How to Dominate Manufacturing in the US | Chris Power, Hadrian

0:00Today I'm sitting down with Chris Power, the founder and CEO of Hadrian. Hadrian has opened six factories since like 2022, latest one being 2.2 million square feet. I want to start this off with what does it actually require to scale manufacturing in the United States?

0:16Chris Power:A lot. Putting aside the raw horsepower and pain, I think actually it is GPU style stations that can be modularly configured inside a factory which gives you standards to automate against and then a very complex organization that can build real manufacturing autonomy and the third thing is a very careful balance of physical ai software and a new american workforce do you want to go into what does it mean to have like a gpu as a factory so most um production lines are set up linearly like a cpu like you can do more things but they're running linearly so you know there's like a 60 minute component that then goes into a two-hour assembly process and you're kind of just running things like this linearly the gpu is more like everything is so standardized and automated that you can run a high mix low volume factory like as as efficiently as a low mix high volume factory and manage peak capacity like you manage gpus it's very complicated to do but it's the only way you get like real high mix low volume performance at a factory scale one of the things that you talked about earlier today was this idea that you have to basically if you want a contract you have to kind of build ahead of the contract and you have to build a factory to show that you're serious enough to actually you know own it how does that work uh well let me give you an analogy so aws uh you know developed the largest cloud computing capacity ever and then and then second started selling it to the market but they already had the largest data store for Amazon engineering and Amazon retail as the first customer.

2:23Chris Power:So it's pretty believable that like, hey, you're running AWS internally for Amazon, you can start selling it to like Stripe. Contract manufacturing in the United States is this big chicken and the egg problem where it makes sense to anybody that, hey, you are designing a new low cost interceptor. Why would any good startup build their own vertically integrated factories at a worse economic position than Hadrian, it doesn't make sense. What does make sense is that there is no starting point of capacity. So Shenzhen, for example, if you start today a new humanoid robot and you have a circuit board, you're not going to start your own circuit board factory because there's a bunch of capacity so it's obvious to you that you tap into it and you can trust it because it's at scale.

3:15Chris Power:The United States has no contract manufacturing capacity. So if you want anyone to take a big bet on you, you have to basically build at least 50%, more like 1.2x the percent of your future sales capacity ahead of having those contracts. Otherwise, you're going to be so far behind the eight ball of both optics and execution, execution being the most important thing because manufacturing, what we sell in manufacturing is trust, that you're never going to escape. okay so we're building a data center okay um we've never built a data center before how how much easier is it for you to rent gpus from me in my data center where it's already half built as opposed to like on a slideshow so american reindustrialization is largely the former problem not because people are uh bullshitting just because the capacity doesn't exist so you have to build it ahead of demand uh which takes a lot of balls and capital there's this funny thing that you mentioned earlier today which was this idea that to actually get some of the contracts that you're going for it just takes two years from the start line and i think this is kind of unique where you both have to project out here's what the contract might be and build the capacity for it and also not know necessarily when it's going to come.

4:43How do you kind of like math that in your brain?

4:45Chris Power:So the math is burn rate and execution risk. Execution risk is the easier one because once you understand that any factory build out or, you know, I buy a 3D printer or a CNC machine or robot, there are so many things that can go wrong that if you say it's going to take eight months to get here i know from experience that it's going to be somewhere between eight and a half weeks to ten and a half weeks point number one point number two is if i told you to build a drone factory and a missile factory you would think that that is binary risk in the sense that to win either of those contracts you would have to deploy capex and human capital but like both to have an equal chance of winning those contracts if the big if if you have a flexible workforce and flexible capex via software the risk mitigation you can do is that 80 of your capex you can switch from program to program like data centers you don't know whether the consumer chat gpt is going to take more demand than anthropic enterprise compute but you do know that they're going to share the same gpu cluster you can get manufacturing 80 of the way there to the flexible risk grade if you are detail-oriented enough to break down every product into the manufacturing method and underneath the manufacturing method, the specific CapEx, which is extremely complicated.

