Complete Series · 3 essays · 2026

Technological Sovereignty

How much control do we actually need over the technology we depend on?

I started with a question about cloud sovereignty: how much control do we actually need?

I was not convinced that sovereignty meant owning every layer of technology. But using the strongest available technology while depending on somebody else raised a different set of questions.

What should remain under our control? What happens when a provider can change the terms? And how do today's sourcing decisions shape the choices we still have tomorrow?

Technological Sovereignty is a three-part exploration of those questions, moving from selective control of technology and information to the leverage created by dependencies, and finally to build, buy and partner decisions in Product strategy.

The argument moved away from ownership as the measure of sovereignty. I became more interested in credible alternatives, the information we cannot easily recreate, and where knowledge and leverage accumulate over time.

Not every dependency is a problem. The question is which dependencies we can afford, what they give us, and what choices they leave open.

Start with Part 1 ↓

The complete exploration

Three essays, from selective control to dependency and leverage to Product strategy

Part 1Selective Control

Rent the Compute. Protect the Context.

The starting question was simple: how much of the technology stack do we really need to control?

This essay argues against treating sovereignty as ownership of every layer. Capacity and capability can often be sourced externally, provided critical dependencies have credible alternatives. But the threshold changes for proprietary context, data and learning that competitors cannot easily recreate.

The distinction is not between foreign and domestic technology. It is between what we can replace and what we would struggle to recover.

Read Part 1 →

Part 2Dependency & Leverage

The Real Risks Behind Technology Sovereignty

If sovereignty is about preserving options, what exactly can a technological dependency take away?

The second essay looks at the leverage a supplier or jurisdiction may gain: the ability to restrict access, change economics, influence which capabilities are available, learn from information we share, or make exit impractical.

It reduces the discussion to a more useful test: If the dependency changed tomorrow, who would have more decisions available to them: us, or them?

Read Part 2 →

Part 3Product Strategy

Build, Buy, Partner... and Then What?

The exploration eventually led me back to a familiar Product decision: build, buy or partner.

Cost, speed and capability matter, but a sourcing choice also changes the organization after it is made. Products grow around providers, knowledge accumulates on one side of a relationship or the other, and switching becomes easier or harder over time.

The final question is not only how to acquire a capability. It is which choices remain available after we have acquired it.

Read Part 3 →

From ownership to strategic choice

I started with a question about cloud sovereignty. Three essays later, I think ownership is still the wrong place to start.

What matters is the consequence of a dependency. Some capabilities are worth renting because external providers can deliver them better or faster. Some information deserves a much higher threshold of control. And some dependencies become strategically significant only after we have built products, processes and expertise around them.

That does not mean we should avoid external technology, build everything ourselves or preserve every possible option. Flexibility has a cost, and commitment can create value.

It means being deliberate about what we control, what we share, where learning accumulates and how much power a dependency gives somebody else.

The question I would now carry into a technology or sourcing decision is not simply whether we own the capability. It is what choices we are preserving, and which ones we are knowingly giving up.