Enough Until Tuesday

There are two ways to look at a shelf of spare parts.

One is to ask why the money is sitting there doing nothing. The other is to imagine the machine that will stop when the shelf is empty.

Both questions deserve an answer. The first keeps a business from filling its storeroom with expensive objects nobody needs. The second keeps an accountant’s tidy improvement from becoming a mechanic’s impossible afternoon.

The difficulty is that the shelf looks the same on the day when the spare part is unnecessary and the day before it becomes essential. The future does not put a red sticker on the correct box.

We have followed goods while they move. Now let them wait.

Waiting is not a single activity. A fan waiting to be sold, a motor waiting to be assembled, a shipment waiting at sea and a replacement bearing waiting for a machine to fail serve different purposes. They may all be counted as inventory in some business’s records, subject to ownership and accounting rules. They are not interchangeable answers to the question What do we need to keep going?

A room full of finished fans cannot repair the only machine that makes their motors. A room full of motors cannot satisfy a customer if every guard is missing. Spare parts in a warehouse across a flooded road can be physically intact and practically unavailable.

Useful inventory is specific. It is a thing that can perform a needed job, in a place from which it can be reached, before the need becomes an unacceptable loss.

A castaway with a supply ship

Robinson Crusoe is often remembered as a man alone, solving the problem of survival through his own effort. Open Daniel Defoe’s novel and look at what he does after the wreck. He goes back to the ship.

He takes food, clothes and tools. He finds the carpenter’s chest and gets it onto his raft. He makes repeated trips to salvage more. His new life begins with objects produced by people who are no longer there. The old world has left him a store of work.¹

Crusoe is resourceful. The tools do not use themselves. But his resourcefulness works on a much better starting point because someone else made a saw, a hammer and a chest before the story of his solitary labor began.

This is a useful correction to the fantasy that independence means beginning with nothing. It is also an incomplete moral portrait of the novel, whose world includes enslavement, colonial possession and Crusoe’s domination of Friday. We need not turn the castaway into an innocent hero to notice the material fact at the center of his survival: previous cooperation, and previous coercion, can remain present in the things a person calls his own.

A modern storeroom is less dramatic than a wreck. Its shelves make a related arrangement across time. Someone produced a thing earlier so that someone could use it later. That gap may last a day or many years. The question is whether the stored work remains ready for the task.

Crusoe cannot eat a promise that a boat will call next Tuesday. Neither can a factory install a part whose arrival has been postponed. When a connection fails, the difference between resources already available and resources expected from elsewhere becomes very sharp.

This does not establish that we should accumulate everything we can imagine wanting. Crusoe’s opportunities are unusual: the supply ship is wrecked, there is no warehouse rent to negotiate, and his choices are made under immediate danger. A modern business has recurring orders, finance costs, alternative suppliers and limited space. The novel gives us a question, not an inventory formula.

What work from the past should we keep close enough to help us through a break in the future?

The days inside the stock

Lena’s shop sells fans. Let us give it a deliberately simple rhythm: ten of the same model each day. A new shipment normally takes fourteen days to arrive after an order is placed.

If Lena waits until the last fan is sold before ordering again, she creates a predictable two-week shortage. The supplier could perform perfectly and the shelf would still be empty. Good replenishment begins before the need becomes visible as zero.

With perfectly steady sales and a perfectly reliable fourteen-day lead time, one hundred and forty fans would cover demand while she waits. If a new order arrives at the exact point those units are used, the cycle can continue without an extra cushion.

The word perfectly is doing a great deal of work.

Sales may rise. The shipment may be late. Stock records may include damaged goods. A customer may buy several units at once. Lena can hold a buffer to cover some of that uncertainty. A buffer of seventy fans gives her seven additional days at ten per day, provided the stock is usable and the selling rate stays as assumed.

Now let the shipment take twenty-one days instead of fourteen. A shop starting with one hundred and forty fans runs out after fourteen days and cannot meet the next seventy units of demand from that stock. A shop starting with two hundred and ten has enough for the full twenty-one-day interval. The extra seventy has bought seven days of continued service.

