The launch cost curve just bent. Here's who wins and who loses.

For three decades, the price of getting to orbit was stubbornly stuck. From the Space Shuttle's roughly $54,000 per kilogram to LEO, through Delta IV Heavy's published $14,000/kg, the cost-per-kilogram trend line looked like a flat highway — small reductions, occasional regressions, no real change. That story ended in the last 24 months. The question now isn't whether launch got cheap. It's what we do with it.

The curve that wasn't

If you plot the cost of getting a kilogram to low Earth orbit across the history of spaceflight, the line is, frankly, embarrassing. The Space Shuttle, across its operational lifetime, averaged something on the order of $54,000 per kilogram to LEO once you account for the program costs the way NASA actually budgets them. The European Ariane 5 came in around $10,000/kg. ULA's Delta IV Heavy, until its retirement, was around $14,000/kg. Russia's Proton was cheaper at roughly $5,000/kg, but that price came with geopolitical strings nobody wanted to keep pulling.

The narrative was that launch was structurally expensive. It was a thing you did rarely, with great ceremony, for a small number of high-value payloads. The whole architecture of the space industry — small constellations of expensive satellites, decade-long design cycles, government customers as the default — was built on the assumption that this flat line would continue.

Then, between roughly 2017 and 2024, the line moved. SpaceX's Falcon 9 reusable configuration now publishes launch prices that work out to somewhere in the $2,500–$3,000/kg range for a dedicated ride-share, depending on orbit and contract structure. That's not aspirational. That's on the invoice.

SpaceX's stated target for Starship is a cost-per-kilogram that begins with a "2" and might, on the company's long-term cost curves, end with "10." I want to be careful with this number because SpaceX's communications on Starship economics have been aspirational, not contractual. The credible near-term claim is sub-$1,000/kg. The credible long-term claim is sub-$500/kg, possibly much lower, if the company actually hits the reuse numbers it keeps saying it's on track to hit. For this piece I'm going to use $200/kg as a working assumption for "what the Starship era might look like" — and I'll flag the assumption when it matters.

What $2,500/kg already bought us

The first thing cheaper launch did was enable Starlink. That's obvious in retrospect but was not obvious in 2018. A constellation of small satellites in low orbit, replaced on a three-to-five-year cycle, with continuous launches — that's only an economic proposition if you can get to orbit cheaply. Every other mega-constellation that has been announced since (Kuiper, Guowang, the Chinese state constellations, the European Iris² project) is a direct consequence of the Falcon 9 cost curve.

The second thing it did was collapse the cost of access for everyone else. Universities, small nations, commercial Earth-observation startups — they all flew on Falcon 9 because the alternatives were either unavailable or an order of magnitude more expensive. The diversity of payloads launched in the last five years is unprecedented. We went from "a handful of large GEO comsats per year" to "hundreds of varied missions of all sizes."

The third thing it did, and this is the one I think is underappreciated, was create a market for second-tier launch. Relativity, Rocket Lab, Firefly, Blue Origin, Stoke, a dozen Chinese startups — they all exist because the Falcon 9 cost curve demonstrated that there was a market at $2,500/kg, and they figured out niches (small payloads, dedicated rides, specific orbits, government customers who want redundancy) where they could compete on something other than price. Most of them are still pre-scale. But the category exists, and the category only exists because of what SpaceX did.

So: the $2,500/kg world is the world we already live in. Starlink works. The launch industry is bigger than it has ever been. CubeSats are a real product category. The academic and amateur smallsat community exists in its current form because of this cost curve.

What $200/kg actually changes

Here's where I want to push back on the consensus narrative. The dominant claim about cheaper launch is that it "unlocks everything" — space-based solar power, asteroid mining, lunar habitation, in-orbit manufacturing, you name it. The implicit model is that launch was the bottleneck, and once you remove the bottleneck, the rest of the pipeline fills up.

I think this is mostly wrong. The bottleneck for almost everything people want to do in space is not launch. It's regulatory approval. It's demand. It's the ground segment. It's the operational complexity of doing anything useful with a satellite once you've put it up there. It's the fact that most space applications are still serving Earth-based customers who don't actually need cheaper space access — they need cheaper, better ground services and software.

