Scaling Direct-to-Device Connectivity: LEO Constraints, Spectrum Limits, and the Shift to Mass-Market IoT & Mobile

Direct-to-device satellite connectivity is nascent and complementary, focused on SOS, messaging, and very narrowband IoT, not media streaming or telemedicine. Scale stands at about half a billion users and revenue of half a billion, concentrated in remote and uncovered areas. Commercial use exists in the U.S. and Canada; Australia is supportive; Japan is under review; China is closed. Only about 10% of realistic countries are considering dual use today. In the U.S., ARPU is $50, and customers pay $3–$5, or $10 on mid-tier plans, for D2D.

Satellite beams span 300 kilometers versus 5 to 10 kilometers for mobile cells, making D2D roughly 20 times less efficient. Per-user throughput is less than a megabit, while market demand is 15–20 Mbps; LEO latency is 20 to 40 milliseconds. D2D spectrum is less than 100 megahertz versus 1 to 2 gigahertz for LEO broadband; players fight for tens of megahertz. Pricing is $1 per GB, while vertically integrated cost can be less than 5 cents per GB. In India, usage exceeds 30 GB per user per month, and terrestrial pricing is 10 rupees per GB, or 10 cents.

Spectrum conflict is the first structural break, steering MNO cooperation and regulatory outcomes. The FCC allows mobile spectrum for satellite; India is not allowing D2D, and border spillover requires geofencing. Time to market is critical with five to six‑year LEO lifetimes, while handsets await NTN’s three to five‑year 3GPP cycle. Uptake is 2028 to 2030; enterprise IoT leads next three years, with automotive the biggest driver. Value concentrates in harmonized spectrum and scaled constellations: 40 megahertz acquired for $17 billion, and Globalstar sold for $11.6 billion on $230 million revenue.