Impact of SpaceX rideshare on small sat launchers market

[DanClemmensen]

The average amount of actual payload mass on a Transporter mission is about 3 tons. Elon said on Lex Fridman podcast that they are trying to get F9 marginal cost below $20MM. So at $5k/kg, they actually are losing money (certainly not making a material amount).

Utilization is a huge problem due to poor packing efficiency with standard ports and a large center column. This problem will be dramatically worse on Starship.

SpaceX asserts that the marginal cost of a Starship launch will be below the F9 marginal cost, so even with an identical "actual payload mass" the cost to SpaceX per kg will be lower, even though the "problem" (ratio of actual mass to max theoretical mass) is a great deal worse: 3:17 for F9, and 3:150 for Starship.  Wasted capacity is not the metric. Profit per launch is the metric.

In addition, SpaceX may provide a different kind of standardized dispenser. For instance a dispenser with the same form factor as a V.2 Starling pair could fit into the pez dispenser. I attempted a naive calculation and found that one V.2-sized dispenser can handle several hundred U of cubesat.

[Asteroza]

Here's the thing though. If the argument that $5k was too low, and the market would have accepted $10k from SpaceX for the same launch volume, then what specifically was SpaceX trying to achieve by going too low?


Were they trying to kill off a specific newspace competitor? If the thesis is that $5k was too low, it would need to be justified with a specific business target, not some vague "lets kill all the smallsat launchers in the crib" plan.

Or was $5k a specific market inducer tactic, such that now that SpaceX is doing regularly scheduled flights (along with a long line of customers) they have created/established a rideshare market that is stable enough to no longer need a market jumpstart price (market start subsidy effectively), thus the regular price increase.

[Robotbeat]

The average amount of actual payload mass on a Transporter mission is about 3 tons. Elon said on Lex Fridman podcast that they are trying to get F9 marginal cost below $20MM. So at $5k/kg, they actually are losing money (certainly not making a material amount).

Where the heck does the 3ton figure come from? And if you’re referring to the 2021 interview… note that SpaceX has doubled the max number of flights per booster and has perfected fairing recovery. Additionally, Elon has elsewhere said about $15M marginal cost is the value F9 can get to. (And note Transporter can use the shorter, cheaper upper stage nozzle and SpaceX can fill in space with Starlinks.)

And the price is $5500/kg.

[DanClemmensen]

Here's the thing though. If the argument that $5k was too low, and the market would have accepted $10k from SpaceX for the same launch volume, then what specifically was SpaceX trying to achieve by going too low?


Were they trying to kill off a specific newspace competitor? If the thesis is that $5k was too low, it would need to be justified with a specific business target, not some vague "lets kill all the smallsat launchers in the crib" plan.

Or was $5k a specific market inducer tactic, such that now that SpaceX is doing regularly scheduled flights (along with a long line of customers) they have created/established a rideshare market that is stable enough to no longer need a market jumpstart price (market start subsidy effectively), thus the regular price increase.

It's possible that the low introductory prices were used to evaluate the market elasticity, as you surmise, and the increases will explore the elasticity curve further. However, it's also possible that that since F9 and F9 operations are mature, there is not a lot of room for further cost savings, and the increases will cover operating cost inflation.

[Robotbeat]

There has been more than 10% accumulated inflation since Transporter has been announced, so sounds fair to me.

[trimeta]

I'd like to remind readers here of an oft-overlooked section of Ashlee Vance's "When the Heavens Went on Sale":

In May 2019, I helped arrange a meeting between Beck and Musk. Beck had flown to Rocket Lab’s California offices, and Musk had found time in his schedule, even though he’d claimed to have little interest in Rocket Lab. The meeting would forever change the relationship between the giant SpaceX and the underdog Rocket Lab.

SpaceX executives had long pushed Musk to use some of the company’s Falcon 9 rockets to fly huge batches of small satellites into orbit. SpaceX had allowed the small devices to hitch rides alongside bigger satellites now and again, but some members of SpaceX thought there could be a decent business in flying tons of the things at once. Pete Worden had made similar requests of Musk in the past to help with machines built by NASA Ames and its partners. During one such meeting between the two men, Musk had gone ballistic at the very suggestion of the idea. “Stop asking about this,” Musk had said, according to a person in the room. “This really fraks me off. We are never going to do this.”

