The Starship’s re-usability strategy appears to be in disarray after SpaceX’s S-1

SpaceX’s recent IPO and Starship rocket test flight provided two key points that provide a real vision for the coming years – and one that will disappoint both the company’s supporters and critics.

Hidden behind the optimistic expectations of AI business benefits and lunar plans is a stark reality: A low-cost Starship could keep SpaceX in business, but it won’t achieve the cost-cutting — or marginal business — Elon Musk is betting on.

SpaceX is a lot of businesses, but right now it’s the only one that’s making a lot of money. Starlink, its satellite communications network, is the company’s public offering network. The top line is amazing; SpaceX’s communications business generated $11.4 billion in revenue last year, the largest share of the company’s revenue.

But underneath, you can see the cheap strategy that scared previous traders away from this model. SpaceX has to replace one-fifth of its satellites each year to keep it operational. It has invested more in its satellite business ($11.4 billion) since early 2023 than it did in Starship and its base ($8.4 billion).

SpaceX’s S-1 filing with the US Securities and Exchange Commission predicts that prices will continue to grow, but it expects that improvements in its technology will reduce it as part of its investment.

Musk has said that Starship is the key to controlling Starlink’s cost, even saying that SpaceX could collapse without the power of a vehicle to replace satellites cheaply. In this context, the note that became known in SpaceX’s S-1 was the first acknowledgment that the restart of Starship is not necessary to launch the new generation of Starlink satellites. But without complete reuse, the cost will increase, making the business less attractive.

“If these improvements are not completed then the launch cost of Starship will not be much lower than Falcon 9, even if the full capacity of 100 tons is achieved (which is not a fantasy),” satellite market analyst Tim Farrar wrote in a letter to clients last week. “Each installation cost can be $100M (ie $1000 per kilogram) while the tempo remains horizontal with the rate at which the second phases can be produced and the first phases can be reprocessed.”

Last week’s test of the third version of the Starship and its propulsion brought these concerns to light. The most recent rocket flight saw challenges with re-use capabilities – re-igniting the Raptor rocket engines on boost and Starship to return to Earth. However, Starship sent several satellites and two test vehicles into space.

This supports SpaceX’s prediction that it will start launching the new generation of Starlink satellites 60 at a time, a twenty-fold increase compared to a single Falcon 9 launch, by the end of this year. At first glance at Musk’s prototype, it might be expected that the initial launch would destroy the Starship. If that’s the case, SpaceX may not be able to rely on as much free satellite revenue as it had hoped, and its plans to launch a data center will be impossible until the rocket is used again.

Starlink’s growth slows down

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

SpaceX’s total market share is based on its ability to provide service to every broadband or cell phone subscriber in the world. This is unlikely, as Starlink does not compete on price with terrestrial fiber. The entire document shows that SpaceX continues to see direct-to-device as a complement, rather than a replacement, for terrestrial mobile providers.

Starlink has more than 10 million subscribers, more than any other satellite communication network. But Farrar indicates that the number of users fell in the first quarter of 2026. Quilty Space, a space consulting company, said earlier this year that SpaceX will end the year with 16.8 million subscribers. That would require the company’s interest rate to double from its current level, which would be difficult given the recent rate hikes.

Growth is important for SpaceX because new Starlink users pay less than previous ones. Starlink’s revenue per user has dropped from $99 in 2023 to $66 in the first quarter of 2026 – a change driven by its growth in new global markets that cannot charge as much as it does in developed countries. Without a growing number of users, each new satellite that is launched is making less money.

Increasing competition threatens Starlink. Amazon’s Leo network is approaching the scale needed to force SpaceX, even as it waits for the Federal Communications Commission to extend its July target of launching 1,600 satellites.

The content of the SpaceX file provides a comprehensive overview of the company and competitors such as Blue Origin. Farrar says that if SpaceX – far ahead of any other company – is seeing less demand, that could indicate the market for space shuttles is smaller than players expect.