What If Expensive Things Aren’t Actually Expensive?

Post 2: What If Expensive Things Aren’t Actually Expensive?

I recently came across a meme on Facebook and an idea associated with Elon Musk’s engineering philosophy has stuck with me. I am paraphrasing the concept rather than quoting him: if something remains expensive after it has been properly designed and produced at enormous scale, perhaps it really is expensive. But if its cost collapses at scale, much of what looked like intrinsic cost was actually inefficiency in the system. This post explores more in depth this concept.

Suppose a product sells for $10,000. It is tempting to think, “That is a $10,000 object.” But why does it cost $10,000?

Sometimes the answer is justified: engineering, certification, warranty risk, intellectual property, regulation, capital requirements and low production volume can all be expensive.

But sometimes the system itself is expensive.

Scale changes more than manufacturing

Scale affects manufacturing, but also purchasing, shipping, utilization, insurance, data, negotiating power, reliability and coordination.

Imagine 10,000 small businesses each paying $2,000 annually for the same kind of service. Individually, none has much negotiating power. Collectively, they represent $20 million of demand.

An organization aggregating that demand would not have manufactured the underlying service. It would have manufactured scale.

Utilization changes cost

Consider a $500,000 machine used four hours per week. The service it provides may appear extraordinarily expensive. If utilization rises to 60 hours per week, the machine did not become cheaper; its capability became cheaper.

People do not really want drills; they want holes, they want to turn screws. Sure a screwdriver manually can do the job to turn a screw but they want speed and efficiency too. Businesses do not necessarily want servers; they want reliable computing. A company may not fundamentally want a warehouse; it wants protected storage and distribution capability.

Ownership is often simply a mechanism for guaranteeing availability.

Tiny improvements can become enormous

Suppose a corporation performs an activity 400 million times per year and someone reduces the cost by three cents.

400,000,000 × $0.03 = $12,000,000.

A three-cent improvement becomes a $12 million annual improvement.

The inverse matters too

Something individually worthless can become valuable at scale.

One database record may be worth almost nothing. Ten million correctly linked records may answer questions no partial dataset can answer.

This gives us two transformations:

Expensive → cheap through scale

Worthless → valuable through scale

Sometimes the important economic properties do not appear until the pieces are combined.

As we continue through this project, I will use several anecdotes and recall situations from some of my past experiences. This is deliberate and I am attempting to incorporate my knowledge, experiences and past to try to make this project unique and relevant.

Please continue to Post 3 if you’d like to continue following along.