The satellite industry’s favorite word right now is resilience. The argument sounds simple enough: if one orbit fails, use another.

That logic has pushed operators, integrators and defense customers toward multi-orbit systems that combine low Earth orbit, medium Earth orbit and geostationary satellites into a single connectivity strategy. One industry paper from Avanti Communications put it plainly, saying that multi-orbit satellite connectivity had taken centre stage as GEO incumbents added NGSO capacity and new LEO systems expanded the market.

Contrivian, a San Francisco connectivity startup, is pushing a different thesis. It argues that resilience for modern internet applications may not come from jumping between different orbits, but from routing traffic across multiple low Earth orbit networks that behave more alike.

The company’s product, Contrivian Constellation, is pitched as a software-defined layer that unifies Starlink, Amazon Leo and other LEO providers into one managed service. In its own description, Contrivian says the platform continuously measures and steers traffic across available paths in real time.

That is a narrower claim than saying multi-orbit is useless. It is also a sharper one. Contrivian is effectively saying that the industry has been solving for the wrong kind of diversity.

The physics problem inside the resilience pitch

The technical dispute begins with latency.

LEO satellites sit close enough to Earth that broadband links can feel closer to terrestrial internet. GEO satellites, by contrast, sit far higher above the planet. That distance creates a latency penalty that cannot be engineered away entirely.

An Avanti comparison table lists typical latency of 25 to 100 milliseconds for LEO, 100 to 200 milliseconds for MEO and more than 500 milliseconds for GEO. Those figures vary by system and routing path, but the hierarchy is not controversial: farther orbits usually mean longer signal travel time.

That gap matters during failover, when a network automatically shifts traffic after congestion, interference or an outage. If an application moves from a low-latency LEO path to a much higher-latency GEO path mid-session, the application does not experience that as abstract redundancy. It experiences it as a different network.

Video calls can freeze. Cloud sessions can stall. Interactive tools can become sluggish. Some connections may reset entirely. The point is blunt: there is no normal TCP-based internet application that improves when latency suddenly jumps by an order of magnitude.

Contrivian wants to redefine resilience

The multi-orbit camp tends to define resilience as orbital diversity. If one orbital layer is congested, jammed, degraded or politically unavailable, another layer can pick up the load.

Contrivian’s counter-argument is that resilience for high-bandwidth, low-latency applications should be provider diversity inside the same orbital layer. Instead of falling back from LEO to GEO, its preferred model is to move between LEO providers whose performance profiles are close enough that applications do not have to adapt to a dramatic step-change.

That is why its marketing focuses on active-active multi-constellation LEO rather than classic multi-orbit failover. On a separate government and special operations page, Contrivian describes its system as active-active multi-constellation LEO SATCOM for government and SOF use cases in denied, degraded and disrupted environments.

The field kit is positioned as portable and ruggedized. Contrivian describes a self-contained terminal designed for contested environments, with multiple LEO networks running simultaneously rather than waiting for one connection to fail before another takes over.

The practical promise is simple: if every available path sits inside a broadly similar latency envelope, live applications have a better chance of surviving a network event without the user feeling the handoff.

The bet depends on a more crowded LEO market

The weakness in Contrivian’s argument is not the physics. It is the market.

For now, there are still only a limited number of mature global LEO broadband options. Starlink dominates the category. Eutelsat OneWeb exists as a more enterprise- and government-oriented LEO network. Amazon Leo, formerly Project Kuiper, is still deploying.

Amazon says Amazon Leo is its low Earth orbit satellite network, powered by more than 3,000 satellites and intended to extend fast, low-latency broadband to locations beyond existing networks. That makes Amazon central to Contrivian’s future-facing pitch, but it is not yet a mature global peer to Starlink.

Blue Origin has also entered the conversation with TeraWave, though it should not be described as a pure LEO broadband clone. Blue Origin says TeraWave will use 5,408 optically interconnected satellites across LEO and MEO, aimed at enterprise, data center and government users. Deployment is scheduled to begin in the fourth quarter of 2027.

Telesat is another possible piece of the future LEO market, but its schedule has moved. Via Satellite reported that Telesat’s Lightspeed service has slipped to 2028 after delays tied to chips used in the satellites’ onboard processors.

That means Contrivian is selling into a market that is still partly theoretical. Its software becomes more valuable as the number of viable LEO networks grows. If the market remains concentrated around one dominant operator, the case for multi-constellation orchestration is harder to prove at scale.

The Pentagon’s concentration problem

The military angle is why the argument matters beyond commercial telecom strategy.

Defense users have powerful reasons to want alternatives to any single commercial satellite provider. A communications architecture that depends too heavily on one company, one terminal ecosystem or one political decision-maker creates obvious operational risk.

Multi-orbit systems emerged partly as a hedge against that kind of dependency. If a LEO service is unavailable, the thinking goes, MEO or GEO capacity can keep the mission connected.

Contrivian’s response is more surgical: concentration risk is real, but it is a provider problem, not always an orbit problem. From that view, adding GEO or MEO into the same automated failover fabric may provide backup connectivity, but it can also introduce a performance shock that breaks the applications the network is meant to protect.

That does not make GEO irrelevant. GEO satellites remain useful for broad coverage, fixed-position services, broadcasting and mission profiles that value persistence more than low latency. MEO systems also have their own role, especially for high-capacity backbone and specialized connectivity.

The real debate is not whether higher orbits should exist. It is whether all orbits should be blended into one supposedly seamless automated service for applications that are highly sensitive to latency.

The software layer is becoming the fight

Contrivian’s argument also points to a larger shift in satellite communications.

As more constellations enter service, the strategic value may move away from the satellite operator alone and toward the orchestration layer that decides which network carries which traffic at which moment.

That is a self-interested forecast from a company that sells orchestration software. It is also consistent with how terrestrial networking evolved. When customers have multiple pipes, the valuable layer often becomes the system that manages performance, cost, failover and vendor lock-in without making users think about the underlying network.

For defense customers, that layer could become especially important. The ideal outcome is not loyalty to one satellite brand. It is the ability to keep operating when one provider is congested, degraded, jammed, unavailable or politically complicated.

Contrivian is betting that the next few years of LEO deployment will create enough real alternatives for that software layer to matter. If Amazon Leo, Telesat Lightspeed and other networks mature on schedule, the company’s case gets stronger. If they do not, the market remains too thin for the full promise of LEO-to-LEO routing to arrive.

That is what makes the argument interesting. The physics is settled enough to make the critique plausible. The market is not settled enough to prove the solution.