Why the next phase of space will be decided by control, integration and capital architecture rather than by hardware alone
Andy Demir | SwissTechnic Perspective | Space · Industrial Strategy · Finance & Capital
Executive Summary
The space economy is still described as an extension of aerospace: rockets, satellites, spacecraft, manufacturing programmes and government procurement. That description is becoming economically misleading. Physical infrastructure remains indispensable, but a growing share of the value that space creates is realised elsewhere: in communications networks, positioning, logistics, defence intelligence, cloud infrastructure, financial services and agriculture, where orbital capability is turned into decisions on Earth.
This changes the strategic question. The decisive distinction is no longer upstream versus downstream, and the answer is not that hardware will inevitably commoditise while software captures everything above it. What matters is who controls enough of the system to convert infrastructure into recurring economic power. A launch company, a constellation operator, a geospatial analytics provider and a cloud platform all participate in the same space economy, yet their economics, bargaining power, capital requirements and strategic value differ fundamentally. The trillion-dollar forecast matters; the architecture beneath it matters far more.
A rocket leaving its pad remains the defining image of the space economy. It is visible, technically formidable and emotionally compelling, compressing billions of dollars of engineering into a few dramatic seconds. Yet the image tells executives remarkably little about where the economics of space are heading.
According to the Space Foundation's estimate, the global space economy reached $613 billion in 2024, up 7.8 per cent, with commercial activity accounting for 78 per cent of the total. The World Economic Forum and McKinsey project that the figure will reach roughly $1.8 trillion by 2035, up from $630 billion in 2023. The composition of that forecast matters more than its size: much of the additional value is expected to arise in supply chains, transportation, food, defence, retail and digital communications rather than in the manufacture of rockets and satellites.
This is where the conventional classification begins to fail. Space is no longer simply an aerospace market expanding into orbit; it is becoming an economic control layer embedded inside other industries. The distinction matters because industries are defined not only by the technology they manufacture but by where customers pay, where margins accumulate, where switching costs form and who owns the relationship with the end user. A satellite may be designed and assembled according to an aerospace industrial model, while the value it produces surfaces in an insurance model, a military intelligence workflow, a container-routing system, an autonomous vehicle or a mobile network. The machine may be aerospace. The market, increasingly, is not.
Measuring the infrastructure is not measuring the economy
The McKinsey and World Economic Forum work draws a useful distinction between the space economy's "backbone" and its "reach". The backbone comprises the technologies traditionally associated with space: satellites, launch systems, ground equipment and the services immediately around them. The reach consists of economic activity elsewhere that depends on those capabilities. In 2023, the two were already of similar scale, roughly $330 billion for the backbone and about $300 billion for reach applications, and that proportion says more than the headline forecast does.
It suggests that space is starting to behave less like a vertically bounded industrial sector and more like electricity, telecommunications, cloud computing or semiconductors: a foundational capability whose economic importance eventually far exceeds the revenue of the companies that build the underlying infrastructure. A semiconductor fab and Uber do not belong to the same industry, yet without chips, GPS and mobile networks, Uber's business model would not exist. The OECD reaches a similar conclusion from a policy perspective, finding that space-based systems now support more than half of the most critical infrastructure and services across OECD countries, including transport, energy, communications, and food supply.
That is why the expected growth of the space economy should not be read as evidence that every part of the traditional value chain will benefit proportionately. Economic expansion and economic capture are different things. The satellite manufacturer may enable the value; the company that embeds satellite-derived information into a customer's daily workflow may capture more of it.
The wrong debate is upstream versus downstream
A popular investment thesis holds that the upstream eventually commoditises while downstream software captures the margin. There is truth in it, but it is easily taken too far, because space infrastructure is not a server rack. Orbital position, spectrum, sensor capability, launch access, radiation exposure, debris risk, national licensing and security classification impose physical and legal limits on substitution that have no equivalent in cloud computing. A synthetic-aperture radar constellation cannot be swapped for whatever satellite happens to be available, as compute capacity can be shifted between data centres.
The more useful distinction, in my view, is between controlled architecture and dependent architecture. A business in an apparently attractive downstream position may still be strategically weak if its primary data source, distribution channel or government customer is controlled by someone else. Conversely, an infrastructure business can remain enormously powerful when it controls scarce capacity, customer access or the ability to integrate adjacent layers.
SpaceX is the obvious illustration. The lesson of Starlink is not that communications services are inherently superior to launch; it is that the company participates across manufacturing, launch, constellation deployment, network operation and customer delivery, so the value accrues to the system as a whole. Rocket Lab is moving in the same direction. It generated $602 million of revenue in 2025, up 38 per cent, with Space Systems contributing about $403 million, or two-thirds of the total, and it ended the year with a backlog of $1.85 billion, up 73 per cent. The company is no longer intelligible as a launch business, and it is not alone: the strongest players increasingly refuse to stay inside a single conventional category.
