Amazon’s Project Kuiper is far behind Starlink and is under time pressure, but Amazon has several things going for it.

In 2019, I wrote that Amazon would be a formidable satellite-ISP competitor. I still think so, but I didn’t expect it would be over four years until they launched the first test satellites. In the meantime, SpaceX has put over 5,000 satellites in orbit and has over two million Starlink customers.

Amazon has permission to launch 3,236 satellites. They must manufacture and launch at least half of them by July 2026 and the remainder by July 2029. Can they do it? After many delays, they have finally launched two test satellites, confirming that inter-satellite laser links (ISLLs) worked at 100 Gbps while sending traffic “in both directions from the internet over an AWS fiber-optic connection to our ground gateway station, up to our satellites, and then down to a customer terminal at our test location.” All Starlink satellites launched since September 2021 will have ISLLs, so by the time Kuiper is complete, July 2026, all or nearly all Starlink sats will have them and they will have a much larger constellation.

Amazon has not launched any production satellites, and they will have to hurry to meet the 2026 and 2029 deadlines. They have signed contracts for 83 launches over a five-year period, which they say will provide capacity for “the majority” of the constellation. SpaceX was conspicuously not one of the vendors, and a shareholder lawsuit pointed out that Amazon had not considered SpaceX as a provider and nearly 45% of the overall value is for launches and engines from Blue Origin, a rocket company founded by Jeff Bezos. Subsequently, Kuiper signed a 3-launch contract with SpaceX.

(Note that Blue Origin has not yet launched their forthcoming New Glenn rocket, which was initially scheduled to fly in 2020. The New Glenn will have greater capacity than SpaceX’s current Falcon rocket but significantly less than their forthcoming Starship).

Amazon’s Project Kuiper is far behind Starlink and is under time pressure, but Amazon has several things going for it:

• In his first letter to stockholders, Jeff Bezos stressed that Amazon was an infrastructure company and that has been borne out by subsequent investments in facilities and services. Amazon will bundle Kuiper access with data storage and cloud computing services.
• Kuiper will offer service-level agreements to non-consumer customers.
• Amazon will be Kuiper’s largest customer. With over 1.5 million employees staffing offices, warehouses, and other facilities, Amazon will use the Kuiper constellation internally as will their fleets of delivery trucks, planes, shipping containers, and perhaps delivery drones someday.
• Amazon is already in the space business with its satellite ground station service.
• The US, Taiwanese, and other governments and militaries will see Amazon as a more reliable supplier of critical infrastructure than Starlink given Elon Musk’s political activism and Tesla’s dependence on China.
• Some potential customers may not approve of Elon Musk’s political involvement.
• Re-usability gives Starlink a large launch-cost advantage, but if Elon Musk can afford Twitter, Jeff Bezos can afford Kuiper.
• Amazon’s New Glenn rocket is designed to be reusable, and eight Chinese private and state-owned entities are developing reusable rockets.
• Kuiper will be launching state-of-the-art satellites and selling state-of-the-art terminals.
• Satellite antennas are expensive, and Amazon has experience designing and manufacturing consumer devices like the Echo and Kindle.
• Amazon has announced three Kuiper antennas.
• A user terminal is more than just an antenna and Dave Täht, Chief Science Officer at LibreQoS, has been calling attention to Starlink’s latency problem for years. It now seems Elon Musk is ready to listen. Amazon should talk with Täht.
• Amazon is already talking with enterprises, governments, schools, hospitals, and mobile operators. They have contracts with Verizon, Vodafone, and NTT and licenses to operate in more than 15 countries, including Brazil, Canada, France, Mexico, and the US.

I could go on, but you get the idea—I think Kuiper will survive despite a rocky start and will eventually offer Starlink healthy competition.

Update Feb 9, 2024:

I listed Jeff Bezos’s wealth and the expected use of the Kuiper constellation among the causes of my optimism. A report on the capital expenditures by the three major cloud service companies puts Amazon’s commitment to invest $10 billion over several years in Project Kuiper in context. In 2023 alone “Amazon’s (relative) CAPEX austerity continues, as the company spent a measly $53.7 billion, a decline of 20%” and it has invested around $380 billion since 2000.