Before the Bell Rings

The message says your fan will arrive between two and four.

For the next few hours, the world has a small, reasonable duty. Someone must bring a box to your door. You have paid. There is a photograph of the fan on the screen. The blades are white, the base looks sturdy, and the description promises three speeds, though at this point you would accept one speed if it arrived soon. The room is hot.

At three fifteen, the doorbell rings. A person hands you the box, takes a photograph or asks for a signature, and leaves. You pull apart the tape. A few pieces of packing material attempt to establish permanent residence in the room. You fit the guard, set the fan on the table and press a button.

Air moves.

It is possible to go through the rest of the day without wondering how many people had to agree, prepare, lend, measure, teach, inspect and wait for that small improvement in your life. In fact, this is part of what you bought. You bought a fan instead of accepting a second job as the coordinator of its manufacture.

There is nothing shameful about wanting the room to be cooler. A person should not need a degree in world trade to choose a household appliance. But the ease of the last step can teach us a misleading lesson about all the steps before it. We begin to think that because the order took seconds, the thing itself was only seconds away from being possible.

The factory did not start digging for copper when you pressed Buy.

For the fan in this book, we will use a simple design: a plastic body and blades, a steel guard, an electric motor with winding wire, bearings, a power cord and a switch. It has no need to contain every technology in the news. An ordinary motor is already enough to reveal how much of the world can be present in one room.

Months before your purchase, a buyer had to decide how many fans people might want. Someone reserved production time. Someone else ordered parts. Materials were shaped, joined and tested. Packed fans left a factory in a larger shipment. Some of them waited in a warehouse near enough to make a two-hour delivery window look natural.

Your order did set work in motion. It helped determine which box left which shelf and where a driver went. It might also contribute to the next forecast. Yet much of the physical effort that made quick delivery possible had already happened. The seller was making a bet about you before knowing your name.

The customer who had not yet arrived

Imagine that a small retailer, whose buyer we will call Lena, plans to sell these fans next summer. She has records from last year. She knows which models sold out and which were still sitting under a discount sign when the weather cooled. She can ask suppliers about prices, ask the sales staff what customers disliked, and watch orders already coming in.

She cannot ask next summer to fill out a form.

Suppose Lena orders too few. Customers may leave for another shop. Some will make do with an older fan. Others will wait through uncomfortable nights. If she orders too many, she has money tied up in goods that need space and may need a price cut. The cartons do not become more charming with age.

Ordering later would give her better information about demand. It would also give the manufacturer less time to assemble, pack and ship. Ordering earlier buys time by accepting uncertainty. This is one of the basic exchanges hidden inside a full shelf.

There are different ways to arrange that exchange. A made-to-order product can wait for a named buyer before much of the work begins. Another business may keep standard parts ready and finish the item after an order comes in. A retailer can hold finished stock nearby. These choices place the point of commitment in different parts of the journey. They also put different limits on what a customer can change and how soon the result can arrive.

If you want a fan tomorrow, it helps if somebody chose its color last month.

Now suppose Lena would like a special blue model. The supplier can offer it, but only if she orders enough to justify a separate production run. Paint or plastic color is only one consideration. There may be a different box to print, a new product code, separate stock to count, and a production line to clean or adjust before the next batch. A small aesthetic choice spreads into work that the buyer never sees.

This is not an argument against choice. It is a way to see its cost. Ten colors do not always require ten factories, but they can require ten judgments about how much of each version to make. A shop can have plenty of fans and still be out of the one that customers want. Counting total units will not reveal that problem.

Even the promise of quick delivery depends on the shape of demand. If requests are spread across a month, a modest warehouse crew may handle them well. If everyone wants the same model on the same afternoon, the same monthly total becomes a very different job. Averages can describe the season while hiding the Tuesday on which the system fails.

Lena’s decision also reaches backward. If she places a large order, her supplier may book motor production. The motor maker may order wire. The wire producer has other buyers, its own equipment and its own doubts about the months ahead. Each person is reading a different part of the picture. Your wish for moving air eventually reaches a person whose immediate concern is whether a machine can finish a batch before maintenance begins.

