Impact of SpaceX rideshare on small sat launchers market

[john smith 19]

I do agree you could in theory dog food it if with your own megaconstellation but then you have to pay for that too and compete against other megaconstellations that aren’t hobbled by also developing their own launch vehicle. I don’t see how anyone gets funding for that when there are already multiple deep pocketed players in the space.

Note the hidden assumption in all this discussion.

That the business model is always a sole mfg/sole operator model.

Realise this is completely different to every other transportation method on the planet.

The market changes beyond recognition when you stop doing spacelaunch-as-a-service and turn it into spacelaunch-as-a-product.

But for that to happen you need to get away from a launch tower that Arianspace described as "The Eiffel Tower on wheels" IE 27 000 tonnes of steel 

So if you do what you've always done you should'nt be surprised you get what you've always gotten.

[M.E.T.]

I do agree you could in theory dog food it if with your own megaconstellation but then you have to pay for that too and compete against other megaconstellations that aren’t hobbled by also developing their own launch vehicle. I don’t see how anyone gets funding for that when there are already multiple deep pocketed players in the space.

Note the hidden assumption in all this discussion.

That the business model is always a sole mfg/sole operator model.

Realise this is completely different to every other transportation method on the planet.

The market changes beyond recognition when you stop doing spacelaunch-as-a-service and turn it into spacelaunch-as-a-product.

But for that to happen you need to get away from a launch tower that Arianspace described as "The Eiffel Tower on wheels" IE 27 000 tonnes of steel 

So if you do what you've always done you should'nt be surprised you get what you've always gotten.

I don’t understand your meaning. Are you saying the launch service needs to be split into its constituent components - with one company providing the launch facilities and another providing the launch vehicle?

No, I don’t think I’m understanding you correctly at all.

[c4fusion]

I do agree you could in theory dog food it if with your own megaconstellation but then you have to pay for that too and compete against other megaconstellations that aren’t hobbled by also developing their own launch vehicle. I don’t see how anyone gets funding for that when there are already multiple deep pocketed players in the space.

Note the hidden assumption in all this discussion.

That the business model is always a sole mfg/sole operator model.

Realise this is completely different to every other transportation method on the planet.

The market changes beyond recognition when you stop doing spacelaunch-as-a-service and turn it into spacelaunch-as-a-product.

But for that to happen you need to get away from a launch tower that Arianspace described as "The Eiffel Tower on wheels" IE 27 000 tonnes of steel 

So if you do what you've always done you should'nt be surprised you get what you've always gotten.

I don’t understand your meaning. Are you saying the launch service needs to be split into its constituent components - with one company providing the launch facilities and another providing the launch vehicle?

No, I don’t think I’m understanding you correctly at all.

I think you are understanding it correctly, I believe that john is talking about how there are a bunch or airlines in the world while there are effectively 4 manufactures (maybe 3 more coming from China, Japan, and Russia) in the world including heavily subsidized ones.

[imprezive]

I do agree you could in theory dog food it if with your own megaconstellation but then you have to pay for that too and compete against other megaconstellations that aren’t hobbled by also developing their own launch vehicle. I don’t see how anyone gets funding for that when there are already multiple deep pocketed players in the space.

Note the hidden assumption in all this discussion.

That the business model is always a sole mfg/sole operator model.

Realise this is completely different to every other transportation method on the planet.

The market changes beyond recognition when you stop doing spacelaunch-as-a-service and turn it into spacelaunch-as-a-product.

But for that to happen you need to get away from a launch tower that Arianspace described as "The Eiffel Tower on wheels" IE 27 000 tonnes of steel 

So if you do what you've always done you should'nt be surprised you get what you've always gotten.

I don’t understand your meaning. Are you saying the launch service needs to be split into its constituent components - with one company providing the launch facilities and another providing the launch vehicle?

No, I don’t think I’m understanding you correctly at all.

I think you are understanding it correctly, I believe that john is talking about how there are a bunch or airlines in the world while there are effectively 4 manufactures (maybe 3 more coming from China, Japan, and Russia) in the world including heavily subsidized ones.

I agree that’s what he’s saying but I don’t follow the logic. Adding more middle men into the supply chain does what? Somehow gets a lot more satellites built?

[john smith 19]

SpaceX is under no compulsion to offer prices that are just above cost. SpaceX will make profit where they can. So don't expect prices to go that low without competition.

