Schema einer Wertschöpfungskette: von Rohstoffen über Vorprodukte, Fertigung und Vertrieb bis zum Kundendienst, darunter der KI-Stapel aus Chips, Rechenzentren, Sprachmodellen und Apps mit Hinweis auf die Gewinnspanne je Stufe.

Value Chain

The value chain describes all the work steps that turn raw materials and ideas into a product ready for sale. At every step, value is added — and each step can be carried out by a different company.

No product is created in a single work step. A T-shirt needs cotton, then yarn, then fabric, then a sewing shop, then a retailer. At each station, the goods become somewhat more valuable than before. This series of stations is called a value chain. The term helps answer a question: who actually earns how much along this chain? And usually the stations don’t belong to one company, but to many different ones.

Who ends up earning the money

Value is distributed very unevenly across the chain. For the T-shirt, the sewing shop often keeps only a small part of the sale price. Considerably more stays with the brand and with retail. Anyone who wants to understand a chain must therefore ask which station is scarce. It is exactly there that high prices can be enforced.

This is especially visible in tech. The chain behind a smartphone includes chip design, chip manufacturing, displays, assembly, and distribution. Assembly is a business with thin profit margins, because many companies can do it. Only a few plants worldwide can manufacture chips using the most advanced process. This scarcity is the reason for the high profits there.

For investors, this is an important tool. If demand for a product rises, not all participants benefit equally. That’s why stock market news often features companies whose names nobody knows. They sit at a narrow point in the chain and supply a component that can hardly be replaced.

From the raw material mine to customer service

A chain is divided into stages. At the beginning are raw materials and intermediate products, called upstream stages. Then come manufacturing and refinement. At the end are sales, delivery, and customer service, the downstream stages. Experts also speak of upstream for the beginning and downstream for the end.

Value creation is not just physical labor. Research, design, software, and advertising also increase the value of a product. This is obvious with apps, since there is almost no material involved. The chain then consists of data centers, programming work, testing, and distribution via app stores. The basic idea remains the same.

It is important to distinguish this from the supply chain. The supply chain mainly describes how materials and goods physically move from A to B. The value chain additionally asks where value is created and where profit ends up. Another common misconception is imagining a straight line. In practice, it’s more of a network in which parts from many countries come together.

The AI chain in economic news

When it comes to artificial intelligence, the term is used constantly. At the very bottom are specialized chips for computation, plus power and cooling. Built on top of that are data centers that rent out computing power. Above them sit companies that train large language models, i.e. programs that learn to formulate answers from huge amounts of text. At the very top are apps and services that make these models usable for customers.

This breakdown explains many headlines. When chipmaker Nvidia reports record numbers, it is earning at a narrow point far down the chain. Whether the companies further up are making money with their apps is an entirely different question. Analysts argue precisely about this: whether the value stays permanently at the bottom or migrates upward.

The chain also plays a role politically. Export bans on certain chips target a single stage in order to slow down the entire chain. And laws like the Supply Chain Act require companies to know under what conditions their intermediate products are made. Anyone who knows the term reads such news much more precisely.

Subscribe free. Unsubscribe the second it sucks.

High-signal news across AI, business, UX, and tech. Every morning.