Everyone at WSBW Described a Bottleneck. In Every Case It Was Someone Else’s Business.

By Nick David, Editorial Lead, SatNews

World Space Business Week produced plenty of good news. Open Cosmos reached unicorn valuation. Aerospacelab reported profitability. Thales Alenia Space and MDA described factories built to turn out one or two satellites a day. On the first morning, a panel of five banks described a sector that has, in Morgan Stanley’s phrasing, migrated “from science to core infrastructure.”

Across four days and more than 50 sessions, though, the people who build and operate satellites kept describing bottlenecks that sit almost entirely outside the satellite.

The launch squeeze started in July

The constraint the week kept circling had been set two months earlier. On July 20, SatNews reported that SpaceX had stopped accepting new commercial bookings for Transporter and Bandwagon rideshare missions beyond late 2028, an account sourced to smallsat operators and mission integrators rather than to the company. Three days later, Bloomberg reported that SpaceX was turning away Falcon customers as it bet on Starship. On Sept. 15, with the conference under way, Space Intel Report said SpaceX had begun telling prospective customers it was no longer taking orders for commercial Falcon 9 launches, a further step rather than the start of the story. Five days before that, Avio chief executive Giulio Ranzo had told an earnings call that on discounted Transporter rideshare, “the party is finished, and they don’t do that anymore.” SpaceX has not publicly confirmed any of it.

On stage, the companies selling launch described the same squeeze from the other side of the table. At the Global Launch Providers session on Sept. 15, Firefly’s Michael Creech told the room he would “wonder which rock you have been living under” if anyone thought supply was adequate, and said Firefly is “fully sold 24 months out.” Blue Origin’s Jordan Charles called it “very much a seller’s market.” Exolaunch chief executive Robert Sproles said customers now book “one to three years in the future for their launch” while building satellites far faster than that.

Where they differed was on how bad it gets. D-Orbit’s Renato Panesi named a window, “the critical moment will be 2029, 2030, maybe 2031,” and said his own company is covered to the end of 2029. A day earlier, Aerospacelab’s Benoît Deper had worried about “the launch window between I would say 29 and 31, that seems to be quite stuck in terms of offering… if not a lot of satellites will be stuck on the ground.” Sharing the stage with Panesi, Exotrail’s Jean-Luc Maria called the situation “totally predictable” and said he is “really, really confident” it resolves.

Nobody on either side of that argument treated the shortage as a technology problem.

Price per kilogram lost its standing

For a decade the industry measured launch in dollars per kilogram. In Paris, the companies that sell launch spent two sessions taking that measure apart.

Creech argued the cheap-launch period was subsidized rather than economic: “this subsidized price per kilogram model was not supporting launch companies. That is leading into the current day crisis of having a supply shortage.” Northrop Grumman’s Nicole Jordan said price per kilogram had been useful once but was not what her customers were raising: “What really matters is the overall cost, the risk, having access to less congested launch sites.” The next morning, Eddie Kato, executive advisor to the CEO at Interstellar Technologies, called it “a misguiding measure.” Arianespace’s David Cavaillolès kept price in the conversation but weighed it against injection accuracy, pointing to the operating life that accuracy bought on a $10 billion telescope.

On that panel’s account, customers are now optimizing first for whether the rocket exists and when it flies.

Factory capacity has outrun the parts

The manufacturing story at WSBW came in two halves, told on different panels by people who never compared notes.

The primes described capacity. Thales Alenia Space’s Rome Space Smart Factory is built for “100 satellites per year,” deputy chief executive Giampiero Di Paolo said; chief executive Hervé Derrey put the objective at “one satellite per day.” MDA Space’s Mike Greenley put his new Montreal line at “up to two satellites a day,” fully operational by the end of this year. Safran’s Jean-Marie Bétermier said private constellation customers now ask for hundreds of units a year, where a couple of years ago orders ran to five to ten.

Component makers described queues. ThrustMe has doubled output three years running, to 400 units this year. York-owned Orbion runs “at about 90 percent our current capacity.” Dcubed is at “100 plus percent,” and for triple-junction solar cells its choice is to “pay a fortune for it or wait 18 plus months for shipment.” CesiumAstro chief revenue officer Trey Pappas laid out why the wait persists: “if an FPGA is going to go to a car company where they’re doing 10 million a year. I’m not doing 10 million a year. And so my problem seems a little less interesting to providers at the moment.”

Honeywell Aerospace’s Lisa Napolitano made the same point from the buying side: even its space volumes are “still a bit low volume” against what its own component suppliers care about. Asked whether customers now request standardized hardware, Beyond Gravity chief executive Barbara Frei-Spreiter said it happens, but “that’s more the exception. That’s not yet the rule.”

New factory space in Rome and Montreal does nothing for a supplier quoting a year and a half on solar cells.

The ground segment became the target

The program gave the ground segment three sessions of its own. It kept surfacing in the sessions that were not about it.

