Latest / The Operations Podcast with Fexingo: Process, People, and Profit in Modern Business / How SpaceX Revolutionized Rocket Manufacturing Operations
Transcript
- Lucas: When most people think about SpaceX, they think about rockets landing on drone ships. The dramatic stuff. But the more interesting story, the one that actually made those landings possible, is how SpaceX decided to build rockets in the first place. Luna: Because before SpaceX, the aerospace industry had a very set way of doing things. Lucas: Exactly. The traditional model, epitomized by Boeing and Lockheed Martin, was built around low volume and extreme documentation. You'd produce maybe one or two rockets a year, each one essentially handcrafted, with thousands of pages of paperwork per component. The supply chain was fractured — engines from one vendor, avionics from another, body panels from a third. And the cost was astronomical. A single Delta IV Heavy launch could run $400 million. Luna: And then SpaceX comes along and says, we're going to build almost everything ourselves, and we're going to produce rockets like they're iPhones. Lucas: Elon Musk has talked about this — he looked at the cost of buying a rocket engine from Russia and realized that the raw material cost was only about 2 percent of the purchase price. That's a first-principles moment. So he decided to build the engines in-house. The Merlin engine now costs a fraction of what any supplier would charge, and SpaceX can produce them in batches. As of early 2026, they've built over 1,200 Merlins. Luna: Wait, 1,200 engines? That's more than the entire US rocket industry produced in the previous two decades combined. Lucas: That's the scale shift. And it's not just about making more engines; it's about how the entire factory floor is organized. Traditional aerospace factories are organized by craft — the welding section, the machining section, the assembly bay. Parts move from station to station, waiting in queues. SpaceX instead organized the factory floor around the product flow. Small teams of about five to eight people own a whole subsystem — the turbopump, the combustion chamber — and they sit right next to the assembly line. Luna: So it's more like a Toyota lean manufacturing cell than a traditional aerospace hangar. Lucas: Exactly. And they iterate incredibly fast. There's a famous example from early Falcon 9 development: engineers redesigned the octaweb — that's the structure that holds the nine engines at the base — something like eight times in one year. Each time, they built a new version, tested it, found a flaw, and fixed it. In traditional aerospace, that would have triggered a year-long review process. At SpaceX, the person who spots the problem can often walk over to the machine shop and have a fix within hours. Luna: I remember reading that the original Falcon 1 almost bankrupted them. Three launches, three failures. But they treated each failure as data, not disaster. Lucas: That's the culture: fail fast, but fail forward. After the third Falcon 1 failure, Musk addressed the team and basically said, 'Look, we have the resources for one more launch. If it fails, SpaceX is over. But what did we learn from the last three launches?' And they applied those lessons. The fourth launch succeeded. That iterative approach is now embedded in their operations. They don't treat a failed test as a black mark; they treat it as a necessary step toward a better product. Luna: And that's completely opposite to how NASA or Lockheed would handle a failure. There'd be months of investigation, hundreds of recommendations, and a culture of blame avoidance. Lucas: Right. And SpaceX can afford that faster iteration because they vertically integrated so much of their supply chain. They don't have to wait for parts from ten different suppliers across the country. The Hawthorne factory has machine shops, welding, electronics fabrication, and final assembly all under one roof. They even developed their own metal alloys for the rocket bodies. That vertical integration gives them what operations people call 'reduced lead time variability' — they control the timeline. Luna: And because they produce so many engines, they can also test them more aggressively. I think they fire every single Merlin engine on a test stand before it ever goes on a rocket — full duration burn. That's unheard of in an industry where you might test every tenth engine. Lucas: And that testing gives them reliability data they can feed back into the production line. If one engine develops a crack at a certain weld point, they can adjust the process for all subsequent engines. It's a closed feedback loop: build, test, measure, improve. That loop, running at high speed, is the operational engine behind SpaceX's launch cadence. As of June 2026, they're launching roughly every two weeks. That's 26 launches a year from a single pad. Each launch uses a Falcon 9 that might be on its tenth flight. The cost per launch? Somewhere around $30 million, compared to $150 million for a comparable Atlas V. Luna: That's a 5x cost reduction. And it's not just the rocket itself — it's the launch operations. They've automated a lot of the pre-launch checks. They have a team that can turn around a booster in under a month. Lucas: The turnaround operations are a whole other dimension. After a booster lands, they do inspections, refurbish the grid fins, replace any worn parts, and mate it to a new second stage. The entire process is designed for speed. They even built a special crane system to swap boosters quickly on the launch mount. The thinking is: every hour the rocket sits on the ground is a revenue opportunity lost. Luna: It's almost like airline operations, but for space. The goal is to get the asset flying again as fast as possible. Lucas: And the next step is Starship, which is designed to be even more like an aircraft. They're building a factory in Boca Chica, Texas, that can produce multiple Starships per week. The manufacturing process there uses even more automation: robotic welding, additive manufacturing for engine parts, and a continuous production flow. If they achieve their goal of one Starship launch per day, the operational model will have completely transformed the economics of space access. Luna: Before we go further, I want to come back to something you said earlier about the culture of learning from failure. That's one of those things that sounds good in theory but is really hard to implement in practice, especially in a safety-critical industry. Lucas: It is, and it's something we think about a lot on this show — how operational culture can either accelerate or block progress. And it's one reason we keep this podcast free of ads or sponsors. We want these conversations to be useful to anyone trying to improve their own operations, without any commercial agenda. If you find value in that, and you want to support the choice to stay ad-free, the link is buy me a coffee dot com slash fexingo. No pressure, but it does help us keep doing this. Luna: Absolutely. It's a small way to keep the conversation independent and focused on what matters: real operational insights. So back to SpaceX — how much of this model do you think is transferable to other industries? Lucas: I think a lot more than people assume. The core principles — vertical integration, rapid iteration, small autonomous teams, closed feedback loops — those are straight out of the lean startup and agile manufacturing playbooks. What's unique about SpaceX is the willingness to go all-in on those principles in an industry that was hostile to them. But any company that manufactures complex products can learn from how they collapsed supply chains and accelerated learning cycles. The key question is: how do you create the psychological safety to fail fast in a safety-critical environment? SpaceX did it by decoupling failure from blame. The postmortem isn't about who messed up; it's about what the system missed. Luna: That is a powerful lesson for any operations leader. And it's one that's easy to say and very hard to do. Lucas: Hard, but not impossible. And the proof is in the launch rate. Every two weeks, a Falcon 9 goes up, and that cadence is a direct result of the operational choices made in a factory in Hawthorne, not just the engineering in the design office. That's the story we wanted to tell today.