No single person has to understand the entire world for the fan to work. That is one of the achievements of trade. People can learn a part of the task deeply, exchange what they make, and rely on others to perform different parts. Yet specialization creates a duty that is easy to lose sight of: somebody must understand enough about the connections to know when one promise threatens another.

The customer hears two to four. Lena hears a delivery date from her supplier. The supplier hears a date from the motor maker. These dates can all look firm on separate screens while resting on the same unresolved problem.

A country name is a short answer

Turn the box over. Somewhere it may tell you where the product was made.

That label has a purpose, including legal rules about origin. It is not a full list of the places that helped make the fan. The country where final assembly happened may have supplied some parts and imported others. The machines that made those parts may have arrived from elsewhere. Product design, testing, software, insurance and transport can involve still more places.

The World Trade Organization explains why this matters for the way trade is measured. Goods can pass through several economies as they are designed, made and assembled. Assigning the whole export value to the last exporting economy can obscure the value contributed earlier. Measures of trade in value added try to make those contributions clearer.¹

We can see the basic problem without a world database. Suppose, in a deliberately small example, a producer buys imported parts for $12 and sells a finished item abroad for $20. The export record can show a $20 item crossing the border. It would be a mistake to conclude that all $20 was newly created in the final exporting country. It would also be a mistake to treat the remaining $8 as the assembler’s profit. That difference may have to cover wages, local services, rent and many other costs.

The example leaves out taxes, exchange-rate changes and several accounting details. Its purpose is to correct a habit of vision. A border can record the full price of an object without telling us how that price was divided among the people who made it possible.

This becomes important when arguments about trade grow heated. A large import figure can sound like money handed intact to a foreign factory owner. A low factory price can sound like proof that almost no work went into the object. Neither conclusion follows. We need to know what is being measured, which transactions are included, and whose income remains after whose bill has been paid.

We also need to resist the opposite romance: the idea that because many countries contribute, everyone gains fairly. Cooperation can create a good thing and still divide the rewards badly. The fan may be affordable partly because production is well organized. It may also pass through firms where workers have little bargaining power. Those possibilities require different evidence and different remedies. A crowded production map does not settle the moral question.

For now, look at the small word made. It contains a great deal of unfinished explanation.

The journey on the screen

A tracking page offers a story with admirable discipline. Order received. Packed. Dispatched. Out for delivery. Delivered.

There are no lengthy arguments about working capital. Nobody interrupts to explain a bearing. The story begins at the point where the seller has something useful to tell you, and it ends at your door.

That is a good design for a customer service. Trouble begins when we mistake it for a complete account of the product’s life.

A scan is a record of an event. It can be very useful: a parcel was accepted, a pallet entered a warehouse, a container was loaded. But the kind of event matters. A shipping label created at a desk does not itself put a carton inside a truck. A planned departure is different from an actual departure. A last observed location is different from a continuously observed position.

GS1’s standards for sharing supply-chain events make these distinctions concrete. They give businesses a common way to record what was involved, where and when an event happened, and its business context. They also distinguish the time of an event from the time the record enters a data system. Common identifiers help connect a product with a box, a pallet or another larger unit that contains it.²

A standard is a way for records to make sense to one another. It cannot make an unrecorded event appear, force every firm to share, or turn a mistaken entry into the truth.

Imagine that the warehouse records a pallet as ready at nine. A broken loading bay delays the truck until noon. If the screen displays the planned movement as though it has already happened, the customer will inherit three hours of false confidence. If the delay is reported honestly, the picture looks worse while becoming more useful.

This is why a beautiful dashboard can be a poor comfort. The bright line across the map may tell you where a ship is. It may tell you little about whether the goods aboard are yours, whether the correct documents are ready, whether unloading capacity is available, or whether a road closure will stop the next leg. Each piece of knowledge can be valuable. None should quietly impersonate the others.

There is another difference: knowing where something is does not grant the power to move it.

