[bumby]

98% wasn't a number I pulled out of thin air. It's hard to know exact figures, but there are some people who publicly track what they can. [1]

From that source, SpaceX has about ~$5.515B in contracts. Of that, about $5.411B is from government contracts (military, NASA). That's about 98.1%.

There's probably a good chance it's higher, if they have govt contracts that aren't allowed to be public knowledge for security reasons.

Also, nobody is claiming SpaceX receives "money for nothing" from the govt. They are a contractor who, at least at these relatively early stages, need government contracts to survive. It's not necessarily a bad thing, and it's the way many nascent industries survive. But we need to call a spade a spade.

[1] https://docs.google.com/spreadsheets/d/1ld7Dz__7_VjqMd2uZNgA...

[Meegul]

That source only lists contracts where the price is known or can be reasonably estimated. Naturally, US government entities which have a semi-public bidding process are much easier to obtain a price estimate from. There's a large amount of foreign government contracts and private commercial contracts that are omitted from that list. Just from 2020 alone, SkySats 16-21, ANASIS-II, SXRS-1, SAOCOM 1B, GNOMES 1, Tybak-0172, SXM-7, and NROL-108.

Edit: I did unreasonably make the assumption that you were making the case the SpaceX was receiving subisidies for "nothing" and I see that you weren't now.

[bumby]

Hell, SpaceX is private so we have limited information across all domains. I acknowledged in my first post that accurate numbers are hard to come by. But we have to work with what information we have.

Foreign govts would still be govt money; many of those you listed are still public projects. Do you have estimates for those other private launches? If not, it's just speculation. Based on the information we have, it seems like it's safe to say the bulk of their launches are for governments. It's not a knock on them, and I'm not saying this is your perspective, but there's this kinda weird sentiment that they are some paragon of free-market capitalism when in fact they are a company highly dependent on government money. But that's exactly what I think is reasonable to expect in a nascent industry.

[s1artibartfast]

As a side note, There are easy esimates for private launches. SpaceX widely publicises their launch pricing, which lines up pretty well for the publicly disclosed contracts.

[panick21_]

> But we have to work with what information we have.

No sorry, if you know the numbers you have are clearly bias and wrong then you don't need to work with them.

[bumby]

Do you know SpaceX's cash flow? If so, please share. Otherwise, I can just claim you are speculating towards your own bias.

If you think the data is bad, at best all you can claim is "we don't know"

[vernon99]

And this is imo what you should have done instead if saying 98%. I agree with other commenters in that this is extremely misleading.

[bumby]

That’s fine. I’m completely ok with someone saying “I don’t know”. But the original post was trying to make a wild claim with zero evidence. I’m at least trying to ground my claim in some actual data while clearly acknowledging the limits of that data. I’m also completely fine with the veracity of that data being brought into question. What I’ve yet to see is anyone trying to refute it being an any additional data to the discussion that is of better quality. Losing Wikipedia launches (with little understanding of the business context and no actual revenue claims) is even more specious than I was using. It’s not a more informed or more rational argument, it’s an emotional bias.

Put differently, why do you think there isn’t an abundance of people chiming in to say “there’s not enough data to make a claim” rather than jumping to defend SpaceX’s honor? If people were truly rational, the strength of their convictions would be proportional to the strength of the data being it.

[s1artibartfast]

The other respondent brought up a good point regarding selection bias. A different way to approach the problem would be to look at total SpaceX launches by customer. This has its own problems, but might paint a better picture.

When talking about funding in general, you also have to consider that SpaceX has significant non launch Equity funding to the tune of $10B, and that contracts can include potential future Revenue instead of funds received.

Edit: If you look at their launches for 2020 onwards[1], Their number 1 customer is SpaceX (themselves), followed by US and other governments, followed by private companies.

https://en.wikipedia.org/wiki/List_of_Falcon_9_and_Falcon_He...

While your borader conclusion is probably accurate regarding income, your spreadsheet only 71 of 216 Falcon 9 and Falcon Heavy launches, for ~$9B total. This is roughly on par with the VC funding they have recieved.

[bumby]

Sure, I can acknowledge there is likely a selection bias based on public reporting. But I would push back that it skews the data by an order of magnitude as suggested. We'd likely see many, many more private launches compared to public for that to be true. Let's be super generous and say that Starlink brings in $1B in revenue; that still means govt funding accounts for 80% of their revenue.

And, yes, we can talk about different ways of accounting. But I wouldn't count the fickle or volatile ones like "potential future revenue" as better than measuring actual contracts. (This is what SpaceX is investing in with their own launches, to be fair, but that's different than actual revenue). People trying to raise capital are always talking about potential revenue and potential market share, but you'll get much more investor money if you have actual purchase orders.

[s1artibartfast]

Like I said, I dont disagree with your broader point, and would agree that 80+% is a reasonable estimate.

$1B income for Starlink makes sense given they have ~1M customers paying ~$100/mo, and a $4-500 hookup cost.

Lasty, you have to consider that the government contract numbers you have are cummulative going back to 2012, while the Starlink revenue is annual (2022 only). If you look at government contracts for 2021, you only get $1.6B. Using those numbers, Starlink internet would be closer to 40% of revenue, with launch contracts at 60%, which starts to paint a very different story.

I think a lot of the feedback you are getting is a kneejerk reaction based on language. E.g. "government funding" had a very different connotation than "government sales". People think subsidies and then their minds short circut.

