Managed Connectivity, Integrated Space Platforms, and Bundled Mission Architectures Reshape Orbital Economics

Managed Connectivity Platforms Replace Standalone Bandwidth Sales

October 21, 2026 marked a clear pivot from selling raw satellite capacity to carrier-controlled service layers: AST SpaceMobile and AT&T demonstrated a seamless handoff of an active AT&T voice call from terrestrial network to AST’s BlueBird direct-to-device satellite network and back, using Ericsson radio/core technology and Nokia mobility software. AT&T’s revenue-sharing agreement with AST through 2030 reinforces the point: recurring value is shifting toward network-controlled mobility, not coverage claims alone, even though nationwide scale is still unproven.

Amazon’s planned Leo system points the same way. After the FCC cleared Amazon’s $11.5 billion Globalstar acquisition, giving it about 25 MHz of spectrum, Amazon is planning up to 5,105 satellites for deployment from 2028 while leaving carriers in control of subscriber, billing, and service relationships. SpaceX is also moving toward managed hybrid service, with Starlink Mobile negotiating access to at least 80,000 U.S. tower sites and backed by Qualcomm Dragonwing and a Layer 1 agreement for a hybrid satellite-terrestrial 5G rollout beginning in 2027. Eutelsat is emphasizing partner-led distribution through AWS-based cloud workflows. The strategic edge is shifting to orchestration, interoperability, and recurring platform control.

Integrated Space Platforms Are Absorbing Critical Infrastructure

Rocket Lab’s Q2 2026 results show where value is moving: total revenue hit $234 million, with Space Systems at $189.5 million, or about 81%, versus $44.6 million from Launch Services. Space Systems grew roughly $91.6 million year over year while launch fell only about $2.1 million, confirming that integrated spacecraft, components, and recurring services are now the main growth engine even as Neutron’s maiden flight slips to 2027 on manufacturing issues.

The same consolidation is playing out in communications infrastructure. SatPort is expanding carrier-neutral ground-segment services through build-to-suit landing stations, design, construction, installation, data management, maintenance, and regulatory support; its 2026 Telesat deal covers globally distributed landing stations while Telesat keeps network ownership and operations. Rocket Lab’s expanded KSAT service adds access to more than 200 antennas plus automated pass scheduling and S-/X-/Ka-band support. SES is bundling GEO, MEO, and LEO assets with gateways, terminals, edge equipment, hosted payloads, and guaranteed-throughput services. The strategic takeaway: buyers are shifting from standalone hardware and capacity to managed, end-to-end platforms, and vendors that control more of the stack are capturing the margin.

Orbital Infrastructure Is Becoming a Data-Product Stack

October 1 and November 4, 2026 marked a clear shift from orbital hardware demos to the infrastructure needed to sell space-derived intelligence. Google moved Project Suncatcher from concept toward prototype by flying four TPUs in orbit on roughly 1 kW of solar power, running Gemma 4, and proving on-orbit AI inference under tight thermal and power limits for about 20 minutes per session. Satlyt’s NASA-funded test was narrower, but still important: one physical spacecraft computer modeled four hosts, four users, and eight services, showing onboard workload planning rather than true multi-satellite execution.

SES’s meoSphere Pathfinder 1 then repeatedly demonstrated a bidirectional optical link between its MEO payload and Cailabs’ ground station in Rennes using a Mynaric Condor terminal, validating acquisition, pointing, tracking, and high-capacity transfer in an operational setting. Downstream, Windward bought Azimut.ai to add visual intelligence, while SkyFi launched a satellite data app on Robinhood.

The competitive center of gravity is moving from selling orbital assets or bandwidth to owning the stack that turns space assets into usable intelligence. Value is concentrating in optical terminals, orchestration software, edge compute, and recurring data products, not commodity capacity alone.

Defense Procurement Is Moving to Bundled Mission Architectures

The Space Force’s June 2026 SAM.gov notice for up to five “seed vehicles” in a proliferated cislunar constellation shows governments are buying mission stacks, not standalone satellites. Targeted for late 2029, the effort is meant to extend military space services beyond GEO with space-domain awareness in complex multi-body orbits, positioning/navigation/timing without GPS-like infrastructure, and resilient communications and battle management; AFRL’s Oracle program is cited as an initial path.

That cislunar logic mirrors proliferated LEO procurement but with fewer, more autonomous spacecraft because of harsher dynamics and longer communications delays. SpaceX’s reported defense awards reinforce the shift: a $2.29 billion Space Force contract for SDN-B, a secure LEO data-transport constellation with an operational prototype due by end-2027, and a separate $4.16 billion Space Force contract for SB-AMTI satellites to detect and track airborne moving targets. Viasat’s Nov. 6, 2026 IDIQ, capped at $450 million with an initial $115 million task order, adds a Mission Operations Center for SDN-B to integrate multiple satellite providers and manage resilient network operations. The buying pattern now favors bundled architectures that combine space segment, protected relay, and operations layers, creating value for vendors that can own integration and mission assurance, not just hardware.