If SpaceX did decide to fly lots of small satellites in one go, it would pose a major threat to Rocket Lab just as the company was hitting its stride. SpaceX’s large Falcon 9 rockets gave it an overall cost and cargo advantage. Someone would need to pay $60 million for a SpaceX launch instead of Rocket Lab’s $6 million, but they’d be able to put entire satellite constellations into place with a single rocket instead of buying them month by month.

By that time, Brian Merkel had left his job setting up Rocket Lab’s US operations and returned to SpaceX. Ahead of the dinner with Beck, Musk asked one of his vice presidents to debrief with Merkel about Rocket Lab and help ascertain how real of a competitor it was. “I said that I couldn’t speak to how successful they would be as a business but that they’re great engineers and that their rockets will launch and fly well,” Merkel said. “I don’t know exactly what was said at that dinner, but people came back afterward and said Elon had come away impressed. I think Pete had conveyed a vision for Rocket Lab that was not that far off from what SpaceX had in general. I think Elon may have used the meeting as a piece of feedback that told him Rocket Lab was attracting a bunch of business and SpaceX should take some of it.”

Part of the reason Beck asked to meet with Musk could be chalked up to ego. Understandably, Beck wanted people to realize that he and Rocket Lab were in the same ultraelite club as Musk and SpaceX. More to the point, he wanted Musk to recognize him as a peer. Beck retained much of his low-key, humble New Zealand spirit, but he’d always harbored grand ambitions. Rocket Lab’s success had started to inflate his self-confidence and given rise to yearnings for adoration. The problem, of course, was that striving to get onto Musk’s radar meant that he just might end up on Musk’s radar, which is historically a horrible place to be.

In August 2019, SpaceX revealed a new plan to begin regular launches for small-satellite makers. It would free up entire Falcon 9s and let various companies buy space on the rockets. If a company wanted to send five hundred pounds of cargo, about the equivalent of what an Electron would carry, it would cost a bit more than $1 million, or $5 million cheaper than Rocket Lab’s charge. SpaceX later put up a record 143 satellites in a single launch via the program.

I'll leave the specifics up for debate, but if we take the order of events laid out here as accurate, the concept that SpaceX started its Transporter program simply because they wanted to give small satellites a cheaper ride to orbit, and that the idea of "this will bankrupt all of our competitors, killing them before they can develop reusable vehicles that could challenge our dominance" wasn't at least one factor in initiating that program, is perhaps a bit naïve.

(The ethics of saying "we're going to bankrupt our competitors by using our first-mover advantage to provide services at a price they can't match, and still make a profit for ourselves so it isn't dumping and is perfectly legal" is a separate question.)

[deltaV]

If someone did a 500kg to LEO full RLV (like a mini version of Stoke's Nova), they could almost certainly eat most of the Transporter market, even if SpaceX cut its prices. I think a small RLV could win even in a Starship world.

A 500 kg to LEO full RLV may compete well with Starship but I think they'd have trouble competing with fully reusable 2-20 tonne to LEO versions of vehicles such as Stoke's Nova, Firefly MLV, Neutron, or Terran R. Those vehicles just seem like the right size to me: big enough to access most of the market but no bigger.

Fortunately the small sat launchers are privately developed so taxpayers won't be left holding the bag when one of us is proven wrong.

[Exastro]

Some questions that may be technically off-topic, but they're short and closely related to the thread topic:

1) What's the impact of SpaceX rideshare on the success of new smallsat startups?  Another way to look at it: if Transporter suddenly disappears, what fraction of smallsat builder/operators will find their business cases no longer close?

2) If you add together the percentages of the small launch market that each smallsat launch startup figured they'd need to capture to make their business case close, how many times larger than the actual market for such launches would that sum be?

3) Closely related: what fraction of today's Transporter customers would be in business and launching on a small-launcher competitor if Transporter had never existed?

4) What fraction of smallsat-launcher startups would have gotten their early financing if SpaceX hadn't already demonstrated that it's possible to make money in the commercial space business?

I don't know the answers to any of these questions, but it seems to me that if we're going to argue about the impact of Transporter on other parts of the commercial space sector then we ought to be interested in those answers.

