Why Elon Musk wants to sell you cell service | Decoder

Summary

• SpaceX’s $8 billion purchase of 800 MHz spectrum signals a serious terrestrial challenge to the incumbent carriers. Following $20 billion of EchoStar spectrum purchases, Walter Piecyk sees the low-band acquisition as the missing ingredient for coverage and building penetration. “They can’t do this by satellites alone.”
• Piecyk thinks an initial ground network would probably cost $10–15 billion, but securing roaming access remains the harder entry constraint. His conditional path is to win 5–10 million customers and reinvest, rather than replicate an incumbent’s entire network immediately. Nilay Patel’s objection remains substantial: satellites, terrestrial infrastructure, roaming and consumer operations add up to “a lot of cost.”
• The proposed consumer advantage is reliable service in everyday coverage holes, rather than emergency satellite texting alone. “It’s the soccer field,” Piecyk says, pointing to suburban locations where opposition to towers leaves gaps. Whether that difference wins customers is unresolved: “Is that enough to differentiate with consumers? I don’t know.”
• SpaceX’s capital allocation could delay mobile competition without removing the threat from telecom valuations. Starlink broadband generates cash; the mobile business would initially consume cash and run negative EBITDA. Piecyk sees wireless as another opportunity within a roughly $1 trillion US telecom enterprise-value pool, but “money needs a narrative,” and space compute might attract Musk’s attention and investment first.
• Verizon is Piecyk’s most exposed incumbent because its premium-network identity has eroded without a replacement growth strategy. “The networks are effectively commoditized,” he argues, despite wireless EBITDA margins around 60%. AT&T’s fiber-and-wireless bundle offers a clearer retention strategy; Verizon faces spectrum constraints, investment needs and reliance on cost cutting.
• A fourth carrier could pressure margins, while incumbent resistance could invite regulatory intervention. Piecyk recognizes that refusing wholesale agreements is the carriers’ right, but thinks the FCC could become involved under its public-interest mandate. He also expects today’s dismissals of SpaceX’s limited spectrum to become awkward if incumbents later need to portray it as competition to justify mergers.
• AI devices could disrupt both carrier economics and Apple’s position, but neither speaker expects a SpaceX phone launch. If agents change how people consume information, handset prestige, subsidies and bandwidth demand could change together. Piecyk’s conditional framing: “Whenever there’s change,” there is an opportunity for disruption and margin contraction.

Deep dive

1. Low-band spectrum makes the terrestrial ambition visible

• Patel opens with the arithmetic: $8 billion for 800 MHz spectrum, following $20 billion of EchoStar spectrum purchases and SpaceX’s plan to launch 15,000 more satellites to support the network. Piecyk describes an evolving strategy: SpaceX bought spectrum, received positive investor feedback, then bought more. The attraction is a US telecom sector with roughly $1 trillion in enterprise value, alongside global ambitions.
• For Piecyk, low-band spectrum solves a ground-network problem: it travels farther and improves coverage and building penetration. Ordinary phones lack the power and dish used for Starlink broadband, so “they can’t do this by satellites alone.” His analogy is Sirius, whose satellite service also uses supplementary terrestrial towers.
• The 800 MHz holding originated with Nextel. The Sprint–T-Mobile merger required divestiture to support a fourth carrier, initially Dish. Dish could not finance its purchase rights, forcing an auction; the spectrum subsequently passed through Grain Management before SpaceX arrived.
• Patel presses the obvious question: if the spectrum was valuable, why did nobody else buy it? Piecyk cites limitations on incumbents’ use, but thinks AT&T or Verizon perhaps should have spent $8 billion defensively. Their enterprise-value losses from concern about SpaceX might exceed what blocking its access would have cost.

