Skip to main content
You're offline. Cached data shown.

The Space Economy Value Chain: Upstream, Midstream, Downstream

Rockets get the headlines, but launch is less than 2% of the space economy. This guide follows the money from the factory floor to the phone in your pocket โ€” which tiers are large, which are profitable, who owns the choke points, and where the value is moving.

By SpaceNexus Team|12 min read|Updated August 2026

Overview

The space economy is worth roughly $600 billion a year โ€” the Space Foundation counted about $613 billion in 2024, three-quarters of it commercial โ€” and McKinsey and the World Economic Forum expect it to reach $1.8 trillion by 2035. Almost none of that is rockets. The industry is best understood as a chain: upstream companies build and launch hardware, midstream companies operate it, and downstream companies sell what the hardware produces. Value concentrates at the ends of the chain that touch scarce capability (upstream niches) or the end customer (downstream), and thins out in the middle wherever capacity is abundant.

The framing matters because it predicts behaviour. It explains why SpaceX built a satellite operator on top of a launch business, why Earth-observation companies keep rebranding as analytics companies, and why the ground segment โ€” the least glamorous tier โ€” is the largest by revenue.

The Value Chain Map

Segment sizes below are order-of-magnitude, drawn from the Satellite Industry Association's annual State of the Satellite Industry report and the Space Foundation's Space Report. Government spending (~$125 billion) runs through every tier as a customer and is not shown as a segment of its own.

Upstream

Components & Subsystems

Tens of $B (inside manufacturing)

What it sells: Solar cells, rad-hard electronics, propulsion, star trackers, antennas, structures

Margins: High for sole-source parts; thin for commodities

Players: Rocket Lab (components), Redwire, Honeywell, Airbus, Mitsubishi Electric, hundreds of specialists

Upstream

Spacecraft Manufacturing

~$17โ€“20B / yr

What it sells: Buses, payloads, integration and test

Margins: Moderate; project risk is the killer

Players: Airbus D&S, Thales Alenia, Lockheed, Northrop, Lanteris (ex-Maxar), York, Terran Orbital, Apex, SpaceX (Starlink, in-house)

Upstream

Launch

~$8โ€“10B / yr

What it sells: Dedicated and rideshare launch, orbital transfer

Margins: Historically poor; reusability changed it for exactly one company

Players: SpaceX (~85% of global mass to orbit), ULA, Arianespace, Rocket Lab, Blue Origin, CASC, ISRO, Mitsubishi Heavy

Midstream

Satellite Operators

~$110B / yr (satellite services)

What it sells: Broadband, broadcast, mobile connectivity, Earth observation capacity, navigation signals (public)

Margins: High once the constellation is paid for; brutal capex before that

Players: Starlink, SES, Intelsat, Eutelsat OneWeb, Viasat, Iridium, Planet, ICEYE, Maxar/Vantor, Amazon Kuiper

Midstream

Ground Segment

~$150โ€“160B / yr (ground equipment, mostly GNSS devices)

What it sells: User terminals, gateways, teleports, ground-station-as-a-service, GNSS chipsets in every phone

Margins: Commodity for chips; strong for terminals and networks

Players: Qualcomm/Broadcom/u-blox (GNSS), Hughes, Gilat, Kymeta, KSAT, SSC, AWS Ground Station, Viasat

Downstream

Data, Analytics & Applications

The largest and least bounded tier โ€” location, timing, imagery analytics, weather, insurance, agriculture, defence

What it sells: Ride-hailing, precision agriculture, maritime tracking, ESG monitoring, wildfire detection, disaster response

Margins: Software margins โ€” but the space part is a small share of the value

Players: Google/Apple/Uber (GNSS-dependent), Planet, Spire, BlackSky, HawkEye 360, Descartes Labs, Palantir, insurers, banks

Upstream: Building and Launching

Upstream is where the physical risk lives. A spacecraft manufacturer carries multi-year programmes, thin supplier bases and acceptance tests that can fail on the last day. A launch provider carries the rocket. Both tiers are capital-heavy, and both have been reshaped by one company.

