GEO isn't dying, it's becoming a niche
For ten years pundits have predicted geostationary's death. Instead the GEO commercial-launch market has stabilised at twenty-five flights a year while LEO exploded around it. The restructuring story.
Around 2015 the analyst class started predicting that geostationary orbit was finished. The argument was straightforward: LEO megaconstellations would deliver lower latency, higher capacity and cheaper bandwidth than any GEO satellite ever could; the GEO commercial-launch market would collapse; the legacy GEO operators (Intelsat, SES, Eutelsat, Telesat, Hughes) would be either acquired, shrunk or simply allowed to wither. The prediction has been very nearly right about LEO and very nearly wrong about everything else.
What actually happened
The commercial GEO launch market peaked in 2014 at about 23 launches a year. It declined gently for several years, dropped to around 8-10 a year in 2018-2020 (the worst of it), and has since stabilised at roughly 12-15 commercial GEO launches a year. The LEO launch market, by contrast, went from approximately 40 commercial flights in 2014 to roughly 250 in 2024, almost entirely driven by Starlink. So the LEO market is now somewhere around twenty times the size of the GEO market, where in 2014 they were roughly comparable.
What didn't happen is the death of GEO. The legacy operators didn't disappear. SES bought Intelsat in 2024, creating the world's largest non-US commercial GEO operator (combined fleet roughly 100 spacecraft). Eutelsat absorbed OneWeb in 2023 and is now operating both a 150-spacecraft GEO fleet and a 600-satellite LEO fleet. Hughes Network Systems is still profitable. SES, even before the Intelsat acquisition, was generating roughly $2 billion in annual revenue. The GEO commercial business shrank, but it did not collapse, and what's left is unusually durable.
What GEO is actually good at
The thing the LEO-will-eat-everything thesis got wrong is the assumption that low latency and high capacity were the only metrics that mattered. They're not. There are at least four things GEO does substantially better than any LEO constellation can.
The first is broadcast. A single GEO satellite can illuminate roughly a third of Earth's surface with a powerful broadcast signal that can be received by a fixed dish for under $200 of consumer hardware. Direct-broadcast television, satellite radio, and most professional video distribution are GEO businesses for a structural reason: when you need to send the same signal to a large fixed audience, broadcasting from one big satellite at 36,000 kilometres altitude is much cheaper than synchronising hundreds of LEO spacecraft to do the same thing.
The second is high-power point-to-point. Modern GEO communications satellites — Boeing 702, Airbus Eurostar, Lockheed A2100, Northrop GEOStar — carry kilowatts of payload power and can dedicate enormous bandwidth to a single beam aimed at a single customer. The Inmarsat-6 satellites, the EchoStar XXIV, the JCSAT-1C, the SES O3b mPOWER (technically MEO not GEO, but architecturally similar): all of these serve customers — airlines, shipping, large enterprises, military fixed installations — who need committed bandwidth to a known location, with service-level agreements that LEO constellations are not yet structurally able to match.
The third is the maritime and aviation in-flight connectivity market. Inmarsat's Global Xpress, ViaSat's Global Aero, SES's mPOWER and Iridium's Certus all anchor on long-cycle airline and shipping contracts that prefer the predictable, well-understood economics of high-capacity GEO and MEO assets. Some of this is migrating to LEO (Starlink Aviation in particular has been aggressive about onboard connectivity), but the migration is slower than the megaconstellation cheerleaders predicted, partly because aviation customers care about regulatory certification of the in-flight equipment more than about peak throughput.
The fourth, and most overlooked, is national-security and government broadband. Almost every advanced military's strategic communications backbone runs through GEO assets — the US WGS, UFO and AEHF constellations, the UK Skynet 5/6, the French Syracuse, the Italian Sicral, the German SatCom Bundeswehr. Governments care about sovereign hardware they own, predictable orbits, hardened links and physical-security profiles that LEO constellations operated by private commercial operators can't easily replicate. Nothing about LEO megaconstellations changes the structural national-security demand for GEO.
The acquisition wave
The most interesting commercial story is what's been happening at the operator level. Five years ago there were nine independent commercial GEO operators of meaningful scale: Intelsat, SES, Eutelsat, Hispasat, Telesat, Inmarsat, EchoStar, ViaSat, Sky Perfect JSAT. Today there are six. ViaSat acquired Inmarsat in 2023 for $7.3 billion; SES merged with Intelsat in 2024-2025 for approximately $3.1 billion; Eutelsat absorbed OneWeb in 2023 in a $3.4 billion all-stock deal that combined GEO and LEO under one operator; EchoStar and DISH (which owns satellite operations) merged in late 2023.
The pattern is unambiguous. The GEO operators are consolidating because the smaller standalone businesses can't sustain capital deployment for both legacy GEO refreshes and new LEO/MEO investments at the same time. The merged entities will be six operators — SES-Intelsat, Eutelsat-OneWeb, ViaSat-Inmarsat, EchoStar-DISH, Hispasat, Sky Perfect JSAT, and Telesat — controlling roughly 95% of the global commercial GEO fleet. Concentration is a sign of maturity, not death.
What's coming over the next five years
The medium-term picture for GEO is steady-state in transponder demand and a slow shift in what GEO satellites carry. Expect:
The next-generation GEO satellites being built right now (Airbus Eurostar Neo, Boeing 702X, Northrop GEOStar-3 software-defined family) are flexible, software-reconfigurable, electric-propulsion-equipped birds that can switch capacity between regions and customers in flight. This makes them more useful per dollar than legacy GEO satellites, which means fewer of them are needed for the same revenue — but the ones that fly will be more capable. Hence the continued ~12-15 launches a year cadence rather than continued decline.
The next consolidation wave will probably be the smaller fleet operators getting acquired by SES-Intelsat or Eutelsat-OneWeb — Hispasat, Telesat (likely separately, though Telesat Lightspeed complicates that), possibly Sky Perfect JSAT in the long run. By 2030 the global commercial GEO fleet might be operated by four players rather than six.
Government and military demand is, if anything, growing. The US Space Force's GPS and SBIRS programs, the next-generation NRO architectures, the UK Skynet 6 contracts awarded to Airbus, the French Syracuse 5 and the European IRIS² programme together represent more institutional GEO and MEO buying than the commercial market has seen in a decade.
The bigger lesson
Most "X is being disrupted" narratives in this industry are at best half right. Yes, LEO megaconstellations took a huge bite out of the consumer-broadband and small-enterprise segments that were previously GEO's growth markets. No, that didn't kill GEO. The market restructured around what GEO is genuinely good at, the operators consolidated into a smaller number of bigger and more durable companies, and the customer base shifted toward government, aviation, maritime and broadcast — segments where the GEO economics still close.
This is what mature industries do. They don't die, they specialise. The same thing is, separately, going to happen to LEO megaconstellations once the broadband-for-everyone phase gives way to the actual customer mix. We'll write that one in 2031.