Can Amazon Keep Funding the Competitive Launch Industry Beyond SpaceX?

Here is where we think most observers are getting the analysis wrong. SpaceX's dominance in the launch and satellite broadband market is built on a genuine engineering marvel: vertical integration from rocket to router. SpaceX builds its own Falcon 9, reuses the first stage, and launches its own Starlink satellites on a 3+ launch per week cadence. When anomalies happen on a launch (as just happened a couple of weeks ago), SpaceX is able to identify, correct, and test on subsequent launches without breaking cadence for months the way a government contract would require. That flywheel is real, it is compounding, and it gives SpaceX an unmatched advantage in deployment cadence. No one disputes this.

But that vertical integration creates a competitive scenario where SpaceX is simultaneously the dominant launch provider and the dominant LEO broadband competitor. This structural market problem was not considered when the FCC established the current milestone rule for Amazon back in 2020. Our observation is that if Amazon relied solely on Blue Origin, another entity within the Bezos empire, we would absolutely agree with the pundits who opine that the FCC should hold firm on the original deadline. That approach would indicate a captive supply chain that failed to deliver. But that is not what Amazon has been doing for the last few years.

Amazon's procurement spans four independent launch providers across two continents (three if you include Arianespace’s equatorial launch site in South America). The company contracted 18 launches with Arianespace, purchased missions on Lockheed/Boeing joint venture ULA's Atlas V and Vulcan Centaur, in addition to securing flights on Blue Origin's New Glenn. Perhaps most telling, Amazon bought launches on SpaceX's own Falcon 9, the rocket operated by its primary broadband competitor. In January 2026, Amazon disclosed 10 additional Falcon 9 contracts and 12 additional New Glenn missions, pushing the total manifest to 102 contracted launches. We cannot view this as simple spectrum warehousing, which the FCC milestone rules were intended to frustrate. It has to be seen through the lens of being the largest commercial launch procurement in history.

The Progressive Policy Institute recognized this dynamic in its January 2026 statement urging the FCC to grant the extension. The group notes that Amazon Leo is dependent on external rocket launch services and that the launch companies Amazon has contracted with have not been able to ramp up operations as quickly as anticipated. That framing captures something important. Amazon's deployment delays are not a function of insufficient capital commitment ($10 billion invested), insufficient manufacturing capacity (30 satellites per week), or insufficient launch orders (102 missions). The delays reflect a market-wide launch infrastructure gap that Amazon is actively helping to close.

It should be noted that Blue Origin has successfully launched the larger and partially reusable New Glenn rocket twice, and recovered the first stage once. That is years behind the schedule originally anticipated by Amazon. While Blue Origin claims acceleration is on the way, the next launch of the largest carrying capacity rocket (e.g. 60+ satellites vs. the 32 launched by Arianespace on the expendable Ariane 64) in Amazon’s contracted fleet is scheduled for mid-2026. Hardly the aggressive pace needed for Amazon to close the gap with the FCC milestones, much less the Starlink lead. ULA, with only a few of the venerable Atlas V rockets left in its inventory, has now had 4 launches of its Vulcan Centaur replacement rocket, but two of those launches have had anomalies with the solid rocket boosters supplied by Northrop Grumman. Neither company has the cadence needed to keep up with Amazon's demand, much less the rest of the industry.

This reminds us less of a failed telecom buildout and more of the early days of cloud computing, when AWS had to build the data center capacity ahead of the then-current market demand. Amazon is today doing something analogous in space: paying for companies to scale launch capacity ahead of commercial or government demand, thus creating the commercial demand signal that Arianespace, Blue Origin, and ULA need to support their own production ramps. The Ariane 64 launch is proof of concept. Arianespace CEO Cavaillolès confirmed that Amazon Leo is Arianespace's largest commercial customer, with Caroline Arnoux, ArianeGroup's business unit director, quoted that two-thirds of the company's approximately 30-launch backlog is commercial. ArianeGroup is already working on next-generation competitiveness improvements specifically because demand from customers like Amazon justifies the investment.

On February 10th, the FCC approved Amazon's second-generation constellation expansion to approximately 7,700 satellites, including Gen2 and polar coverage systems. Analysts reading the tea leaves are speculating that because the new approval’s deployment milestones stretch to 2035, the FCC was showing confidence in Amazon's long-term trajectory even while the interim Gen1 deadline remains under review. The timing, two days before the Ariane 64 launch, could therefore be read as a deliberate signal rather than coincidence. Maybe.