Asked in a rapid-fire round on Sept. 15 whether the bottleneck to scaling sits in orbit or on the ground, the connectivity operators on stage pointed to terminal cost and the absence of standards. AsiaSat’s Roger Tong did the arithmetic out loud: Starlink’s roughly 10 million terminals at $500 each comes to $5 billion of ground investment, which makes the space-segment cost debate look small. SpaceX reported 12.0 million Starlink subscribers at the end of June, so his terminal estimate is, if anything, conservative.

On a broader connectivity-architecture panel that afternoon, Airbus’s Sarah Casenove singled out the ground segment and its cyber protection as the most underappreciated part of the system, and said that of the five biggest cyber incidents on a satellite system, “three of them were made through the [ground] segment.” The panel dedicated to ground infrastructure in contested environments was blunter. Kratos’s Phil Carrai: “ground sites are now targets.” The attention used to go to the space layer, he said; now it is on the complete system. GMV’s Enrique Fraga Moreira described the resulting change in posture: “we changed from protecting infrastructure to protecting the mission.” He also noted that a backup site offers no protection against a cyber attack, because “you will contaminate both of them.”

The final day’s ground-station-as-a-service panel made the access point directly. RBC Signals’ Christopher Richins said that for large antennas, “given the small number of large antennas that are available in the world today, it’s more of a political issue than it is an issue of dollars.”

Spectrum is the next queue

Mario Maniewicz, the outgoing director of the ITU’s Radiocommunication Bureau, said satellite items now make up about 85% of the WRC-27 agenda, up from less than half 10 years ago. Filings for large non-geostationary constellations are up 50% in four to five years, and the number of satellites inside them has doubled. Narrowband filings for IoT and remote sensing are up fourfold, with satellite numbers up sixteenfold.

Vodafone’s Tim Boddy called spectrum “the kind of oil that drives the industry.” Orange’s Stéphane Cugnon de Sévricourt described it as “a single entry point”: of the five components of a satellite service, the one that gates the rest. Airbus declared itself “fundamentally against” telco pressure to reallocate X-band.

Maniewicz expects WRC-27 to produce a regulatory framework for direct-to-device service in at least some mobile bands. Until then, the partnerships are being built against a rule that has not yet been written: GSMA’s Barbara Pareglio, on the same stage, counted about 151 mobile operators with satellite services planned or live, of which perhaps 50 are live.

Who is actually paying

Underneath all of it, the buyer has changed.

Seraphim’s Mark Boggett gave the figure that makes it concrete: of his portfolio, only two companies describe themselves as defense companies, “and yet 80% of revenue across 80% of the companies is coming from defense.” Aerospacelab is turning its pipeline toward a dual-use majority. Ovzon’s Per Norén described its newest terminal, thermos-sized and small enough to sit on a drone, as a product of what the company had learned in “the last four years” of operating in Ukraine. In the Earth observation market presentation, Annekatrien Debien of Novaspace, which organizes WSBW, said defense is no longer just a customer but is rewriting the market’s rules.

Commercial demand drew more mixed readings. Satlantis’s Juan Tomás Hernani called Earth observation “a five billion business that more or less, despite all the expectations, tends not to be growing that much, that aggressively, with the exception of defense.” Later that morning, Novaspace’s market presentation projected the value of the commercial constellation segment up 325% over the decade, on 85% growth in volume.

The customers nobody designed for

Two buyers from outside Europe and North America said the industry’s model does not fit them.

Dr. Zolana Rui João, general manager of Angola’s National Space Program Management Office (GGPEN), told the room his country was spending “$30 million, approximately every month, renting foreign satellites” 18 years ago. The retail LEO playbook of credit card and self-installation, he said, “will never work” where “people live under $10 per month in most rural villages.” His verdict on the growth story was that LEO systems are connecting people who are already connected, which is “not solving the problem.”

During questions at the SAR panel, a delegate who introduced himself as director general of Nigeria’s space agency said that running an RFP for new satellites had shown him the real restrictions on technology sharing: what emerging nations are offered is images and capacity, not the capability to build their own. One panelist replied that their company had done technology transfer and co-development with approved partners. The moderator closed the exchange by admitting “that’s a really complicated question that you’re not going to get a satisfactory answer here.”

What it adds up to

The capital is largely there. Novaspace estimates public budgets will approach $200 billion before 2030, and the factories are built and largely funded. The exception came from the ground: Richins named “access to capital” as a primary limit on scaling ground networks to thousands of satellites, because building antennas still tends to need fresh equity. The technology arguments that dominated the last decade, GEO against LEO and vertical against horizontal integration, were still being had in Paris. They remain unresolved and look increasingly beside the point.

What the week described is a sector that has worked out how to build satellites but not how to secure the launch slots, spectrum, ground infrastructure and components needed to use them, each of which sits in a queue someone else controls.

---

About the Author

A storyteller at heart, Nick David covers space policy, satellite markets, defense, and the technologies reshaping how humanity operates beyond Earth. With a background in creative direction, brand strategy, and editorial storytelling, he brings a modern lens to complex subjects and a relentless curiosity about what comes next.