>>11286773I believe Gwynne and Elon indirectly confirmed that the cost of making a Starlink sat is ~$2-250k. Due to their vertical integration capabilities coupled with a modular and assembly line design, their full stack of 60 Starlink sats would cost them $12M minimum and $15M maximum. Factor in their F9 fleet, wherein they can use preflown boosters for their launches and assuming a $1M refurbishment cost after every launch and another $250k for fuel, their launch costs of putting up 60 Starlink sats comes to be in the ballpark of $13.25M minimum and $16.25M maximum.
If you add in say an additional $1M for launch and logistics, you'd still be looking at max of $15M minimums and max of $18M maximums. So taking that into account, a fleet of 4,000 satellites (which I believe is the confirmed minimum number to be able to service the world at like 100/100 packages or even 50/50 packages), their combined build and launch costs would be: $1Bn minimum and $1.2Bn maximum.
At 4k satellites, if their primary market is financial systems (massive exchanges) only they'd easily be able to pull in $1-2Bn in annual revenue, which means they'd break even. Since USAF has shown a lot of interest in Starlink, ferrying encrypted traffic over Starlink with a new-gen internet protocol, they'd probably be able to get another $1-2Bn/year out of the US military for being able to offer low latency and high bandwidth communication channels. Finally, if they service the tourism and transportation industries (cruise ships, freighters, and airlines) with this minimum 4K sat fleet, they'd likely be able to get another $1-2Bn/year out of that.
Overall leading to a $3Bn/year minimum YoY revenue and $6Bn/year maximum YoY revenue. If they double that to 8k satellites, they'd be able to service the aforementioned markets and then focus on third world population markets--and even with very low cost reasonable bandwidth offerings (25/25 or 50/50), the demographic is large enough for +$1Bn/year rev.
