One of the things that the great investor Mike Maples is known for pointing out is how great founders live in the future. Often it is difficult for people who aren’t in the world of startups and founders to understand what he means by that. Sometimes, though, founders commit their vision for the future to writing.
In a recent blog post, Stoke Space CEO Andy Lapsa lays out his vision for the future of the space sector, and justifies why Stoke Space needs to exist. It is one of the most interesting founder thesis documents that I’ve come across, so I’m looking at it today.
Zooming out a level, Stoke Space is developing a fully reusable medium-lift launch vehicle.

Such a rocket could compete with SpaceX, Blue Origin, Rocket Lab, the United Launch Alliance (ULA), and a consortium of Firefly and Northrop Grumman. Uniquely among this group, Stoke Space is going right to full reusability, and right to orbit. It is not building an intermediate launch vehicle that will operate commercially.
In January 2025, the startup announced a monster $260 million Series C investment. It’s sufficiently well-capitalized to have a good chance at making the founder’s vision for the future come true.
Economics
The blog post begins by breaking down the costs of operating a rocket, in a section entitled “The economics of rapid reusability”. Given how this drives the firm, this seems like a good place to start.
In Andy’s view, there are two primary drivers of launch costs:
costs of manufacturing the launch vehicle
fixed costs of everything else (he says “factories, test facilities, launch complexes, labor, etc.”)
I don’t disagree that these are the main costs of building a rocket company, but I also don’t love how he thinks about them. Putting on my “MBA hat” for a moment, I’m on board with the “everything else” he’s discussing, but I’m unconvinced they are all actually fixed costs.
In particular, I really think labor is going to be a variable cost more like launch vehicle manufacturing.
I’m also unconvinced the facilities expenses are truly fixed in his case, since Stoke has designed its engines in-house and could switch to selling those if it wanted to exit the rocket business. This depends on how a firm structures its facilities, but from the way he mentions real estate, it sounds more like manufacturing overhead to me.
Regardless of my perspectives on his approach to managerial accounting, I’m aligned with him on the importance of flight cadence in changing the cost structure of a space launch program. They say a picture’s worth a thousand words, and he’s got a great graphic from a USAF publication to show this:
The key takeaway from the chart is that flying more frequently enables the costs of the launch infrastructure and R&D effort to be spread out among more revenue. This lowers the cost of launch (or increases the launch services provider’s profit margin — or both).
Andy convincingly claims that flight cadence is just as important as reusability, and that nobody else has built a company with this as the critical idea. He sees the ideas as historically linked; launching more than 20 rockets per year requires reusability in his mind. His evidence for this is the performance of SpaceX vs ULA, Arianespace, Rocket Lab and a couple other companies.
To put a finer point on it, the limiting factor of Falcon 9 rocket flights today is second stage manufacturing (second stages of Falcon 9 rockets are single-use, unlike the first stage). That’s not a problem that can be solved with the Falcon 9 or Falcon Heavy, but it is at least in theory a problem that can be solved with SpaceX’s Starship.
The takeaway here is that any commercially viable competitor to Starship must also be 100% reusable.
I think one of the strongest pieces of evidence in support of this hypothesis is that it’s basically the direct opposite of Astra’s vision while it was publicly traded. The business idea behind Astra was to build a smaller, cheaper, disposable rocket and fly it so much that a single failure wouldn’t matter. The firm’s valuation didn’t take off after going public, because it made incorrect assumptions about customers’ tolerance for loss of payloads.
Stoke also has an insight on this relationship between flight cadence, reuse, and reliability — but they’re taking that insight in as different a direction as possible.
The Cycle
The next section of the thesis expands on the relationship between cost and reusability. It’s called “Cost, availability, and reliability: A virtuous cycle”.
Andy’s point of departure is that those three things (cost, availability, reliability) matter more than anything else to customers of any transportation system, be it the “L”’s Brown Line, a United Airlines flight out of LaGuardia Airport, or a ULA Vulcan rocket.
What SpaceX has done over the past twentyish years is change the limiting factor on launch services from cost to rocket availability. As Andy discussed in the prior section, and points out again here, full reusability should mitigate the specific second stage availability issue that SpaceX is currently encountering.
