The SpaceX Femtocell Announcement Is Not What It Seems

On its 2Q earnings call, SpaceX announced its intent to deploy femtocells in the U.S. market, using Starlink connections and its acquired spectrum from EchoStar (65 Megahertz (MHz) in the ~2 Gigahertz (GHz) band), which was originally positioned to allow Direct-to-Cellular (D2C) services from Starlink satellites, but did come with rights to be used in terrestrial networks. SpaceX has now announced it plans to add a small cellular radio to some existing Starlink terminals, making them low-power cell sites that will connect to smartphones and other devices the same way cellular base stations do. This means the addressable market for the new service is anyone with a modern smartphone. By doing so, SpaceX will eliminate the biggest challenges previously slowing down femtocell deployments:

1. Getting power to new terrestrial small cells has been a major challenge, because optimal network location does not always mean power availability. In the case of SpaceX, the power problem is solved because the small cells are positioned in existing Starlink terminal locations or even within the satellite terminal itself.
2. Backhaul is another issue, but again this is solved with the Starlink satellite connection.
3. SpaceX also vertically integrates its own hardware and software stack, and it can design and integrate a small cell quite economically, especially when silicon and Radio Frequency (RF) components are commoditized.

A key difference with traditional small cell deployments is the location of the small cell itself. In terrestrial networks, small cell locations are calculated by network planning teams where coverage and capacity are needed. SpaceX’s proposal means that the network location is decided by the location of the Starlink end user, which may not agree with network requirements at all. To understand this implication, it is necessary to dive deeper into coverage calculations.

From both theoretical and technical perspectives, the SpaceX proposal is feasible and even credible. However, this translates into a completely different scenario in the real world and to understand its implications, we need to dive into the small cell specifics and compare against a typical terrestrial coverage scenario. A typical residential small cell has a coverage radius of 20 Meters (m), which translates to 1,256 Square Meters (m2) of coverage, ignoring obstructions and other coverage blockers. According to ABI Research forecasts, there are currently 3.8 million Starlink residential broadband connections in the United States (excluding maritime and aviation) and if we assume every single one of these Starlink terminals is equipped with a small cell, then the estimated theoretical coverage of this network would be 0.05% of U.S. land mass and 1.8% of U.S. urban land mass.

However, these estimates would be much lower because the location of the small cell itself would most likely not be ideal (e.g., deployed in a living room instead of a rooftop). For nearly 100% coverage of the U.S. urban land mass, there need to be ~220 million small cells in the United States alone, and that does not guarantee coverage everywhere because there are buildings with multiple levels where a comprehensive network would need multiple small cells per building. Coverage is a costly and complicated affair even for a greenfield terrestrial network operator: Rakuten Mobile must now spend a significant amount to densify its network across Japan after KDDI decided to terminate their agreement for national roaming.

SpaceX is planning to use this announcement as a distraction to create Fear, Uncertainty, and Doubt (FUD) in U.S. telcos and perhaps forge a national roaming partnership that goes both ways: a U.S. telco subscriber will be able use the Starlink D2C network in remote areas whereas potential U.S. Starlink mobile subscribers will be able to connect to the U.S. telco network. Starlink already has a partnership for D2C in the United States (T-Mobile), but the question remains whether other U.S. telcos would be willing to form a more comprehensive partnership, beyond D2D  with Starlink that could potentially cannibalize their own subscription business. Or perhaps this was a reaction to the AT&T, T-Mobile, and Verizon partnership to solve the rural connectivity problem in May 2026.

A more pertinent question is why SpaceX should deploy femtocells because it currently serves the same market through its Starlink fixed broadband satellite service and Wi-Fi. It already connects smartphones to the Internet and with Wi-Fi calling, it can also provide voice and Short Message Service (SMS). Adding a small cell would only augment its value proposition marginally. ABI Research expects this announcement to be masquerading a different priority that will be exposed in the future, if at all.