SpaceX, the company best known for landing rockets on small boats and occasionally launching cars into space, has decided that disrupting the telecom industry is the next logical step. Last week, SpaceX agreed to purchase a big chunk of low-band spectrum - 800 MHz, to be exact - for $8 billion. This is on top of the nearly $20 billion it already spent on EchoStar spectrum last year. Combined with the FCC's approval to deploy 15,000 next-gen Starlink satellites, SpaceX is now officially serious about competing with AT&T, Verizon, and T-Mobile. Or at least serious enough to make their stocks twitch.
To understand this gambit, we turn to Walter Piecyk, a partner and analyst at LightShed Partners who covers tech, media, and telecom. Piecyk has been predicting this move for months, and he joined The Verge's Decoder podcast to explain why SpaceX is buying up spectrum like it's going out of style and why the existing telecom giants are, in his words, "circling the wagons." Spoiler: they're panicked.
The spectrum in question is the old Nextel spectrum, which Sprint had to divest when it merged with T-Mobile. It's been sitting around, unloved, until Elon Musk showed up and bought it. "Nobody else wanted it," Piecyk notes. But SpaceX sees it as key to building a terrestrial network to supplement its satellite coverage. Because satellites are great for beaming broadband to a dish on your roof, but they're not so great at penetrating buildings or connecting to a phone with limited power. Hence the need for low-band spectrum, which travels further and provides better coverage.
This is where the telecom incumbents come in. They could have bought this spectrum themselves - Piecyk suggests Verizon or AT&T could have snagged it for, say, $8 billion. Instead, they're now losing more in enterprise value over SpaceX fears than it would have cost to just buy the spectrum and sit on it. "I don't think even these companies were taking Elon seriously and thinking that he would actually build the network," Piecyk says. Oops.
But can SpaceX actually pull this off? Building a terrestrial network is expensive - Piecyk estimates $10 - 15 billion for the initial build - and SpaceX will need to convince one of the big three carriers to let them roam on their networks as an MVNO. The big three are not exactly eager to help. "They're not going to give them a wholesale agreement," Piecyk says. "So they're going to overcharge them for that." Still, Piecyk thinks SpaceX can get in the market sooner rather than later, perhaps via roaming agreements, and then build out capacity over time. "You can't just turn on a network overnight and start selling service on that," he says. "The point is it's going to take time."
Meanwhile, the existing carriers are doing their best to downplay the threat. T-Mobile, which partnered with SpaceX on direct-to-device connectivity, has seen its relationship with SpaceX sour. "T-Mobile executives have been throwing shade at SpaceX incessantly over the past couple of months," Piecyk says. They've also formed a joint venture with other carriers to pool satellite spectrum, which Piecyk compares to cable companies deriding fixed wireless as "cell phone internet" before losing customers to it. "There are some similarities now in terms of how these executives are downplaying what they can do," he says. "Is it going to be costly? Is it going to be complex? Sure. But so is catching a rocket with chopsticks."
Verizon, in particular, is vulnerable. Piecyk notes that Verizon made bad bets on millimeter wave spectrum and media acquisitions (remember Tumblr?), and its network advantage has evaporated. "The networks are effectively commoditized," he says. "These networks are basically the same. And they're getting 60 percent margins." That's why AT&T's strategy of bundling fiber with wireless looks smarter - it's harder for SpaceX to compete with a high-value, low-churn fiber network. Verizon, meanwhile, is cutting costs and firing people to make EBITDA look good, but Piecyk warns that "at some point you have to show and have a strategy for some type of revenue growth." Good luck with that.
The interview also touches on the broader telecom landscape: the failed promises of 5G, the missed opportunity to have four carriers in the US, and the looming paradigm shift in mobile devices. Piecyk notes that if AI agents and new form factors reduce our dependence on phones, it could disrupt the telcos' subsidy models. "Whenever there's change, there's an opportunity for disruption and for margins to contract," he says. And could SpaceX make a phone? Piecyk doubts it, but he acknowledges that in the narrative game, it's a possibility. "Why not throw that in the hopper?" he says. "The bigger issue is someone's going to build something that may resonate."
Ultimately, Piecyk believes SpaceX's move into mobile is a matter of when, not if. "At some point, enough money will go back into Starlink Mobile to pose a real threat to AT&T, Verizon, and T-Mobile," he says. "And that is just a matter of time." The telecom industry, which has long been a cozy oligopoly, is about to get a rude awakening. Or, as Piecyk puts it, "They're having the meetings. They are certainly not as loud about the next round of investment as they would normally be in the cycle. Now they have a competitor."