No statistical model has proved that seventy is the right buffer for Lena. We chose the numbers to make the mechanism visible. If daily demand jumps to twenty, the cushion disappears more quickly. If the delay lasts a month, seven days of protection will not cover it. If the shop sells several models, stock of an unpopular version may not solve a shortage of the preferred one.

A statement such as We hold three weeks of stock is therefore only the beginning of an answer. Three weeks at what rate of use? Across which products? In which places? Is the stock already promised to someone? Can it still be sold safely? Does the estimate assume that normal demand continues during a disruption that may itself change demand?

Days of cover translate objects into time under stated assumptions. They do not make those assumptions true.

There is another distinction. Some inventory exists because replenishment takes time even when nothing goes wrong. Goods in transit are often called pipeline inventory. Some is held as a cushion against uncertainty. More pipeline stock and more safety stock are not the same thing, though a business may need both.

A classic result associated with John Little connects the average number of items in a stable system with the rate at which they pass through and the average time they spend there. In plain terms, longer time in the system requires more items within it if the same flow is to continue.²

Suppose a steady supply stream is one hundred fans per day. Add ten days to the average journey, and the stream contains an additional one thousand fans in transit once it has adjusted to that longer journey. At an assumed cost of $20 each, that is $20,000 more in goods occupying the pipeline. Who must finance it depends on payment and ownership terms.

Those fans may be perfectly good. They still cannot cool a room while they are halfway across the ocean.

Why importing can make a storeroom bigger

Distance can affect stock through more than delay. It can also change the size and timing of orders.

If placing and receiving a shipment involves fixed work, buying very small amounts very often may be expensive. A business may combine demand into larger orders. That can lower the handling or transport cost per unit while raising the average stock it holds between deliveries.

George Alessandria, Joseph Kaboski and Virgiliu Midrigan studied inventories and international trade with this problem in view. Their 2010 research connects delivery lags and economies in transactions with less frequent imports and larger inventories. The result is a reminder that access to a cheaper or more specialized overseas source can require additional goods, and money, to be held while the system operates.³

The best source is therefore not always the one with the lowest unit price at the factory gate.

Imagine two offers for Lena. One supplier charges less per fan but requires a large order and a long wait. The other charges more but delivers smaller amounts quickly. The first may still be better. It depends on sales, storage, finance, reliability and the cost of being wrong about demand. Comparing only the invoice price discards most of the decision.

Suppose the cheaper offer requires Lena to buy a thousand units while she is confident she can sell only six hundred before the season changes. If the remaining four hundred can be sold later at a good price, the bargain may survive. If they need deep discounts or become obsolete, the low unit price can be an expensive achievement.

The reverse mistake is to assume that short lead times justify any premium. A business that sells steadily and has inexpensive storage may gain little from paying for constant small deliveries. The important comparison is between whole workable systems, not between one heroic number on each quote.

There can also be room for a mixture. Lena might use a lower-cost source for predictable demand and a faster source for changes. But the faster source needs actual capacity, and relying on it only when everyone else is desperate can be costly. Flexibility bought in advance differs from flexibility begged for after the queue has formed.

These choices connect the purchasing desk to the warehouse. If the buyer receives credit for a discount and the warehouse receives blame for the extra stock, the organization has divided one decision into two misleading stories. A saving that exists only because another department carries the cost needs to be recalculated.

What the accounts do and do not say

Inventory is often described as waste on paper and protection in practice. The phrase catches a genuine tension but can misstate the accounting.

Under IAS 2, the international accounting standard on inventories, inventory is measured at the lower of cost and net realizable value. When it is sold, its carrying amount is recognized as an expense with the related revenue. Write-downs and losses have their own treatment. Stock is not automatically treated as a current expense merely because it is sitting on a shelf.⁴

In simpler language, useful stock can appear as an asset. If its expected recoverable value falls, the accounts must address that decline. The standard does not decree that every full warehouse is a failure or that every empty one is wise.

