SpaceX's satellite direct-to-device service can fill coverage gaps, but large-scale mobile service will still require terrestrial networks

Overview

The report examines whether SpaceX can expand Starlink's satellite direct-to-device capabilities into a mass-market mobile communications service. Bernstein recognizes the size of the market and the technical improvements offered by V2 satellites, but concludes that a satellite-only network will struggle to meet mainstream users' requirements for high speeds, stability, indoor coverage, and continuous connectivity. The eventual model is more likely to combine supplemental satellite coverage with incumbent carriers' terrestrial networks.

Core views

The report first distinguishes SpaceX's overall investment thesis from the viability of its mobile D2D business. Bernstein previously attributed SpaceX's primary potential to achieve a multi-trillion-dollar valuation to its leading position in space launch and the AI opportunity represented by orbital data centers; meanwhile, SpaceX's current revenue and profit still primarily come from its Starlink connectivity business. The firm remains optimistic about the growth potential of global consumer and enterprise broadband but has consistently been more cautious about the Mobile direct-to-device business. SpaceX as a whole remains rated Outperform with a target price of $248, while this report focuses on explaining why its mobile business faces significantly greater challenges than fixed broadband. The mobile communications market is large enough to motivate continued investment by SpaceX. The report estimates that the US consumer broadband and mobile markets will generate more than $400B in combined annual revenue and approximately $160B in EBITDA in 2026, with the mobile market that SpaceX hopes to enter accounting for more than $300B in revenue; approximately 55 major broadband and mobile markets worldwide generate more than $1T in combined annual revenue. On its Q2 earnings call, SpaceX reiterated its plans to develop a mobile business, aiming to launch service by the end of 2027, and has already partnered with mobile network operators in more than 30 countries. The company has also proposed placing Starlink broadband and future mobile services under a single account, offering users bundled packages similar to those of incumbent carriers. The report therefore believes terrestrial carriers cannot ignore this initiative, but commercial success must extend beyond filling coverage gaps in remote areas or signal dead zones. Users' expectations for their primary mobile network create a very high competitive barrier. AT&T, Verizon, and T-Mobile effectively cover nearly the entire US population, and consumers expect their phones to work continuously in homes, office buildings, on highways, in subways, and while in transit, with network handoffs that are secure, seamless, and reliable. Global fixed broadband and mobile data traffic continues to grow annually, driven primarily by mobile video, so network capacity requirements are also rising; the report also notes that annual US wireless voice traffic exceeds 2 billion minutes. Incumbent carriers must invest tens of billions of dollars annually to maintain coverage and capacity. Current D2D can provide valuable supplemental coverage in remote areas but still struggles to achieve the throughput, reliability, and indoor coverage expected of a primary wireless service provider. The foremost limitation of satellite-only D2D is wireless link loss caused by distance. Terrestrial base stations are generally located hundreds of meters to several kilometers from a handset, potentially approximately 1 kilometer in urban environments, while a low-Earth-orbit satellite may be approximately 530 kilometers from the handset. Free-space path loss means that, absent interference, received power declines with the square of distance; the mid-band spectrum used by Starlink, including PCS, also causes received power to decline with the square of frequency. Building obstructions further weaken an already constrained satellite link. The report estimates that the attenuation of PCS signals through concrete walls can be 50% higher than for the 700 MHz low-frequency band. V1 Mobile may support text messaging in indoor environments with some obstruction, but high-uplink-load applications such as video calls require greater throughput and bandwidth and raise the satellite-side noise floor, requiring the handset to transmit a stronger signal. As a result, link reliability is insufficient to support mass-market service. V2 Mobile will improve but not eliminate these limitations. The report expects the V2 satellite receiver to be approximately 4 times the size of V1, with higher beam density, a reduction in orbital altitude from approximately 525—535 kilometers to approximately 325—350 kilometers, and higher-gain antennas. These changes can significantly improve free-space path loss and reception capability. However, weak signals are more susceptible to errors and transmission failures, and antenna gain can only amplify a signal, not guarantee the restoration of its clarity; packet reconstruction and error-correction capabilities also have limits. Low-Earth-orbit satellites also move rapidly across the sky, requiring handsets to perform continuous handoffs and dynamic tracking. Incorporating the electronic beam-steering capabilities of Starlink fixed-broadband terminals into handsets constrained by size, power, and thermal management is significantly more difficult. Handset power consumption is the second practical constraint. Satellite uplinks generally require handset modems to transmit at higher power than under strong terrestrial connections, increasing battery consumption and heat generation. Under normal circumstances, the screen, CPU, and applications are the primary sources of handset power consumption, while uplink transmission accounts for a relatively small share; however, in an indoor satellite D2D scenario, signal transmission becomes a non-negligible component of power consumption. Maintaining a continuous satellite connection throughout the day could shorten battery life; when building obstructions or interference occur, service may be interrupted, and the handset may increase transmission frequency and power to reconnect or maintain the link, further consuming battery power. Beyond technology, spectrum regimes also constrain the speed and scope of deployment. Mobile communications spectrum is generally licensed separately by each country to terrestrial carriers. In most markets, satellite service providers cannot independently offer D2D service and must obtain spectrum-sharing arrangements or partner with local mobile network operators. This means the satellite ecosystem remains substantially dependent on incumbent carriers. The report therefore concludes that satellite direct-to-device service is more likely in the near term to serve as a complementary layer to terrestrial networks, primarily for remote and underserved areas, emergency communications, and the elimination of coverage gaps, rather than disrupting or replacing the wireless communications industry. If SpaceX is to provide a comprehensive mobile service capable of competing with nationwide carriers, the report identifies three paths: build, acquire, and partner. Building would require obtaining sufficient nationwide spectrum and constructing or leasing a large number of macro sites and small cells to meet high-density urban coverage needs. It is technically feasible but capital-intensive and highly time-consuming. The acquisition path would involve purchasing a carrier with an existing network and spectrum, with the report citing T-Mobile as an example. The partnership path would involve signing an MVNO agreement with a carrier willing to open its network, with SpaceX paying the counterparty network-access fees. Comcast's and Charter's existing MVNO arrangements with Verizon demonstrate that this model already has industry precedent. SpaceX has also discussed using installed Starlink user terminals to build distributed small cells, but the specific plan remains unclear. Considering cost, regulatory, and technical constraints, Bernstein continues to believe that partnership is the most reasonable approach, while neither satellite-only D2D nor a traditional greenfield build offers an obvious comprehensive solution.

