Does Starlink have all it needs to be a bonafide mobile carrier? (summary)
Roger Entner argues, with seven vital boxes already ticked. It's not whether but when.
On May 19, Roger Entner of Recon Analytics published a piece in Light Reading arguing that the question is no longer whether Starlink can become a stand-alone US mobile carrier, but when one of the three nationwide operators breaks ranks and hosts it. His frame is a checklist of seven operating capabilities. By his count, all seven are now in the public record.
It’s a striking claim, because the conventional wisdom for most of the past year has gone the other way. AT&T’s John Stankey said in September 2025 that Starlink’s spectrum holdings, even with EchoStar added, weren’t enough for a “robust terrestrial replacement.” T-Mobile’s Srini Gopalan said in late April that T-Satellite is being used less than the carrier expected and that an MVNO with Starlink doesn’t fit T-Mobile’s TAM-expansion criteria. And at MWC 2026 in Barcelona, SpaceX president Gwynne Shotwell and Starlink SVP Michael Nicolls leaned hard into the carrier-partner framing, calling Starlink Mobile (the new name for direct-to-cell) a hybrid-network component, not a replacement.
Entner is reading the same room and reaching the opposite conclusion. Here’s the checklist, unpacked.
The seven boxes
1. Spectrum. Entner says the FCC on May 12, 2026 approved SpaceX’s purchase of 65 MHz of nationwide direct-to-device spectrum under tech-neutral performance obligations. This is the load-bearing piece — the only US midband allocation of its kind dedicated to D2D. The bulk of that spectrum traces back to EchoStar: in 2025 SpaceX agreed to buy EchoStar’s AWS-4 and H-block licenses for roughly $17 billion in cash and stock, plus its unpaired AWS-3 licenses for about $2.6 billion in SpaceX stock.
2. Power. Entner cites a March 2025 FCC waiver that lifted the power flux density limit by 770 percent, enabling 4G-class consumer service over Starlink Direct-to-Cell. PFD limits cap how much signal a satellite is allowed to put on the ground in a given band; raise the ceiling and the link budget for ordinary smartphones starts looking like terrestrial LTE.
3. Capacity. Entner points to an April 30, 2026 NGSO spectrum-sharing rules overhaul that lets up to eight Starlink satellites operate in the same area-and-frequency cell. NGSO means non-geostationary satellite orbit — the regulatory category Starlink lives in. Allowing multiple satellites to share the same cell at the same time is how a LEO operator stacks capacity over a hotspot.
4. Satellite count. A September 2025 SpaceX filing covers a 15,000-satellite direct-to-cell constellation. For scale: Starlink today is already the largest constellation ever flown, with more than 10,000 satellites in orbit and roughly 65 percent of all active satellites worldwide. On the D2D side specifically, SpaceX had something like 650 dedicated direct-to-cell satellites as of the start of this year.
5. Brand. SpaceX filed for “Starlink Mobile” and “Powered by Starlink” trademarks on October 16, 2025, covering voice, audio, video and data over satellite and cellular infrastructure. Five months later, Shotwell and Nicolls used the MWC keynote to rename the direct-to-cell service to “Starlink Mobile” outright.
6. Network identity. This is the one most readers won’t have seen before. Entner notes that SpaceX has been the assigned operator of mobile network code 901-08 since February 2024. An MNC is the half of the IMSI that says which network a SIM belongs to. Without one you cannot exist as a stand-alone mobile operator in the international numbering system. Starlink has had its since before T-Mobile launched T-Satellite.
7. The Gen2 platform. This is the technology layer that converts Starlink Mobile from a messaging-and-light-data overlay into native cellular voice via the 3GPP 5G non-terrestrial networks standard, which Entner expects to reach consumer-grade operationalization in 2027. Deutsche Telekom signed on at MWC for the future “V2” edition of Starlink Mobile, and Qualcomm has said its current modems already support the N256 band Starlink will use for it.
Why this matters: the JV is interop, not a moat
Four days after the May 12 spectrum approval, the three nationwide carriers announced a satellite joint venture. The market read it as defensive. Entner argues it is not — that the JV is a technical interop platform standardizing the satellite-to-cellular interface, not a purchasing pool and not a binding commitment to deny Starlink wholesale. The Q1 denials from T-Mobile and AT&T still hold, but they hold individually, not collectively, and that is the cliff he is pointing at.
The MVNO, in this view, is the onramp. The destination is Starlink running its own network on its own spectrum at consumer scale — the MNO endpoint. Entner’s structural observation is that Starlink is the first US wireless entrant to begin the journey already owning the spectrum that makes the destination reachable. He contrasts this with cable mobile, which started as an MVNO and chased CBRS for years to acquire any owned spectrum at all.
The three CEOs
Entner reads the situation through how three executives are positioned on it.
John Stankey at AT&T sits closest to a deal substrate-wise — the May 12 order also gave AT&T 30 MHz of 3.45 GHz and 20 MHz of 600 MHz plus the Boost hybrid MVNO host role plus a Tesla in-vehicle IoT contract that the Tesla-SpaceX corporate envelope makes easier to retain if AT&T is the host — and farthest principal-wise.
Srini Gopalan at T-Mobile holds the deepest operational SpaceX relationship through T-Satellite and the new SuperBroadband bundle. But T-Mobile carries the heaviest growth-rate compression in Entner’s analysis, because its premium multiple is anchored on net-add leadership in exactly the cohorts where Starlink ISP already over-indexes.
