Impact of SpaceX rideshare on small sat launchers market

[toren]

I've found two bases for analysis to be useful...
 Valleyites also think in terms of 'platforms' that generate a market and allow the platform provider to gain benefits from it.

Which is so much nicer (and legally safer) than saying "Establish an effective monopoly" or "walled garden" but the effect is the same.

There's no inherent right for an incumbent or startup to have the market structure or economics work out to their benefit. I'll point out that the most successful of the small sat launchers, Rocket Lab, is on the same track - moving downstream into the satellite platform market, while moving towards reuse. Almost make you think Elon is on to something.

You might also consider rule 3
3) Do the simple stuff first so your investors see you are doing something with their money. Then hope or pray that something turns up to solve the difficult parts of the problem you glossed over in your sales pitch.
That's less of a SX thing (excepting F9 S2 recovery and reuse of course) and more of a general rule of thumb with startup types. Radian and Hermeous come to mind for example.

Considering SpaceX's outcome versus the hypothetical 'Elon gave up' alternative, I'll take it. That world has space dominated by Chinese launch with the US stuck with old space and pork barrel NASA projects.

Yes, there are startups in many domains (mine was software and services) that commit the sins of 'underpants gnomes' business plans and 'science project' engineering. One way around that is to find founder(s) who can be trusted to have the knowledge, experience and integrity to deal with the inevitable unknown unknowns. Musk has scored repeatedly over the years, and has a following of investors who have profited accordingly and are willing to keep backing him. Putting him and SpaceX into the same bucket as (for instance) half-assed SPACs funded by naifs is making a mistake in kind.

[RedLineTrain]

SpaceX isn’t a launch company. They’re a megaconstellation company with a launch side project.

Important to note that the ground segment is probably the majority portion of that "megaconstellation company" description.

[XRZ.YZ]

BTW the global launch market is more like $4-5B per year, not $10B.

I’m not sure how much SLS is counted in there, but SLS is $2B per year, for one launch per year. Kind of nuts when you think about it. So take away SLS and the launch market is like $3B per year.

Anyway, yeah, we need to find more launch demand. Starlink and other megaconstellations need to be as big as possible.

I suggest do not believe any of the space industry report on market size.
For example the largest section of satellite service is satellite TV.
In those reports they count all revenue of satellite TV companies.
But actually the majority of payment for satellite TV service is for content, not for distribution of the content (via satellite).

[XRZ.YZ]

There's no inherent right for an incumbent or startup to have the market structure or economics work out to their benefit.

So for the public good of society, regulators and others need to create way to create favorable conditions for startups (apply to lots of industries) as perfect free market only exist in text book.
There usually has barriers to entry and in aerospace the barrier to entry is usually extremely high.
In such case market structure is inherently favoring establishment.

[XRZ.YZ]

also a frequent used disclaimer in financial industry

past performance does not guarantee future results

[toren]

There's no inherent right for an incumbent or startup to have the market structure or economics work out to their benefit.

So for the public good of society, regulators and others need to create way to create favorable conditions for startups (apply to lots of industries) as perfect free market only exist in text book.
There usually has barriers to entry and in aerospace the barrier to entry is usually extremely high.
In such case market structure is inherently favoring establishment.

Which perfectly explains SpaceX. 

[M.E.T.]

The broader cloud that casts a shadow over this entire discussion, which I still don’t think gets fully appreciated by many, is the absolutely tiny size of the commercial launch market.

Robotbeat rightly pointed out that the $10B I have been using as a ballpark figure for illustrative purposes is probably closer to $4B.

When evaluating the sustainability of healthy competition, one is tempted to present Apple and Samsung or perhaps Intel and AMD as examples where competition arose after heavy investment by a rival to challenge the original dominant player in each market.

But consider that both the smartphone and semiconductor chip markets are ~$500B in size, respectively. So billions invested are more than warranted given the potential returns.

By contrast, for a small launch company, success after years of development and catastrophic risk will maybe get you 25% of the $4B non-Starlink launch market. That’s a measly $1B annual revenue.

And in the low cost launch market of tomorrow, as a small launcher you probably have to launch 300-400 times a year just to get to that $1B.

Is the whole effort worth it?

Edit

I feel a necessary perspective shift is to look at this from the point of view of an investor in a small launch start up in 2023, rather than from the point of view of what may be a public good for the market as a whole in 2023. An investor doesn’t care about improving competitive market outcomes. He cares about return on capital invested.