2. A foothold costs less than reproducing an incumbent overnight

• Piecyk argues Dish got a network built and had wholesale agreements, but Patel disputes that it could compete nationally, citing a PR email insisting, “Our network does exist,” despite Verge testing that could not find it in advertised cities. Piecyk says the phone he used mostly roamed onto AT&T and T-Mobile, and that the theory was to get customers first and backfill the network.
• The more favorable precedent is John Legere’s T-Mobile: acquire customers while promising improvement, then let investment catch up. Piecyk thinks SpaceX could similarly backfill capacity. But wholesale access is the principal launch obstacle; alternative roaming arrangements might provide entry sooner, at prices that cannot support a high-margin business.
• Patel’s cost objection is cumulative: 15,000 additional satellites, a terrestrial build and potentially overpriced roaming. Piecyk distinguishes entry costs from maintaining enormous established networks. Against claims of $100–200 billion, his estimate is probably $10–15 billion in initial terrestrial capex; incumbents’ annual spending of $10–15 billion supports much larger customer bases.
• Piecyk’s conditional bull case starts with spectrum spending of $10 billion, $20 billion, “maybe 30 billion,” an initial build, then 5–10 million customers that justify reinvestment. This is not 50% share overnight. Competitive pressure might eventually make Verizon accept wholesale revenue, while spectrum sales or FCC auctions provide further capacity.

3. Everyday coverage is the pitch, but macro towers remain necessary

• Piecyk shifts the satellite use case from an injured person texting in the Utah desert to “the soccer field.” Suburban opposition to cell sites leaves ordinary coverage holes. New constellations could provide a fuller cellphone experience than Apple’s basic messaging through Globalstar, but he preserves the commercial uncertainty: “Is that enough to differentiate with consumers? I don’t know.”
• Patel’s consumer test is long battery life and fast service wherever people go. He also challenges another round of IoT promises after similar pitches around 4G and 5G. Piecyk sees possible IoT revenue, broadband cross-selling and Grok distribution, with incumbent EBITDA margins around 60% providing room to compete at lower margins. Customer acquisition, fulfillment, subsidies and service still require substantial operations.
• Patel describes SpaceX’s proposed femtocells beside Starlink dishes, using satellite backhaul, then questions reliability and backup requirements. Piecyk allows that small cells could contribute, but expects macro towers for coverage, power and penetration. “It’s hard to see the network getting built without macro towers.”
• Existing tower REITs, led by American Tower, have space for additional antennas, making entry easier than erecting an entirely new tower estate. Piecyk also sees a possible shortcut in EchoStar’s bankruptcy: radios already installed on towers and compatible with purchased spectrum could be acquired cheaply. SpaceX has not shown interest in that route, he notes.

4. T-Mobile’s fraying partnership echoes an incumbent mistake

• Patel asks whether T-Mobile, Starlink’s original direct-to-device partner, is being discarded. Piecyk will not confidently assign responsibility: he does not know whether SpaceX is leaving T-Mobile or vice versa. He suspects the transition from Mike Sievert to a new CEO changed the relationship, but explicitly presents SpaceX’s possible frustration as a guess.
• Patel sees carriers circling the wagons through wholesale resistance and their own satellite collaboration. Piecyk hears echoes of cable executives dismissing fixed wireless as “cell phone internet,” before customer losses became painful. Investors are extrapolating from SpaceX’s execution elsewhere; whether applying that record to telecom is fair remains something “time will tell.”
• Patel questions why SpaceX would attach itself to an easily valued telecom business when rockets, Mars and AI attract expansive dream-based valuations. Piecyk rejects the either/or: connectivity adds another narrative and revenue opportunity. “They need money, right? And money needs a narrative.” Capturing part of telecom’s value would also shrink incumbent valuations, rather than simply transfer the entire pool.

