SpaceX and the Market for Launch

I was having a conversation with an Air Force guy who told me inside DOD launch is considered a solved problem because of the huge improvements in cost introduced by SpaceX through the Falcon9. This sentiment also anticipates the entrance of Super Heavy into the market, driving future improvements. This has been the conventional wisdom across the sector, not just in DOD, for about five years, give or take. You’ll frequently hear investors quote the figure of 150-ish launch companies and the inevitability of a lot of those guys getting gobbled up or simply disappearing.

The thrust being: “Don't do launch. It’s saturated. SpaceX is going to eat your lunch.”This view is both true and false. The high-level advice that comes out is wrong because of a lack of some important nuance. A more accurate version would be: “Don’t try and replicate what SpaceX did 15 years ago, or they are going to eat your lunch”. A lot of rocket companies founded in the wake of SpaceX’s success around 2012-2015 were variations on the theme of “We’re like SpaceX except for XYZ”—where XYZ might be a toroidal aerospike nozzle, electrical fuel pumps, a carbon fiber fuselage, 3D printed engines etc.

'Copycats + some variation' emerging is totally normal in the wake of a big new thing and isn't itself bad. Likely a lot of these are great ideas that could result in better rockets. However, they all fail to recognize just how big a lead SpaceX has and that it was mostly non-technical factors that drove its success. The SpaceX we know today was built by a ruthlessly lean development and management culture, lots manufacturing innovations, political/contracting savvy, a lot of cash, and tons of hard work. Their starting products, Falcon1 and early Falcon9’s, were also conservative projects that contained no new technology. And don't forget that steward of all success, luck. Yet, with all that behind them, SpaceX still came within a hair of failure and had to be actively bailed out by the Air Force and NASA between the Falcon1 and Falcon9. But when you’re a hammer, everything looks like a nail. Smart rocket engineers, emboldened by what they were seeing, felt they could do better. They thought the solution to competing with a rocket company was to build a better rocket. And they have PhDs and years of experience building space stuff.

Investors, perpetually in a rush to follow someone else's success, thought these guys must know what they were talking about and gave them a lot of money. Hence the glut of ‘SpaceX clones’ in the mid-2010s, some of which are still chugging along at the time of this writing. Rocket Lab, Astra, Relativity, and loads of others at various stages of development are in this basket. To the degree they survive as launch providers, it will be because the DOD and NASA don't want SpaceX to be the only provider in the market. They’ll fight for the contracts the U.S. government throw out to keep them afloat while SpaceX dominates the growing commercial market. It’s a zombie existence and not a fate to be envied. A SpaceX competitor needs to have something WAY stronger than a variation on the theme that barely worked for SpaceX over a decade ago. New entrants are competing with one of the best capitalized and most dynamic hardware companies in the world.

SpaceX had to beat companies that were hobbled by an addiction to cost-plus contracting and who had grown technically stagnant. So, fundamentally, we can agree that a lot of these ~150 companies are doomed. However, I disagree with underlying assertions that “the status quo is good” and “there is no way to compete with SpaceX”. We are nowhere near facilitating a human future in space. The Falcon9 has halved the historic cost for mass to orbit compared to peers in the US. It’s still better than the Russians or Indians though that’s a much closer thing. Price per kg/orbit can be sliced a few different ways and what you get quoted from SpaceX varies a lot depending on what you’re asking for. Government customers tend to pay more; rideshare customers tend to pay less. $5000/kg is normal though it might be as low as $2000/kg in some situations.

SpaceX is a lot more transparent about its pricing than the legacy players, but the devil’s in the details. Elon has tweeted that the fuel costs, and thus the eventual marginal cost, for Superheavy/BFR/Starship points towards $10/kg. I have quibbles with this. The fuel costs for the Falcon9 are currently around $10/kg, and yet the price remains $5000/kg. A little bird tells me that costs to SpaceX per Falcon9 launch hover at slightly less than half of price rather than less than one precent of price. A bigger, and fully reusable rocket, is fundamentally better from the physics up. But Elon has to assume a crazy level of demand (megatons of lift to orbit yearly) to justify the economies of scale necessary to drive that cost. This is to say nothing of the actual price they elect to charge. Super Heavy may cost 10 to 30 billion dollars to develop. That would be an incredible achievement and vastly better than SLS, but it’s still a lot of money investors are going to want to see back. Also, given how much the Super Heavy is intended to lift per rocket (100 tons), it may take decades for the demand to emerge that can drive economies of scale. If they only launch 3 times a year (which would satisfy a year's worth of Earth’s demand for space launch), it’s not going to get cheap fast.

