Relativity Space: General Thread

[TrevorMonty]

Interview with Tim Ellis on the topic of building backlog (that is, signing launch contracts) for unflown rockets. As one might expect, he strongly disagrees with Peter Beck's perspective, saying that "Deciding not to build a backlog is taking a business strategy that has failed pretty epically in history across other products, which is, ‘build it and they will come,’ without actually validating that your pricing and your product capabilities are something that solves the customer problem such that they’re willing to put up material cash up front."

Although I did find this part telling:

Ellis said that signing these LSAs also gives a greater look into a customer’s technical plans and what their requirements are. Relativity was able to see the need for a second launch provider for all the forthcoming telecom constellations — such as OneWeb’s, a company that announced a launch agreement with Relativity in June 2022 — because it was in conversation with prospective customers, Ellis said.

“There was this need and we saw that early,” he said. “Why we were able to see that early is because we were actually talking with customers, and we were actually working to sign launch deals with them.”

It doesn't seem to me like signing contracts was necessary to understand that there would be demand for a non-SpaceX launch provider among the companies directly competing with Starlink.

Interesting interviews with two different prespectives. In end it comes down to execution, both companies need to launch reliably and on schedule. Neither RLV is likely to hit flight rates constellation companies require for 2-3 years after maiden launch, which maybe reasoning behind Becks remarks. Once RLV has demonstrated a few reliable launches and recoveries then both launch company and customer has confidence that launch expectations will be meet.

RL can currently sell Electron and be reasonably confident on meet launch schedules, unexpected failures aside. SpaceX even more so with F9R. With Terran R Relativity is selling powerpoint RLV with some HW in development, it would be a brave customer to hitch their wagon to that horse without an alternative.

[edzieba]

It doesn't seem to me like signing contracts was necessary to understand that there would be demand for a non-SpaceX launch provider among the companies directly competing with Starlink.

I don't understand. If SpaceX provides the lowest price to orbit and meets all my technical and schedule requirements, why would I pick another launch provider? They'll take my money and launch my satellites. My money is very unlikely to affect their bottom line much, because if I don't use them then someone else will.

The same reason every business buys from multiple providers: redundancy. Supplier goes out of business, supplier has capacity constraint issues, supplier is bought out by another entity and slams pricing up, supplier is situated in a nation that is now subject to new export controls, supplier gets in a spat with a customer, supplier suspends product/service for any reason, supplier has longer lead times than another supplier, etc. Don't put all your eggs in one basket, even if one basket may be a bit cheaper than the others.
Plus all the aforementioned issues of bargaining power, different services available, etc.

[whitelancer64]

Sounds like a huge business risk. Seems like if you’re doing a ton of flights, it would be worth buying a few launches from all the lower cost launch providers if you can.

Sure, but how many customers are that big ("a ton of flights")? I know of two other than SpaceX: USSF (NSSL) and Amazon (Kuiper). Those two customers may sustain a credible competitive market for the rest of us. But I don't think even those two can continue to pay much more than twice the price of SpaceX service for very many launches.

There are a dozen or so other planned megaconstellations. The ones from China are likely to only use Chinese launch providers, but the rest are up for grabs.

[DanClemmensen]

Sounds like a huge business risk. Seems like if you’re doing a ton of flights, it would be worth buying a few launches from all the lower cost launch providers if you can.

Sure, but how many customers are that big ("a ton of flights")? I know of two other than SpaceX: USSF (NSSL) and Amazon (Kuiper). Those two customers may sustain a credible competitive market for the rest of us. But I don't think even those two can continue to pay much more than twice the price of SpaceX service for very many launches.

There are a dozen or so other planned megaconstellations. The ones from China are likely to only use Chinese launch providers, but the rest are up for grabs.

The only currently-viable megaconstellation is Starlink. It uses a single launch provider.  OneWeb has used multiple providers, but it has also gone bankrupt and gone through other financial contortions, so I feel its viability is questionable. It's not clear to me how many megaconstellations will actually launch. The largest of the "planned" constellations appears to be another Greg Wyler special.