Funny how so many people don't seem to understand that fairly simple idea. 

[john smith 19]

I don’t understand your meaning. Are you saying the launch service needs to be split into its constituent components - with one company providing the launch facilities and another providing the launch vehicle?

No, I don’t think I’m understanding you correctly at all.

Actually that's exactly what happens with every other tranpsortation system.

Ships have docks, airliners have airports.

Ship builders rarely own shipping docks. Airliner mfgs don't run airports.

Their finances are not coupled to how much business in moving goods they can attract, but in how many ships or planes they can sell to operators.

But while launch facilities are so tightly coupled to the vehicles they launch, and most of those are expendable anyway that will never happen in space launch and the business will remain (relatively) tiny compared to air transport or shipping.

And SS won't change that. 

[john smith 19]

I agree that’s what he’s saying but I don’t follow the logic. Adding more middle men into the supply chain does what? Somehow gets a lot more satellites built?

It's called competition and it's the only truly effective way to lower prices to customers in any market.

[M.E.T.]

I don’t understand your meaning. Are you saying the launch service needs to be split into its constituent components - with one company providing the launch facilities and another providing the launch vehicle?

No, I don’t think I’m understanding you correctly at all.

Actually that's exactly what happens with every other tranpsortation system.

Ships have docks, airliners have airports.

Ship builders rarely own shipping docks. Airliner mfgs don't run airports.

Their finances are not coupled to how much business in moving goods they can attract, but in how many ships or planes they can sell to operators.

But while launch facilities are so tightly coupled to the vehicles they launch, and most of those are expendable anyway that will never happen in space launch and the business will remain (relatively) tiny compared to air transport or shipping.

And SS won't change that. 

For that you need launch vehicles to be as generic as airliners. Basically SpaceX should build Starships for everyone like Boeing builds 747’s. And Blue Origin needs to build New Glenns like Airbus does A380’s. And the two rockets need to be as mutually compatible as airliners are when it comes to fuelling, loading, boarding and landing. Else generic launch facilities simply won’t work like airports do.

That’s not a realistic prospect until we have Expanse-type Epstein drives or equivalent.

[edzieba]

https://twitter.com/sciguyspace/status/1628801563824103425?s=46&t=NUvXH8X_iuI_twjBjZBghA

This revenue breakdown illustrates my point quite clearly.

In 2022, even with total market dominance and a record cadence, SpaceX launch revenue was only $2.4B. Across all launch categories - Small, Medium, Heavy and Crewed.

Let’s say you can double that if you count Starlink launches as revenue generating. So call it $5B.

That’s basically the pot of gold everyone is fighting over. And no newcomer is going to get all of it, or even most of it, as SpaceX will continue to claim the bulk of it, even with new competition emerging.

And that’s before they start dropping prices to hold on to as much of it as possible.

It’s a fool’s errand for the new arrivals.

Notice those numbers are pure (estimated) gross revenue. They are not net profit, i.e. do not include cost of doing business or any past investment - despite pretty much every post in this thread slagging small launch revolving around the notion of 'paying back' investment (as if that was how investment worked).

[M.E.T.]

https://twitter.com/sciguyspace/status/1628801563824103425?s=46&t=NUvXH8X_iuI_twjBjZBghA

This revenue breakdown illustrates my point quite clearly.

In 2022, even with total market dominance and a record cadence, SpaceX launch revenue was only $2.4B. Across all launch categories - Small, Medium, Heavy and Crewed.

Let’s say you can double that if you count Starlink launches as revenue generating. So call it $5B.

That’s basically the pot of gold everyone is fighting over. And no newcomer is going to get all of it, or even most of it, as SpaceX will continue to claim the bulk of it, even with new competition emerging.

And that’s before they start dropping prices to hold on to as much of it as possible.

It’s a fool’s errand for the new arrivals.

Notice those numbers are pure (estimated) gross revenue. They are not net profit, i.e. do not include cost of doing business or any past investment - despite pretty much every post in this thread slagging small launch revolving around the notion of 'paying back' investment (as if that was how investment worked).

Surely that makes it even worse for your argument? Profit will be far less than the revenue. So if a small launcher can access say $1B of the revenue pie, that still needs all costs deducted from it before a business case can be justified.

