OPINION
Anyone who has done a jigsaw puzzle on a rainy weekend knows the feeling of finding the last piece. It’s under the couch, or the dog has it, and when you finally hold it up you announce to the room that the puzzle is done. Then someone points out that the other 999 pieces are still in the box.
On Oct. 8, SpaceX agreed to buy Grain Management’s nationwide block of 800 MHz airwaves, and Elon Musk went on X to call it “the last critical piece of the spectrum puzzle” SpaceX needs to provide complete phone coverage in America. Carrier stocks took it seriously. AT&T, Verizon and T-Mobile each fell somewhere around 6% to 7% in overnight and early trading, according to Benzinga and 24/7 Wall St.
So I went looking for the rest of the puzzle, and I found it mostly where puzzles usually are, in the government’s paperwork.
The pieces still in the box
Start with the deal itself. Grain’s announcement says the sale is “subject to FCC approval and other customary closing conditions.” The piece Musk is holding up hasn’t been handed over yet.
Then there are the satellites meant to use it. On Oct. 6, the FCC approved SpaceX’s plan for a 15,000-satellite direct-to-device constellation. The order gives SpaceX until Oct. 7, 2032, to have half of them operating and until Oct. 7, 2035, for the rest, according to Broadband Breakfast and a summary of the order by Exterra. SpaceX has said it will start launching this second generation in mid-2027 on Starship. None of the 15,000 is in orbit yet. The roughly 650 phone satellites up there today belong to the first-generation system, so all 15,000 pieces are still in the box, and each one has to be put in place by a rocket.
There is also a permission slip. The 800 MHz block isn’t one of the bands where the FCC currently allows satellites to provide “supplemental coverage from space” to ordinary phones. When the agency approved Grain’s purchase of these same licenses in July, it put off ruling on a waiver for satellite use until a satellite partner applies. SpaceX is that partner now. It still has to apply, and the FCC still has to say yes.
The part I keep coming back to is the report card. That July order set performance targets for whoever ends up offering satellite service on this spectrum. The first is due Nov. 30, 2029, a signal meeting the FCC’s quality bar 70% of the time across 90% of each license area. The last is due Nov. 30, 2036, by which point it has to be 90% of the time. And in a footnote, the FCC defines service availability as what a user receives “when the user is outdoors.”
SpaceX’s stated reason for buying this block is indoor coverage. Its own statement says the low-band airwaves will let Starlink Mobile’s signal pass “through obstacles, such as walls.” The government’s test, meanwhile, will be taken in the yard.
Put all of that in a row and “complete phone coverage” looks less like a finished puzzle and more like a very good corner. Even the spectrum SpaceX already owns comes with a wait. Under this week’s FCC grant, SpaceX can’t start operating in the 2 GHz airwaves it bought from EchoStar until the second step of that deal closes.
Why Musk may be right about the piece that matters
Here is the best case for taking him at his word, and it’s a good one. Rockets and satellites are things SpaceX builds itself, and it has a record of building them that most of its competitors would envy. Spectrum is the one input it can’t manufacture. It has to be bought from someone who holds a federal license, and this block sat largely unused for years, through a Justice Department divestiture order, a Dish option that expired and a long FCC review. Read literally, Musk’s sentence is about the spectrum puzzle, and on that narrow point he may simply be correct. The FCC also handed SpaceX something carriers have long assumed it would never get, permission to sell mobile service in the U.S. directly without leasing airwaves from a carrier partner. And the investors who sold carrier shares that night may have read the same FCC orders I did and concluded that SpaceX’s remaining obstacles are engineering ones, which it has a habit of clearing.
That is a fair point, and it may turn out to be the right one. The rockets may well fly on schedule. But a statement about owning the right airwaves is not the same as a statement about phones working in basements in 2027, and the second one is what most people heard. The paperwork says that’s a question for the next decade.
Somewhere a puzzle sits on a card table with one piece in place and a confident man standing over it. I don’t know if he’ll finishing it but I’m pulling up a chair to watch.
The FCC’s July order on these licenses, with the satellite targets and the outdoor footnote, is DA 26-653. Our news report on the deal and the block’s history is SpaceX Buys the 800 MHz Block the FCC Drew for Nextel in 2004.
Sources: FCC Wireless Telecommunications Bureau, DA 26-653, WT Docket 25-178 (July 1, 2026); Grain Management announcement (Oct. 8, 2026); Via Satellite, SpaceX statement (Oct. 8, 2026); Benzinga, Musk’s post and overnight trading (Oct. 8, 2026); Techaeris (Oct. 9, 2026); 24/7 Wall St. (Oct. 9, 2026); Broadband Breakfast, FCC approval of 15,000 satellites (Oct. 7, 2026); Exterra, summary of FCC order DA 26-1078; Broadband Breakfast, SpaceX’s 2027 launch plan; 5G World Pro (Oct. 9, 2026).