Lena may see that her shipment has stopped at a port. That is better than guessing. But she cannot create a free berth with a more detailed map. She cannot order a customs officer to ignore a problem. She cannot make a driver appear by refreshing the page.

Visibility helps when it changes a decision. An early warning may let her tell customers the truth, send stock from another branch, postpone an advertisement, or choose a different route for the next shipment. If the information arrives after every option has closed, it may still help explain the failure. Explanation and prevention are different services.

The people doing the work also need information that the customer never sees. Which cartons can be stacked? Which load needs care? Which product code was replaced? Is a shipment delayed, or has the customer canceled it? A perfectly readable consumer tracking page can coexist with badly connected business records behind it.

The right ambition is therefore larger than putting a dot on a map. It is to give the person who can still act a trustworthy account of what has happened, what has not, and what remains uncertain.

Paper can stop steel

The fan is a physical object. Its movement depends on nonphysical permissions.

A truck can reach a gate while a shipment remains unable to leave. An invoice may use a description that does not match the other records. The quantity may be wrong. A required safety or origin document may be missing. Someone may need to establish who is entitled to collect the goods. The problem is then neither a lack of roads nor a shortage of horsepower.

It is tempting to call all such work bureaucracy, usually while waiting for somebody else to finish it. Some requirements are needlessly complex. Others help stop unsafe products, prevent fraud, collect public revenue or keep pests from arriving with cargo. We should judge a rule by what it does and how well it does it, rather than by whether it uses paper or a screen.

The WTO’s Trade Facilitation Agreement includes provisions for processing documents before goods arrive and for reviewing formalities to reduce unnecessary complexity. The aim is to make movement and clearance faster while retaining legitimate public purposes. An agreement creates duties and procedures; it does not prove that every border now operates smoothly.³

For Lena, the practical lesson is less grand. If a document can be checked while the fans are still at sea, discovering the problem then is better than discovering it after a truck has been paid to wait. Good preparation can remove a delay without asking a worker to move faster or a safety inspector to look away.

Suppose the shipment contains two models, one for each of two electricity systems. The cartons look similar. If the paperwork identifies them badly, the error can travel farther than the boxes. A warehouse may book the wrong stock, a shop may promise the wrong item, and a return may be treated as a new customer complaint rather than evidence of an earlier mix-up. The first mistake acquires several different names as it moves through the system.

Clear records do more than satisfy a distant official. They let the next person perform the right task.

This also explains why changing one part of a system can take longer than buying a new machine. A common product code, a reliable process for correcting records, or an agreement about who reports a delay may require several businesses to change how they work. Each has other customers. Each fears being blamed for a mistake it did not make. The technology may be available while cooperation remains expensive.

A promise to digitize can therefore mean two very different things. It can mean that the same confused process now produces electronic confusion. Or it can mean that the people involved have agreed what an event means, who records it, and how the record will be corrected. Only the second is a serious improvement.

The money has its own journey

Before the fan can cool your room, somebody has to pay for work that has not yet earned a final sale.

The factory buys materials. Workers need wages. Electricity must be paid for. The carrier expects payment under agreed terms. Meanwhile, the retailer may not receive money from you until much later. Different firms can borrow, use their own cash, request deposits, or wait for their customers to pay. Someone must cover the interval.

Working capital is a useful name for the resources tied up in keeping that ordinary cycle going. It sounds like a topic for people who enjoy spreadsheets. It is also a topic for the worker whose payday comes before the retailer’s next busy weekend.

Consider another simple example. A supplier spends $8,000 to complete an order and expects to receive $10,000 after delivery. Leave other costs aside for a moment. The order may offer a positive margin. That does not mean the supplier has the $8,000 today. A profitable promise can arrive at an unhelpful time.

Now suppose payment is due sixty days after delivery. The buyer is pleased: it has time to sell some fans before paying for them. The supplier must finance a longer wait. If the buyer then asks for another month, the terms have moved a cost backward along the chain even if the price per fan stays the same.

The buyer’s improved cash position is real. So is the supplier’s new problem.