The pressure to reduce stock often comes from other concerns: cash tied up, space used, damage, theft, obsolescence and the wish to earn more from a given amount of capital. These are real concerns. They should not be disguised as a law of nature requiring the smallest possible buffer.

Suppose a business removes $50,000 of stock. It may free cash that can pay down debt, improve equipment or keep the business alive. That is a genuine benefit. If the same reduction leaves it unable to cope with a common delivery delay, the decision has also changed risk. A complete evaluation must include both.

Now suppose a business holds $50,000 of parts that no longer fit any machine it owns. Calling them resilience will not make them useful. The stock may need to be sold, repurposed or written down. Preparedness is not a magic word that protects bad purchasing from inspection.

This is why I dislike the argument that an unused reserve was obviously unnecessary. A fire extinguisher can go unused because there was no fire. That does not tell us, by itself, whether keeping it was a mistake. Equally, the existence of fires does not prove that a room full of expired extinguishers is sound preparation.

The decision must be judged with the information and alternatives available before the outcome was known. What was the reserve meant to cover? How likely or harmful was the shortage? Could a cheaper method provide similar protection? Was the reserve maintained? What changed after it was purchased?

Hindsight tends to make both thrift and preparation look simpler than they were. In a quiet year, the buyer who cut stock can appear brilliant. After a disruption, the buyer who kept extra can appear prophetic. Neither result alone tells us who reasoned better.

We should reward the quality of the decision and the learning that follows, not only whether the next event happened to flatter it.

A lean system needs a real world

Just-in-time production is often reduced to the instruction keep nothing. Toyota’s own account describes making and delivering what is needed, when it is needed, in the amount needed. It places that approach within a broader production system that also emphasizes detecting problems and stopping to address them.⁵

There is a difference between a carefully coordinated flow and simply refusing to pay for a buffer.

Imagine a factory that removes stock because it has reduced defects, made replenishment more reliable and learned to change production smoothly. It may need fewer parts waiting between steps because the process works better. The lower stock is evidence of an improvement elsewhere.

Now imagine another factory that removes the same quantity because a manager has announced a target, while leaving unreliable suppliers, long repair times and poor records untouched. The shelves may look equally lean in a photograph. The operating conditions are not equal.

A small buffer makes problems visible sooner. That can be useful if people have the skill, authority and time to fix them. If they are instead punished for exposing problems, the organization may get better at hiding them. A quality fault becomes a rushed shipment. A maintenance need becomes an extra shift. A supplier’s warning becomes a more optimistic date on the buyer’s screen.

The issue is not whether efficiency is good. Wasting material, effort and space helps nobody. The issue is whether the apparent efficiency depends on pushing uncertainty into places the performance measure does not examine.

A factory can have very little stock because its supplier holds a great deal. A retailer can have a short cash cycle because suppliers wait a long time to be paid. A customer can receive fast service because workers have unstable hours. Some of these arrangements may be agreed and fairly compensated. We still need to know where the burden went before celebrating its disappearance.

The whole network matters. Reducing inventory at one firm can lower total waste, leave total stock unchanged or increase it elsewhere. The outcome depends on how others respond. A local target is not a complete account of the system.

This is also why redundancy should be chosen rather than sprayed over every activity. A duplicate machine may be worthwhile at one crucial step. Elsewhere, reliable maintenance and a common spare part may provide better protection. A second warehouse might reduce exposure to one hazard while adding transfer work and record errors. A shorter supplier lead time could reduce the need for a large stock reserve.

A serious design compares these possibilities. It does not begin by declaring either emptiness or abundance virtuous.

The spare that has to stay ready

A spare part can fail while waiting.

Materials age. Batteries lose charge. Metal corrodes under the wrong conditions. Packaging can be damaged. A part may be in excellent physical condition but no longer match the equipment after a design change. The person who knows where it is may have left, taking an unwritten map with them.