6:30Chris Power:But if you are capable of doing that, then you can make bets that are not risky to you that are extremely risky to everyone else. Do you want to go through one of those bets that is extremely risky? If you were to just look at building a new$300 million casting facility and you had no contracts, and the capex takes two years to get here, which means you're in production in year three, best case. If you were looking at that linearly

7:06Chris Power:for castings for turbine blades, for argument's sake, you'd have to be damn sure that the demand is still going to be there in year four. If you break down the capex that is linked to the process that is a subset of the output, which is a, let's call it a Manel casting. And you do the same thing for munitions, 40 % of which are castings that are never going to be additive because it wobbles too much. And there are some unique bits of CapEx for sure, but 80 % of it can be shared. and you are willing to have a 20 % larger CapEx bill that buys you 100 % flexibility, what you're actually doing is deploying 300 million CapEx against three demand signals, not one linear one.

8:07Chris Power:But that is down to the level of detail of, hey, both tonnage of pour by alloy and for like a molding line uh you know envelope size of average component by floor space all this can be done it's just stupidly detail oriented it's very precise your previous factories were like 20 000 square feet and then 100 000 square feet and then you jump to like 2.2 million how do you kind of think through what risk you're willing to take on what timeline in order to just build the capacity that you see demand for a few years out so i i think there is um people think about risk very wholly and i'll give you a couple of examples um firstly for large companies capex risk is binary what do i mean by that um what i what i mean is you may think traditionally that if you were building a large circuit board facility to win a large supply chain contract for drones like from our friends as soren may need to buy a lot of circuit boards that you would really like to have the contracts before you bought a billion dollars in capex and if you don't get the contracts, your company's dead.

9:38Chris Power:Well, that is true. What is more real is you have a 0 % chance of growing revenue unless you spend the billion in capex because no one's going to give you a production contract unless you've already started executing. And the second instance is you bought all the capex and if you don't close the contract, your company dies. But at least now you have a more than 50 % chance of executing the revenue. So it is binary in the sense that you need to deploy billions of dollars ahead of contracts and it doesn't mean that that is not a company-killing event. It just means that you've increased your probability to more than 50 % and if you don't deploy a billion dollars in CapEx, you're 100 % dead or you're a$100 million acquisition.

10:30Chris Power:and until you're at i think roughly 50 billion in scale every capex bed is a company killing event but what is more true is that if you don't do it you're 100 dead and that's how that's how we think about it i remember two and a half years ago when i first interviewed you we were having some conversation afterwards i asked you like what do you think the probability is that hadrian makes it and you said something along the lines of 20 and i i remember thinking that's fucking wild like that's crazy that you think that but since then you know you've continued to if you're saying like a 50 billion scale suddenly you're not going to die every time right now you're still on that like curve um yeah because each each year to grow you have a time constant that your capex build is two years outside of the revenue curve so every bet you make you're dead so the risk doesn't trade until you stop growing how do you think the risk has changed over the past two and a half years from like has has the like risk of you know death gone from you know very very high at 80 percent to like 30 percent or how does that change it's straight at 80 percent crazy it's not because we're risky it's because the fundamental construct is that if you want to continue having venture returns and a relatively decent cost of capital, which is required for the re-industrialized mission.

11:57Chris Power:Therefore, you are always growing your risk balance sheet, let's call it CapEx or R &D, by double the sane amount. And the only reason the risk retires is if you slow your growth to 20 % because then your cash flows catch up to your CapEx bets. now you can get extremely good at uh forecasting tracking like we're not yoloing billion dollars capex it's very intentional but if you if you deploy two billion in capex now today that gives you the opportunity to grow to five billion in revenue in 2028 now if you don't grow to 5 billion revenue in 2028 you don't have any venture returns if we also don't grow 5 billion revenue by 2028 the capex bill is such that you know you're turbo fucked so the only option is to do both at once so until such time as you're growing 20 % year over year then the risk doesn't retire there's this amazing line that jeff bezos once talked about where there's this idea of two-way doors and one-way doors and two-way doors you can go in and make a decision and then walk that decision back if it doesn't work one-way doors you basically go and if the thing doesn't work you've committed yeah but isaiah from valor has this other idea where he's like in order to move fast i've had to sprint through some one-way doors yes how do you think about sprinting through one-way doors in order to maintain a high pace I think about it in two modes.

13:40One is, and it's mostly about time.