A few things do change at $200/kg that don't change at $2,500/kg:

• In-orbit refueling and assembly become economic. If you can launch a structure in pieces and assemble it in orbit, you sidestep fairing-size constraints. That's a $200/kg problem, not a $2,500/kg problem.
• Large-scale debris remediation becomes affordable. Active debris removal has been technically feasible for a decade. It has not been economically viable at $10,000+/kg because the cost of removing a defunct satellite exceeded the cost of replacing it. At $200/kg, the math changes.
• Cargo to lunar surface becomes a real logistics problem rather than a science project. NASA's CLPS program, the commercial lunar landers, the long-term Artemis logistics — all of these work at $200/kg in a way that they don't at $5,000/kg.
• Mass to GEO stops being a binding constraint for space-based infrastructure. If you can build a 20-tonne structure in GEO at $200/kg instead of $14,000/kg, you can do things that were previously impossible — large antennas, structural radiators, on-orbit data centers.

So cheaper launch does change the picture, but selectively. It is not a universal unlock. It is a specific unlock for a specific category of missions, and the category is roughly: "things where the bottleneck is genuinely mass-to-orbit rather than anything else."

Who wins, who loses

The companies I think benefit most from the cost curve dropping further are not the established primes. It's the new operators who have been quietly building around an assumed $500/kg future — smallsat constellation operators, in-orbit service companies (refueling, assembly, inspection), space situational awareness startups with serious capital, and the ground-station-as-a-service category. These companies are pre-positioned for a world that doesn't fully exist yet but is increasingly plausible.

The companies I think are most exposed are the traditional primes that built their accounting models and program structures around $10,000/kg. Their cost accounting, their program management overhead, their customer relationships — all of it was designed for a market that no longer exists. ULA's Vulcan, Ariane 6, H3, Long March 5/6 in their current forms — these are all costed for a world that is going away. Some of them will adapt. Some of them will end up as the equivalent of the regional aircraft manufacturers of the 1970s — technically competent, structurally expensive, slowly squeezed out by companies that can move faster.

The companies I think have the most interesting strategic position are the ones who are explicitly building for a $200/kg future and have already shipped something operational in that direction. There aren't many. Worth watching.

What this means for the rest of us

For the amateur and hobbyist community — and I think this is the most underrated part of the cost-curve story — the practical effect is that space is becoming more accessible, not less. More frequent launches mean more chances to spot a satellite pass. Lower-cost constellations mean more public Earth-observation data, more radio satellites to listen to, more opportunities to actually see the equipment in operation rather than just read about it.

If you have never tracked a Starlink pass across the sky, you can do it tonight with a free app. If you've never listened to a NOAA weather satellite downlink, you can do that with a $30 SDR dongle and a piece of wire. If you've never pointed a camera at a rocket launch, the next time a Falcon 9 goes up from Florida or Vandenberg, there are apps that will tell you exactly when and where to look. The cost curve didn't just change the industry. It changed what you can do from your backyard.

That's the part of the story I want to keep coming back to. The launch cost curve is one of the most underappreciated economic events of the last decade, and its downstream effects are going to be uneven, surprising, and largely determined by people who aren't the major launch companies. The companies that adapt fastest — not the companies that are biggest — are the ones that will define the next ten years of space.

I'll be writing more about this. In the meantime, if you want to read the closest thing to a serious book on the orbital economy as it stands today, the Wiley title linked below is the most useful single volume I've found on the subject. It's not cheap and it's not breezy, but it's the one I'd hand to a smart friend who wanted to actually understand what's going on.

And if you want a low-cost way to start interacting with the sky yourself — not reading about it, actually doing it — the smart telescopes in the DWARFLAB and ZWO lines are doing genuinely interesting things in the sub-$1,000 range. They're not traditional telescopes, they're more like pointing a small computer at the sky and letting it do the work. The DWARFLAB Dwarf 3 in particular is the closest thing I've seen to a "press a button and see the Andromeda Galaxy" device, and at the price point it's now available, it's a real option for someone who wants to participate rather than spectate.

That's it for this week's essay. Next Saturday: what the commercial space station transition actually looks like, and why the post-ISS era is going to be weirder than most people think.

— Atlas Renner, Editor-in-Chief, SpaceOrbitals