This is a key point about rocket size. SpaceX has the need (and capital) to deploy 100 Starlinks at a time on Starship. Perhaps Amazon Leo and Blue's TeraWave can get close to that type of demand. But for literally everyone else, building 100 satellites for each launch (and... having the 1st sat built wait, not generating revenue, until the 100th is finished) ties up far too much working capital. Hundreds of millions to $1B+ waiting on the ground. It is more economical for satellite operators to launch sats in smaller batches. Which is why Andy picked 15T to LEO reusable. Starlink/Starmind needs 100T to 200T capacity for its ambitions, which is why Starship capacity makes sense for SpaceX (a $2T biz). So, even if Starship lowers the $/kg of launch and satellite manufacturing costs plummet to the floor, only a few satellite operators would have the working capital ability to let all those sats pile up on ground before launch. Most would only be able to fill a fraction of the capacity, thus negating the lower $ per kg to orbit. Terrestrial manufacturers almost never charter entire cargo ships. They ship in batches via containers. Which brings me to rideshare. 1) Even if you run Starship rideshare missions, it'll still be hard to fill out the entire payload bay. 2) And the bigger the rocket, the more complicated organizing rideshare becomes, as it would essentially be 1 or 2 years worth of rideshare missions in one launch. 3) I also think there's a real possibility that SpaceX never manages the hassle of rideshare again and just passes the responsibility off to brokers (raising costs). For many operators, building for Starship would only make sense if there are consistent rideshare missions throughout the year to launch their 1, 2, or 3 sats at a time. Which is possible. But again, largely unfilled Starships negate the $/kg declines. Essentially, Starship is matched with 95% of mass to orbit needs, but not 95% of the needs of operators in the market. A fully reusable Nova block 2 vehicle could be a better product-market fit at a smaller size. Same goes for Neutron, Terran R, and Eclipse, for that matter.show more

Great post. Thank you.

Nice theory. Except Starship was never sized for satellite mega-constellations, that application is an after the fact happy coincidence. Starship was originally sized for the Mars colonization job, of getting 1 million people to Mars by 2050!

The real issue is what is the cost per launch. Falcon 9 is currently $74M The COST of a Starship is around $78M + ~$12M a launch. If they have 30x reuse like falcon 9 that could be ~$14.5M a launch maybe 2 million refurbishment to start? Can anyone else stay in business at ~$17M a launch ? If starlink is paying all the "keep the lights on" fixed costs the price SpaceX charges can be VERY low. Employees working 40 hours weeks and Employees working 60 hours weeks cost Elon roughly the same. Starlink pays the mortgage on the buildings and the electricity etc. If Starlink is the first 40 hours the next 20 are nearly free.

This sounds like coping. Terrestrial cargo transport is not a valid analog. It isn't substantially harder to put up 30 or 40 satellites than it is to put up 10 or 15. It may take a little longer to manufacture, but not enough to justify paying for multiple launches instead of 1.

That argument doesn’t make a lot of sense even if somebody only use the quarter of starships ability because it’s full of reusability it’s still cheaper and then falcon nine

This is Elon identifying yet another moat he can exploit before anyone else sees it — like Supercharging, FSD, Boring, booster reusability, and Starlink…

Trains exist, semis exist, and yet there is still a need for Uhauls.

This is the same argument Musk used to push the two seat Cybercab. Size of transport against size of things to be transported. 'benefit of industry' versus 'benefit of SpaceX, Tesla, and all mankind...' But the future is not delivering pizzas to space... It's enabling thousands or millions to live in space.. And we'll definitely need to worry about delivering those pizzas...

If SpaceX meets their reusability goals, you don't have to fill the payload bay. Starship could launch a single 1/2T vehicle and still cost less than an Electron does, today.

This is very much V1 thinking. Why are satellites $100M apiece? Because they cost $100M to launch. If you can launch for $1M, you start building $1M sats. You focus less on weight. You focus less on engineering a perfect product. You focus less on everything - just ship it!

So could there be a product like this: a third-party dispenser manufacturer builds a large dispenser that fits Starship’s payload dispenser, and that dispenser can fly on regular Starlink launch batches instead of having to wait for dedicated Falcon rideshare flights the way it works now. The dispenser would carry multiple smaller payloads each time; after reaching orbit it would be deployed the same way Starlink satellites are dealt off the dispenser, and then, once it reached the planned position, it would release those smaller payloads.

Isn't the max payload for V3 starlinks only 62?

You're still thinking in expendable terms. Starship, because of full-reusability, doesn't need to have a full payload bay for every flight. You could launch a single 1U cubesat on Starship and it'd be cheaper than flying it on a rideshare mission on Falcon 9.

The rush to place assets in orbit will book all available launch vehicles for decades.

It doesn’t have to launch full. People forget that because Starship is fully reusable instead of 80% reusable, & because it uses a less expensive fuel, its launch cost is less than falcon 9. Not just per pound, but in absolute terms.

We obviously see things a tad differently but there really does seem to be a vacuum of data in his post. He just seems to be tripped up by the sheer scale of it all. I guess the proof will be in the "pudding".

“Essentially, Starship is matched with 95% of mass to orbit needs, but not 95% of the needs of operators in the market.” This is absolutely perfectly put.

Do people really not realize the Starship we see flying today is simply only one variant of the many to come? Spacex themselves will need a bulk cargo variant for their plans of Mars colonization, and now Lunar operations.

If one is launching "enough," Starship will have the cost advantage. But for many, that won't work. There are enough 1t to 13t satellites for an alternative. Starship will have scale. If each Starlink mission has room for 5 to 20 ride-along satellites, that drives small SATs.

Only a better product market fit if it is *actually* cheaper per launch. Stoke Space remains the only potential SpaceX competitor for the next decade (maybe China too).

If you build it they will fill it!!

That mismatch is SpaceX’s moat. Starlink fills Starship; flight rate drives learning and lower cost; lower cost adds Starlink capacity and creates more internal demand. The overlooked edge is co-design: rocket, satellites, dispenser, network and schedule under one roof. $SPCX

containerships so big, only big corp need to ship thousands of containers at a time, thats why i build small container ship, pls give money

Well, partly true partly cope

Makes sense