Dan Schulman at Verizon, Entner argues, sits on the lightest reputational anchor — he did not address SpaceX on Verizon’s Q1 call — and therefore has the lowest cost of reversal.
The numbers Entner puts on it
This is the part the piece monetizes. If a Starlink Mobile activation happens inside a three-year window from the IPO, Entner models sustained equity reallocation of $55 to $120 billion of compression across nine incumbent actors, against $10 to $22 billion of uplift to SpaceX, EchoStar and the tower companies. He splits scenarios — $55 to $105 billion compression in an MVNO-host outcome, $70 to $120 billion in a bypass outcome where Starlink builds OEM-direct distribution and selective tower acquisition. The three exceptions to the one-sided math on the incumbent side are EchoStar, the tower companies, and AST SpaceMobile under a JV-anchored scenario.
The trigger he is watching for is the SpaceX IPO disclosure window, which he says opens within weeks. SpaceX is reportedly preparing to raise up to $75 billion at a $1.75 trillion valuation, with Starlink as the cash-cow segment — roughly $10.6 billion of 2025 revenue and $5.8 billion EBITDA, a 54 percent margin.
What to watch for
A few inflection points worth keeping an eye on:
Whether one of the three MNOs breaks ranks on a Starlink wholesale deal before the IPO prospectus crystallizes the optionality story
Whether SpaceX bids in the upcoming upper C-band auction the FCC must hold by July 2027, which Entner has flagged elsewhere as the real tell on terrestrial-carrier intent
Whether the carrier JV publishes a technical specification — if it does, Entner’s “interop standard, not moat” reading gets confirmed
Whether T-Mobile’s “less usage than expected” assessment of T-Satellite is a verdict on D2D as a category or a verdict on the messaging-only product before Gen2 voice arrives
The piece works because it reframes the debate. The argument for the past year has been “can Starlink build a mobile business?” Entner says the operating capabilities are already on the table, settled in FCC orders, trademark filings, ITU codes and 3GPP standards work. What remains is which of the three nationwide carriers decides to be the one inside the tent when it happens.
RESOURCES
Starlink’s path to mobile is a checklist, and most boxes are already checked — Roger Entner’s May 19 analysis in Light Reading, the source piece
Recon Analytics — Entner’s firm; home of the Customer Pulse platform underpinning the cohort analysis
Roger Entner — analyst bio, ex-Nielsen, Yankee Group and Ovum
Turning Starlink into a global mobile carrier ‘one of the options,’ Musk says — September 2025 setup piece on the EchoStar deal and MVNO question
SpaceX files for ‘Starlink Mobile’ trademark — the October 16, 2025 filing that is checklist item 5
SpaceX cleared for Starlink capacity boost — FCC authorization for 7,500 additional Gen2 satellites
At MWC, SpaceX execs tout Starlink V2 — and a key carrier partner — the Starlink Mobile rebrand and Deutsche Telekom signing
Starlink D2D use is less than expected, says T-Mobile — Gopalan’s Q1 comments ruling out an MVNO with Starlink
T-Mobile and Starlink connect on SuperBroadband for biz customers — the 5G+satellite business bundle referenced in the CEO section
Here comes Starlink, the next telecom giant — April 2026 piece with the SpaceX IPO / $75 billion / $1.75 trillion valuation figures
GLOSSARY
D2D / DTC / Direct-to-Cell / Direct-to-Device — satellite service delivered to ordinary, unmodified smartphones, with no special antenna or handset
MNO (Mobile Network Operator) — a carrier that owns spectrum and runs its own network (AT&T, Verizon, T-Mobile)
MVNO (Mobile Virtual Network Operator) — a reseller that runs its brand and billing on top of an MNO’s network and spectrum
MNC (Mobile Network Code) — the digits in a SIM’s IMSI that identify which network it belongs to; required to exist as a stand-alone operator. SpaceX holds 901-08
IMSI (International Mobile Subscriber Identity) — the globally unique number on a SIM that identifies the subscriber and home network
LEO (Low Earth Orbit) — the orbital regime Starlink uses, roughly 550–1,200 km up, enabling low latency
NGSO (Non-Geostationary Satellite Orbit) — the FCC regulatory category covering LEO and MEO constellations
PFD (Power Flux Density) — the regulatory cap on how much radio energy a satellite may put on a square meter of ground in a given band
3GPP NTN (Non-Terrestrial Networks) — the 3GPP standards track that lets satellites participate natively in 4G/5G networks, including voice
FCC (Federal Communications Commission) — the US regulator that licenses spectrum and authorizes satellite operations
AWS-3 / AWS-4 / H-block — specific US spectrum bands EchoStar is selling to SpaceX; AWS = Advanced Wireless Services
CBRS (Citizens Broadband Radio Service) — a shared US midband (3.5 GHz) tier that cable mobile operators have used to acquire owned spectrum
JV (Joint Venture) — here, the unnamed satellite interop entity announced May 16 by the three US carriers
TAM (Total Addressable Market) — the universe of potential customers a business could reach; Gopalan’s criterion for which MVNOs T-Mobile will host
Gross adds / Net adds — gross adds = new lines activated in a period; net adds = gross adds minus disconnects. The metric T-Mobile’s premium multiple rests on
SCS (Supplemental Coverage from Space) — the FCC’s regulatory framework for direct-to-cell satellite services on terrestrial spectrum