Analysis framework

Bernstein first assesses commercial attractiveness based on the size of the mobile and broadband markets, user requirements, and the intensity of network investment. It then analyzes the performance ceiling of satellite-only D2D through free-space path loss, frequency bands, obstructions, antennas, link budgets, and handset power consumption. The report subsequently compares three market-entry approaches—building, acquiring, and MVNO partnerships—and evaluates the capital, time, and execution challenges of each in light of spectrum licensing and existing network coverage.

Methodology notes

• (Method outside the taxonomy)

  Wireless link budget and free-space path loss analysis

  The report assesses the signal strength ultimately received by the satellite by examining handset transmission power, propagation distance, frequency, building obstructions, antenna gain, and other factors, and uses this assessment to determine the viability of services such as text messaging and video calls.

• Competitive and Strategic FrameworkValue chain analysis

  Comparison of network-entry paths through building, acquiring, or partnering

  Across key components including spectrum, satellites, macro sites, small cells, user terminals, and carrier networks, the report compares the costs and constraints of SpaceX building infrastructure itself, acquiring an existing network, or purchasing access through an MVNO.

• Industry/Sector Analysis FrameworkSupply-demand framework

  Analysis of data traffic growth and network capacity requirements

  The report links data traffic growth driven by mobile video to coverage, capacity, and annual infrastructure investment requirements, illustrating the network performance and sustained capital investment required for mass-market mobile service.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• SpaceX (SPCX)

  The subject of the report; Starlink Mobile offers an opportunity to enter the large mobile market, but a comprehensive service will likely require terrestrial carrier networks.

  Strengths

  A leading position in space launch, an existing Starlink connectivity business, V2 satellite technology upgrades, and an established foundation of partnerships with carriers in more than 30 countries.

  Weaknesses

  Satellite-only D2D is constrained by distance, uplink performance, handset power consumption, antennas, indoor coverage, capacity, and spectrum licensing.

  Comparison

  D2D can supplement AT&T's, Verizon's, and T-Mobile's coverage in remote areas, but its current performance and reliability struggle to match their nationwide terrestrial networks.

  Risks

  If SpaceX cannot obtain terrestrial network and spectrum access at a reasonable cost, its mobile business may struggle to expand from filling coverage gaps into a mass-market service.

• AT&T (T)

  Both a nationwide US carrier that SpaceX plans to compete with and a potential type of terrestrial network partner.

  Strengths

  A mature terrestrial cellular network covering nearly the entire US population; rated Outperform with a target price of $25.

  Comparison

  Its existing coverage, capacity, and indoor service capabilities constitute the performance benchmark that SpaceX's satellite-only D2D must confront.

  Risks

  If SpaceX establishes a scalable mobile service, it could intensify competition in connectivity services.

• Verizon (VZ)

  A nationwide US carrier that provides network access to Comcast and Charter through existing MVNO agreements.

  Strengths

  Its mature terrestrial network demonstrates an MVNO partnership model that SpaceX could emulate; rated Market-Perform with a target price of $47.

  Comparison

  The report cites its partnerships with Comcast and Charter to show that the virtual operator path already has established industry precedent.