13:46Chris Power:And I'll give you a practical example. And we try to teach the same all the time. So let's say you're building a automated welding line. And you have to deliver a working prototype in month 10, but your CapEx lead times are nine months. and to be 100 % accurate around what robot is it, what's the station look like, is it the right world, it's going to take you three months. Best team in the world, it's going to take you three months. Okay, that doesn't work because then you're in month 13. So how do you hit the timeline while minimizing the risk hedges of what you can't control, which is time?

14:31Chris Power:the answer is if you break down that automated welding station maybe it's a million dollars maybe 800 grand is single skew things like the particular machine and the rest of it is a robot arm a couple other things if you're running a big platform like hadrian what you can do is say hey guys you need to spend three weeks out of four weeks being 100 sure that this big block of capex is 100 % because that you can't change that but let's hash in a rough idea of what robot arms are going to fit and if you're wrong it's going to cost us 50 grand because we have to reuse the robot arm for something else but if you miss the timeline the company's dead because the customer cancels the contract so you got to think about it as capex capex and long engineering projects are all one-way doors and you're better off spending 30 more to bake in hedges because the one thing you can do is shit we ordered all the capex and we got some of the engineering predictions wrong so it cost us three weeks because we screwed up an api integration or okay that happens all the time but then you're three weeks late what you can't do is order nine months worth of capex it arrives and on nine months and day one you realize you're wrong and then you're at 18 months risk so you have to balance execution timelines against bets but if you're dealing with the thing about software or sales or whatever is it's bad to miss a month it's bad to miss a quarter but you can recover from it because you can observe the problem and then pivot and your correction time is maybe you burnt 10 million in runway, but you can pivot.

16:31Chris Power:CapEx is a whole different ballgame. And if you don't have 100 % certainty that the CapEx you're buying is correct, then you have to hedge. And then you have to financialize that hedge across different trade spaces. So that's the way we think about it. um the second most important way is one of my favorite lines of all time is sometimes you just gotta roll a hard six what does that mean it means that sometimes actually most of the time you just gotta be right and in businesses where you're dealing with long correction timelines you just have to be really fucking good and one of my this is one of my favorite tv shows of all time is battle cycle like to go and you're fighting an asymmetric war and you don't know what's going on and everyone's like well you know we gotta we're gonna go raid this fuel depot and all and the opposing force is 10 times bigger and you know you're gonna risk a lot of fighter jets going getting this you know fuel for the Battlestar and everyone's like well we're gonna lose you know it's it's very risky it's super asymmetric it's never gonna work but the option space is basically you don't do it you don't win and you run out of fuel or you just win and you got to roll a hard six and you gotta you know like in in in Dungeons and Dragons it's a natural 20 so mostly the right attitude is there's no way to manage this you just got to roll a bunch of hard sixes and if you once you understand that that is the business that's fine but there are no one-way or two-day doors there are there are only one-way doors and then you got to get very good at doing that and understanding the difference between something that you can fix later and something that you're not going to be able to fix later I'll give you a great example.