5. Mobile consumes cash before it can subsidize other ambitions

• Patel offers a deliberately simple allocation test: a dollar invested in mobile might return $1.10, unlike money committed to speculative xAI returns. Piecyk’s correction matters: Starlink broadband generates cash, while mobile would initially be cash negative and run negative EBITDA. Building the network and signing customers must precede extracting that dollar of profit.
• Piecyk therefore sees capital allocation as a genuine constraint. Higher expected returns, investor excitement or the next space-compute catalyst could take priority over wireless. So could Musk’s attention: “What does Elon happen to care about the most in that moment?” Piecyk also rejects the idea that connecting Musk’s own businesses is the principal motive; he sees a revenue business.
• Delay helps incumbent reported results, but Piecyk thinks the stock-market threat persists over one, two or three years. SpaceX bulls could argue that an initial build and 5–10 million customers would put pressure on incumbents, potentially leading a company like Verizon to accept a wholesale deal or sell spectrum.
• Asked how investors would react to SpaceX’s results, Piecyk says there is insufficient history to know. Believers may look past quarterly numbers, while skeptics dismiss progress. The practical issue is whether SpaceX can keep raising cash or issuing stock when investment across its growth initiatives consumes more than its operating businesses generate.

6. Verizon lacks a clear answer after the 5G payoff disappointed

• “5G didn’t deliver,” Piecyk says. Network-dependent collision avoidance was implausible, and remote surgery could not plausibly transform businesses generating hundreds of billions in revenue. LTE, by contrast, had an identifiable application: video. Patel’s better 5G pitch is thousands of people livestreaming Taylor Swift from stadiums. Piecyk says the underlying need was spectrum for data growth, cheaper than building equivalent network capacity.
• Another investment cycle remains possible: Piecyk says the FCC chair wants a $100 billion spectrum auction, requiring carriers to find a convincing 6G revenue story despite leverage and new competition. AT&T has a clearer existing strategy: fiber plus wireless creates valuable, low-churn customers, supporting the prospect of 2–3% revenue growth and dividends.
• Piecyk faults Verizon’s spectrum decisions, including millimeter wave, and thinks it needs another $10–15 billion of EchoStar spectrum. Patel challenges the emphasis on media mistakes: AT&T’s Time Warner acquisition was vastly larger than Verizon’s Yahoo-related bets. Piecyk concedes that point, placing more weight on spectrum, competition and the erosion of Verizon’s premium-network positioning.
• “The networks are effectively commoditized,” Piecyk argues. Verizon can no longer rely on being the best, and staff cuts improve near-term EBITDA without supplying a revenue strategy; he expects operational consequences. Patel identifies its retreat from FiOS as another mistake. Piecyk thinks more fiber investment, potentially easing pressure from spectrum-constrained fixed wireless, is probably sensible despite FiOS’s cost and union issues.

7. The fourth-carrier debate returns to competition and merger credibility

• Patel contrasts blocking AT&T–T-Mobile, whose breakup fee helped finance T-Mobile’s renewal, with allowing Sprint’s acquisition under a remedy relying on Dish. Should regulators simply preserve four carriers? Piecyk says former FCC chair Wheeler would agree; Carr instead points to prices he sees falling and competition from fixed wireless.
• Piecyk favors more competition to encourage innovation and lower prices, especially with wireless margins approaching 60%. Refusing SpaceX a wholesale agreement is “their right,” but he thinks sufficiently aggressive resistance could invite FCC involvement under its public-interest mandate, potentially under a different administration.
• His warning about future merger arguments is pointed: save executives’ claims that SpaceX cannot compete with only 80 megahertz of spectrum. When those same companies want consolidation approved, he expects them to reverse course and insist SpaceX is a meaningful competitor.

8. AI devices could redistribute value beyond the carriers

• Patel raises a conditional demand shift: Meta’s “Muse charm” or similar agent interfaces might reduce phone-based video consumption and change bandwidth requirements. Piecyk sees an Apple risk if carrying an iPhone becomes less important. “If that happens,” the carrier subsidy system and the economics around devices could change too.
• Neither expects Musk to add a handset business. Patel’s present-day constraint is Instagram: a practical SpaceX device would look like a midrange Android phone with the Play Store. He cites Amazon’s new tablets as conceding to existing application expectations. Piecyk agrees, while observing that merely discussing a phone could add another financing narrative.
• Piecyk’s broader point is that consumers spend $1,000 on devices still offering basic swipe interactions. A different interface might resonate, changing carriers’ role and margins. Wireless operators previously surrendered considerable value to Apple; another paradigm shift could redistribute that value again, without guaranteeing who captures it.