Three more satellites went up on Tuesday to end the dead zone. The dead zone ends when about ninety are up there, and that is the whole invoice.

On Tuesday morning a Falcon 9 lifted off from Cape Canaveral carrying three more of the strangest satellites anyone is building right now — BlueBirds 11, 12 and 13, each one unfolding in orbit into a phased-array antenna roughly the size of a tennis court. The pitch that goes with them is genuinely remarkable, and I want to give it its due before I do the thing I always do. The promise is that your phone — the one in your pocket, unmodified, no dish, no special handset — can talk directly to one of these satellites and pull down something close to broadband speed, in the middle of an ocean or a desert with no tower for a thousand miles. That is not a trick or a rounding of the truth. AST SpaceMobile has demonstrated it. The physics works. Which is exactly why it is worth turning the renderings over and reading the invoice on the back, because the reason this may or may not become a business has almost nothing to do with the part that looks like magic.

Here is the number the launch coverage tends to skip. Tuesday's flight brings the operational Block 2 constellation to somewhere around a dozen satellites, on top of the BlueWalker 3 prototype that proved the concept. A dozen satellites is enough to demonstrate a call. It is nowhere near enough to sell one. Continuous coverage — the kind you would actually pay a monthly fee for, where the signal is simply there when you need it — requires a full shell of satellites arranged so that one is always overhead. For the service AST wants to sell in the United States, that means something on the order of ninety spacecraft, and by some designs more. The company is targeting roughly forty-five launches worth of satellites across this year and is still building toward that shell. So the honest description of Tuesday is not "the dead zone is closing." It is "the dead zone closes at satellite number ninety-ish, and we are at about a dozen."

The number that decides it isn't speed. It's how many.

You can see the gap in the company's own timeline. AST has moved its commercial direct-to-device service to early 2027, with a beta running before that, and a recent assessment of the coverage available today put the plain problem plainly: with the satellites now in orbit, the windows in which a given spot on the ground can reach one are intermittent, and the gaps between them can span hours. That is not a criticism of the engineering; it is arithmetic. A handful of satellites in low orbit are each overhead for only a few minutes at a time, and between passes there is nobody home. "Broadband from space, direct to your phone" and "coverage that arrives for a few minutes every few hours" are the same system at two different points on the same deployment curve, and the entire business lives in the distance between them. That distance is measured in satellites, and satellites are measured in dollars.

This is the part I find genuinely clarifying about AST as a case study, because it strips the space business down to its actual mechanic. Nothing here is bottlenecked on a physics breakthrough. The beamforming works, the giant antenna works, the link to an ordinary handset works. What stands between a successful demonstration and a service you can charge for is not a discovery. It is a manufacturing-and-launch schedule, financed to completion, executed before the money or the patience runs out. The rocket, as ever, is the easy part. The hard part is the roughly eighty more satellites you have to build, insure, launch and replace — and the spreadsheet that has to survive all of it.

Follow the money

So let us follow it. AST finished its most recently reported quarter with about $3.5 billion in cash, which management says is enough to fund more than a hundred BlueBird satellites — meaningfully, enough to reach a usable shell without necessarily going back to the market for more convertible debt this year. That is the strong version of the balance sheet, and it is a real answer to the first question anyone should ask a pre-revenue space company: can you afford to finish? Against it, set the outflows. The company guided capital expenditure in a single recent quarter to a range of roughly $575 million to $650 million, driven by the timing of launch payments. Building and lofting satellites the size of a tennis court is not cheap, and the bill arrives in large, lumpy pieces tied to each flight.

