Direct-to-device is the only satellite service that reaches a phone with no signal and no new hardware, and Starlink currently dominates it [13]. The measure of it is whether the message leaves and how long that takes, not how often anybody opens anything.
Permission and hardware in orbit are the scarce inputs behind that message, and different companies hold them. The 2 GHz mobile satellite band is a 30 MHz pairing that lets a handset keep a connection where no mast reaches [3], and cutting it in three leaves roughly 10 MHz paired per slice [18]. SpaceX holds the hardware, and already sells texts, data and calls to unmodified phones through mobile operators, including Deutsche Telekom, but cannot sell the service under its own name in the EU because it holds no spectrum there [7].
A four-carrier consortium would be buying the first input and renting the second. The band changes hands in May 2027 [4], and IRIS2 does not enter service until 2029 [11], which leaves about two years in which the European third exists as a licence and nothing else [19]. TNW's read is that the likeliest outcome is a European consortium holding the spectrum and leasing capacity from an American constellation until a European one exists [16]. The domestic supply side is thin against that: Sateliot is seeking 150m euros to beam 5G to phones, Eutelsat is still waiting for growth to return in 2027 [14], and Amazon has asked the FCC for 5,105 satellites dedicated to phones [15]. Any such lease would be negotiated while Washington lobbies against the EU Space Act's prior-authorisation regime, and in a week when four American space companies pulled out of Macron's Paris summit under White House pressure [17].
That is why the ownership test bites before the auction does. The consortium settles whose logo is on the bill for a service Starlink already reaches European phones through, given Deutsche Telekom's existing agreement to offer Starlink's direct-to-device service [9].
The grid to hold against your own capability has permission down one side and delivery capacity across the top. Hold both and you set the price. Capacity alone makes you a wholesaler, which is where SpaceX sits in the EU today [7]. Permission alone makes you a retailer carrying a margin and a renewal date, and that is the box the consortium would be buying into. Neither one leaves you a reseller of a reseller.
TNW's read is that one joint bid would decide the auction before it opens: the four serve most of the mobile customers in the EU's largest markets, so the reserved third would have an obvious winner [12]. Starlink's European roadmap stays a separate question. The open third stays open to SpaceX and Amazon [5], the wholesale route through carriers already works [7], and Bloomberg's own sources say no decision has been taken on the consortium or on whether it will bid, while the rules themselves still have to pass the European Parliament and the Council [2][6]. The reserved third buys, until a European constellation flies, the right to put a European name on somebody else's satellites [16][19].