[Robotbeat]

The investor MIGHT care about improving competitive outcomes if they also are investing in OneWeb or Kuiper or similar… or even just satellite manufacturing which is significantly larger than the launch business (factor of 5?).

There’s a reason why RocketLab rapidly entered the satellite manufacturing business.

[Robotbeat]

Astra poo pooed reuse. As did RocketLab originally. Vega/Ariane especially. Etc, etc.

[john smith 19]

There's no inherent right for an incumbent or startup to have the market structure or economics work out to their benefit. I'll point out that the most successful of the small sat launchers, Rocket Lab, is on the same track - moving downstream into the satellite platform market, while moving towards reuse. Almost make you think Elon is on to something.

Which is not a reply to my observation. It's more a semi-random comment. 

Considering SpaceX's outcome versus the hypothetical 'Elon gave up' alternative, I'll take it. That world has space dominated by Chinese launch with the US stuck with old space and pork barrel NASA projects.

Musk was never going to give up. I think he tends to view every setback as a dress rehearsal for success.

Yes, there are startups in many domains (mine was software and services)

Why am I not surprised?

that commit the sins of 'underpants gnomes' business plans and 'science project' engineering. One way around that is to find founder(s) who can be trusted to have the knowledge, experience and integrity to deal with the inevitable unknown unknowns.

Isn't that kind of the job description of VC's? To spot companies with the resilience to overcome obstacles (especially when "Plan A" fails) ?In fact when Musk founded SX in 2002 he had no experience of rocket design (although degrees in Physics and Economics are I think quite a good start). that changed when Gwynn Shotwell (Ex Microcosm) came on board.

Musk has scored repeatedly over the years, and has a following of investors who have profited accordingly and are willing to keep backing him. Putting him and SpaceX into the same bucket as (for instance) half-assed SPACs funded by naifs is making a mistake in kind.

Yes he has. But I still think it's a  pretty good rule to apply when talking to any team at a startup with origins in the Valley.    The implication that no "professional" investor would touch such companies or invest in them is nonsense. The Internet bubble of the early 2000's (and others before and since) is proof of that. Uninvested money will always find some way to get invested.  .

It takes a very strong operation to say "I don't like any of these options and we're leaving it in the bank" when the prospect of getting in on the ground floor of the next DEC is still out there. Even if everyone of them is poorly/under staffed and their ambitions are delusionally optimistic 

[john smith 19]

Astra poo pooed reuse. As did RocketLab originally. Vega/Ariane especially. Etc, etc.

Again RL's response has been instructive. I haven't re-checked but I'd be prepared  Beck ruled out reuse for Electron IE that it was impossible at that scale. With the margins in the design he couldn't see how they could make it work. This is a conservative approach.

Now that RL has hard data on what actually happens in flight either some (all?) of their estimates have turned out to be very conservative or they've realised they can make it work. Note they are not going the F9 route of a powered landing. Data has created new options and allowed them to redefine their goals, which is how things are supposed to work.

Arianes views are driven as much by non-economic forces and a rather narrow viewpoint as their business. They also didn't want any more competition in the market. The French aerospace community seems to lost its ability to think creatively. Very disappointing.

[edzieba]

Something like Stoke, however, could easily give Starship a run for its money on smaller payloads (up to medium lift). The heatshield tech is metallic so in principle should have much lower turnaround costs on a per launch basis.

We keep circling back to this. Starship does not need to compete on a launch-for-launch basis for smaller payloads, because the huge majority of smaller payloads are perfectly happy to be aggregated.

Out of curiosity, how many small satellite developers have you spoken to about this? I'm just laughing because I had yet another conversation with a smallsat developer last week who was hoping that the small launchers are successful precisely because they're not happy aggregating on Transporter missions. They're far from the first I've heard that from.

~Jon

This is rather the key: Transporter missions only eat up the market for smallsats that can tolerate going to a Transporter orbit (which means either a Starlink shell orbit, or trying to corral enough smallsats that all want to go to a different orbit. See: the headaches involved in getting SmallSat Express together). That means university or private test payloads that have loose orbital requirements, but not payloads that have specific orbit requirements (e.g. SAR).
There are rideshare tugs as an option (and some have even worked successfully!), but once you start adding the cost of a tug stage to get from a shared dropoff orbit to the target orbit, you start seeing total mission costs approach dedicated smallsat launchers but still requiring you to deal with the headaches of a rideshare, e.g. scheduling, materials/outgassing limitations, etc.
With a really big constellation where you are packing enough satellites into the same or similar orbits that you can start to fill up a medium or even heavy lift vehicle for a dedicated launch, small launch vehicles then start to become less attractive again. But that leaves a large range where you can have a large constellation of less tightly packed vehicles that end up best launched on smallsat launchers than rideshares. And it blunts some of Starship's advantages in cost/kg if your total plane mass is low to moderate, as Starship then needs to compete on cost/launch instead of cost/kg.

