Vertical Integration: Owning the Step That Decides the Product
Every company faces the same recurring question about each step in producing what it sells: do we own this step, or buy it from someone who specialises in it?
For most of the last few decades the default answer was to buy. Specialists achieve scale, competition among them lowers prices, and outsourcing keeps a company focused on whatever it does best. This reasoning built the global supply chains that make modern manufacturing possible.
The opposite move — vertical integration, owning more of your own supply chain — was treated as old-fashioned. It has come back, and the reasons are worth understanding, because the argument for it is not that the earlier reasoning was wrong.
What integration is actually for
Owning a step in your own chain buys two specific things, and it is worth being precise about which one you are after.
Control over what differentiates your product. If the component that determines whether your product is good is bought from a supplier who sells it to everyone, you cannot be meaningfully better than your competitors on that dimension. Owning it means you can make choices nobody else can copy quickly.
Insulation from a fragile supply. Where a component comes from very few sources, depending on them means accepting their pricing, their timelines, and their decisions about who to prioritise in a shortage. Ownership converts an external risk into an internal cost.
These are different motivations and they justify integration in different circumstances. The first is offensive and pays off when the step genuinely determines product quality. The second is defensive and pays off when the supply market is concentrated enough that market discipline has stopped working.
Neither justifies integrating everywhere. A component available from twenty competing suppliers, which no customer notices, is one you should buy — owning it means taking on cost and complexity to control something that does not affect the outcome.
What it costs
The trade is always the same: flexibility for control, and the cost side is systematically underestimated because it arrives later.
Fixed costs replace variable ones. A supplier relationship can be scaled down or ended. A factory cannot. Integration is excellent when volumes are high and stable, and it becomes an anchor when demand falls, because the cost stays whether or not you use the capacity.
You take on a business you may not be good at. Manufacturing, logistics and chip design are deep specialisms. Deciding to own one means competing with organisations that do only that, and doing it at smaller scale.
You lose the benefit of others' scale. A specialist serving the whole industry runs at volumes no single customer can match, which is usually why their costs are lower.
Focus fragments. Every integrated step is one more thing that can go wrong and one more area demanding management attention.
Internal suppliers face no competition. This one is quiet and corrosive. An in-house unit that has no alternative customer and whose customer has no alternative supplier lacks the pressure that keeps external suppliers sharp. Costs drift, and it is difficult to see, because there is no market price to compare against.
Why the boundary moves
The most useful thing to understand about this decision is that the right answer changes over time, in a fairly predictable direction.
Early in an industry, the pieces do not fit together well. Standards do not exist, suppliers are immature, and the interfaces between components are unsettled. Integration wins, because coordinating across company boundaries is harder than coordinating internally.
As the industry matures, interfaces standardise, suppliers specialise, and a modular ecosystem becomes more efficient than any integrated company. Buying wins.
Then something shifts — a new technology, a supply shock, a component becoming the thing customers care about — and integration wins again for that specific piece.
This is why the same industry oscillates between integrated and modular structures over decades, and why arguments about which is correct in general are unresolvable. The question is not which approach is better; it is where this industry currently sits and which specific step is in play.
The practical default that survives this: integrate where the product is genuinely decided, buy where many suppliers compete, never depend on a single external source for anything critical, and re-examine the boundary when the industry has changed enough that the original reasoning no longer holds.
That last point is the one most often skipped. Integration decisions are made once and then treated as permanent, when the conditions that justified them frequently expire — leaving a company owning a factory for something that has since become a commodity anyone can buy.
