Launch slots, not launch prices, now set the pace for satellite programs

Ars Technica reports a growing crunch in launch availability even as orbital launch rates and price competition hit record levels ^[1][2][3]. For anyone running a satellite program, that inverts the planning assumption of the last decade: the scarce input is no longer dollars per kilogram, it is a confirmed seat on a vehicle that exists.

The macro numbers still look healthy. An average of 270 orbital rockets have launched from Earth annually over the past three years, more than a three-fold increase from a decade ago ^[1], which puts the mid-2010s baseline below 90 launches a year ^[1]. Prices have never been more competitive and access never more rapid ^[2]. Caleb Henry, director of research at Quilty Space, who talks to satellite operators regularly, describes the customer side differently: "What we have is an industry in panic" ^[4].

The specifics are what matter for operators. Amazon's LEO constellation has had to throttle back satellite production because few of the dozens of launches it booked half a decade ago are ready ^[5]. That is the failure mode worth internalising: a long-dated launch contract is an option on a vehicle reaching operational status, not a reservation of capacity, and when the vehicle slips the cost lands on the factory floor. Telesat, meanwhile, has been approached by other companies asking whether it would share some of the 11 Falcon 9 launches it has booked for its new constellation; the answer, per the report, is no ^[6]. Booked manifest has become a strategic asset that holders do not sublet. On an earnings call last Monday, AST SpaceMobile officials said launch availability is now the pacing item for deploying its constellation ^[7].

The demand side gives no relief. The Commercial Space Federation predicts demand for thousands of satellite launches annually in less than a decade ^[8], and Analysys Mason forecasts that more than 37,000 satellites will need to be launched between 2023 and 2033 ^[9] - roughly 3,700 satellites a year across that span ^[2]. Set against a current cadence of 270 launches a year, that implies about 14 satellites per launch on average, every launch, worldwide ^[3]. Quilty Space has charted the corresponding shift in industry conversation, from how many launch companies the US market can support to who can execute fastest against surging satellite demand ^[10]. The report notes there are reasons to believe conditions get worse over the next two to four years ^[11].

The operating consequence is that manufacturing throughput and launch throughput have to be planned as one system. A production line sized to a manifest that does not materialise burns capital storing finished hardware, and a program that under-books rides cannot buy its way out at short notice when incumbents refuse to share.

Watch whether Amazon's booked launches start clearing and production resumes at rate ^[5], and whether AST SpaceMobile still calls launch the pacing item at its next earnings call ^[7]. Watch also for any softening in the Telesat position on shared rides, which would be the clearest signal that slot-holders see slack returning ^[6].