Launch is roughly $8โ€“10 billion a year of revenue on well over 250 orbital launches, and SpaceX carries the large majority of the world's mass to orbit. Reusability drove the price of a kilogram to LEO from tens of thousands of dollars to about $3,000 on a full Falcon 9, and rideshare put a 50-kilogram satellite in orbit for under $400,000. The strategic consequence: launch stopped being the bottleneck of the industry. Our launch cost guide has the vehicle-by-vehicle numbers.

Manufacturing splits into two businesses that barely resemble each other. Bespoke GEO communications satellites and government spacecraft sell for hundreds of millions each in single digits per year. Constellation buses sell for low single-digit millions in hundreds โ€” a production-line business with automotive-style supply chains, where the customer (Starlink, Kuiper, the Space Development Agency) often designs the bus itself.

Components are the quiet profit centre. Radiation-hardened processors, triple-junction solar cells, precision reaction wheels and optical inter-satellite terminals each have a handful of qualified suppliers worldwide and lead times near a year. Rocket Lab's acquisitions of solar, separation-system and software vendors were a bet on exactly this tier. Our supply-chain course covers how those bottlenecks work from the buyer's side.

Midstream: Operating

Satellite operators own the assets in orbit and sell capacity: broadband, broadcast, mobile backhaul, Earth-observation tasking, machine-to-machine links. Satellite services run at roughly $110 billion a year. The economics are a constellation's worst enemy and best friend: enormous capex before the first dollar, then high-margin recurring revenue on a paid-for asset. Legacy GEO operators (SES, Intelsat, Eutelsat) are living through the transition from broadcast to broadband; Starlink โ€” now well past $10 billion in annual revenue โ€” showed that a vertically integrated LEO operator can reach profitability, and Amazon Kuiper is spending to prove it can be done twice.

The ground segment is the tier nobody photographs and the largest by revenue โ€” on the order of $150โ€“160 billion โ€” because it includes every GNSS chipset in every phone, car and drone on Earth, plus user terminals, gateways, teleports and the ground-station-as-a-service networks (KSAT, SSC, AWS Ground Station) that rent antenna minutes to satellite operators. Flat-panel terminals are the current battleground: whoever makes a cheap, reliable electronically steered antenna sets the cost of every LEO subscriber.

Downstream: Data, Services, Applications

Downstream is where space stops looking like space. A ride-hailing app, a precision-agriculture platform, a maritime insurer pricing risk from AIS and radar imagery, a hedge fund counting cars in parking lots โ€” all consume a satellite input that is a small fraction of the value they create. This is the tier McKinsey and the WEF expect to drive most of the growth to $1.8 trillion, and it is the reason Earth-observation companies reposition as โ€œinsightsโ€ companies: raw imagery is a commodity with falling prices, an answered question is not.

The downstream tier has two structural features. First, the space input is often free or public โ€” GPS, Galileo, Copernicus Sentinel data, NOAA weather โ€” so the customer relationship, not the satellite, is the asset. Second, government is the anchor buyer for the newest capabilities (radar, RF sensing, hyperspectral) years before commercial demand matures, which is why defence and intelligence agencies appear on every EO company's revenue chart. Track the deals on the procurement tracker.

Enablers: Capital, Insurance, Regulation, Talent

Capital

Venture, growth equity, SPAC-era public listings, and โ€” increasingly โ€” sovereign and defence-adjacent funds. Investment concentrates upstream in hardware rounds and downstream in software; the middle is financed by operators' own balance sheets and export-credit agencies. Live data on the funding tracker.

Insurance

Launch and in-orbit insurance is a ~$500โ€“600 million-a-year premium pool that has paid out more than it collected in several recent years after a run of GEO failures. Rates rose, capacity shrank, and many constellation operators now self-insure โ€” a quiet transfer of risk from underwriters to balance sheets.