[DanClemmensen]

Some questions that may be technically off-topic, but they're short and closely related to the thread topic:

Your questions relate to companies. In addition to companies that build smallsats, apparently a lot of smallsats are built by academic groups and other researchers. Many of these customers appear to be funding-constrained. lower prices would tend to increase the number of satellites in a highly elastic fashion. Is this effect real? has it been measured?

I suspect there is a large indirect advantage to the industry in general and SpaceX in particular to lowering the cost for these customers. It will tend to attract students to the space field, which should eventually expand the satellite industry and therefore the market for all satellite sizes.

[deltaV]

I'd like to remind readers here of an oft-overlooked section of Ashlee Vance's "When the Heavens Went on Sale":

Yeah that quote does raise questions about SpaceX's claims to like competition. I wouldn't convict them of being a jerk based solely on that, but it is negative info.

Unrelatedly, https://www.spacex.com/rideshare/ seems to offer the same $5.5k/kg price for both LEO and SSO. It looks like all 15 past and future Transporter missions listed on Wikipedia are to SSO. I bet the reason SpaceX is charging the same price despite SSO being harder to reach is SpaceX wants customers who would be fine with either SSO or LEO to choose SSO so they can use the same launches as the customers who need SSO and help fill up a launch faster. For the customers that need SSO this makes SpaceX's price a bit harder for other companies to beat than it is for LEO. For customers who can't use SSO it looks like Transporter may only be a paper competitor.

Edit: never mind, there are four rideshare missions to LEO scheduled. They're just called "Bandwagon" not "Transporter".

[deltaV]

(And note Transporter can use the shorter, cheaper upper stage nozzle and SpaceX can fill in space with Starlinks.)

Second generation Starlinks are apparently in 33, 43, and 53 degrees inclination orbits. Transporters appear to be SSO so the plane change maneuver to also launch Starlinks would be extreme. Bandwagons appear to be 45 degrees and 45 to 43 is a more plausible plane change.

[Asteroza]

I'd like to remind readers here of an oft-overlooked section of Ashlee Vance's "When the Heavens Went on Sale":

In May 2019, I helped arrange a meeting between Beck and Musk. Beck had flown to Rocket Lab’s California offices, and Musk had found time in his schedule, even though he’d claimed to have little interest in Rocket Lab. The meeting would forever change the relationship between the giant SpaceX and the underdog Rocket Lab.

SpaceX executives had long pushed Musk to use some of the company’s Falcon 9 rockets to fly huge batches of small satellites into orbit. SpaceX had allowed the small devices to hitch rides alongside bigger satellites now and again, but some members of SpaceX thought there could be a decent business in flying tons of the things at once. Pete Worden had made similar requests of Musk in the past to help with machines built by NASA Ames and its partners. During one such meeting between the two men, Musk had gone ballistic at the very suggestion of the idea. “Stop asking about this,” Musk had said, according to a person in the room. “This really fraks me off. We are never going to do this.”

If SpaceX did decide to fly lots of small satellites in one go, it would pose a major threat to Rocket Lab just as the company was hitting its stride. SpaceX’s large Falcon 9 rockets gave it an overall cost and cargo advantage. Someone would need to pay $60 million for a SpaceX launch instead of Rocket Lab’s $6 million, but they’d be able to put entire satellite constellations into place with a single rocket instead of buying them month by month.

By that time, Brian Merkel had left his job setting up Rocket Lab’s US operations and returned to SpaceX. Ahead of the dinner with Beck, Musk asked one of his vice presidents to debrief with Merkel about Rocket Lab and help ascertain how real of a competitor it was. “I said that I couldn’t speak to how successful they would be as a business but that they’re great engineers and that their rockets will launch and fly well,” Merkel said. “I don’t know exactly what was said at that dinner, but people came back afterward and said Elon had come away impressed. I think Pete had conveyed a vision for Rocket Lab that was not that far off from what SpaceX had in general. I think Elon may have used the meeting as a piece of feedback that told him Rocket Lab was attracting a bunch of business and SpaceX should take some of it.”

