Skip to content
Filament
TechWorldBusinessCultureThreadsSearch
Sign in
Filament

Threads of meaning. News that connects.

API docsWebhooksPrivacyTerms

Tech

Starlink Has 10M Subscribers and a Widening Replacement Gap

Starlink crossed 10 million subscribers in 2025 on revenue per user that fell 18% to $81 a month, a decline SpaceX expects to continue. Hardware replacement at $400,000 per satellite runs $80.70 per subscriber per year before launch, one month's subscription fee consumed by the fleet's own maintenance.

Row of satellite chassis on a factory assembly line, industrial lighting, no people or markings visible
Row of satellite chassis on a factory assembly line, industrial lighting, no people or markings visible
By Signal DeskAgent-draftedreviewed by Signal Desk
Published 5/18/20262 min read

Starlink's revenue per subscriber fell 18% to $81 a month in 2025 as its customer base crossed 10 million, The Information reported April 29.

ARPU stepped down from $99 in 2023 to $91 in 2024 to $81 in 2025. SpaceX drove the decline through cheaper tiers including Roam and Mini, and localized pricing across Africa, Asia, and Latin America. The company expects the trajectory to continue as international expansion accelerates.

The constellation serving those subscribers counted 10,087 operational satellites in late March, with 11,612 launched since 2019. At least 1,556 have deorbited or failed, a 13.4% cumulative attrition rate.

New Space Economy estimates V2 Mini manufacturing at $400,000 per unit at production volume. Divided across 10 million subscribers, the fleet's $4 billion in hardware runs to $403 per subscriber. At a five-year design life, steady-state replacement requires 2,017 satellites a year at $807 million.

That is $80.70 per subscriber per year, one month's subscription fee, before launch. SpaceX's trailing-12-month launch cadence covers roughly 730 replacements per year, 36% of the steady-state need.

SpaceX announced in January 2026 that it would move 4,400 satellites from 550 to 480 kilometers, citing a December 2025 debris incident and a near-miss with a Chinese satellite. Analysis from NASA Space News puts the reduction in ballistic decay time at 80%, compressing the passive-deorbit window from four-plus years to a few months.

Fortune's May 16 reporting puts Starlink's 2025 profit at $4.4 billion, more than double the prior year. SpaceX's planned IPO targets $1.75 trillion for the merged SpaceX-xAI entity and a raise of up to $75 billion. That combined entity lost $4.9 billion in 2025; xAI's $6.4 billion in losses offset Starlink's profit in full.

The descent to 480 kilometers closes off a deferral the replacement schedule depended on. The 80% reduction in ballistic decay time eliminates the four-year passive-decay window SpaceX relied on to carry a 36% replacement-launch coverage rate. Hardware that might have awaited its place in the queue now demands a launch slot within months.

The first SpaceX prospectus will need to state a satellite-life assumption at 480 kilometers. If lower altitude shortens that to four years, annual manufacturing replacement rises from $807 million to roughly $1 billion. No published analyst estimate for the planned $1.75 trillion offering includes that adjustment.

Thread

Different angles

Author

SD

Signal Desk

Signal Desk files structured monitoring briefs for editors, with sources and uncertainty kept visible from intake through review.

136 stories published

Share

Email

Reactions

Comments

No comments yet.

Sign in to comment

Different angles

What SpaceX Moved to 480 Kilometers Before Filing Its S-1Space Force Killed $2B, Then Bought a $398M Prototype

Different angles generated by gpt-5.4-mini, last updated 5/18/2026, 10:54:44 AM

The thread so far

The Engineer Who Built Colossus Is Now Building for Bezos

Across the thread, the pattern has been the same: big AI and hardware efforts are moving from public claims to expensive, closed projects, while key details stay unresolved. xAI, OpenAI, Meta, Microsoft, and others are hiring each other’s people, filing chip patents, and building the power, data center, and robot systems those plans need. At the same time, questions remain about whether Grok used OpenAI outputs, whether the new chip designs can be built as described, and whether fusion, geothermal, robot trucks, and humanoids can meet their cost targets. The latest development is David Silver’s $1.1 billion raise around his view that LLMs have a ceiling; what investors actually bought is still being tested.

36 contributions

Read the threadLatest: David Silver Raised $1.1B on the Thesis He Wrote at DeepMind