Space is becoming part of the data economy
A satellite does not create commercial value merely because it produces an image, a signal or a measurement. Value emerges when that output changes a decision. An Earth-observation image becomes economically significant when it alters an insurer's view of catastrophe exposure, flags an agricultural problem before crops are lost, reveals military movement or shifts a commodity trader's view of inventory. Navigation becomes valuable when it disappears inside transport, logistics, aviation and mobile applications; satellite communications become strategic when they form part of military command architecture, maritime connectivity, disaster response or ordinary consumer broadband.
This changes the unit of competition. Companies no longer compete only on resolution, payload performance, launch price or satellite lifetime, but on how quickly and reliably orbital capability enters an operational workflow. A company that sells raw data may therefore be less defensible than one whose output has become difficult for the customer to remove. That is why APIs, analytics, workflow integration, AI models, secure networks and customer-specific decision systems are becoming part of the moat.
The implications of today's enthusiasm for downstream space software are uncomfortable. If the analytical layer becomes attractive enough, infrastructure owners can move down into it; if orbital data becomes valuable enough, cloud providers can move up towards it. The supposedly protected software company can find itself squeezed from both directions at once.
Capital should follow control, not labels
For investors, the label "space technology" says remarkably little about a company's economics. One business may need years of engineering spend before its first revenue; another depends on government milestone payments; a third sells recurring software subscriptions; a fourth carries the capital expenditure of an entire constellation while competing for consumer subscribers. Applying a single sector multiple to all four makes little sense.
Even the size of the sector depends on who is measuring it. The OECD's Space Economy at a Glance 2026, published in September, puts the global space economy at close to $600 billion in 2025, below the Space Foundation's figure for the previous year. The gap is not so much an error as a reminder that sophisticated institutions still draw the boundary of "space" in different places.
A more useful investment discipline asks where recurring revenue actually originates, who owns the customer and the data, and who can deny access to the system. It asks how much capital must be committed before each additional euro of revenue appears, and who captures the benefit when one part of the stack becomes cheaper. These questions reveal bargaining power in a way the upstream–downstream label never will.
Sovereignty prevents this from becoming a normal technology market
There is a further reason the aerospace analogy no longer tells the whole story: space is becoming sovereign infrastructure. Governments depend on orbital systems for communications, navigation, Earth observation, missile warning, intelligence and military command, so the same infrastructure that produces commercial efficiency also carries national-security consequences. Tellingly, the OECD notes that independent orbital launch capability remains concentrated in just 12 countries, plus Europe through ESA.
Europe's recent decisions illustrate the shift. At the November 2025 Ministerial Council in Bremen, ESA member states subscribed a record of around €22.3 billion, including a 70 per cent increase for the agency's core technology programme and a new component dedicated to resilience and security. In parallel, the proposed EU Space Act would replace a patchwork of national rules with common European requirements on safety, resilience and sustainability.
Market access is therefore no longer determined by technical performance and price alone. Licensing, cybersecurity, spectrum, trusted supply chains, data residency, sovereign procurement and political alignment increasingly decide which companies may operate where. A constellation can be commercially global and politically regional at the same time, and capital models that assume frictionless global scaling may prove optimistic.
Europe may be approaching the question from the wrong end
Europe's space engineering capability is not seriously in doubt. The harder question is how much of the economic system Europe controls once the hardware leaves the factory. Consolidation can improve programme economics, remove duplication and strengthen negotiating power, but manufacturing scale alone does not deliver control of the customer or of the data economy around the asset.
The proposed combination of the Airbus, Thales and Leonardo space businesses, known as Project Bromo, makes the issue timely. The new company would employ around 25,000 people and generate roughly €6.5 billion in annual revenue, and its industrial logic rests on achieving the scale to compete with American and Chinese systems. The deal still awaits European Commission clearance, which Leonardo does not expect before the second half of 2027, and rivals including OHB and Indra have warned that it could reduce competition in Europe's satellite market.
The board-level question is therefore larger than whether Europe needs bigger satellite manufacturers: which parts of the space economic architecture does Europe intend to own? Manufacturing and launch sovereignty matter, but so do secure communications, cloud integration, geospatial intelligence, AI processing, spectrum control, data distribution and the commercial applications built on top of them. Owning the factory while renting the economic control layer from someone else is not technological sovereignty. It is sophisticated dependency.
The counter argument deserves serious attention
There is a strong case against the view that value will continue to migrate away from physical infrastructure, and telecommunications history supports it. Infrastructure businesses did not disappear when applications proliferated. Fibre networks, mobile networks, cable systems and data centres became highly valuable wherever operators controlled scarce capacity, customer relationships or regulated territory.
