Is SpaceX Really a Rocket Company? Looking at the Revenue Numbers

I was a bit skeptical when I first saw the headline of a Barron’s article.

** “SpaceX is the next Amazon.” ** Amazon started as an online bookstore and is now a retail giant spanning logistics and cloud computing, while SpaceX is a rocket company, right? Since the industries are completely different, I wondered what the basis for such a comparison could be, so I dug into their financial statements.

However, as I looked deeper, the truly interesting point wasn’t whether “the two are similar,” but the fact that calling SpaceX a “rocket company” is no longer an accurate description at all.

Is SpaceX Really a Rocket Company?

Looking at the S-1 filing for SpaceX, which went public in June 2026, the numbers are quite clear.

Out of $18.7 billion in total revenue in 2025, the [Starlink](/en/posts/starlink-direct-to-cell-uplink/) (satellite internet) division accounted for $11.4 billion, or 61%. By the first quarter of 2026, this share had reportedly risen to 69%. In contrast, the rocket launch business for external customers stood at around $4.1 billion, a growth of only 8% year-over-year. In fact, while the Falcon 9 was launched 165 times in 2025, only 43 of those were for external customers, with the remaining three-quarters used to launch the company’s own Starlink satellites.

What’s even more striking is the profitability structure.

Starlink generated $4.4 billion in operating profit in 2025. Compared to the company's overall GAAP net loss of $4.9 billion, this means that Starlink is effectively the one keeping the company afloat. The rocket launch business itself is known to be a loss-making segment. So, if we were to summarize SpaceX in one line, it would be closer to: “A company that supports a subscription-based satellite internet business with its rocket manufacturing capabilities.” The rocket is the flashy, visible face of the company, but the real body that earns the money is the over 12 million Starlink subscribers paying their monthly fees.

Does It Really Resemble Amazon?

Barron’s likely brought up Amazon because of the valuation narrative.

When Amazon went public in 1997, it was an online bookstore with only 256 employees. Its revenue at the time of its IPO was around $15.7 million, and while its stock price jumped nearly 50 times its IPO price during the dot-com bubble, it plummeted by over 90% from its peak when the bubble burst. Yet, it survived and is now a company with a market capitalization of over $2 trillion.

SpaceX also saw a rollercoaster ride, starting with a market cap of $1.7 trillion on its first day of trading in June and briefly surpassing Amazon within a few days. Both companies share a narrative structure of “the crazed expectations of the dot-com era” and “the rewards for those who ultimately survived.”

However, there is a difference worth noting here.

Amazon was loss-making at the time of its IPO and took nearly 10 years to turn a profit. In contrast, SpaceX already had a clearly profitable business unit called Starlink at the time of its IPO—even if the company as a whole is in the red, the nature of those losses is different, as they stem from “deliberate spending” items like Starship R&D costs or losses related to the xAI acquisition. In fact, SpaceX’s current situation is more similar to Amazon during the Amazon Web Services (AWS) era in the early-to-mid 2010s. The structure where a high-margin business like AWS supported the company when retail margins were thin mirrors the current SpaceX, where Starlink covers the low margins of the launch business.

I heard one analyst point out, “Its revenue is one-twentieth of Amazon’s, but it’s receiving an Amazon-level valuation,” which is an accurate comparison. However, since this is based on “total company revenue,” the story might change if you isolate the high-margin business unit.

It is also worth breaking down why the company reported a net loss. Of the 2025 capital expenditures, $1.27 billion was invested in AI, and it reportedly poured $7.7 billion into AI in the first quarter of 2026 alone. SpaceX absorbed and merged with xAI, another one of Elon Musk’s companies, earlier this year, meaning that the previously loss-making AI business was brought into SpaceX’s financial statements. Therefore, a significant portion of the current losses is closer to the result of taking on an entire AI startup that was a separate business, rather than any weakness in the rocket or satellite business.

This is the point that makes SpaceX’s valuation even more complex—the boundaries are becoming increasingly blurred as to whether investors are buying a rocket company, a satellite internet company, or an AI company.

Amazon Web Services and Starlink: Overlapping Pictures

The decisive turning point for Amazon’s current status was not actually retail, but its cloud business, AWS. As the high-margin cloud infrastructure business supported the thin-margin online shopping mall business, the company’s overall profitability moved onto a completely different trajectory.

What is happening at SpaceX now is structurally similar. The high-margin subscription business called Starlink is supporting the low-margin (and even loss-making) rocket launch business.

However, the difference is that it took Amazon over 10 years for AWS to establish itself, whereas for SpaceX, this structure was already complete before its IPO. This suggests that the market is pricing in the “next stage” in advance.

Volatility Issues Due to Low Float

One of the reasons the stock price jumped and dropped by 20% within a few days of its IPO is that the number of tradable shares was very small. Since only about 4% of the total shares were released for trading at the time of the IPO, it created a structure where there were many buyers but few sellers, making the price easily volatile.

Research firms like CFRA reportedly issued sell ratings immediately after the IPO, citing that “the growth strategy is overly ambitious and capital-intensive.” Since this initial volatility is often due to pure supply and demand issues rather than a change in the company’s fundamentals, it feels a bit early to judge the company’s value based solely on stock price movements.

Why a Rocket Company Solves Local Governments’ Water Worries

Another example of SpaceX’s influence recently occurred in Brownsville, Texas.

As SpaceX requested to remove a 445-acre site near its Starbase launch facility from the city limits, it agreed to contribute up to $220 million to the city’s water and sewage infrastructure projects in return. Brownsville currently relies on the Rio Grande for 70% of its water, and this investment is said to allow the city to reduce its reliance on the river by increasing the capacity of desalination facilities and water treatment plants.

While it may look like a heartwarming story of local cooperation on the surface, looking at it from another angle, it means a single private company has the bargaining power to influence a local government’s city planning boundaries and water and sewage budgets. In fact, it is said that SpaceX rejected the city council members’ requests for “more time” during the negotiation process, and the motion was passed after a special Saturday meeting that lasted over six hours, so it seems it wasn’t a decision the local community accepted willingly. As the company continues to expand around Starbase, it seems highly likely that these types of negotiations will recur in the future.

While the valuation debate will eventually be settled by the market over time, I personally thought it was worth questioning the perspective of viewing this company solely through the lens of a “space exploration company.” If I were to give it a name closer to the reality of SpaceX right now, I think it would be more along the lines of “A satellite internet company, also makes rockets as a side job.”