Trade-finance arrangements divide such risks in different ways. Under payment in advance, the buyer gives up money before receiving the goods. Under open-account terms, the seller ships before getting paid. A letter of credit can add a bank’s payment commitment when the specified conditions and documents are satisfied. These methods have different costs and requirements; a bank’s involvement does not itself inspect the fan or erase every commercial risk.⁴

The point is not that every small buyer needs the most elaborate instrument. It is that payment terms are part of the product’s journey. They determine who must trust whom, who needs cash first, and who bears the cost if the calendar changes.

The Asian Development Bank’s 2023 survey estimated $2.5 trillion in unmet demand for trade finance in 2022. That figure describes a survey-based estimate of financing requests not met, rather than a count of boxes that definitely failed to ship. It is large enough to show why a story about trade cannot be told entirely through ships and roads.⁵

For our supplier, a delay at sea can create several bills at once. The goods are still tied up. The buyer may pay later because delivery happened later. Another production run may need materials before the first one has returned cash. If the supplier has access to affordable finance, it may bridge the gap. If not, a transport problem can become a production problem even when the factory itself is undamaged.

This is one reason I am wary of praise for companies that squeeze every supplier as hard as possible and then announce their devotion to reliable supply. A strong buyer can obtain a low price by pushing risk onto a weaker firm. The low price may be genuine. The reliability claim still has to survive the weaker firm’s bad month.

There is a difference between removing a cost and persuading somebody else to carry it until they collapse.

We should not assume that every late payment proves abuse, or that every supplier is well run. Disputes, errors and poor performance happen. Yet a buyer who wants dependable partners has an interest in understanding whether its terms make dependable performance possible. The financial health of the factory is not unrelated to the delivery date merely because it appears in a different department’s report.

The last mile and the years before it

The phrase last mile directs our attention to the end of the trip. Here the work becomes personal. A driver has to find a building, deal with traffic, carry the parcel and meet the person waiting for it. The final handover is easy to judge. Either the box arrives or it does not.

Earlier work is harder to see. Nobody sends you a cheerful notification when the motor maker maintains a machine well enough to avoid a shutdown. The technician who catches a defect before a batch leaves does not appear in the delivery photograph. You will probably never learn the name of the person who noticed that two records disagreed and corrected them before they caused a week of delay.

This creates an odd imbalance in our gratitude and our anger. The visible person inherits credit or blame for a long chain of decisions. Sometimes that person has made the mistake. Often the cause lies much farther back, where a forecast was ignored, a shipment was booked too late, or an impossible promise was offered because it looked good beside the Buy button.

The customer still deserves an honest answer. Explaining complexity should not become a way to avoid responsibility. A seller cannot demand trust for a delivery promise and then declare that the people behind the promise are unknowable whenever something goes wrong. The seller’s job includes deciding which commitments it can reasonably make and how it will respond when they fail.

But responsibility becomes more useful when it follows the actual cause. If a driver lacks enough time for the route, scolding the driver will not add hours to the day. If a warehouse has the wrong product records, another reminder to be careful will not repair the database. If a buyer repeatedly changes orders after materials are bought, the supplier’s difficulty may begin in the buyer’s own office.

The world that produces the fan is neither a flawless machine nor an inexplicable miracle. It is a large collection of people making judgments under limits. Many of those judgments are good. The room becomes more comfortable because people have learned to do difficult things together, repeatedly, without knowing the person who will enjoy the result.

That achievement deserves more than the discovery that supply chains can break. Of course they can break. A household meal can be delayed by a missing spoon, and we are discussing a system that joins factories, currencies, ports and laws. The more interesting questions concern why it works so often, which risks make it fail, and how to improve it without throwing away what makes it useful.

Start with the dates. Who committed before you did? Start with the records. What happened, and what has merely been planned? Start with the money. Who pays while everybody else waits?

Then ask Lena whom she buys from.

She can open a file and give you a name. The supplier is established, the contact answers messages, and the last shipment arrived on time. That is useful knowledge. It may also be the edge of what her business knows.

The supplier buys its motor from someone else.

Sources and notes