578 — the Number Amazon Needs Before It Can Charge a Single Customer

ULA just set a date.

November 13. That's when the Vulcan rocket flies again — nine months after a solid rocket booster burned through its own nozzle mid-flight and put the whole program on ice.

But here's the part that matters. The payload isn't a Pentagon spy satellite. It's 40 broadband satellites for Amazon Leo, the company's answer to Starlink.

And this isn't a one-off. It's the first of 38 Vulcan flights Amazon booked to build its constellation. The largest commercial launch deal ever signed.

Let me explain…

Amazon has poured more than $10 billion into Leo. The plan: 3,232 satellites in low orbit, beaming internet to the ground. A direct rival to SpaceX's Starlink network.

So far? About 392 satellites in orbit. Starlink has over 11,000.

That gap is enormous. And it got worse this year.

When Vulcan's Northrop Grumman GEM 63XL booster failed on February 12, ULA couldn't fly. Not for Amazon. Not for the Space Force. Not for anyone.

The Pentagon moved six missions to SpaceX. Amazon's launch schedule stalled. The FCC had told Amazon to get half of its constellation — roughly 1,600 satellites — in orbit by July 2026. Amazon didn't come close. The FCC waived the deadline in June, keeping the final cutoff at July 2029.

In other words, Amazon's biggest hardware bottleneck wasn't the satellites. It was the ride.

Now, I know what you're thinking. Amazon has other rockets. Ariane 6. New Glenn. True. The total deal covers 108 launches across four providers.

But New Glenn blew up its launch pad in May. Ariane 6 has its own cadence limits. Vulcan was supposed to carry the bulk of the fleet.

So this November 13 launch is more than a test. It's a proof point. Vulcan will fly in its most powerful setup yet — six strap-on solid boosters, the VC6L config — for the first time. If the redesigned nozzles hold, ULA can start burning through that 38-flight backlog.

If they don't? Amazon's three-year sprint to 3,232 satellites gets even harder.

There's a magic number here: 578. That's how many satellites Amazon needs in orbit before it can turn on commercial service. At 392 today, they need about 190 more. Four or five successful Vulcan flights could get them there by mid-2027.

That's the race. Not rockets against rockets. Satellites against a deadline.

Amazon can build up to 5 satellites a day in its Kirkland factory. The factory isn't the bottleneck. The launch pad is. And on November 13, we'll find out if that bottleneck just broke open… or if it's still stuck.

Worth watching.

I need to speak with you directly.

I want you to see the most important warning of my career …

And one of the most important warnings any American family will receive before the next great financial crisis hits this country.

I'm Dr. Martin D. Weiss. I'm a Wall Street Journal, BusinessWeek, and New York Times best-selling author.

And for 55 years I have run Weiss Ratings … the only 100% conflict-free, independent financial rating agency in America.

We are paid by no one we rate. We answer to no one but our subscribers.

And in the 55 years since I issued my first public warning before a financial crisis, I have publicly predicted six major American crashes, before they happened.

What I'm seeing now is a $3 trillion problem.

And the biggest danger right now is NOT just another stock market crash.

The biggest danger is NOT only a major bank failure.

Rather, it's a chain reaction of crashes and failures that could cause a global freeze in the financial markets. And it starts here:

Go here now see the most important warning of my career

Understand what actually grounded Vulcan. On February 12, less than 30 seconds into the USSF-87 mission, a nozzle failed on one of the Northrop Grumman GEM 63XL solid rocket boosters — the second such nozzle failure in just four Vulcan flights. A joint ULA–Northrop–Space Force team found “several potential contributors” and changed the nozzle's design and production process. Notably, ULA is flying the fix rather than waiting, with more testing running in parallel. That's a calculated bet… return to service now and validate as you go. Watch whether the redesigned nozzle holds through the full booster burn on November 13.

Follow why so much rides on Vulcan specifically. Amazon's launch contract spreads across four providers, but the alternatives are constrained: Blue Origin's New Glenn lost its pad to a May explosion, Ariane 6 has limited cadence and flies from French Guiana, and ULA's reliable Atlas V has only six rockets left — all reserved for Boeing's Starliner. That leaves Vulcan to carry the bulk of Amazon's ULA manifest. The deployment speed of a would-be Starlink rival now hinges largely on one returning rocket. Watch whether ULA can build a launch cadence, not just a single return flight.

Here's the pressure behind the whole effort. The FCC requires Amazon to deploy its full first-generation constellation — more than 3,200 satellites — by July 2029, after waiving an earlier half-by-2026 milestone that Amazon badly missed. With about 392 satellites up and more than 11,000 flying for Starlink, Amazon is racing both a regulator and a dominant rival at once. The factory can reportedly build satellites far faster than rockets can lift them, so every grounded month burns runway. Watch the launch cadence against that 2029 wall; that ratio decides whether Leo becomes a real competitor.

Watch not just November 13, but how fast the second and third Vulcan flights follow.

Here's why one launch doesn't settle it. A successful return-to-flight proves the nozzle fix works once. What Amazon actually needs is a steady drumbeat — flight after flight, 40 satellites at a time — to close an enormous gap with Starlink. ULA has said a second Vulcan for Amazon is already assembled, which is the kind of signal that matters.

The math is unforgiving. Amazon needs roughly 190 more satellites just to switch on commercial service, and more than 2,800 beyond that to finish by 2029. At 40 per Vulcan flight, that's a lot of launches in not much time.

Why the stakes run beyond Amazon. Vulcan also carries national-security payloads that piled up during the grounding, and the Space Force wants flight data on the redesigned nozzle. A clean, repeatable Vulcan unclogs both the Amazon manifest and the Pentagon's backlog.

The risk is real on cadence, not just this flight. Even if November 13 succeeds, ULA has to ramp from a standing stop to a rhythm it has never sustained — and new rockets rarely hit their stride immediately.

So watch two markers into 2027: the gap between Vulcan flights shrinking, and Amazon's in-orbit count climbing past the 578 it needs to go commercial. Those, not the first liftoff, decide the race.

Let's keep this simple.

Imagine a factory that can churn out cars far faster than there are trucks to ship them to dealers. The cars pile up in the lot. It doesn't matter how fast you build if you can't deliver.

That's Amazon's problem in space.

Amazon can reportedly build satellites at a remarkable clip — by some accounts, several a day at its factory near Seattle. Building them was supposed to be the hard part… It isn't.

The hard part is getting them to orbit. A satellite does nothing sitting in a warehouse. It only earns money once a rocket carries it up and it starts beaming internet. And rockets are scarce, expensive, and — as this year proved — prone to sudden grounding.

So the bottleneck moved. For a while, the industry's limit was how fast you could manufacture spacecraft. Now, for Amazon, it's how fast you can launch them. The factory races ahead… the launch pad lags behind.

Here's why that's such a bind. When your one main rocket gets grounded for nine months, your finished satellites just wait. The clock keeps ticking — toward a regulator's deadline, and against a rival adding thousands of its own satellites in the meantime.

And this is exactly why SpaceX has such an advantage. It builds its own rockets and launches them at will, dozens of times a year. It never has to wait in line for someone else's ride. Amazon, by contrast, has to buy its launches — so when its suppliers stumble, Amazon stumbles too.

That's the quiet lesson of the whole Amazon-versus-Starlink race. Owning the factory isn't enough… In space, whoever controls the launch controls the pace.

So the number to watch isn't how many satellites Amazon can build. It's how many it can actually fly.

Follow the launch cadence, not the factory output — in orbit, delivery is the whole game.