For this reason, a reserve has a life of its own. It needs records, suitable storage, inspection and a way to keep it aligned with what it protects. Some stock can be rotated through ordinary use so the oldest usable items leave first. Other parts may remain dormant for years and need periodic checks. The correct practice depends on the object; there is no single shelf-life rule for everything in a storeroom.

Consider a water pump serving a small business. The owner has bought a replacement seal because a failed seal could stop the pump. That is a sensible beginning. But does the seal fit this model? Is the material suitable? Is the required tool available? Can somebody safely carry out the repair? Has the pump been modified since the spare was ordered?

The stock has protected one dependency. It may have left several others untouched.

Now make the pump part of a public water system. The same questions grow in consequence. A spare part without access to the site, trained staff or reliable power may offer less protection than its purchase announcement suggests. Residents do not receive water from the procurement receipt. They receive it when the whole service works.

This is where preparedness can become politically awkward. Buying a visible object is easier to announce than maintaining a modest routine. The replacement unit can be photographed. The year’s worth of checks, training and record corrections cannot be gathered into one satisfying ribbon-cutting ceremony.

Yet the routine may determine whether the purchase was worth making.

A good reserve budget must therefore include the cost of keeping the reserve usable. If every available dollar is spent on buying stock and none remains to store or inspect it, the headline quantity can become an expensive fiction. The same principle applies to emergency equipment, backup vehicles and supposedly ready production capacity.

Readiness is an activity that continues after acquisition.

When the inspection changes the plan

In Giant Hubs, a documentary about Rotterdam uploaded by Free Documentary in December 2019, a diver inspects the area around a ship’s propeller. The visit is expected to fit around the vessel’s work in port. The inspection reveals a leak and damage that cannot be resolved by the hoped-for quick intervention. The ship needs docking for repair.⁶

The diver has done his job and the original plan has failed.

That combination is worth holding in mind. A good inspection can make a schedule worse by revealing that the schedule was built on a false assumption. The earlier departure was never safely available under the newly discovered conditions. Continuing to treat it as the proper target would mean asking the evidence to get out of the way.

The film’s scene also shows the limit of preparation by parts alone. Before inspection, the team does not know the full extent of the problem. A warehouse cannot guarantee the right repair if the diagnosis is incomplete. Tools, parts and skill become useful in relation to an actual fault.

We do not need the narrator’s dramatic language or a precise leak rate to understand the reversal. Finding the problem has value even when the immediate result is delay. The alternative is not necessarily an on-time departure with no cost. It may be an undetected defect carried into the next voyage.

This is a broader lesson for supply chains. A quality hold, a damaged-container report or a failed test can look like the event that caused a shortage. Sometimes it is the event that made an existing problem visible. Blaming the person who reports it confuses discovery with creation.

There are still questions to ask. Could the fault have been detected earlier? Was maintenance adequate? Were the right inspection tools available? Was a reasonable repair option missed? A safety explanation should be examined, not used as a blanket excuse. But the examination must begin with the possibility that stopping was the competent act.

A business that wants usable reserves needs this kind of honesty. If expired, damaged or incompatible stock is quietly counted as available because reporting the truth would create trouble, the apparent buffer exists only in the report. The shortage has already begun; the spreadsheet is merely late to hear about it.

Who gets the last box

Inventory does not decide who may use it. People do.

Suppose Lena has fifty fans left and knows that the next shipment is late. One customer wants twenty for a resale opportunity. Several ordinary customers want one each. A local care organization asks whether any can be reserved for rooms that are becoming uncomfortably hot. The stock count gives Lena a quantity. It does not give her a policy.

Different allocation rules serve different aims. First come, first served is simple but rewards the person who learns earliest or can get to the shop fastest. A purchase limit can spread access but may be hard to enforce and may inconvenience legitimate bulk buyers. A reservation for a priority use can protect people with greater need, but someone must define and verify that need. Raising the price can reduce demand and reward supply, while excluding people with less money.