Reusability in his mind should also increase reliability, though his evidence here seems very anecdotal. He points out that space launch has historically had a 2-8% failure rate since 1980, but SpaceX’s failure rate over their past 300 launches is closer to 1%. That’s an interesting point, and mathematically correct. I wish he went into more detail about why he thinks it is happening, and why he thinks other institutions will be able to replicate the behavior.
Andy pointed out airplanes have extensive acceptance tests; he did not point out they also have extensive maintenance regimes once placed in service. I see no reason why this wouldn’t also be necessary for reusable rockets, and expect this would change the cost structure significantly from the first chart in his post — regardless of whether or not the rocket is reusable.
I suspect part of why he’s right relates to SpaceX seeing a larger sample size for flight hardware, and capturing more data about vehicle performance. But if his underlying idea as to why this is the case is different, that doesn’t feel immaterial to me. I feel like there might have been more to say here.
At the end of this section, he’s connected full reuse to reliability and availability (which he sort of handwaves away, but is the easiest argument to back up observationally through SpaceX). He’s already covered cost in the first section.
This part of the thesis is succinct, and I’m on board with the conclusions.
Timing
The last part of the thesis, entitled “The time is now”, covers Andy’s view on the critical question of “why now?”.
He starts by pointing out that the cost of launch as a fraction of mission costs has spiked radically over the past fifty years as satellites have become cheaper. A $200 million dollar launch used to be a rounding error for an average government customer with a school bus-sized spacecraft; now a startup might see launch costs as 30-70% of the deployment costs for their small satellite.
This creates, in his mind, a strong incentive to reduce launch costs now where none used to exist. I found myself nodding along enthusiastically as I read this for the first time.
He has a view that space is now an extension of the global economy, which only works if there’s sufficient competition. This is an interesting point of view. I agree it’s an extension of the developed global economy. I’m less than convinced that it is an established extension of the developing global economy — which is quite possibly the part of the world with the most growth potential in the space sector, if the costs can be structured correctly. That impacts how much space I think there is for competition in the space economy, including in the launch sector.
Andy’s constant references to SpaceX make it clear that that’s who he sees as his greatest competition. It’ll be interesting to see from a strategy perspective if SpaceX maintains its leading role in the medium-lift launch space after Starship comes online.
Certainly his constant references to SpaceX answer “why now?” well. SpaceX has created an impressive new generation of talent, and has solved the difficult and adjacent problems at least once. They are so close to solving the final set of problems, which have already been solved in other contexts, on Starship that the timing of Stoke Space is compelling.
The necessary technologies all exist. The talent is excited. The market has changed, and is ready now for this service. These things together present a compelling picture about why Stoke Space is on time to compete in the medium-lift launch space.
Final Thoughts
Everybody who’s been paying attention to Stoke Space knows that Andy Lapsa has remarkable problem-founder fit on the technical side. One simply doesn’t raise as much money as he has without it.
Andy holds a PhD from the University of Michigan, and spent time building engines at Blue Origin (as an individual contributor and manager, in both the design and test departments) before spinning out and founding Stoke Space.
What this post does really well is demonstrate his views on the economics of the industry he’s building a behemoth to compete in.
I’ll happily admit that I don’t love all of the intermediate steps he outlines in the post. However, despite those differences of opinion as to his methodology, I agree generally with all of the conclusions he arrives at.
The other thing this post shows is just how nuanced his thought process is. This bodes very well for his ability to communicate with stakeholders as his company grows and takes the next steps towards fully reusable medium-lift launch vehicles in a potentially quite competitive market.
I’m very excited to see how right he actually is, both about his intermediate steps and his overall conclusions, over the coming decade.
Stepping back from Stoke Space, and even the space sector, I wish more founders publicly shared their internal thesis. Beyond being great marketing material, I think these sorts of posts are inspiring!
Furthermore, posts like this help me understand how the CEO thinks, how he sees the past, and how he’s building a future he wants to live in. Those seem like the sorts of things most founders should want the public to understand about them.