[DanClemmensen]

Something like Stoke, however, could easily give Starship a run for its money on smaller payloads (up to medium lift). The heatshield tech is metallic so in principle should have much lower turnaround costs on a per launch basis.

We keep circling back to this. Starship does not need to compete on a launch-for-launch basis for smaller payloads, because the huge majority of smaller payloads are perfectly happy to be aggregated.

Out of curiosity, how many small satellite developers have you spoken to about this? I'm just laughing because I had yet another conversation with a smallsat developer last week who was hoping that the small launchers are successful precisely because they're not happy aggregating on Transporter missions. They're far from the first I've heard that from.

~Jon

This is rather the key: Transporter missions only eat up the market for smallsats that can tolerate going to a Transporter orbit (which means either a Starlink shell orbit, or trying to corral enough smallsats that all want to go to a different orbit. See: the headaches involved in getting SmallSat Express together). That means university or private test payloads that have loose orbital requirements, but not payloads that have specific orbit requirements (e.g. SAR).
There are rideshare tugs as an option (and some have even worked successfully!), but once you start adding the cost of a tug stage to get from a shared dropoff orbit to the target orbit, you start seeing total mission costs approach dedicated smallsat launchers but still requiring you to deal with the headaches of a rideshare, e.g. scheduling, materials/outgassing limitations, etc.
With a really big constellation where you are packing enough satellites into the same or similar orbits that you can start to fill up a medium or even heavy lift vehicle for a dedicated launch, small launch vehicles then start to become less attractive again. But that leaves a large range where you can have a large constellation of less tightly packed vehicles that end up best launched on smallsat launchers than rideshares. And it blunts some of Starship's advantages in cost/kg if your total plane mass is low to moderate, as Starship then needs to compete on cost/launch instead of cost/kg.

Yes, there is a market for dedicated launches of small satellites: that has never been at issue. The two questions are:
    * How big is this market after you remove those that will use cheaper service based on aggregated payloads?
    * What will the price be to use Starship for a small-mass dedicated launch?

[Robotbeat]

Astra poo pooed reuse. As did RocketLab originally. Vega/Ariane especially. Etc, etc.

Again RL's response has been instructive. I haven't re-checked but I'd be prepared  Beck ruled out reuse for Electron IE that it was impossible at that scale. With the margins in the design he couldn't see how they could make it work. This is a conservative approach.

Now that RL has hard data on what actually happens in flight either some (all?) of their estimates have turned out to be very conservative or they've realised they can make it work. Note they are not going the F9 route of a powered landing. Data has created new options and allowed them to redefine their goals, which is how things are supposed to work.

Arianes views are driven as much by non-economic forces and a rather narrow viewpoint as their business. They also didn't want any more competition in the market. The French aerospace community seems to lost its ability to think creatively. Very disappointing.

it would be funny to insist that reuse isn’t “possible” at electron scale. Hilarious to me how people have apparently no idea what the word “possible” means. (And New Shepard is at Electron’s scale, does VTVL reuse, and could be made into a pop-up reusable firsts stage…)
(Even Full reuse actually is possible even at Electron scale.)

But regardless, expendable needs to die in commercial launch (there’s a reasonable case for emergency or military or robotic planetary missions for some expendable stages). It’s good that it dies, and that companies that stand in the way of reuse are crushed, so that room is created in workforce and capital markets for companies pursuing a long-term useful approach like reuse. Creative destruction.

[Herb]

I feel like some of the views here are rather short sighted.

Once cost per kg has been lowered sufficiently, it becomes possible to start getting into the territory of space-based solar power generation. Who knows how it'll go for sure, but there'll certainly be lots of interest from various governments across the world in dabbling.

And then we're also running into solar flare season soon (happens every 11 years give or take). Depending on how things go, a large amount of Earth's LEO satellites could take a major blow, meaning there'll possibly be a rush to replace what's been lost.

