Problems of Scale in E-Commerce

We are starting to see indications that e-commerce may not scale much farther.

Scalability is something we take for granted most of the time, but we know there are limits. We might encourage more people to go to an especially good restaurant, for example, but if twice as many people show up on any given day, the restaurant runs out of tables and people are turned away.

Few people have given much thought to the capacity of e-commerce. Consumers and manufacturers alike are planning on a greater proportion of sales happening through digital channels with ever-lower friction. In less than five years, it is widely assumed, more than half of sales may happen online. This consensus view was formed without examining the question of whether such a thing is possible.

The problem is not in the online selling process itself. There are few reasons to doubt the ability of Internet data centers to handle twice as many online transactions. The strain that is seen now is in the delivery of the tangible merchandise that is being ordered. Size makes a difference for this purpose. There may be no practical limit on the number of books and similar small articles that can be pulled from a warehouse and sent to purchasers’ homes. It is a different story for clothing, appliances, furniture, and groceries; items that weigh multiple kilograms or are too large to fit in a mailbox are harder to deliver, and this is where problems are occurring.

We’ve seen several years in a row of missed Christmas deliveries. Items ordered on Cyber Monday or earlier should, in theory, arrive in plenty of time for Christmas, but instead, 1 to 2 percent of these shipments are lost in the rush and arrive late.

UPS, which carries almost half of these packages in the United States, has no solution ready. Instead, the company is introducing surcharges for many December deliveries. Surcharges make good economic sense. If there is a peak demand for a service that exceeds the capacity of the system, the peak customer ought to pay more. However, e-commerce customers are already paying a premium for delivery. Regardless of how shipping charges are disguised, they are fees that in-store purchasers don’t pay. Higher shipping charges are not a solution, though. It does not take a large surcharge to defeat the advantages of purchasing online.

Home delivery is not necessarily a convenience for the recipient either. Someone may have to be home to receive the shipment. That requirement is an extra appointment that no one really wants. If we really are to order half of our stuff online, then someone will have to be home almost every day to receive those shipments. In most households, that would be impossible.

If home delivery is a problem, then maybe local pickup is the answer. Already you can get e-commerce purchases delivered to a local store. Store delivery is free for many online purchases from the likes of Sears and Home Depot. Free store delivery might seem like a gimmick, but it makes some sense. Most of the costs of delivery occur in the last ten miles to a customer’s home, with a worker driving a truck some distance just to deliver the one package to the one point.

In-store delivery may face problems of scale too. The last time I arrived at my local Home Depot store to pick up an online order, it took workers two minutes to find the package. That two minutes could be ten minutes if they had twice the number of packages, an hour if they had five times as many. Waiting an hour at a counter would be an obvious inconvenience, but still possibly a better answer than home delivery.

There has been a lot of talk about drones as a possible answer for package delivery. Drone delivery is only an experiment at this point, an option for lightweight orders if you live within walking distance of a particular Amazon warehouse. Along the same lines, one pizza delivery restaurant is experimenting with driverless vehicles for pizza delivery. If that works, it could be also be an answer for package delivery. You probably have to be home to receive a robotic delivery, but if the delivery is something you can arrange on short notice after you actually are home, it wouldn’t be such an inconvenience. Then again, maybe someday households will be able to buy robots capable of receiving deliveries from robotic vehicles.

Some of the delivery tactics might be workable in only a few special situations, but they add up. Every solution matters when you are facing a capacity issue. E-commerce volume is expected to increase again this year, and no one should be surprised if the delivery mechanisms already in place aren’t quite enough to deliver all of the physical products. The undelivered or delayed shipments are the first sign that we are approaching the limits of e-commerce capacity. The second sign will be an increase in shipping charges for online shopping. If there is not enough capacity to deliver everything everyone wants, then those who are willing to pay extra will be the ones who can still get things delivered.

Fish Nation Information Station | Rick Aster’s World | Rick Aster