[JayWee]

I don’t understand your meaning. Are you saying the launch service needs to be split into its constituent components - with one company providing the launch facilities and another providing the launch vehicle?

No, I don’t think I’m understanding you correctly at all.

Actually that's exactly what happens with every other tranpsortation system.

Ships have docks, airliners have airports.

Ship builders rarely own shipping docks. Airliner mfgs don't run airports.

Boeing did work like that in the beginning, until anti-trust broke it apart.

https://en.wikipedia.org/wiki/United_Aircraft_and_Transport_Corporation

[trimeta]

The only argument they could have is if they’re using it to enable their own megaconstellation without helping fund their competitor. Also, if their tech is potentially truly transformative like Stoke’s. Or if they’re just content over lower returns plus enabling humanity to become more spacefaring.

So to categorise:

1. Develop launch as a support service for your own constellation - that’s basically Blue Origin (if you fudge the lines between Blue and Amazon). But at much higher cost to Amazon than if they went with SpaceX. And Blue is not going for dedicated small launch, so this doesn’t support the small launch business case.

2. New tech which is a category all on its own, and yes, better tech than SpaceX is indeed the one use case that I agree with. A big assumption to base one’s business case on, though.

3. Atruism rather than profit as a motivation. In that case business cases become irrelevant - as long as continued subsidization keeps flowing in from somewhere.

For what it's worth, I'd put Rocket Lab into your first category here. They don't need to launch profitably, because launching regularly (even if it's barely breakeven on a per-launch basis) gives them free launches to test out their own spacecraft components. So it's not exactly "a support service for their own constellation," but is instead a support service for their components business.

Still, this is consistent with your idea that the launch market overall is small, and whatever piece of it you think you can take from SpaceX isn't worth the investment necessary to do so.

[Robotbeat]

Much cheaper per-launch cost than Starship. Potentially a competitive per-kg cost, too. Which means SpaceX will need to lower the price on Starship. Both of those are pretty awesome for the customer.

SpaceX is under no compulsion to offer prices that are just above cost. SpaceX will make profit where they can. So don't expect prices to go that low without competition.

SpaceX isn’t the only game in town. Electron, LauncherOne, Vega, SSLV and PSLV all exist a lower per launch cost for smaller payloads and you’ve got Firefly, ABL, and Relativity coming soon-ish. Even if Stoke gets their cost down crazy low to let’s say $1M, they will need to cover the costs of running the company too. Even running lean but at a high cadence that’s probably $10M/mo. That’s 10 launches a month … just to break even. That was effectively Astras plan and we see how that worked out.

Stoke's design could get to below $300k per launch, their capacity is about a factor of 10x per launch than Astra, and Astra's plan relied on magic cost reductions of aerospace hardware, not on reuse (which doesn't require lower hardware cost, just recovering the rocket and having low refurb costs).

The approach is good for hundreds per month. And with megaconstellations, that's feasible.

Reuse is a near step-function increase in capacity. You can see that with Falcon going from like a dozen launches per year to now over 60 (and possibly up to 100 this year) just with first stage reuse (and fairing). With equally-rapid upper stage reuse, you could be doing thousands of flights per year.

The dumb idea was to try to do massive constellation launches with expendable smallsat launchers, literally the most expensive possible approach. Stoke is neither expendable nor a smallsat launcher (it's a medium lift launcher) and they're going with an approach that could allow an edge over SpaceX's approach (a more reusable and robust actively-cooled metallic heatshield which should require less turnaround time and maintenance than the Shuttle-like Starship tile system), so the same logic does not apply.

[imprezive]

The only argument they could have is if they’re using it to enable their own megaconstellation without helping fund their competitor. Also, if their tech is potentially truly transformative like Stoke’s. Or if they’re just content over lower returns plus enabling humanity to become more spacefaring.

So to categorise:

1. Develop launch as a support service for your own constellation - that’s basically Blue Origin (if you fudge the lines between Blue and Amazon). But at much higher cost to Amazon than if they went with SpaceX. And Blue is not going for dedicated small launch, so this doesn’t support the small launch business case.

2. New tech which is a category all on its own, and yes, better tech than SpaceX is indeed the one use case that I agree with. A big assumption to base one’s business case on, though.

3. Atruism rather than profit as a motivation. In that case business cases become irrelevant - as long as continued subsidization keeps flowing in from somewhere.

