Ask most people to describe a housing development and they will describe a building. Ask them to describe a factory and they will describe a different thing entirely, in a different industry, run by different people, financed on different terms.
That separation is a convention. It is not a fact about the world.
A modern building is an assembly of manufactured components. How fast it goes up, what it costs, and whether it is finished on the date somebody promised are determined less by what happens on the site than by what happened in the factories that supplied it. Construction depends on manufacturing efficiency.
Manufacturing, in turn, depends on technology. Not in the brochure sense — in the sense that a production line without accurate demand signals, without process automation, without instrumentation telling you where the waste is, runs at a fraction of what it could. And technology depends on operational execution, because software installed into a business that has no process to attach it to changes nothing at all. Most failed digital transformations are that sentence, written expensively.
Three industries, one chain
So the chain runs: technology enables manufacturing, manufacturing enables construction, construction delivers the asset. Each link constrains the next. And each link is owned, financed and operated by people who mostly do not talk to the people either side of them.
That is where the inefficiency lives. Not inside any one of the three — each is competitive and each has spent decades optimising itself — but in the seams between them, where nobody has the standing to make a decision.
A developer cannot tell a factory to change its tolerances. A factory cannot tell a technology vendor what to build. A software company has no reason to care whether a building opens in March or September. Every party is behaving rationally, and the aggregate is slow, expensive and unpredictable.
Why the seams stay open
The obvious question is why, if this is visible, it has not been solved.
Partly because it is genuinely hard. Operating across three industries means being competent in three sets of regulation, three labour markets, three financing structures and three cultures. Most people who are excellent at one are actively bad at the others, and the failure mode of trying is being mediocre at all three.
Partly because the returns arrive late. Integrating a value chain does not pay off in the first quarter or the first year. It pays off when the second project runs on the systems the first one built, and the third runs on the systems the second corrected. That requires patience that most capital structures do not permit.
And partly because complexity puts people off. A business that requires a diagram to explain is harder to raise money for than one that fits in a sentence. That is a real cost — and it is also why the opportunity is still there. Complexity is a barrier, and barriers protect whoever is willing to cross them.
What being positioned across it actually looks like
It does not mean owning everything. Vertical integration for its own sake produces a conglomerate, and conglomerates are usually worth less than the sum of their parts for good reasons.
It means being close enough to each link to see where the next one is being starved. Knowing that the schedule slipping in month nine was caused by a specification decision in month two. Knowing which manufacturing tolerance is the one that matters to the installer. Knowing whether the software being proposed solves a problem the operation actually has.
Those are not strategic insights. They are operational ones, and they are only available to someone who has stood in all three places.
The businesses that will be difficult to compete with over the next decade are the ones that can move a decision across those boundaries in a day instead of a quarter. That is not a technology advantage or a capital advantage. It is a structural one, and structural advantages are the only kind that survive a cycle.