Profile: Starlink — LEO Satellite Broadband Network

Starlink delivers broadband internet via a dense LEO constellation at 25–60ms latency — structurally lower than geostationary alternatives — across six plan families: Residential (fixed-address home), Roam (portable land), Boats (recreational marine), Mobility (vehicle/fleet), Fixed Site (commercial ground), and Maritime (commercial vessel). The service's structural moat rests on SpaceX's vertical integration: it owns the launch vehicle (Falcon 9), the satellite factory, the ground network, and the subscriber relationship simultaneously, which compresses marginal satellite deployment cost below any third-party competitor.

Starlink reached 10.3 million consumer subscribers across 164 markets as of March 31, 2026 — a 105% year-over-year gain confirmed in the SpaceX S-1 filing — making it the fastest-growing ISP by subscriber count in history. Growth has shifted from saturated rural North America and Europe toward emerging markets, maritime, and aviation verticals, which now contribute a disproportionate share of revenue relative to their subscriber count. A majority of subscribers now reside outside the United States.

Starlink launched as a private beta in late 2020 with roughly 10,000 users and reached general availability progressively from 2021 onward, scaling subscriber count from 1 million (2022) to 2.3 million (2023) to 4.6 million (2024) to 9 million+ (end-2025) to 10.3 million (Q1 2026). The service has iterated through three satellite generations — V1, V2 Mini, and the pending V3 — while continuously restructuring its plan architecture, hardware lineup, and pricing tiers.

Starlink holds a commanding first-mover position in LEO consumer satellite broadband with no current direct peer at scale: the next largest LEO constellation (Eutelsat OneWeb) has approximately 656 satellites versus Starlink's 10,000+, and serves only enterprise and government through partners, not consumers directly. The first credible consumer-scale competitor, Amazon Leo (formerly Project Kuiper), entered enterprise beta in April 2026 with commercial availability targeted for mid-2026 and an FCC deadline requiring 700+ satellites by mid-2026. GEO incumbents Viasat and HughesNet compete on price and no-hardware-upfront models but are structurally disadvantaged on latency.

Starlink's revenue model is a hardware-plus-recurring-subscription structure across six plan tiers spanning consumer ($55–$130/month residential) to enterprise maritime (up to $1,265+/month). The Connectivity segment generated $11.39 billion in FY2025 revenue at a 63% adjusted EBITDA margin — the only profitable segment in SpaceX — and funds deep losses in the launch and AI segments. SpaceX's stated strategy, per the S-1, explicitly prioritizes subscriber volume and geographic expansion over ARPU maximization; blended ARPU declined from $99 (2023) to $66 (Q1 2026), but volume growth drove absolute revenue up ~50% annually.

Starlink's financial inflection is real and documented in a public S-1: 63% adjusted EBITDA margins at $11.4B revenue in 2025 is a structurally strong business. The forward tension is between two forces running in opposite directions — subscriber volume growth (which requires lower ARPU to unlock emerging markets) and ARPU recovery (which the May 2026 price increases began). The central long-run variable is Starship: SpaceX's S-1 explicitly ties V3 satellite deployment to Starship operationalization in H2 2026, and V3's larger capacity is required to sustain improving unit economics as volume scales. If Starship's deployment cadence slips, the current Falcon 9-based constellation faces increasing depreciation headwinds as it ages, while Amazon Leo builds toward a genuine second-mover competitor position.