For what it's worth, I'd put Rocket Lab into your first category here. They don't need to launch profitably, because launching regularly (even if it's barely breakeven on a per-launch basis) gives them free launches to test out their own spacecraft components. So it's not exactly "a support service for their own constellation," but is instead a support service for their components business.

Still, this is consistent with your idea that the launch market overall is small, and whatever piece of it you think you can take from SpaceX isn't worth the investment necessary to do so.

I agree and I think people are too quick to not factor in Rocket Lab in these conversations because the question is really is there room for two small launchers in the US.

Much cheaper per-launch cost than Starship. Potentially a competitive per-kg cost, too. Which means SpaceX will need to lower the price on Starship. Both of those are pretty awesome for the customer.

SpaceX is under no compulsion to offer prices that are just above cost. SpaceX will make profit where they can. So don't expect prices to go that low without competition.

SpaceX isn’t the only game in town. Electron, LauncherOne, Vega, SSLV and PSLV all exist a lower per launch cost for smaller payloads and you’ve got Firefly, ABL, and Relativity coming soon-ish. Even if Stoke gets their cost down crazy low to let’s say $1M, they will need to cover the costs of running the company too. Even running lean but at a high cadence that’s probably $10M/mo. That’s 10 launches a month … just to break even. That was effectively Astras plan and we see how that worked out.

Stoke's design could get to below $300k per launch, their capacity is about a factor of 10x per launch than Astra, and Astra's plan relied on magic cost reductions of aerospace hardware, not on reuse (which doesn't require lower hardware cost, just recovering the rocket and having low refurb costs).

The approach is good for hundreds per month. And with megaconstellations, that's feasible.

Reuse is a near step-function increase in capacity. You can see that with Falcon going from like a dozen launches per year to now over 60 (and possibly up to 100 this year) just with first stage reuse (and fairing). With equally-rapid upper stage reuse, you could be doing thousands of flights per year.

The dumb idea was to try to do massive constellation launches with expendable smallsat launchers, literally the most expensive possible approach. Stoke is neither expendable nor a smallsat launcher (it's a medium lift launcher) and they're going with an approach that could allow an edge over SpaceX's approach (a more reusable and robust actively-cooled metallic heatshield which should require less turnaround time and maintenance than the Shuttle-like Starship tile system), so the same logic does not apply.

As an engineer I love the idea and want to see it fly. However $300k at thousands of flights per year is crazy unrealistic. The launch infrastructure required and impact to air traffic would be insane. It’s also not clear that’s better for a megaconstellation that can do a plane per launch in a larger vehicle. And aside from that who would that constellation be? Starlink, Kuiper, and OneWeb all seem have made their bets already.

[Dmitry_V_home]

https://twitter.com/sciguyspace/status/1628801563824103425?s=46&t=NUvXH8X_iuI_twjBjZBghA

This revenue breakdown illustrates my point quite clearly.

In 2022, even with total market dominance and a record cadence, SpaceX launch revenue was only $2.4B. Across all launch categories - Small, Medium, Heavy and Crewed.

Let’s say you can double that if you count Starlink launches as revenue generating. So call it $5B.

That’s basically the pot of gold everyone is fighting over. And no newcomer is going to get all of it, or even most of it, as SpaceX will continue to claim the bulk of it, even with new competition emerging.

And that’s before they start dropping prices to hold on to as much of it as possible.

It’s a fool’s errand for the new arrivals.

In my opinion, SpeceX has moved from the concept of selling launches to selling services to the end consumer. In this case, the launch vehicle and satellite are just a tool, not a commodity.

[Robotbeat]

The only argument they could have is if they’re using it to enable their own megaconstellation without helping fund their competitor. Also, if their tech is potentially truly transformative like Stoke’s. Or if they’re just content over lower returns plus enabling humanity to become more spacefaring.

So to categorise:

1. Develop launch as a support service for your own constellation - that’s basically Blue Origin (if you fudge the lines between Blue and Amazon). But at much higher cost to Amazon than if they went with SpaceX. And Blue is not going for dedicated small launch, so this doesn’t support the small launch business case.

2. New tech which is a category all on its own, and yes, better tech than SpaceX is indeed the one use case that I agree with. A big assumption to base one’s business case on, though.

