SpaceX's next move isn't Mars. It's your phone signal.

Those signal bars on your screen? SpaceX thinks that arrangement is up for negotiation.

On June 26, the FT reported that Shotwell told IPO investors the company is evaluating a Starlink retail mobile service and potentially building its own terrestrial cellular network. [FT, June 26, 2026] Three things followed: T-Mobile's CEO admitted Starlink direct-to-cell usage was "lower than expected." [Light Reading, April 29]; AT&T, Verizon, and T-Mobile formed a joint venture to respond together [AT&T, May 14]; Oppenheimer called SpaceX a "1.6 trillion U.S. communications market" disruptor [Reuters, June 3]. SpaceX had already spent ~20B on EchoStar's mobile spectrum. [CNBC, Sept 8, 2025]

But the real question isn't "when will SpaceX become a carrier." It's two deeper ones.

One: A space company with consumer subscribers is a different asset class.

Rocket Lab's revenue comes from contracts—every dollar requires a new bid, a new execution. SpaceX's $11.4B Starlink revenue comes from 10.3 million people paying monthly. Next month's revenue is nearly predictable.

Contract-based companies grow arithmetically. Subscription-based ones can grow geometrically. Starlink's ARPU dropped from 91 (2024) to 66 (Q1 2026), but its user base grew from 4.6M to 10.3M. ARPU down 23%, revenue up 58%. That math only works with subscriptions.

One nuance: consumer reach itself isn't the premium. Proven, scaled consumer reach is. AST SpaceMobile also does direct-to-device, but without millions of paying users, it's priced as a project-based company.

For investors: if a space company demonstrates recurring revenue—subscriptions, data-as-a-service, long-term industry contracts—its valuation shifts from "contract multiple" to "recurring revenue multiple." The gap between those two multiples may be the largest source of alpha in space.

Two: The same playbook is running in China — just different actors.

SpaceX's "be your own carrier" playbook requires three things:

• Spectrum. SpaceX spent ~$20B buying it from EchoStar.

• Consumer access. Starlink terminals sell directly to end users.

• Capital market fuel. A $1.35T valuation stock pays for spectrum and acquisitions.

It happens in a similar way to a different extent in China, by telecom, device provider, and listed tech firms.

Investor implications:

Private markets: 10 companies in the IPO queue, zero approvals. The bottleneck isn't a lack of good companies — it's the absence of a proven recurring revenue model. Changguang Satellite raised 5B. LandSpace raised 1.3B. Capital isn't the problem. The first Chinese space company to prove recurring revenue at scale will find the IPO window widening. Policy follows precedent.

Public markets: Alpha in space isn't about who launches more rockets. It's about who completes the contract-to-subscription transition. Rocket Lab vs. SpaceX has already demonstrated the valuation gap.

One open question: will China take SpaceX's path or carve a different one entirely? The answer depends on the first success case walking itself out.

No matter what, the second half of the AI race partly belongs to whoever controls the interface—and space infrastructure is a critical piece of that. More to come.

LU FRONTIER STRATEGIES

© 2026 LU Frontier Strategies. All rights reserved.