18:40Chris Power:So usually when you're constructing a factory, you want to have a very good design and layout. And the layout determines your construction costs because some parts of the layout you need very expensive foundations and electrical, clean rooms, et cetera. If you're moving fast, you're not going to have the perfect layout. You want to really think it through because whatever the layout, like once you land the CapEx, it's stuck and a lot of factory layout determines your operating efficiency in weird ways travel time sure but like where people communicate and stuff like that what the way one of the ways we think about it is okay the one of the longest poles in factory bring up is literally where the electrical drops and where the foundations are now if you want to aim for perfection maybe you don't have to lay 20 inch concrete vibration isolated foundations everywhere but if you want to kick off construction tomorrow you're better off spending 25 million dollars on foundations instead of 15 million dollars so that you're buying your layout team four months worth of simulation time and they can put the capax wherever they want and that is a calculation of terminal efficiency value of that factory um so we over invest in things like we're going to do electrical drops everywhere and the reason why we do that is you can pay an electrical contract to do electrical drops everywhere and you're not going to use 30 of them that's kind of counterintuitive but one they can just do mass electrical drops in two weeks bang them everywhere and then a you know a manufacturing engineer can connect an electrical drop to a machine so you want to arbitrage the very expensive lock-in things while not waiting for long lead decisions like foundation build outs or capex selection and you have to think about the calculation of risk in a kind of different way there's this idea in warfare of like asymmetry where if you have like take china for example where they have like 232 to one shipbuilding capacity to the united states if you tried to just front you know gun it and build a bunch of shipbuilding capacity in the united states you couldn't beat them but if you can figure out like how to take out their shipbuilding capacity and stuff you can kind of like even the playing field yeah left of um is there such a thing in manufacturing as like asymmetric manufacturing no fuck there's asymmetric products like i think you can design a thing that's got to cost a kilo of a thousand you know there's no asymmetric asymmetric manufacturing there's scale flex scale flexibility and agility which we can do um there's no asymmetric manufacturing method where you magically going to make a tomahawk casting for a dollar you know it doesn't exist when you built the first factory uh in 2022 i got to imagine that there's a whole bunch of things that you did back then that you would not do today if you were to build the same factory would have been the biggest learnings of how to build factories really efficiently over the past like four years i'll break the answer into two which is factory development and then um like factory design and the first one is really easy and i'll tell you a story so um one of the big things we do at our one of the parts of opus is called mother brain and effectively we take over all the capex software and reprogram and you need some decent apis into the capex or at minimum you need a text file you can read but it's brutal engineering okay so you're a small startup and you've got six million in capex to spend so we buy two types of machines makes sense and you call the president of the german capex vendor because your team sent them an email saying can we please have the api documents and they think you're a hacker because no one's ever asked for the api to the machine before and you explain it to them and they give you the api documents and you read them and it looks all fine when you get to the factory floor what ends up happening is both brands of machines the api that reports off the hardware is completely different so to give stochastic inconsistent results and one of the capex brands you hit all the apis just to test it and 50 of the api calls for a return like nix implemented yet in german and you realize that actually what you're doing and you know you try to build software on top of it and one of them takes 12 weeks and the other one takes two weeks so what you realize is actually what you're doing is dealing with a bunch of unknown unknowns so you build every single plan against the fact that everyone's going to fail against you and you start building capability engineering systems around it so you never expect like the plc of the world robot until you've literally integrated it you have no idea what reality looks like which is fine as long as you build boatloads of hedges in your timeline and you plan around the fact that the industry is fucked in terms of factory design the biggest lesson is that people design factories around the success case and the success case is like 90 yield smooth flow etc i'm not talking about like maintenance downtime just one of the things we do which i think is you can only do with software powered factories but makes complete sense is you have a ton of slack capacity that's just based around failures so the analogy i'll use is like okay so you've got a five-lane freeway and a bunch of waymos and some of them are going a screw up at some point right and you've seen traffic jams right you hit the brakes everything backs up what you actually want to do is as soon as someone has a blown tire there's a there's an off-ramp or there's a slip lane on the freeway where they can just wait but it doesn't hold everything back most factories are designed with the capex the people and the let's call it business processes that if one of the cars blows a tire there's no off-ramp and i'm not saying you trash the car you need to send it back and fix the tire but you're you're sacrificing this like kind of industrial engineering efficiency for like you know what you would have wished when that car blew a tire man i really really wish i had an off-ramp so i could put that to the side and fix it while not holding up five lanes of traffic at rush hour now what it requires is that every station inside the factory is so standardized that you can switch things out like a GPU and that you have high enough efficiency that it doesn't hurt your margins or whatever.