Starlink is of course a unique case, SpaceX being both the launch provider and the constellation operator, with the synergy apparently being crucial to each of these two components. Maybe this is the only viable model? We don't know as much about the Chinese plans, but I suspect the structure will be functionally equivalent.

[TrevorMonty]

Sounds like a huge business risk. Seems like if you’re doing a ton of flights, it would be worth buying a few launches from all the lower cost launch providers if you can.

Sure, but how many customers are that big ("a ton of flights")? I know of two other than SpaceX: USSF (NSSL) and Amazon (Kuiper). Those two customers may sustain a credible competitive market for the rest of us. But I don't think even those two can continue to pay much more than twice the price of SpaceX service for very many launches.

There are a dozen or so other planned megaconstellations. The ones from China are likely to only use Chinese launch providers, but the rest are up for grabs.

The only currently-viable megaconstellation is Starlink. It uses a single launch provider.  OneWeb has used multiple providers, but it has also gone bankrupt and gone through other financial contortions, so I feel its viability is questionable. It's not clear to me how many megaconstellations will actually launch. The largest of the "planned" constellations appears to be another Greg Wyler special.

These constellations really need more RLV launch options. Without at least two RLVs customers aren't going to get competitive pricing. F9R is still sold at its ELV price because it doesn't have any direct competition.

[DanClemmensen]

Sounds like a huge business risk. Seems like if you’re doing a ton of flights, it would be worth buying a few launches from all the lower cost launch providers if you can.

Sure, but how many customers are that big ("a ton of flights")? I know of two other than SpaceX: USSF (NSSL) and Amazon (Kuiper). Those two customers may sustain a credible competitive market for the rest of us. But I don't think even those two can continue to pay much more than twice the price of SpaceX service for very many launches.

There are a dozen or so other planned megaconstellations. The ones from China are likely to only use Chinese launch providers, but the rest are up for grabs.

The only currently-viable megaconstellation is Starlink. It uses a single launch provider.  OneWeb has used multiple providers, but it has also gone bankrupt and gone through other financial contortions, so I feel its viability is questionable. It's not clear to me how many megaconstellations will actually launch. The largest of the "planned" constellations appears to be another Greg Wyler special.

These constellations really need more RLV launch options. Without at least two RLVs customers aren't going to get competitive pricing. F9R is still sold at its ELV price because it doesn't have any direct competition.

OK, upcoming LVs that might compete on $/kg with F9 appear to be Starship, New Glenn, and Terran-R (this thread!) with projected first launches in 2024, 2024, and 2026. Anyone else?  How close to the current F9R price will they need to be? How low can SpaceX drop the F9R price and still be profitable?  The F9R business model seems to depend on using Starlink to fill the manifest with high-volume deferrable launches. Will the competitors need an equivalent customer? These are not intended to be snarky or rhetorical. I'm genuinely interested in your responses, especially as they relate to Terran-R.

[Robotbeat]

Neutron or Firefly MLV could as well compete with F9R $/kg; it all comes down to flightrate for the most part.

[Asteroza]

I vaguely remember from some other thread the estimated marginal cost of F9R is $15 million? That's probably the price floor for semi-reusable rockets, even if SpaceX itself won't go there.

[xyv]

Assuming $15M is the marginal cost (and I have no reason to agree or disagree with that) that is the equivalent of "COGS" - cost of goods sold.  That is the incremental cost to make another widget and the price has to cover everything else.  A decent margin product can be priced at 50 points - 2X COGS.  When your back is against the wall you can accept lower margins but saying that $15M is the price floor is nowhere near the case.

[DanClemmensen]

Assuming $15M is the marginal cost (and I have no reason to agree or disagree with that) that is the equivalent of "COGS" - cost of goods sold.  That is the incremental cost to make another widget and the price has to cover everything else.  A decent margin product can be priced at 50 points - 2X COGS.  When your back is against the wall you can accept lower margins but saying that $15M is the price floor is nowhere near the case.

The prior post explicitly said "marginal cost", not "price". As you quite properly infer, we should all be careful to use the same metric when making comparisons. If SpaceX can hit $15 M marginal cost, then other companies will need to hit a number close to this to get the same margins if competing on price. The challenge for the competition is that SpaceX is now a big company with a major revenue stream, they have economies of scale, and F9 is a mature product. This means they can afford to reduce their margins.