Regulation

Spectrum (ITU, FCC), launch and re-entry licensing (FAA), remote-sensing licences (NOAA), export control (ITAR/EAR) and a growing debris-mitigation rulebook. Regulation is a moat for incumbents and a schedule risk for everyone. Follow it on the regulatory radar.

Talent

The chain runs on a few hundred thousand specialised engineers, and demand exceeds supply for RF, GNC, flight software and manufacturing roles. The jobs board syncs thousands of open roles daily โ€” a real-time read on where the chain is hiring.

Who Captures the Value

Three positions capture disproportionate value in the chain:

  • โ€ขVertical integrators. Own two or more tiers so the margin does not leak to a supplier or a customer. SpaceX (launch + operator + terminals) is the canonical case; Rocket Lab (launch + components + spacecraft) is building the same shape from the other end.
  • โ€ขScarce-capability suppliers. Sole- or dual-source components with year-long lead times. Small revenue, strong pricing power, and the reason big primes keep acquiring small specialists.
  • โ€ขCustomer owners. Downstream businesses that own the relationship and treat satellite data as one input among many. They capture software margins on a commodity input โ€” and they can switch suppliers.

The squeezed middle is generic capacity: undifferentiated launch, undifferentiated imagery, undifferentiated buses. Prices fall, and the value moves to whoever does something specific with the capacity. See how individual companies stack up on the comparison pages.

Where the Chain Is Moving

Direct-to-device

Satellites talking to unmodified phones (Starlink's D2C service with T-Mobile, AST SpaceMobile, Skylo) collapse the terminal tier and hand the customer relationship to mobile carriers. The ground segment shrinks; the operator tier grows.

Defence as anchor tenant

The Space Development Agency's proliferated architecture, Golden Dome, and allied programmes are buying constellations, launches and analytics at a scale that sets prices for the commercial market too. Government share of the chain is rising for the first time in a decade.

In-space economy

Servicing, refuelling, manufacturing in orbit and lunar logistics add tiers that do not exist yet at scale. Starship-class lift makes them plausible; the first real revenues are government (NASA CLPS, Space Force servicing demos).

Frequently Asked Questions

What is the space economy value chain?

It is the sequence of activities that turns capital and engineering into revenue from space: building components and spacecraft, launching them (upstream); operating satellites and the ground networks that talk to them (midstream); and turning the resulting data, signals and connectivity into products people pay for โ€” navigation, broadband, imagery analytics, weather, defence intelligence (downstream). Finance, insurance, regulation and workforce sit alongside as enablers of every tier.

How big is the space economy?

Estimates converge on roughly $600 billion a year. The Space Foundation put the 2024 global space economy at about $613 billion, roughly three-quarters commercial and one-quarter government. McKinsey and the World Economic Forum estimated $630 billion in 2023 and project $1.8 trillion by 2035, most of the growth coming from downstream applications rather than rockets.

Which part of the space value chain makes the most money?

By revenue, the ground segment and downstream services dwarf launch and manufacturing: launch is under 2% of the total. By margin, satellite operators with paid-off constellations and downstream software businesses do best. Launch is the most visible tier and, historically, the least profitable โ€” SpaceX changed that for itself through reusability and by owning its largest customer, Starlink.

Why is launch such a small share of the space economy?

Because a launch is a one-time cost that enables years of revenue. A $70 million Falcon 9 mission can deploy satellites that earn hundreds of millions over their lives. Falling launch prices have made the share smaller still even as the number of launches has grown โ€” the value migrates downstream to whoever uses the capacity.

Where should a new space company position itself?

Where a scarce capability meets a paying customer. Upstream niches with few qualified suppliers (radiation-hardened electronics, propulsion, optical terminals) command pricing power. Downstream, the winners sell an outcome โ€” crop yield, ship location, insurance risk โ€” not "satellite data". The most crowded, hardest tier to enter is launch, followed by generic Earth-observation imagery.

Related Guides