Then there's the likelihood that many companies will be seeking to start up other projects with the reduced cost of sending things up, such as space stations / messing with asteroids / moon missions and what have you.

With the increase of the number of missions which only SpaceX would be able to do with Starship (And they would likely go for it, since these types of projects would likely come from governments and... with Starlink, SpaceX needs to build up/maintain relations with those entities for licensing to sell their internet), I don't find it too far fetched that anybody seeking to build constellations of their own would be looking at a long, long wait to do anything/get anywhere.

As for current rideshare with falcon 9's, it's not easy to say. Companies looking to build up their service quickly may still go for small LV because going by rideshare will stall things for a significant amount of time (change of orbital inclination can be extremely slow). I would kind of equate it to taking the bus (Falcon 9) and having to walk a hundred miles afterwards, or paying for a limo (small LV) that gets you exactly where you need to go. Costs a high premium, but there's sure to be quite a few that still go for it. I imagine the market will be quite small for small LV in the end though, all things considered, but that's considering perfect market conditions. Some small LV's will survive by latching onto certain governments that won't care about the cost (because let's face it, having the ability to head into space is a Prestigious endeavor).

[nicp]

Maybe SpaceX should resurrect Falcon 1e.
Not gonna happen of course…

[john smith 19]

it would be funny to insist that reuse isn’t “possible” at electron scale. Hilarious to me how people have apparently no idea what the word “possible” means. (And New Shepard is at Electron’s scale, does VTVL reuse, and could be made into a pop-up reusable firsts stage…)
(Even Full reuse actually is possible even at Electron scale.)

True, but that does ignore a couple of little details. 1) Shepard uses LH2 as a fuel. That buys a substantial boost in available Isp. Of course LH2 is a well known PITA to handle safetly and use effectively, but this leads to detail #2, that I think it's fair to say Bezos put a smideon* more funding in Blue than Beck was able to raise

Unqualified absolute statements are a massive bugbear in the business. Would you have accepted "We can't make reuse work at this size with with kerolox and with the vehicle usually maxed out for payload mass. There's just no mass left over"?
Note since then Electron has had at least 1 upgrade. It is therefor (like F9) not the same design that took their first payload to orbit. And AFAIK it has also turned out that Electron is not maxed out on every launch either. And the continuing improvement in Lithium Ion batteries (as tracked by the National Renewable Energy Laboratory) is frankly bloody amazing (about a 10% year on year improvement over 20+ years. Outside of semiconductors where else do you so those sorts of sustained numbers?)

But regardless, expendable needs to die in commercial launch (there’s a reasonable case for emergency or military or robotic planetary missions for some expendable stages). It’s good that it dies, and that companies that stand in the way of reuse are crushed, so that room is created in workforce and capital markets for companies pursuing a long-term useful approach like reuse. Creative destruction.

I don't think anyone who wants to see more, better, cheaper space launch would argue with you for a minute.  From now on anyone who says "I want to build a Kerolox ELV about F9 sized but it will not be reusable  in any way shape or form" should be shown the door, although I suspect that won't stop some Old Space organisations pitching such things to their local government, along with a hugely convoluted excuse explanation of what they are not doing this. 

But remember that from 1956 to 2017 fully expendable ELVs were SOP everywhere.

What it really took was someone with a) A huge need for better space lift b)The funding (or the ability to attract the funding) to develop it c)The absolute rock-solid belief that better was possible and to keep pursuing that goal.

*In Billionaire land smidgeon is about $200m

[Blackjax]

There are rideshare tugs as an option (and some have even worked successfully!), but once you start adding the cost of a tug stage to get from a shared dropoff orbit to the target orbit, you start seeing total mission costs approach dedicated smallsat launchers

What data are you looking at to reach this conclusion?
(not agreeing or disagreeing, just interested in where the conclusion came from)

We have 3ish tug companies out there trying to reach regular operations but have we seen their pricing stabilize into a regular competitive market?

[su27k]

This is rather the key: Transporter missions only eat up the market for smallsats that can tolerate going to a Transporter orbit (which means either a Starlink shell orbit, or trying to corral enough smallsats that all want to go to a different orbit. See: the headaches involved in getting SmallSat Express together). That means university or private test payloads that have loose orbital requirements, but not payloads that have specific orbit requirements (e.g. SAR).
There are rideshare tugs as an option (and some have even worked successfully!), but once you start adding the cost of a tug stage to get from a shared dropoff orbit to the target orbit, you start seeing total mission costs approach dedicated smallsat launchers but still requiring you to deal with the headaches of a rideshare, e.g. scheduling, materials/outgassing limitations, etc.