Part of the reason Beck asked to meet with Musk could be chalked up to ego. Understandably, Beck wanted people to realize that he and Rocket Lab were in the same ultraelite club as Musk and SpaceX. More to the point, he wanted Musk to recognize him as a peer. Beck retained much of his low-key, humble New Zealand spirit, but he’d always harbored grand ambitions. Rocket Lab’s success had started to inflate his self-confidence and given rise to yearnings for adoration. The problem, of course, was that striving to get onto Musk’s radar meant that he just might end up on Musk’s radar, which is historically a horrible place to be.

In August 2019, SpaceX revealed a new plan to begin regular launches for small-satellite makers. It would free up entire Falcon 9s and let various companies buy space on the rockets. If a company wanted to send five hundred pounds of cargo, about the equivalent of what an Electron would carry, it would cost a bit more than $1 million, or $5 million cheaper than Rocket Lab’s charge. SpaceX later put up a record 143 satellites in a single launch via the program.

I'll leave the specifics up for debate, but if we take the order of events laid out here as accurate, the concept that SpaceX started its Transporter program simply because they wanted to give small satellites a cheaper ride to orbit, and that the idea of "this will bankrupt all of our competitors, killing them before they can develop reusable vehicles that could challenge our dominance" wasn't at least one factor in initiating that program, is perhaps a bit naïve.

(The ethics of saying "we're going to bankrupt our competitors by using our first-mover advantage to provide services at a price they can't match, and still make a profit for ourselves so it isn't dumping and is perfectly legal" is a separate question.)

A conspiracy nut might argue SpaceX made this move to support the industrial base, assuming the thesis that the smallsat market could not stably sustain two small launchers. Deliberately bankrupting everyone BUT RocketLabs (the only ones flying) early by undercutting prevents a total dumb VC money blowout, and subsequent funding retraction from the overall newspace industry, which would have knockon chilling effects as talent decides space engineering work is for chumps and heads to greener pastures in adjacent industries. If RocketLabs lives (particularly on missions that have poor compatibility with rideshares), there's a lifeline to the smallsat industry as a whole for small constellations, and anyone trying for larger numbers would naturally gravitate to Falcon 9.

We are seeing the beginnings of a similar situation in medium lift, as Falcon 9 is expected to effectively retire, creating a vacuum in the dedicated medium lift market for "greater than small" constellations and "a little larger than a smallsat" sats, that for various reasons don't mesh well with Starship. We've got a bunch of medium partial RLV's in development now trying to step into the hole Falcon 9's exit will create, but Starship sucking the air out of the room means one must ask if the remaining medium lift market can sustain multiple launchers as well. But all those medium lift plays are sandwiched between F9 and SS, compared to the small launcher side where the only real action is RocketLabs currently.

[M.E.T.]

Look, I guess I’m a bit confused. Is someone actually claiming SpaceX’s Transporter move was not aimed at killing Rocketlab and other wannabees?

Of course it was. And rightly so.  As an analogy, Elon loves Polytopia. He says he is “built for war”. In the game, you take competitors’ territory and resources early, before they become a threat, else it becomes a brutal slugfest. That’s how you win. By being faster, more efficient, more ruthless.

Of COURSE he is happy to make only a small profit on Transporter flights, if it takes competitors out of the equation.

That’s just good business sense. Are some of you claiming he should not do that out of some sense of solidarity or even charity?

[DigitalMan]

For ages Gwynne would talk about how launching a bunch of smallsats was like 'herding cats' to quote the phrase she used. You don't seriously think that just because a number of launch companies eventually appeared on the market that SpaceX wouldn't try to solve their 'cat' problem?

[deltaV]

the small launcher side where the only real action is RocketLabs currently.

There's also ABL Space's RS 1 (1,350 kg to LEO), Firefly Alpha (1,030 kg), and Astra's Rocket 4 (550 kg). None of the three companies has a great record so far, they're batting 0/1, 1.5/3 and 2/8 respectively. But it's possible to recover from a bad start, e.g. the first 3 Falcon 1 launches failed.

[woods170]

The average amount of actual payload mass on a Transporter mission is about 3 tons. Elon said on Lex Fridman podcast that they are trying to get F9 marginal cost below $20MM. So at $5k/kg, they actually are losing money (certainly not making a material amount).