3. Atruism rather than profit as a motivation. In that case business cases become irrelevant - as long as continued subsidization keeps flowing in from somewhere.

For what it's worth, I'd put Rocket Lab into your first category here. They don't need to launch profitably, because launching regularly (even if it's barely breakeven on a per-launch basis) gives them free launches to test out their own spacecraft components. So it's not exactly "a support service for their own constellation," but is instead a support service for their components business.

Still, this is consistent with your idea that the launch market overall is small, and whatever piece of it you think you can take from SpaceX isn't worth the investment necessary to do so.

I agree and I think people are too quick to not factor in Rocket Lab in these conversations because the question is really is there room for two small launchers in the US.

Much cheaper per-launch cost than Starship. Potentially a competitive per-kg cost, too. Which means SpaceX will need to lower the price on Starship. Both of those are pretty awesome for the customer.

SpaceX is under no compulsion to offer prices that are just above cost. SpaceX will make profit where they can. So don't expect prices to go that low without competition.

SpaceX isn’t the only game in town. Electron, LauncherOne, Vega, SSLV and PSLV all exist a lower per launch cost for smaller payloads and you’ve got Firefly, ABL, and Relativity coming soon-ish. Even if Stoke gets their cost down crazy low to let’s say $1M, they will need to cover the costs of running the company too. Even running lean but at a high cadence that’s probably $10M/mo. That’s 10 launches a month … just to break even. That was effectively Astras plan and we see how that worked out.

Stoke's design could get to below $300k per launch, their capacity is about a factor of 10x per launch than Astra, and Astra's plan relied on magic cost reductions of aerospace hardware, not on reuse (which doesn't require lower hardware cost, just recovering the rocket and having low refurb costs).

The approach is good for hundreds per month. And with megaconstellations, that's feasible.

Reuse is a near step-function increase in capacity. You can see that with Falcon going from like a dozen launches per year to now over 60 (and possibly up to 100 this year) just with first stage reuse (and fairing). With equally-rapid upper stage reuse, you could be doing thousands of flights per year.

The dumb idea was to try to do massive constellation launches with expendable smallsat launchers, literally the most expensive possible approach. Stoke is neither expendable nor a smallsat launcher (it's a medium lift launcher) and they're going with an approach that could allow an edge over SpaceX's approach (a more reusable and robust actively-cooled metallic heatshield which should require less turnaround time and maintenance than the Shuttle-like Starship tile system), so the same logic does not apply.

As an engineer I love the idea and want to see it fly. However $300k at thousands of flights per year is crazy unrealistic. The launch infrastructure required and impact to air traffic would be insane. It’s also not clear that’s better for a megaconstellation that can do a plane per launch in a larger vehicle. And aside from that who would that constellation be? Starlink, Kuiper, and OneWeb all seem have made their bets already.

No crazier than Starship, and in answer to your second question: Kuiper or anyone else who wants a megaconstellation to compete with Starlink. Kuiper has secured launches on other rockets, but that just means they’re not resting on their laurels, not that they wouldn’t consider other launch vehicles for later iterations. After all, even OneWeb is launching on Falcon 9.

[chopsticks]

As an engineer I love the idea and want to see it fly. However $300k at thousands of flights per year is crazy unrealistic.

Likely is unrealistic, but not any more so than what some of the diehard Starship fanatics on here say what that vehicle will surely do.

[Robotbeat]

As an engineer I love the idea and want to see it fly. However $300k at thousands of flights per year is crazy unrealistic.

Likely is unrealistic, but not any more so than what some of the diehard Starship fanatics on here say what that vehicle will surely do.

Exactly. And I’m one of those. But there are no first principles arguments saying it can’t be done. Stoke has a big challenge with paperwork per launch, tho. Something I’d like to see the FAA take seriously so that we don’t have to rely on just really big launchers to get costs down.

As far as what is realistic… I thought SpaceX would do just 37 launches last year, but Falcon’s reuse has enabled massive launch rates that would be unrealistic except with insane resources with expendables.

Effective reuse just *demonstrably* changes the whole game and you can’t rely on normal experience to judge what is realistic once it’s deployed.

[imprezive]

The only argument they could have is if they’re using it to enable their own megaconstellation without helping fund their competitor. Also, if their tech is potentially truly transformative like Stoke’s. Or if they’re just content over lower returns plus enabling humanity to become more spacefaring.

