Are Chinese and American Space Networks Too Ambitious? A Reality Check of Mega-Constellations

The year 2025 has been a remarkable time for space with three new Low Earth Orbit (LEO) mega-constellations (1,000 satellites+) routinely launching into operational orbit. The first is U.S.-based Amazon Project Kuiper, which launched its first batch of 27 satellites in April, and another batch of 27 satellites in June, bringing total satellites in orbit, to date, to 54. China has also launched its fourth batch of the Guo Wang/Xingwang satellite Internet constellation, moving to a total of 34 commercial satellites in orbit. Chinese commercial space network Spacesail/Qianfan also launched its fifth batch of satellites into orbit, bringing the current network size to 90 satellites. Finally, SpaceX’s satellite network Starlink is closing in on 8,000 commercially operational satellites in orbit to date. Several more launches are planned by these networks before the year is out.     

While evolving trade policies are increasingly shaping network deployment plans, especially for operators navigating global supply chains or launching across borders, there appears to be no delay in chasing LEO supremacy. Expectations have been set as well. Each network has set lofty deployment goals (e.g., full or near-full deployment by 2030) and are required by the ITU to hit deployment milestones (e.g., 10% network deployment within 2 years, 50% within 5 years, and 100% within 7 years of regulatory approval). While there are few reinforcement mechanisms in place, operators could lose international recognition and protection for unused spectrum and orbital resources beyond what has been deployed. With some of these networks expecting a network size of tens of thousands of satellites by 2030, and billions of dollars of Capital Expenditure (CAPEX) and annual Operational Expenditure (OPEX), this begs the question: are these networks too ambitious?

LEO satellite networks, especially large mesh networks made up of thousands of satellites (mega-constellations), have generated a lot of buzz and enthusiasm around the industry. Watching the market’s launch patterns, however, suggest that these new networks face immense challenges to becoming commercially successful. The launch market, while radically more affordable and rapidly evolving, is nowhere near the level needed to see these networks reach full deployment by 2030. The significant deployment and maintenance costs of these networks also raise questions about the viability of their business cases.

Let’s take China’s Guo Wang network as an example. The network at full size is planned to reach 12,992 satellites in orbit by 2030. Each launch typically brings online 9 to 10 satellites, and so far, has only had three launches in 2025. While more launches are expected, around another 140 launches are needed between now and the end of 2026 to bring 10% of the network (1,299 satellites) online. This is roughly the number of orbital launches completed by the entire world in 2021 and around 3X the volume of launches done by the China Aerospace Science and Technology Corporation (CASC), China’s leading launch provider, in 2024. Assuming the same mission characteristics, the company will need to achieve roughly 1,439 orbital launches in 5 years to reach its 2030 goal. At this rate, even the company’s short-term goal of reaching 260 satellites (~25 more launches) by the end of 2025 seems like a long shot.

Likewise, with Spacesail/Qianfan, the network’s 15,000 satellite goal by 2030 will require 828 orbital launches in batches of 18 satellites. The company’s short-term goal of 648 satellites (31 more launches) by the end of 2025 also seems awfully optimistic. Note that this is assuming 100% launch success rate, does not account for replenishment, and is using the current variety of Long March series rockets, which has a reported launch price of US$27 million per mission. This puts the total launch costs for the initial deployment of Guo Wang at around US$40 billion and US$22 billion for Qianfan. Amazon has secured launch contracts with a variety of partners, including ULA, Blue Origin, Arianespace, and SpaceX for nearly all of its constellation (~90 launches), yet it still needs to secure more launch capacity for full network deployment—which is estimated at US$20 billion for the entire network. Major milestones are required by mid-2026 (323 satellites—10% deployment) and mid-2029 (1,618 satellites—50% deployment), but with Amazon’s launch diversification with domestic and international partners, the company minimizes risk against not meeting its network milestones.

While the pockets backing these systems are no doubt deep, launch cadence and capacity remain vital bottlenecks for the industry. SpaceX has been breaking records in the launch sector; however, it should be noted that SpaceX’s launch capacity is largely in service to support Starlink’s deployment, which is constantly replenished and upgraded. What’s more, if satellite network operators remain concentrated on only domestic launch capabilities for deploying their network, they risk capacity bottlenecks as new competitors fight for rideshare space on rockets. This is all without even looking at the bill for procuring the satellites, managing, and maintaining the network.

LEO networks are certainly brimming with promise, poised to usher in an age where you could Zoom call from a mountaintop or game with lightning speed in the middle of nowhere. (Just imagine the bragging rights!). However, it's becoming clear that the hype might be outrunning reality. Beyond their aggressive launch timelines, the financial realities of operating LEO networks are slowly being exposed to the market. It turns out that getting these networks into orbit and running is only the first hurdle. Consider Eutelsat's OneWeb constellation: it incurred a substantial €64.3 million in operating costs for just 6 months, ending June 2024. Roughly a tenth of the group’s €606.2 million revenue for the same period. Such high expenses inevitably prompt concerns about whether these services can be priced competitively and, ultimately, the sustainability of the LEO network business model. Perhaps this is why SpaceX remains so tight-lipped about its financials.

ABI Research estimates that Starlink’s satellite Internet subscriber base will grow to 6.7 million subscribers, excluding Direct-to-Cellular (D2C) subscriptions (smartphone messaging), and generating US$7.4 billion in service revenue for 2025. Despite this, another US$3 billion to US$4 billion will go to launch and billions more to manufacturing, operations, Research and Development (R&D), and more during the year. Therefore, it appears unlikely that even the space industry “behemoth” is making significant profit (if any at all) at this stage, and the company’s revenue streams must expand to maintain the network’s momentum. This is likely why SpaceX is expanding into cellular backhaul, D2C, and Internet of Things (IoT) markets, and perhaps why it initially planned such a massive constellation of 42,000 satellites. These moves would help it maintain market momentum and continue to market its rockets effectively. At its core, SpaceX has always been about “the rockets.” The satellites? They are just a convenient way to keep the lights on at SpaceX and advance Elon Musk’s long-term goal of getting to Mars.

To achieve more favorable economics and higher launch frequency for these networks, larger rockets capable of launching bigger batches (e.g., Starship and Long March 9) and reusable rockets (e.g., Falcon 9 and China's Tianlong-2) are proving essential. Despite these challenges, players are now entering the market with the “reverse philosophy”—putting the network and connectivity business first. However, the market is quickly learning (or soon will) that building a global network, whether on Earth or in orbit, is a strategic goldmine, nationally intertwined, and eye-wateringly expensive. But hey, it's all worth it to connect every corner of the planet with a solution that's accessible by everyone.