Space may follow the same path. Launch capacity remains difficult to create, spectrum is finite, orbital congestion is rising, and government requirements can narrow the pool of eligible suppliers. Some sensors and constellations have characteristics that cannot readily be substituted. If congestion, regulation or geopolitical fragmentation makes access scarcer rather than cheaper, infrastructure may well regain pricing power.
So the simple claim that hardware loses and software wins does not hold. My reading is different: businesses that control only one replaceable piece of the system are vulnerable, wherever that piece sits. The strategic premium belongs to control, integration and scarcity.
What to watch through 2027
Over the next six to twelve months, three developments will say more about the industry's direction than launch counts. The first is vertical integration: the more often launch providers, satellite manufacturers, data businesses and defence companies acquire adjacent capabilities, the stronger the evidence that industry leaders no longer consider a single-layer position sufficient. The second is sovereign market architecture, as regulation, secure-connectivity programmes and defence procurement reveal whether the global space economy stays broadly interoperable or splits into partially separate American, European, Chinese and allied ecosystems.
The third is where AI processing takes place. If more interpretation moves onto the sensor, the spacecraft or edge infrastructure rather than terrestrial analytical platforms, today's assumed division between orbital hardware and downstream software will shift again.
I put the probability that integration continues to deepen through 2027 at roughly 75 per cent. Confirmation would come from further acquisitions across adjacent layers, larger sovereign programmes that buy integrated capabilities rather than individual components, and continued movement by launch and constellation operators into services and software. The view would be weakened if specialised single-layer suppliers consistently earned superior margins and bargaining power without integrating. That is entirely possible, but it is not where the current evidence points.
What boards should reconsider now?
The first mistake would be to approve a space strategy built on a market-size forecast. A $1.8 trillion economy tells a board that demand may expand; it says almost nothing about which part of that economy the company can defend. The second would be to confuse technological importance with economic capture, since a component can be indispensable yet still occupy a weak bargaining position.
The third would be to underestimate capital architecture. Space companies live on different clocks. Engineering, certification, government budgets, constellation deployment, customer adoption and investor liquidity rarely align, and a strategy can be technically correct yet fail because the balance sheet runs out before the market arrives.
Over the next 90 days, leadership teams with material space exposure should map their business without relying on conventional industry categories and instead trace the flow of control. Start with the end customer and work backwards: identify who owns the relationship, who controls the data and physical access, which capabilities can be substituted, where government permission is required, where capital becomes trapped and where recurring revenue finally appears. Then ask the uncomfortable question: if the space economy triples but our position in the architecture stays the same, do we actually become more valuable? For some companies, the answer will be yes. For others, market growth will simply build a larger system in which someone else controls the economics.
A different way to understand space
For decades, space has been treated as a place the aerospace industry goes. That mental model made sense when governments financed missions, prime contractors built spacecraft and value was measured in programmes completed and hardware delivered. It is becoming less useful. The next space economy will still demand extraordinary engineering, from rockets, satellites, ground systems, electronics and materials to the industrial competence to make machines survive unforgiving environments, but those capabilities increasingly form the foundation of something larger.
Space is becoming part communications infrastructure, part intelligence architecture, part data economy, part defence system, part sovereign infrastructure and part digital platform. That does not make aerospace less important; it makes space economically larger than aerospace. The strategic question for the coming decade is therefore not who can reach orbit most efficiently, but who will control what happens to the value once they get there.
Questions for the C-Suite and Board
- Are we investing in a space asset, or in a defensible position within a wider economic system?
- Which part of our current advantage comes from technology, and which from control of customers, data, regulation or scarce infrastructure?
- If an infrastructure owner moves downstream, or a cloud platform moves upstream, what remains defensible in our position?
- Which capability would we regret letting another company or another country control five years from now?
- Does our capital structure survive the time required for the strategic thesis to become economically real?
References
- Space Foundation, The Space Report 2025 Q2, July 2025.
- World Economic Forum and McKinsey & Company, Space: The $1.8 Trillion Opportunity for Global Economic Growth, April 2024.
- OECD, The Space Economy at a Glance 2026, September 2026.
- Rocket Lab, Fourth Quarter and Full Year 2025 Financial Results, 26 February 2026, and SEC filings.
- European Space Agency, CM25 Ministerial Council outcomes, November–December 2025.
- European Commission, proposal for an EU Space Act and Vision for the European Space Economy, June 2025.
- Reuters and Financial Times reporting on Project Bromo (Airbus, Thales, Leonardo), October 2025 to June 2026.
About Andy Demir
Andy Demir is the editor of SwissTechnic and writes on strategy, international growth and capital in aerospace, defence, space and advanced technology. Drawing on a career in commercial leadership, the analysis examines how technology, sovereignty and industrial capability shape competitive power and long-term value.
