7 ms·
Those charts show each of the newest satellites buys ~400 paying subscribers (per additional satellite). Build, launch, operating for one satellite--while low f
by forgot-im-old 3y ago
Those charts show each of the newest satellites buys ~400 paying subscribers (per additional satellite). Build, launch, operating for one satellite--while low for SpaceX--is too expensive for that to make sense.
If they try to do more than 400 users per satellite, the service throughout plummets.
- s1artibartfast 3y agoI get 500 customers/satellite (2,300,000/4,500) since 2021, and 1000 new customers per satellite for Q3 2023. Using the lower number of 500 at $120/month, that is $7,200 per customer, or $3.6 Million per satellite over a 5 year lifetime. 4600 operational satellites (out of 5700) over 134 launches equals 34/ launch or $124 million revenue. or $250 million/launch if we want to use thee 1000 users/satellite number. Space X sells commercial falcon 9 launches for $ 67 million (to GTO opposed LEO) and this already includes a profit margin. It still isn't obvious to me.
- kortilla 3y agoThis math is completely wrong with even a trivial napkin check. They announced 2 million users with fewer than 4k sats in orbit (presumably even fewer serving). Let’s assume 2 mil by 4k for simple math, which yields 500 users per sat. However, the vast majority of the sats are not where the bulk of their customer base is (US and Europe) at any given moment. That means that the 5% of sats over user dense areas at any given point in time are capable of carrying 100% of the current load. Stated differently, they have sats capable of serving the US user density sitting over every other country in the latitude bands of those orbits just waiting for licensing. I.e. User count could 10x on the current constellation if the growth is outside of the existing markets.