[bumby]

You're probably right; I could have phrased it better. I think we pretty much agree, given what data is available. My main point is that their business model is predicated government contracts early before they can transition to other income streams. That's not really crony capitalism and it's part of the intent of the government to spur new industries. But I would also hold it as distinct from pure free-market capitalism that can exist outside of government money. (Which, ironically, is a point many people hold against legacy aerospace). And I think that latter point is relevant because SpaceX (and Musk himself) is often held up as some sort of capitalistic ideal and any counterpoint to that makes people bristle. In all, however, I think that hybrid approach is ultimately beneficial to both SpaceX and the public.

[s1artibartfast]

I think we prettymuch agree to, but I have an obvious quibbling problem.

> I would also hold it as distinct from pure free-market capitalism that can exist outside of government money. (Which, ironically, is a point many people hold against legacy aerospace). And I think that latter point is relevant because SpaceX (and Musk himself) is often held up as some sort of capitalistic ideal and any counterpoint to that makes people bristle. In all, however, I think that hybrid approach is ultimately beneficial to both SpaceX and the public

I think a lot of people dont really get this nuanced distinction. Rockets and spaceships have always been never been built by the government. Most of the work was always done by publically traded companies. NASA, whomever, would put out a contract, and then be involved in high level diretion of the design process.

The difference with SpaceX is that 1) they not publically traded (who cares?) and 2) they dont take design direction from NASA for their rockets, and built what they wanted. Thats it!

The cool part is that they built somthing more ambitious buy cutting NASA out of the design process, and relied on the fact that it will be so good that NASA will want it when they saw it.

My quibble is that this is very different than the classic case of the government spuring new industires. The government was already buying rockets from established manufactures for a half centry. SpaceX just came in and ate other companies lucnhes by being more focused and more efficienct.

[bumby]

This isn’t new. It’s exactly how the airplane was invented over 100+ years ago.

The Army didn’t tell Orville and Wilbur how to build. They dangled the carrot of a lucrative military contract to incentivize their innovation. It’s what transitioned the effort from hobbyists to an actual industry. There was no “industry” until the govt put up money because the govt was really the only entity that could bear that kind of risk. It’s the same with the initial days of SpaceX.

[panick21_]

Sorry but this is complete nonsense.

You can literally just go threw the list of their launches on Wikipedia an notice that nothing close to even 80% of them are for the government.

And that is outside of their Starship business, that is by now a large part of its revenue.

The numbers you present are ridicoulus, they have easily more money just from human space flight from Polaris and Axiom.

[bumby]

The bulk of their own launches are for their own infrastructure. That doesn't bring revenue at the launch date. They aren't paying themselves to launch their equipment; it's an investment in future revenue. Even looking at optimistic projections for 2023 Starlink revenue puts would still mean govt contracts dwarf Starlink revenue by 4:1.

>And that is outside of their Starship business, that is by now a large part of its revenue.

Starship is bringing in investment money, not revenue that I'm aware of. I don't know what revenue it would be bringing without actual launches. Maybe I'm wrong and you can correct me here, but I think you are conflating some business aspects.

Now if you're saying that the private money will be a bigger factor in the future, I agree. That's the whole idea behind govt money being used to foster along nascent industries: the govt props them up early until a viable private enterprise can exist later.

[panick21_]

I meant to say Starlink rather then Starship. They have over a million subscribers now. Plus larger clients in shipping and air travel, plus the military.

Honestly I don't get your issue. You can just go to wikipedia and look at the list of the launches. So clearly the 98% is wrong, this isn't a question, its not speculation.

[bumby]

This doesn’t pass a simple sanity check.

The original claim was “US government can't be more than some low double digit percentage”.

Let’s say it’s by “low double digit percentage” we’re generous and say 20%. (I interpreted their stance to be lower, but they didn’t specify).

We can are there is incomplete data but we have reasonably good data on the the govt side: U.S. govt contacts are at least $5.4B, likely higher. So that means we expect the complementary private side to be 5x, or over $22B. The absolute most optimistic Starlink estimates about $2B through 2023. Do you think the other private streams dwarf Starlink by 10x? I just don’t see how we get there given the govt is still there #1 launch customer after Starlink, and this is the most generous interpretation to steelman the case. The only way is true is if there is some unknown customer launching nearly 3x that of the US govt.

The only way the numbers work out is if the govt is the major revenue stream or there’s some secret launch customer that dwarfs the US space agencies.

It’s a claim based on wishful feeling, not on data.

[bumby]

Fair enough, but it’s still a multiple smaller than their govt revenue.

[s1artibartfast]

>Even looking at optimistic projections for 2023 Starlink revenue puts would still mean govt contracts dwarf Starlink revenue by 4:1.

As I just point at elsewhere, you get 4:1 by comparing cummulative contract revenue with an annual Starlink revenue.

[bumby]

No disagreement; this actually is proving my point. I'm pretty clear about using the larger business cycle as my point. I'm not taking a snapshot in time, I'm saying relying on govt contracts early is necessary for their business to survive. Others are either twisting or misinterpreting my point.

To be absolutely explicitly clear: their business model requires using govt contracts/funding/sales early before they can transition to other viable means of revenue. They could not survive as a company without those early govt contracts, but this does not mean they are dependent on govt contracts into perpetuity.

A dependency on govt contracts early and an increasingly private revenue stream can both be true. This isn't bad, but it also isn't the often-lauded ideal of free-market capitalism.

[bryanlarsen]

Starlink revenues are already over $1B per year, and they're just getting started.

[bumby]

That's based on forecasts and not real numbers (because SpaceX is a private company), but it's a good point.

[bryanlarsen]

They've announced that they had 1 million users in December. It's likely that ARPU is over $1000/user.

[bumby]

Assuming a 100% renewal rate

[bryanlarsen]

Assuming they are adding faster than losing.

[bumby]

I think we can agree that starlink is growing and will continue to be a bigger share of their revenue. That doesn’t negate my point, though.