The two largest European satellite operators are abandoning planned geostationary satellites and pushing the capital toward low Earth orbit, a clear signal that the GEO build-out era is winding down even at its most committed players. SES has canceled two software-defined satellites inherited from Intelsat, mirroring Eutelsat’s earlier decision to scrap its Flexsat Americas program and redirect more than 100 million euros toward LEO growth.
The cancellations were disclosed alongside SES’s Q1 2026 earnings, reflecting the operator’s absorption of Intelsat. The dropped orders were part of a software-defined fleet that Intelsat had committed to before the merger.
Fleet rationalization, not retreat
SES framed the cancellations as housekeeping after a merger that produced the largest GEO fleet in commercial history. The operator cited the need to optimize across a larger, more resilient satellite fleet and reduce unnecessary duplication. Other software-defined orders remain in place, giving the operator what it calls ample flexibility for future demand.
The review weighed new builds against keeping existing hardware aloft longer, including consideration of extending the life of existing satellites with in-orbit services. SES has outlined life-extension missions scheduled between 2026 and 2029, work that increasingly competes with new orders for capital.
The numbers behind the pivot
SES reported strong growth on a year-over-year basis. Most of that growth came from the Intelsat integration rather than organic GEO demand, and management has been candid that legacy video distribution remains under pressure.
Eutelsat’s quarter told the same story from the opposite direction. GEO revenues declined while LEO connectivity sales, drawn almost entirely from OneWeb, showed strong growth. The GEO drag was almost fully offset by the LEO surge.
Those two trend lines explain the capital reallocation more clearly than any executive narrative. Money flows where revenue grows.
From one-off constellations to continuous build
SES CEO Adel Al-Saleh has been signaling this shift since the deal closed. In an interview with SpaceNews last July, Al-Saleh said the combined company would move away from launching one next-generation constellation at a time and instead build a continuously expanding fleet in medium Earth orbit. According to the publication, SES is considering a MEO constellation potentially numbering in the hundreds of satellites.
SES currently operates MEO satellites through its O3b mPower program. The O3b mPower spacecraft are under a Boeing contract, with additional launches scheduled. Each mPower satellite offers significantly higher throughput than the original O3b fleet.
The company has guided to capital expenditures of roughly 600 to 650 million euros annually through 2028, excluding its IRIS² commitments. That budget is now being weighted toward MEO expansion, ground infrastructure and emerging segments like hosted payloads and direct-to-device, rather than incremental GEO capacity.
Eutelsat’s parallel calculation
Eutelsat reached the same conclusion months earlier, canceling Flexsat Americas and citing increased vigilance on GEO spend while focusing capital expenditure on LEO. The operator has ordered hundreds of LEO satellites from Airbus to replenish the OneWeb constellation, a commitment that dwarfs anything it could realistically sustain alongside a parallel GEO renewal cycle.
The Flexsat decision was notable because the program had been pitched as a flexible, software-defined answer to changing demand patterns in the Americas. The fact that even a flexible GEO platform failed the capital test suggests the issue is structural rather than architectural.
What the GEO retreat actually means
GEO is not dying. Both operators continue to invest in life-extension missions, software-defined replacements for existing slots, and government services where guaranteed coverage and hardened spectrum still command premium pricing. The expectation is that the GEO market will stabilize as demand consolidates around video distribution, government communications and guaranteed broadband.
What is dying is the assumption that GEO fleets should grow. For two decades, the rhythm of the industry was simple: operators ordered replacement satellites slightly larger and more capable than the ones they retired, and the global fleet expanded. Starlink broke that rhythm. Project Kuiper is reinforcing the break. When a single LEO constellation can add more usable capacity in a quarter than the entire GEO industry adds in a year, the math of speculative GEO orders stops working.
SES and Eutelsat are doing what financially disciplined incumbents are supposed to do. They are sweating existing assets, killing speculative orders, and concentrating new spending in the segment where growth is actually showing up on the income statement.
The competitive backdrop
Al-Saleh has been blunt about the pressure from below, citing Starlink’s growth and Kuiper’s deployment. The strategic response has been to build scale through consolidation rather than debt-funded organic expansion.
SES emerged from the Intelsat deal with a large GEO fleet. That scale provides cash flow to fund the MEO build-out and, through the OneWeb partnership and Lynk Global stake, exposure to LEO and direct-to-device markets without the capital intensity of building a dedicated LEO constellation from scratch.
For Eutelsat, the calculation is more constrained. The OneWeb merger left it with a LEO asset that requires significant replenishment spending, and the company has less room to fund parallel GEO expansion. Cutting Flexsat Americas was less a strategic choice than a financial necessity.
What to watch next
The interesting question is what SES does with the software-defined satellite orders it kept. The operator has signaled flexibility on timing and configuration, which means those builds could be deferred, redirected to different orbital slots, or eventually canceled as well if MEO economics continue to improve.
Also worth watching: the pace at which GEO life-extension services scale. Northrop Grumman’s Mission Extension Vehicles and similar offerings turn aging GEO satellites into multi-year revenue extensions at a fraction of replacement cost. Every successful life-extension mission is, in effect, a deferred or canceled new-build order somewhere.
The Intelsat-SES merger and Eutelsat’s OneWeb integration were both bets that consolidation would produce the cash flow needed to compete in a multi-orbit world. The cancellation of software-defined satellite orders is the first hard evidence that those bets are reshaping how satellite capital actually gets allocated. Fewer GEO satellites. More LEO. And the era of speculative geostationary expansion, for now, is over.