With the new 1-ton class smallsat LVs, you'll have to rideshare as well. The only difference between them and Transporter is whether you rideshare with 10 other smallsats or 100 other smallsats.

[su27k]

Well yeah, and your comments are your personal opinion too (unless you work for SpaceX sales, in which case you shouldn't participate in this discussion anyway), so I don't see the point of this comment.

True, but they are based on logic and SX's past behavior. This is how companies that offer unique capabilities tend to operate. Your view is based on what?

Actually my views are based on SpaceX's past behavior as I explained in previous reply, and I have actually provided some examples of such behaviors.

It's real simple. You want more. You pay more. You want more than anything any other supplier can supply, then you pay the SX price, which will be negotiated with you probably under an NDA. You don't like their price you don't launch. Your choice.

Yes, they'll charge different prices for different market segments, that's hardly news.

Really? Some people seemed to be thinking they'd just charge one price regardless of the payload.

Really, different pricing for different segment is obvious to those who have been following SpaceX's past behavior. But I agree that an alarming amount of people don't seem to grasp this, this has led to faulty claims such as "Starship is too big for the market" or "How can Starship get any contract when FH got so few contracts".

No, not hope, I expect their pricing will be reasonable, since that's how they have behaved in the past.

I rarely leave my own quotes in a posting but I've decide to make an exception in your case because I think it's pretty clear what I think their pricing will be. Note particularly those word "You want more than anything any other supplier can supply,"
 Likewise I've dumped your strawman argument because it's exactly that. A strawman.

It would help if you read what I write, not what you think I wrote.   

We've already got a pretty good idea of how SX price payloads up to FH size and that's what I expect them to continue to do. The fact they continue to win commsat launch business says they are competitve with A6. They continue to launch FH so the deals they offer there seem competitive (or for NSS simply the only game in town, and we have very little visibility of their pricing in this segment).

By now SX has a pretty good idea of what it needs to make a profit on its flights and a pretty good idea of what its customers can pay.

Where the monopolist behavior comes in is where the payload is beyond FH. With DIVH gone what's the alternative? SLS? I did not say that SX's monopoly pricing would be unreasonable. I think they do have an internal "pricing model" which takes into various specific tasks, and the some kind of multiplier which allows for payload specific stuff that's not specifically identified, again this is WRT to beyond FH payloads.

But if you don't like it your options are very limited.
 1) Re-negotiate the contract. So instead of SX doing stuff for the customer the customer does them itself, like using the payload engines to do orbit raising or plane changes 
2) Re-design the payload to use other launchers or SH with its public pricing. Since you went with SH to begin with for either the volume or the mass to orbit (IE beyond FH, which they will probably phase out ASAP) this rather defeats the original purpose. 
3) Don't launch.

The point about a monopoly is that there is  no "market price" It's their price or nothing. As far as you know SX's prices have been reasonable, but without any effective competition that's their choice, and their choice alone. It can change. 

It won't be "SpaceX price" or nothing as I explained in the previous reply, which you dismissed as strawman but it really is not.

There's nothing special about Starship capability beyond FH, it's true that this will be a unique capability only available from SpaceX, but that's not a new situation, we already have situations where SpaceX is the only game in town. For example, if you want to send 2 metric tons to TLI today with a price less than $50M, Falcon 9 is the only choice that is close to your price point.

Of course Falcon 9's official list price is $67M, so does this mean you're out of luck? Not at all, because your option 1) to 3) are not the only choices, there is in fact an option #4:

4) Negotiate with SpaceX to get a discount.

This has nothing to do with fancy words such as "monopolies" but has everything to do with the oldest human behavior on Earth: Making a deal. You go to SpaceX sales and tell them "We want to buy a Falcon 9 for $40M for our lunar mission, I know your listed price is $67M but you know from public information that my revenue for the entire mission is only $80M, I truly cannot afford a $67M launch. If you give me a discount and only charge $40M, you can still get a healthy profit of say $10M if we make this deal, how about it?"

The numbers may not be exact, and obviously the sales pitch is illustratory, but if you have been following CLPS you'd know something like this has likely already happened, multiple times. So yes, SpaceX can offer a unique capability and be reasonable about the price at the same time, whether it's a "monopoly" or not really doesn't matter.