Personally, I think Starship is going to work and is going to have a huge impact on the long-term reality of human presence in space. But for the next ~20 years, it's going to cost a bit less than the Falcon9 per kg, perhaps as little as half (2500 to 1000 $/kg). Unless someone comes along and exerts competitive pressure, why would they charge less? As for costs, I would believe $500/kg (which would be transformative and awesome!), but I’m not believing $100/kg, much less $10/kg. It may be the case that $10/kg is simply not in the cards ever, regardless of scale. At the end of the day, rockets blow up occasionally. But keep in mind those prices would be cool as hell! Remember, the space shuttle was closer to $20,000/kg. From the Air Forces perspective, $5,000/kg or $500/kg makes little difference as long as it keeps us commercially and militarily competitive with China. However, if what you are interested in are the prospects for independent branches of humanity off the Earth, then you really are talking about lifting megatons of material to locations in deep space. Each human you send to settle would need something around 100 tons of stuff. The up front investment in habitats, an industrial base, agriculture, consumables, and all the paraphernalia of permanent human life will demand a lot of upmass. And if we’re talking about millions of tons, $2,000/kg and $500/kg are exactly the same thing as $20,000/kg: prohibitive. Unfortunately, you can't count on demand for space settlement itself driving economies of scale for an expensive launch system. Space colonization is simply not going to be profitable. For it to start, space access will have to already be cheap. All the plausible ways people propose to make big money in space, even the more speculative stuff like asteroid mining, can be done with robots. Humans would literally make them worse by imposing all our squishy, breathing, pooping, needs onto an otherwise robotic system. In order for us to get to a future where space is for normal people, not just the super-rich or nation states, the price for upmass needs to start below $100/kg and then proceed sharply down from there. We simply can’t count on SpaceX delivering that. They might be able to do it in the long term (30+ years) but it’s not happening fast and the fundamental danger, infrastructure needs, and costs of rockets may make those prices impossible. Starship is necessary for the cool future. It represents an important step, but it is also insufficient for the cool future. The second part of the conventional wisdom I disagree with is the impossibility of competing with SpaceX. I do think that it can be done. SpaceX is scary to compete with, but it has also done some big favors to potential launch entrants. No one now doubts that there is money to be made in space or on launch. It may seem silly, but that was the major barrier before ~2012. The only people who took bets on early-stage capital intensive space startups were rich crazies like Elon or Paul Alen, not mainline venture capitalists. The Money Goons now know the market is real and poised for growth, and a lot of that is directly the result of SpaceX’s trailblazing. However, the companies who could now emerge and stand a chance to compete won't look like SpaceX did in 2009. I have a lot of beef with Spinlaunch’s technical approach, but they do look different. They have a fundamentally distinct technology base with separate physical and engineering constraints. I find them more plausible than Rocket Lab as a long term challenger to SpaceX despite the fact that Rocket Lab is already flying customers to space. I’m super confident Rocket Lab’s technology will work and scale. At the same time, I have serious doubts Spinlaunch’s tech can reach orbit with any payload. However, if I’m wrong (always an option) and Spinlaunch’s technology does work, they have a real chance to compete on price while Rocket Lab is screwed even granting their rockets do everything they want. They'll always be a rocket company competing against another rocket company with a bigger rocket. ULA and those who preceded them were on cost plus contracts and had no incentive to innovate or improve price performance. Because of that there was an unrecognized opportunity for someone to enter the market and do rockets competently. That is now gone. Future technologists, entrepreneurs, and investors have to look much further ahead on the technology possibility frontier if they want to make an impact getting humanity into space. All the non-technical business competency SpaceX brought to bear is now simply the price of admission—you have to do all that AND have a truly innovative concept. If we want to see breakout success that can push the industry forward the same way SpaceX did, we have to try weirder stuff than SpaceX brought to the table: laser/maser thermal rockets, railguns, nuclear powered trebuchets, high altitude blimp spaceports, 10km long space guns*, nuclear propulsion, rotating detonation scramjets, etcetera, etcetera... If you can’t build a plausible story around something like that, and you are trying to compete in launch, you may have a serious problem. *I’m a particular fan of this one.