SpaceX’s recent IPO and Starship rocket test flight delivered two big data points that offer a realistic vision for the coming years — and one that may disappoint both the company’s boosters and its critics.

Hidden behind the fantastic expectations for AI enterprise profits and plans for a moon base is a more grounded reality: An expendable Starship could keep SpaceX in business, but doesn’t achieve the cost reductions — or frontier business models — Elon Musk is betting on.

SpaceX is many businesses, but right now only one is producing significant revenue. Starlink, its satellite communications network, is the tent pole of the firm’s public offering. The top line is fairly incredible; SpaceX’s connectivity business generated $11.4 billion in revenue last year, the bulk of the company’s earnings.

But underneath, you can see the capital expenditure treadmill that scared previous entrepreneurs away from this model. SpaceX needs to replace about a fifth of its satellites every year just to maintain its current level of service. It has invested more in its satellite business ($11.4 billion) since the beginning of 2023 than it has building Starship and its launch infrastructure ($8.4 billion).

SpaceX’s S-1 filing with the U.S. Securities and Exchange Commission predicts costs will continue growing, but expects that improvements to its technology will allow it to reduce them as a percentage of its revenue.

Musk has said that Starship is the key to keeping Starlink’s costs under control, even saying that SpaceX could go bankrupt without the vehicle’s ability to replace those satellites cheaply. In that context, a note that stood out in SpaceX’s S-1 was the first acknowledgment that full reusability of Starship isn’t necessary to launch the new generation of Starlink satellites. But without full reusability, the cost will go up, making the business less attractive.

“If this reusability is not achieved then the cost of launch on Starship may not be much lower than Falcon 9, even if the full 100 ton capability is realized (which is by no means a foregone conclusion),” satellite market analyst Tim Farrar wrote in a note to clients last week. “The cost per launch may be as much as $100M (i.e. $1000 per kg) while tempo remains constrained by the rate at which second stages can be manufactured and first stages can be refurbished.”

Last week’s test flight of the third version of Starship and its booster bore those concerns out. The newest rocket’s maiden flight saw issues with a key capability for reusability — relighting the Raptor rocket engines on both the booster and Starship in order to make a controlled return to Earth. Starship did, however, deploy a set of dummy satellites and two test vehicles in space.

That helps square SpaceX’s prediction that it will begin launching a new generation of higher-throughput Starlink satellites 60 at a time, a twentyfold increase in capacity compared to a single Falcon 9 launch, later this year. At first glance a classic example of Musk’s timelines, it may actually be an expectation that initial launches will expend the Starship. If so, SpaceX might not be able to count on as much free satellite cash as expected, and its plans to launch space data centers will become untenable until the rocket is reusable.

Starlink growth slows

At the same time, SpaceX’s S-1 shows that Starlink’s growth is slowing.

SpaceX’s total addressable market calculation is based on its ability to offer service to every fixed-broadband subscriber or mobile handset in the world. That’s unlikely, though, because Starlink isn’t competing on price with terrestrial fiber. The rest of the document suggests SpaceX continues to see direct-to-device as a complement, rather than a replacement, for terrestrial mobile providers.

Starlink has just over 10 million subscribers, more than any other satellite communications network. But Farrar notes the rate of user growth fell over the course of the first quarter of 2026. Quilty Space, a space consulting firm, projected earlier this year that SpaceX would end the year with 16.8 million subscribers. That would require the company’s quarterly growth rate to roughly double from where it is now, which may be difficult after recent price increases.

Growth matters for SpaceX because its new Starlink users are paying less than previous ones. Starlink’s average revenue per user has fallen from $99 in 2023 to $66 in the first quarter of 2026 — a change propelled by its expansion into new international markets where it can’t charge as much as it does in developed economies. Without a fast-growing user base, each new satellite launched is making less money.

Increased competition also threatens Starlink. Amazon’s Leo network is approaching the scale required to put pressure on SpaceX, although it is waiting for the Federal Communications Commission to extend a deadline that requires it to launch 1,600 internet satellites by July.

Data in the SpaceX filing presents a gloomy growth forecast for the company as well as rivals like Blue Origin. Farrar says that if SpaceX — much further ahead than any other company — is seeing slowing demand, that may signal the market for space broadband is smaller than the players anticipated.