SpaceX is preparing to go public at a valuation exceeding $1.7 trillion, with its Starlink satellite internet network serving as the primary engine behind that staggering figure.
The connectivity segment posted $11.39 billion in revenue in 2025, a 50% increase year over year, with operating profit reaching $4.42 billion and adjusted EBITDA of $7.17 billion at a 63% margin.
That margin trajectory has been striking, climbing from 41% in 2023 to 50% in 2024 before reaching 63% in 2025, signaling strong operational leverage across the network.
Starlink now supplies more than 60% of SpaceX’s total $18 billion in 2025 revenue, generating cash that funds both the Starship program and the company’s xAI ambitions.
Subscriber growth has been close to vertical, expanding from 2.3 million users in 2023 to 10.3 million by the end of March 2026, spanning more than 150 countries worldwide.
Beneath that growth, however, sits a softer and increasingly worrying number, with average revenue per user falling from roughly $99 a month in 2023 to $66 by the end of March 2026.
The decline is largely a product of geographic mix, as most new subscriber growth is coming from lower-income regions where GDP-adjusted pricing pushes average revenue well below early-adopter levels.
SpaceX raised certain plan prices by between $5 and $10 a month in May 2026, but the structural pressure on per-user revenue is unlikely to reverse as the subscriber base continues expanding into cost-sensitive markets.
The earliest Starlink customers were remote homes, vessels at sea, and aircraft operations with no viable alternative, giving the company significant pricing power that the mainstream market simply does not replicate.
Competition is now tightening on two fronts, with new satellite entrants such as Amazon developing rival networks while fiber, cable, and fixed wireless providers extend coverage into areas that were once effectively Starlink’s exclusive territory.
The satellite replacement cycle adds another layer of financial complexity, as Starlink’s satellites are built to last around five years before reentering the atmosphere and burning up, unlike fiber infrastructure that can serve for decades once installed.
With more than 10,400 satellites in orbit as of June 2026, roughly 2,000 reach end-of-life annually and must be rebuilt and relaunched just to maintain the existing network, not expand it.
SpaceX deorbited 472 satellites between December 2024 and May 2025, and the connectivity division spent approximately $4.2 billion on capital expenditure in 2025, with a large share of that representing recurring maintenance rather than growth investment.
Government and defense contracts offer a meaningful counterweight to residential pricing pressure, with Starshield, the secure military version of Starlink, already holding a $537 million defense contract through 2027.
The FAA is also deploying roughly 4,000 Starlink terminals to upgrade the network managing U.S. airspace, while the direct-to-cell service allows ordinary phones to connect via satellite through carrier partners such as T-Mobile.
These government, defense, and carrier-facing businesses tend to be considerably less price-sensitive than residential broadband customers, and their growth could meaningfully lift blended revenue per unit over time.
The central question for investors is whether Starlink can sustain its profitability as the user base scales into lower-value customer segments, satellite replacement costs compound, and competition intensifies on multiple fronts.
At a valuation approaching 100 times revenue, SpaceX is asking investors to underwrite a combined communications, artificial intelligence, and space infrastructure thesis priced with very little margin for error.
