How to Own a Commodity

Vertical integration is a very under-explored modality of technological progress - Peter Thiel

Everyone says not to start a commodity business. But the truth is, the best businesses are commodity businesses that really own the commodity. The trick is to build a structural cost advantage in production, or differentiate the product enough for customers to no longer view it as a commodity.

Why build a vertically integrated business

The magic of vertical integration is that it takes something perfectly fungible — aluminum, oil, electrons — and makes it yours. That’s how you turn a commodity into an empire.

The prize is clear: a large, durable business with 5+ of Hamilton Helmer’s ‘7 Powers” of businesses with a long term competitive advantage.

✅ Scale Economies

✅ Cornered Resource

✅ Network Economies

✅ Process Power

✅ Branding

🤔 Counter-Positioning (maybe to get there)

🤔 Switching Costs (if your customers bought infrastructure that only works with you)

Opportunities for integration

Opportunities to start an integrated business are a subset of the circumstance that enable any new business.

New technology

Netflix: Streaming was only viable once broadband became cheap.

Alcoa: Aluminum was more expense than gold until the Hall-Héroult smelting process industrialized production.

New demand curve

Amazon Retail: Bezos started Amazon because the internet was growing 2,300% / year

FedEx: Globalization made overnight shipping essential.

Standard Oil: Industrialization, later automobiles.

Certainty (safety, scarcity, reliability)

Standard Oil: Low quality kerosene killed people. Standard was just that.

Meter: WiFi always fails at the worst time. Meter guarantees reliability.

Moderna / mRNA vaccines: First COVID vaccine candidate created in two days.

Distribution / logistics

Coca Cola: Global bottling and distribution enabled the same drink to be served everywhere.

Walmart: global distribution enabled cheap imported goods everywhere

Amazon Retail (internet)

Structurally lower cost basis (control of inputs)

Saudi Aramco: Controls the world’s cheapest oil to produce.

SpaceX: Made their own engines (Elon tried to buy a Russian ICBM first).

Alcoa (owned electrical generation in Tennessee and Quebec Canada)

Regulatory or institutional openings

19th Century Railroads: Granted right-of ways, eminent domain by the government and owned all the land on their routes.

Electric utilities: Traditionally owned generation and delivery to customers.

Half baked integration

Most companies that fail to vertically integrate do so because they can’t reach the scale they need to realize the benefits of integration, don’t integrate the things that accrue the most value, or are in a market that commodifies faster than they can become dominant.

Integration without scale

Not enough scale

GoPro: Tried to launch a a video ecosystem but had too small of a user base.

Rent the Backyard (my previous company): We solved scarcity but priced it like a commodity before reaching scale.

Industry commodifies before you reach scale

SunRun / Solar City: Responsible for solar panels on roofs across America by creating the “solar lease” funded with government incentives. SolarCity tried to go into the solar panel market in 2016 but the product was undifferentiated and Chinese manufacturers were dominant.

Uber: Divested self driving team in 2020. Now, you can book a Waymo via the Uber app in Atlanta and Austin but Waymo has its own app with >25% market share in San Francisco…

Integration in the wrong place

WeWork: Financing innovation turned out to be raising money to burn on riskier than expected leases. Owning buildings wasn’t the bottleneck.

BlueApron: Unit economics were so broken vertical integration of food prep couldn’t help.

Integration with distraction

Juicero: Integrated hardware + proprietary juice packs + logistics. Not enough product differentiation.

Lyft Bike: Doesn’t lead to more Lyft rides. Now Waymo has more market share in San Francisco.

How to integrate

Vertical integration is easy to romanticize or dismiss, but its logic is surgical: it only works when you own a step of the value chain that captures an outsized share of value, or when you change the terms of competition.

Find the bottleneck

Scarce input, distribution channel, reliability guarantee, etc.

Control it

Integrate upstream or downstream to surround that choke point.

Decide how to compete

Option A: Develop a lower cost basis

When everyone gets the same price for a product, the only way to win is to offer your product on a lower cost basis. Economies of scale are most straightforward, but technology or process advantages can get you there too.

Option B: Break the commodity perception

If customers don’t view you as a commodity, you having pricing power. This can mean wrapping your product in a brand that signals trust, engineering reliability guarantees others can’t match, or building an ecosystem that locks in demand.

Follow where value accrues

Early on, this usually means chasing fat margins around scarcity, reliability, or a new technology. As the product commodifies, margins move downstream toward distribution, logistics, and trust, or upstream into scarce inputs and capital intensity.

Owning the commodity

Owning a commodity isn’t about fighting commodification, it’s about bending it to your advantage. The world’s great industrial empires didn’t escape commodity markets; they mastered them by grabbing the choke points where value accumulated, then relentlessly scaling or differentiating themselves.