25:44Chris Power:But that is one thing that we've learned very well. The second thing we've learned is that there is a limit to software complexity that doesn't right match the physical world. And I'll give you an example. So in manufacturing, you've got... yeah production and inspection and you've got some let's say they're components and they're components that have stable processes at some yield rate and the important thing about yield is you can predict it because if you can predict yield and it's 20 you just make 10 you get eight and it flows through um and then you have new product introduction where you're you're machining something you're welding something you're inspecting it you're not 100 sure that that's the right process and you you know you need to make one to okay now you end up with this big queuing system so something goes from manufacturing to inspection and what you want to do in new product introduction is have the highest rate of the the lowest possible impedance between inspection data and manufacturing because the thing that creates cost or late deliveries to the customer is the not the cycle time of the manufacturing of the inspection but the the back and forth so if you make something i want inspected in an hour so the feedback loop happens like this so in software land this is pretty easy like queuing theory so if you scan an mpi component into the queue it goes to the top of the inspection queue and there's a you know this this is very easy software in reality it never works like that so what you do is you have a physical set of capex that's just for the fast line and there's this odd property where that's not just because humans are dumb it's not because the the it's you literally need to separate your fast response part of the factory it can be there can be a rope between it and the interesting about that is software gen software tends to design itself towards a physical layout of systems so you need to kind of like bake the factory to the natural software workflows and that you want to match that as close as possible so it's interpretable you know and the rest of it is like you know you just gotta fucking go like hell on kind of the same thread of sprinting through one-way doors what's been the best example of something that went absolutely horribly wrong should have killed the company and somehow you're able to like pull a rabbit out of a hat and stop it okay so one very clear example is um this was a couple of years ago when we had we had a lot of money for a series a company but an aggregate a lot not a lot of money and we're landing we're landing capex two months before customer contracts hit which is what you do you know um ideally you know i i owe you a bit of compute and i switch on the gpu two months before you don't want a day before but yeah and we we bought all this new equipment incredibly reputable vendor and about a couple of weeks in about four months before we had a bunch of contract deliverables for a very early customer um that was very important but more importantly we were coming up to a fundraiser you couldn't miss a delivery quarter because the whole bet is you know obviously the thesis is real but can you execute against it and two of my top guys came running up to the office and they said like hey and i won't name the vendor and there's you know there's some rust okay and we think it's terminal okay and we've called the vendor and they think you know it's not a problem there's not a problem now our guys are like the best in the world so you got to assume that they're like yeah days at least if not weeks ahead of you know the vendor's self-management and it turns out that within within 48 hours we'd assess the fact that a subcomponent on these in these bits of capex um they they had switched out the subcomponent vendor in this country that produced the capex so you know the car analogy would be we kept buying toyota camrys but they'd switched out the tire vendor which is fine people people do it all the time and we worked and because of the serial numbers like we had 40 machines and the serial numbers were like 600 to 800 which means this is a bash problem right it was propagating really fast so we realized and you know the response plan from the vendor and the warranty was basically like well we'll get it done by thanksgiving and thanksgiving would have killed the company because their fix-it plan was we're going to deconstruct machines one by one we're going to replace this stuff they didn't really understand that you had to replace it completely because it's rust you know and the guys were right i spent 48 hours on it myself and i agreed with them um and then i spent a week straight yelling at the president of this company and told them that i would if they didn't fix it in the next four weeks the company would die which was true that's not a bullshit statement because the first thing is they didn't realize how bad the problem was and what the impact would be um you know there's like eight weeks of downtime and then there's eight weeks of downtime kills the revenue curve of the company because it it's like missing a sales quarter in sass like suddenly your burn rate quadruples because you got all these sales people in no error um and then force them to fly in 30 engineers from this foreign country and get the whole thing done in four weeks through pure just like you know i was going to headlock these guys so that's a real operating problem um that would have killed the company uh there are there are many others it happens about once a quarter but it's manufacturing this is the game what's the most recent company killing event nothing nothing in the last three months but there's one every month but i think this is the point is you're making you you're hard rolling sixes every quarter and there's always a company killing event once a month as a matter of fact i recently interviewed keller from zipline great and yeah he's great but by the way i i think keller is probably the best deep tech ceo top five i was recently interviewing him and his entire timeline was basically you know like five years from now or something he wanted to have a million deliveries a day on his network.

32:35And then Uber came in like a month before the interview and basically pulled in the entire timeline by two and a half years and said, I'm going to take all of your capacity. And this completely radically changed like how they are thinking about scaling their business because that's just one customer and there's going to be a hell of a lot more in two and a half years. When you're thinking about not only like pulling in timelines, but also resetting your expectations upwards. so you go from like your projections where i'm going to make a billion dollars in revenue three years from now and suddenly you're like actually i'm gonna make like five billion three years from now and i need to make a billion next year with these like long lead times on not only being able to like buy capex but also get contracts for that yeah how do you like manage that so so for us this

33:22Chris Power:actually happened so the founders fund let out series c i think it was a year ago and then t-ray priced at the c2 and we just raised a series d and the main reason was because every month we're raising the forecast and it wasn't like a sales thing it was just like yeah we barely have a federal sales team they're great but there's eight of them and yet there's the yeah every quarter there's two billion more in pipeline and we're not asking for it so you got two problems there you've got a burn rate of talent problem more opus engineering to make things more scalable more forward deployed engineering to build more contracts but your rate keeps going up like this so you can't you can't plan and you've got the capex problem so the way we think about it is in aggregate you know are we sure that we're going to close this contract that if we close it we're going to need 50 more robotics engineers no no one does highly competitive bidding processes competitors of all sides but we do know that we have a 95 percent statistical chance across 10 contracts that we're going to need at least 100 robotics engineers so it kind of doesn't matter so you aggregate everything and then you then you always go like you then you bet into the slope because in a fast-growing business you can always stop hiring for a quarter, you can never restart hiring.

