap. Redundancy only ever pays off on the day something goes wrong, and on every other day it reads as inefficiency. The manager who runs lean gets promoted for the savings. The cost of the missing buffer lands later, on someone else, and gets blamed on the storm rather than on the decision to carry no umbrella.

Watch how this plays out when the storm actually comes. When the pandemic scrambled orders and then demand for cars roared back, automakers discovered they had squeezed their chip inventories to nothing. There was no cushion. Factories that could build everything except the one twenty-cent component sat idle. Billions in trucks parked half-finished for want of a part smaller than a fingernail.

Except Toyota. The company that invented lean was, for months, the one that didn’t run out.

The reason is worth sitting with, because it is the exception that proves the rule rather than breaking it. After the 2011 earthquake and the Fukushima disaster tore through its supply lines, Toyota learned that chips take far too long to reorder to survive a shock. So it did something that violated its own famous doctrine. It told suppliers to stockpile two to six months of semiconductors. It deliberately built the fat back in, for one category where the downside was ruinous. When the 2021 shortage hit, that drawer of chips was why Toyota kept building while rivals stalled.

Then, by that August, even Toyota’s buffer ran dry and it too cut production. Which is the point, not a footnote. A hedge does not make you invincible. It buys you time, and time is exactly what a company with zero inventory does not have.

Now the exhibit I’ve been saving.

On the morning of 23 March 2021, a container ship called the Ever Given, a quarter-mile of steel loaded higher than most people picture, caught a gust of wind in the Suez Canal and turned sideways. Wedged bow to bank, it corked one of the narrowest joints in world trade. Around a tenth of everything that moves by sea passes through that canal. For six days it passed through nothing. Hundreds of ships stacked up at both ends. The value of goods stuck in the queue ran close to ten billion dollars a day.

One ship. One crosswind. Six days.

The world economy did not have six days of slack, because it had spent thirty years congratulating itself for removing that very slack. Every container that wasn’t already sitting in a warehouse was a container someone had decided, sensibly, on a spreadsheet, in a calm quarter, not to pay to store. The savings were real and they were banked long ago. The bill arrived all at once, in a canal, in the shape of a boat.

I don’t think the answer is to go back to warehouses full of everything. That waste was real too, and lean genuinely made the world richer. The trouble is narrower and harder. We can measure efficiency to the cent, this quarter, on this manager’s watch. We cannot measure the fragility we’re buying with it until the day it comes due, and by then the person who booked the savings is three jobs away.

So the next time a system runs beautifully, cheaply, with nothing to spare, ask the awkward question. Is it robust, or has it simply not been tested yet?