Now the other side of the ledger, which is where the vertigo sets in. AST's revenue guidance for the whole of 2026 is roughly $150 million to $200 million — and a big share of that is not consumer subscriptions at all but payments from mobile-network partners and government contracts. In the most recent quarter, actual revenue was about $15 million. So the shape of the thing, stated without garnish, is a company spending on the order of half a billion dollars in a quarter to build toward a service that is booking, so far, tens of millions. The company points to more than $1.2 billion in contracted commercial revenue commitments from its carrier partners, and to internal projections that 2027 revenue could approach a billion dollars once the shell is up and the service is live. Those are real, encouraging figures. They are also, every one of them, on the far side of the deployment gap — promises that convert to cash only after the eighty-odd remaining satellites are flying.

A dozen satellites is enough to demonstrate a call. It is nowhere near enough to sell one. The whole business lives in the distance between those two facts, and that distance is measured in satellites.

This is the honest tension in the story, and it is not resolved by rooting for the company. AST has enough cash to make a very serious run at the shell, a genuinely differentiated technology, and roughly three billion subscribers reachable in principle through more than forty-five carrier agreements, AT&T and Verizon and Vodafone and Rakuten among them. It also has a burn rate that eats the cash pile at a real clip, a revenue line that does not meaningfully turn on until a constellation that is mostly still on the ground is mostly in the sky, and a history — common to every ambitious constellation — of timelines that move to the right. More than one analyst looking at the same capital plan has concluded that additional dilution is likelier than not before this is finished. The cash answers "can you afford to finish?" with a qualified yes. It does not answer "and will the finish line still be where you left it?"

Two bets on the same phone

It sharpens the economics to notice that AST is not the only company trying to turn your ordinary phone into a satellite terminal, and that its rival made the opposite engineering bet — which is really an opposite financing bet in disguise. SpaceX's Starlink, working through T-Mobile, is already sending texts to unmodified phones over a large fleet of small satellites, with data to follow. AST is building a small fleet of enormous ones. The trade-off is exactly what it looks like: many cheap satellites get you basic service sooner and a coverage shell you top up continuously; a few giant satellites get you far more capacity per spacecraft — the antenna is the point — but each one is expensive, slow to build, and painful to lose. One approach spreads the cost thin and starts with texting; the other concentrates it and starts with something closer to real bandwidth. Both are legitimate. But they price differently, they fail differently, and the market will not politely wait for AST to complete its shell before Starlink starts signing up the same rural and maritime customers with a good-enough service that already exists.

• The demonstration: an ordinary, unmodified phone pulling broadband-class speeds from a satellite. Real, and proven.
• The scoreboard number: not megabits, but satellites. Continuous U.S. coverage needs on the order of ninety; the constellation is at roughly a dozen.
• The cash math: about $3.5 billion on hand, capex near $575–650 million in a single quarter, 2026 revenue guided to roughly $150–200 million, and the big carrier and 2027 numbers all sitting on the far side of the deployment gap.
• The competition: Starlink is already texting to phones with many small satellites while AST builds a few giant ones — a different bet on the same customer, live now.

Does the math close?

I try to keep one discipline when a space company is this easy to root for, and it is this: the moment the business case leans on the word "inspiration" — or its cousins, "connecting the unconnected" and "no one left behind" — I start reading more carefully, not less, because inspiration is what gets written on the slide where a number should be. AST does not actually need the inspirational case; that is what makes it interesting. It has a differentiated product people would genuinely pay for, contracts that put a dollar figure on that demand, and a balance sheet that can plausibly fund the build. What it has not yet shown is the thing that decides every constellation: that it can get from a dozen satellites to ninety, on schedule and inside its cash, before a well-funded competitor or a slipped timeline changes the arithmetic underneath it.

So watch the count, not the launch. Tuesday's three satellites were a good day and a small fraction of the answer. The question that decides whether AST becomes a business or a beautifully engineered cautionary tale is not whether the next call from the middle of the ocean connects — it will. It is whether the seventieth, eightieth and ninetieth satellites get built and flown before the cash runs low or the patience does, and whether, when they are finally all up there, the monthly fees add up to more than it cost to put them in the sky. The physics closed years ago. The spreadsheet is still open, and the spreadsheet is the story.