  Risks

  SpaceX's entry into the mobile market could create competition, while partnership arrangements would also impose access and capacity requirements on network operators.

• T-Mobile (TMUS)

  A nationwide US carrier; the report identifies it as an example of SpaceX's potential path to acquiring an existing network and spectrum.

  Strengths

  It already owns a nationwide network and spectrum; rated Market-Perform with a target price of $220.

  Comparison

  Acquiring an incumbent carrier could bypass some of the timing obstacles associated with a greenfield build, but the report does not view it as clearly superior to partnership.

• Comcast (CMCSA), Charter (CHTR)

  The two companies' MVNO agreements with Verizon provide an existing reference for SpaceX's partnership-based entry into the mobile market.

  Strengths

  They have established commercial models for bundling broadband and mobile services while using a carrier's network.

  Weaknesses

  They depend on partner carriers for underlying mobile network access.

  Comparison

  Their model resembles the MVNO path that the report considers most likely for SpaceX; CMCSA and CHTR are both rated Market-Perform, with target prices of $28 and $150, respectively.

• American Tower (AMT), Crown Castle (CCI), SBA Communications (SBAC)

  Their communications infrastructure is relevant to potential SpaceX plans involving terrestrial macro sites, small cells, or leased sites.

  Strengths

  AMT and CCI are rated Outperform.

  Comparison

  SBAC is rated Market-Perform; the report does not provide specific target prices or quantitative business comparisons for the three companies.

Key data

• US consumer broadband and mobile market revenueMore than $400B in 2026Estimated combined annual revenue from US consumer broadband and mobile services
• US consumer broadband and mobile market EBITDAApproximately $160BAnnual profit pool estimated by the report
• US mobile market revenueMore than $300BThe primary market SpaceX plans to enter
• Revenue of major global broadband and mobile marketsMore than $1TCombined annual revenue from approximately 55 major markets
• SpaceX partnership coverageMore than 30 countriesPartnerships have been established with local mobile network operators
• Planned Starlink Mobile launch dateBy the end of 2027The plan reiterated by SpaceX on its Q2 earnings call
• V2 receiver sizeApproximately 4 times that of V1Combined with higher beam density to improve reception capability
• V1 Mobile orbital altitudeApproximately 525—535 kilometersEstimated range used in the detailed link analysis
• V2 Mobile orbital altitudeApproximately 325—350 kilometersThe lower orbit helps reduce free-space path loss
• Comparison of distances to terrestrial base stations and low-Earth-orbit satellitesApproximately 1 kilometer versus approximately 530 kilometersUsed to illustrate the structural distance disadvantage of satellite connections
• Difference in frequency-band attenuation through concrete wallsPCS can be 50% higher than 700 MHzMid-band spectrum faces greater losses in environments with building obstructions
• US wireless voice trafficMore than 2 billion minutes annuallyUsed to illustrate the continuous availability requirements of mobile networks
• SpaceX rating and target priceOutperform, $248Bernstein's rating on SpaceX

Impact & implications

The report believes SpaceX's intention to enter the mobile market is credible and that incumbent carriers need to take its potential competition seriously. However, D2D will primarily expand coverage in remote areas and emergency scenarios in the near term rather than replace nationwide cellular networks. If SpaceX seeks to offer a mass-market mobile service, spectrum, indoor coverage, capacity, and capital investment requirements will force it to access terrestrial infrastructure. An MVNO partnership is therefore more likely to become the primary path than a satellite-only solution, a self-built nationwide network, or a major acquisition.

Risks

• The extremely long propagation distance from handsets to low-Earth-orbit satellites causes significant free-space path loss, potentially limiting uplink speeds and connection reliability.
• High transmission power, repeated reconnection attempts, and building obstructions could accelerate handset battery consumption and increase heat generation.
• Handset antenna size, dynamic tracking of low-Earth-orbit satellites, and continuous handoffs constrain the capacity and performance of satellite D2D.
• Spectrum in each country is primarily licensed to terrestrial carriers, making the speed and scope of SpaceX's deployment dependent on spectrum-sharing arrangements or local partnerships.
• Building a nationwide terrestrial network would require substantial spectrum, sites, time, and capital, with annual investment potentially reaching tens of billions of dollars.

What to watch

• Whether SpaceX can launch Starlink Mobile service by the end of 2027 as planned.
• V2 Mobile's actual throughput, indoor coverage, and power consumption with an orbital altitude of approximately 325—350 kilometers, a larger receiver, and higher-gain antennas.
• Whether SpaceX ultimately chooses to build, acquire, partner through an MVNO, or combine multiple approaches, as well as the related strategic details.
• Whether SpaceX can secure spectrum-sharing arrangements in various countries and expand its partnerships with mobile network operators.
• Whether the plan to use installed Starlink user terminals to build distributed small cells can provide a viable supplement to terrestrial coverage.