So to categorise:

1. Develop launch as a support service for your own constellation - that’s basically Blue Origin (if you fudge the lines between Blue and Amazon). But at much higher cost to Amazon than if they went with SpaceX. And Blue is not going for dedicated small launch, so this doesn’t support the small launch business case.

2. New tech which is a category all on its own, and yes, better tech than SpaceX is indeed the one use case that I agree with. A big assumption to base one’s business case on, though.

3. Atruism rather than profit as a motivation. In that case business cases become irrelevant - as long as continued subsidization keeps flowing in from somewhere.

For what it's worth, I'd put Rocket Lab into your first category here. They don't need to launch profitably, because launching regularly (even if it's barely breakeven on a per-launch basis) gives them free launches to test out their own spacecraft components. So it's not exactly "a support service for their own constellation," but is instead a support service for their components business.

Still, this is consistent with your idea that the launch market overall is small, and whatever piece of it you think you can take from SpaceX isn't worth the investment necessary to do so.

I agree and I think people are too quick to not factor in Rocket Lab in these conversations because the question is really is there room for two small launchers in the US.

Much cheaper per-launch cost than Starship. Potentially a competitive per-kg cost, too. Which means SpaceX will need to lower the price on Starship. Both of those are pretty awesome for the customer.

SpaceX is under no compulsion to offer prices that are just above cost. SpaceX will make profit where they can. So don't expect prices to go that low without competition.

SpaceX isn’t the only game in town. Electron, LauncherOne, Vega, SSLV and PSLV all exist a lower per launch cost for smaller payloads and you’ve got Firefly, ABL, and Relativity coming soon-ish. Even if Stoke gets their cost down crazy low to let’s say $1M, they will need to cover the costs of running the company too. Even running lean but at a high cadence that’s probably $10M/mo. That’s 10 launches a month … just to break even. That was effectively Astras plan and we see how that worked out.

Stoke's design could get to below $300k per launch, their capacity is about a factor of 10x per launch than Astra, and Astra's plan relied on magic cost reductions of aerospace hardware, not on reuse (which doesn't require lower hardware cost, just recovering the rocket and having low refurb costs).

The approach is good for hundreds per month. And with megaconstellations, that's feasible.

Reuse is a near step-function increase in capacity. You can see that with Falcon going from like a dozen launches per year to now over 60 (and possibly up to 100 this year) just with first stage reuse (and fairing). With equally-rapid upper stage reuse, you could be doing thousands of flights per year.

The dumb idea was to try to do massive constellation launches with expendable smallsat launchers, literally the most expensive possible approach. Stoke is neither expendable nor a smallsat launcher (it's a medium lift launcher) and they're going with an approach that could allow an edge over SpaceX's approach (a more reusable and robust actively-cooled metallic heatshield which should require less turnaround time and maintenance than the Shuttle-like Starship tile system), so the same logic does not apply.

As an engineer I love the idea and want to see it fly. However $300k at thousands of flights per year is crazy unrealistic. The launch infrastructure required and impact to air traffic would be insane. It’s also not clear that’s better for a megaconstellation that can do a plane per launch in a larger vehicle. And aside from that who would that constellation be? Starlink, Kuiper, and OneWeb all seem have made their bets already.

No crazier than Starship, and in answer to your second question: Kuiper or anyone else who wants a megaconstellation to compete with Starlink. Kuiper has secured launches on other rockets, but that just means they’re not resting on their laurels, not that they wouldn’t consider other launch vehicles for later iterations. After all, even OneWeb is launching on Falcon 9.

I think you’re missing the forest for the trees here. The crazy isn’t the technology it’s going into a relatively small and fixed market against 5+ US competitors with billion+ dollar deep pockets.

[Robotbeat]

Competitors that intentionally crash at least part of their rockets at the end of each flight.

Imagine being an airliner startup competing with single-use airliners. Doesn’t matter how much money they have… there’s literally not enough money in the world to crash 20 million airliners each year, which is the number of annual commercial airline flights (the most common airliner is the 737 and it costs $100 million apiece… $2 Quadrillion. That’s 20 times the global annual GDP of $100 trillion). Reusable lets you compete in a way nothing else does.

And since the rocket is modestly sized and you should get it back after each flight, you shouldn’t necessarily need a ton of capital to make it work. Or even necessarily a lot of people.