[Robotbeat]

“Haha, gotcha! Oh, wait, you already said cost, not price.” —people who usually don’t acknowledge that they just didn’t read very closely.

[Paul451]

“Haha, gotcha! Oh, wait, you already said cost, not price.” —people who usually don’t acknowledge that they just didn’t read very closely.

Pot/kettle. If you had read what xyv actually wrote, they were saying that marginal cost is not "the price floor for semi-reusable rockets", as Asteroza claimed.

Marginal cost is the simple additional cost of production of a unit, without factoring in other costs. Hence the gap between that $15m figure and the launch price therefore isn't SpaceX's actual profit-margin. The figure of merit is "average total cost", aka "unit cost".

[Robotbeat]

It is the price floor. If they sell more even at just above that marginal price, they make more money. Otherwise it’s not the actual marginal price.

[Paul451]

It is the price floor. If they sell more even at just above that marginal price, they make more money. Otherwise it’s not the actual marginal price.

Again, that's not what "marginal cost" refers to. That's "unit cost". They are different things.

Which was the point xyv was trying to make, which you didn't read properly. Hence my amusement at your snarky comment.

[Robotbeat]

Yes, it is what marginal cost means. It means how much an additional launch costs. Unit costs includes fixed costs and is a function of number of launches even without any learning effects. You can still make more money (or lose less money) by selling ADDITIONAL launches above the marginal cost.


It’s kind of silly to use unit costs alone because for a launch company with one launch per year, the unit cost is equal to all the fixed costs plus marginal costs for one rocket. If your fixed costs are $250 million and marginal is $15 million, your unit costs are $265 million. No one will buy at that *price*, and by offering services there, your company will die. If you’re launching 144 times per year, your unit cost is just $16.7 million, however.

So every starting commercial launch company has to look at marginal costs plus addressable market size. Offering at unit costs WHEN YOURE STARTING is a quick way to never get any business.

Therefore marginal cost is a much more appropriate figure for price floor. If you argue “marginal cost + (fixed costs divided by total possible addressable market size in launches) = unit cost”, then sure, unit costs would work there, too, but that’s a very uncommon definition of unit costs and would anyway be very close to marginal cost. In the case with growing launch companies and a growing launch market, marginal cost is a better proxy for the price floor than what is typically meant by unit cost.

(And marginal cost includes things like additional manufacturing capacity if that’s required for another launch.)

[Paul451]

Chris! Please listen to my words.

I have no interest arguing business economics with you.

You made a snarky, insulting comment at xyv's expense, based on YOU misreading & misunderstanding THEIR comment.

You owe THEM an apology. Or at least an acknowledgement.

That is all I am saying.

(The irony of you accusing THEM of doing what YOU did, and continue to do, is just personally amusing.)

[TrevorMonty]

Thread is Relativity General Discussion. Nothing in title about F9R costs. If you want to discuss Terran R costs vs F9R, use or create a different thread.

[FutureSpaceTourist]

https://twitter.com/relativityspace/status/1745136045732344233

📚We wrote a book! Add it to this month’s book club: “Good Luck, Have Fun: The Journey to Launch the First 3D Printed Rocket.” A story paying homage to the first chapter of our story: Launching Terran 1.

✔️Highs.
✔️Lows.
✔️Everything in between from 2016 to our 2023 space journey.

Join us on memory lane as we reflect on our first 7 years — from an idea first scribbled on a receipt to a team that's pushing the bounds of what is possible in aerospace.

💫https://store.relativityspace.com/products/good-luck-have-fun-the-book

Sneak peek #1. 👀 Aeon 1 x 9

Sneak peek #2. 👀 Blue. Flame. Rocket. Club.

Thank you, #Terran1, for giving us the muscle, grit, and perseverance to enter the next chapter in our story: Launching #TerranR. 🫡

From store.relativityspace.com

P.S. 👀
*Only 500 copies available for purchase
*Available in the U.S. only

[Asteroza]

ITAR restricted book?

[lightleviathan]

Looks like an awesome book!
I hope I'll be able to snag a copy before they're all sold out haha