No rule makes the shortage painless. A rule chosen in advance can at least make the decision less arbitrary and easier to explain.

For ordinary consumer goods, a shop has substantial room to make commercial choices within applicable rules. For essential public supplies, allocation can carry much greater duties. A hospital, water utility or disaster agency cannot sensibly treat every request as equivalent to every other request. The purpose of the reserve has to guide access to it.

The important point here is that more stock and fairer access are related but separate goals. A country can have adequate supplies in total while some communities cannot obtain them. A warehouse can be full while distribution fails. A reserve can protect an organization and leave its workers or neighbors exposed.

Even private ownership of stock does not make the wider effects disappear. If every buyer reacts to a warning by demanding several times its usual quantity, available goods can move toward those with cash and storage space. People who cannot afford to stock up then face a thinner market. Some buying in advance is reasonable preparation. The combined result can still intensify the immediate shortage.

That is one reason public communication during disruption matters. A vague assurance that everything is fine can destroy trust when shelves empty. An alarm without a plan can encourage panic buying. Useful information says what is affected, what is available, what people can reasonably expect, and how scarce supplies will be handled.

A person asked to be patient deserves more than a slogan about resilience.

The cost of having enough

Let us return to Lena’s seventy extra fans. They occupy space and tie up money. If she borrows to hold them, there may be interest. They can be damaged or stolen. A new model can make them harder to sell. These costs do not vanish because the stock prevented a shortage once.

On the other side are the costs of being unable to supply customers. Some are lost sales. Others are less easy to count: damaged trust, an urgent purchase at a much higher price, wasted staff time, or the discomfort of people who needed the product. For an essential spare part, the consequences can include a larger service interruption far beyond the item’s own selling price.

A sensible decision compares these effects over the disruptions the reserve is meant to address. It also asks whether another action would work better. Could Lena obtain a more reliable lead time? Reduce the number of models? Improve records? Arrange a transfer from another shop? Could the supplier hold a shared pool of parts that serves several buyers more efficiently than many separate reserves?

Pooling can save stock when people’s needs do not peak together. It can disappoint when everybody needs the same thing at once or when transport from the central store fails. A shared reserve needs an allocation agreement as much as a warehouse. Otherwise, the first disruption becomes a dispute about whose protection was purchased.

A local reserve can be reached quickly but may be exposed to the same flood or fire as the facility it protects. A distant reserve can escape that event but take longer to arrive. The right balance depends on the hazard and the service. There is no virtue in putting the backup generator in the same room that is expected to flood merely because the inventory system prefers one address.

All these choices can sound like a long argument for spending more. Sometimes the better design costs less. Correcting an inaccurate stock record can prevent both unnecessary orders and surprise shortages. Using a common approved part can reduce the number of different spares required. Better maintenance can lower the frequency of urgent replacements. A reliable supplier may permit a smaller buffer than an unreliable one.

The aim is to buy useful time, not to buy objects for the pleasure of counting them.

And time bought should be used. A seven-day buffer is an opportunity to respond, not permission to postpone every decision for six days. If the disruption will last longer, the reserve gives people room to arrange alternatives, repair equipment, change production or allocate supplies carefully. Consuming it without acting simply moves the crisis one week forward.

This is the difference between a reserve and a delay in admitting the problem.

At the end of the day, the warehouse manager has counted the fans that can actually be sold, separated the damaged units, checked the incoming order and marked the point at which the current plan will no longer work. The room contains fewer reassuring assumptions and the same number of physical cartons.

That is progress.

The decision may be to keep the extra stock. It may be to reduce it while improving replenishment. It may be to reserve a different part, in a different place, because that is where a short interruption would cause the greatest harm. Whatever the decision, enough must refer to a service, a period and a set of conditions.

Enough until Tuesday is a practical promise. It becomes a responsible one when somebody also knows what to do if Tuesday arrives without the truck.

Sources and notes