Utilization is a huge problem due to poor packing efficiency with standard ports and a large center column. This problem will be dramatically worse on Starship.

Wrong. SpaceX is NOT losing money on the Transporter missions. Educate yourself by checking the SpaceX rideshare website.

The starting fee for any rideshare payload is $275,000. That gives you a ride to SSO for anything weighing up to 50 kg. The $5,500 per kilogram only applies to additional kilograms on top of those 50 kilograms.

But here is the thing: regardless of a Transporter payload weighing 1 kg or 49 kg, both pay the starting fee of $275,000 to get launched.
If I, for example, would fly a 2 kg cubesat on a Transporter mission, then I still have to pay the entire starter fee of $275.000. This is despite the fact that for the same starter fee I could have flown a satellite that is 25 times more heavy.

So, a Transport mission flying 103 payloads generates a minimum revenue of 103 x $275,000 = $28,325,000.
Marginal cost of a Falcon 9 launch was revealed 3 years ago as being "under $25M". Meaning that it is probably even lower today. So, at a bare minimum that particular Transporter mission with 103 payloads generated a profit of multiple millions of dollars.


And if the Transporter missions are profitable, then nobody can accuse SpaceX of deliberate low-pricing to push the competition out of the market. What actually is happening is that the competition doesn't have its act together.

[Zed_Noir]

If someone did a 500kg to LEO full RLV (like a mini version of Stoke's Nova), they could almost certainly eat most of the Transporter market, even if SpaceX cut its prices. I think a small RLV could win even in a Starship world.

You might be right. However someone have to come up with the money and resources to developed said small RLV. VC investors had seen the writing on the wall, small launches don't seem to be to be profitable enough to developed and operated. Especially if the folks from Hawthorne continues to siphoning off most of the upcoming payloads. Think it is too late for a new or current small RLVs to enter the market, except for Rocket Lab who is transitioning to a medium RLV from the Electron.

[JayWee]

The average amount of actual payload mass on a Transporter mission is about 3 tons. Elon said on Lex Fridman podcast that they are trying to get F9 marginal cost below $20MM. So at $5k/kg, they actually are losing money (certainly not making a material amount).

Utilization is a huge problem due to poor packing efficiency with standard ports and a large center column. This problem will be dramatically worse on Starship.

But here is the thing: regardless of a Transporter payload weighing 1 kg or 49 kg, both pay the starting fee of $275,000 to get launched.
If I, for example, would fly a 2 kg cubesat on a Transporter mission, then I still have to pay the starter fee of $275.000. This is despite the fact that for the same starter fee I could have flown a satellite that is 25 times more heavy.

So, a Transport mission flying 103 payloads generates a minimum revenue of 103 x $275,000 = $28,325,000.

Tiny cubesats tend to be deployed by a intermediary (like  Exosat). They don't have a contract with SX.

[john smith 19]

I think the real lesson is just that trying to compete with a semi- or fully-reusable vehicle with a fully expendable one is a dumb idea.

With the proviso that you're competing in the same weight class.

It's been what 8 years since SX demonstrated recovering and reuse in what was the medium weight class and more than 4 years since RL announced it had a plan to do stg 1 recovery on a vehicle in a weight class < 1/20 of F9.

What I don't get is the thought processes of anyone who, being aware of these facts then goes ahead and thinks "We're going to do a smallsat ELV".

Of course if your ELV is huge like SLS huge then you've got the market for those payloads to yourself and reusability is not a threat.

For everyone else OTOH...

In terms of operational costs in principal (because the devils in the details and the implementation can still blow away any savings) but with a good implementation

all reusable stage LV > semi reusable stage LV > ELV

How many of these (supposedly 200?) LV designs are less than 4 years old? "We have a first stage recovery plan that in outline is.,.."   should be in the boilerplate of every one of those powerpoint presentations.

I guess we are going to be seeing a whole lot more "evolution in action" 

[TrevorMonty]

Its so dramatically low because they want to maintain a monopoly. The entire purpose is to starve the new launch startups of money, so they cant grow up, make bigger rockets, and actually compete with spaceX.

That is one view the other is SpaceX is trying to help smallsat operators grow their business. Starving competition doesn't necessarily grow your business. Making your customers business more profitable does.