34:52Chris Power:On the CapEx side, once you've got multiple programs, at least for our business, which is factories as a service, we can start aggregating CapEx and robotics and stuff to shared Lego bricks and start Ford buying them. So you're kind of pulling CapEx from inventory. Now, 20 % of the CapEx, you can't do this because if there's one weird 3d printer that this one contract is going to use it's not an aggregate but look we standardize around fanic arms for everything small medium large bing bing bing in in specific use cases that is inefficient at the company level what it means is that we're buying 300 200 and 100 and then our programs and capability engineering teams can pull them and plug them into whatever they need and you can massively compress the risk and timeline that is actually very hard to execute um because you need to take a program to a contract forecast to a simulated which which arm are you using that's not not trivial but the math is really easy and scale has this interesting property of making that easy but the the challenging thing is the lead times to adjust to that so you know what is recruiting recruiting is an incredibly strong firewall to make sure that under pressure, no one's going to hire a C plus engineer.

36:15Okay.

36:15Chris Power:Let's say we have that. Then it is marketing and outbound sales. If one recruiter can produce four great hires in a highly competitive environment, and it takes 90 days to hire a great recruiter. By the time you understand that you need a double recruiting for next year your slack time is six months all right hire a recruiter 90 days they so you have to kind of just bet on scale and it has this interesting property where it looks like a risk but it's actually it's actually not you're never going to fire people you're just going to like pause hiring for a month at max and let it catch up but that was intentional because the properties of all these complicated capex and long lead time businesses is everything gets easier the bigger you get operating the company gets harder every week because it's complicated but the risk decisions get much more stomachable the bigger you get sam what was talked about this idea of like emergent properties of scale yes and that's real yes and they're not predictable until you're in it and then you understand what it actually means i remember this story where this guy wanted to start a like machine shop or something and he wanted to buy it like a laser and he wasn't able to get any financing for this laser so he had to buy it outright with his own cash.

37:33Chris Power:Jim. Actually, yeah. And then suddenly, as you kind of scale, and you're generating a lot of revenue, lots of people line up to like fund it with debt and other mechanisms. How do things kind of shift from zero million in revenue to like 1000 million in revenue on the your ability to like scale things? The operation of the company gets harder, but the fundamentals of the market get far easier um material suddenly gets cheaper debt is easier being at the front of the capex q is easier because you're everyone's biggest customer because the market adjusts every business wants to grow and if you're the fastest growth engine the market will correct to you it's just getting there is very hard would have been the biggest things in this business where at the beginning they were just impossibly difficult and then over the past couple years they've gotten ridiculously simple well like um it used to be setting up a new line of machines was like a fucking hell on wheels um and it is hell on wheels for some of the teams now especially when we're doing new things but like we've got a playbook you know it's going to roll out on some cadence you know Right now you're at something like 725-ish employees and over the next like four months you're going to almost double that.

39:01I think hire something like 600 people. How do you design the company where it can actually absorb that new influx of hires?

39:07Chris Power:So there's a couple of things. The first thing is that our org design is built around modular capabilities and programs with extremely strong engineering leadership. that roughly have a maximum headcount of 30 to 40. So the problem that most people face is accidentally hiring bad people fast or hiring great people that can't integrate into the company. And in very large software teams, like you have a thousand backend engineers, objectively true. The way we run is that like the weld engineering team and the backend software team for weld engineering together are going from 30 to 70, which for two competent people, the chief weld engineer and the chief software engineer, not that bad.

40:06Chris Power:So we have 80 of those. And if you get the leadership right and you have the first five engineers have very high standards and you just tell them like, don't hire anyone that is not up to your bar. You can scale the output and capability of the company without running into these weird bottlenecks where like all of a sudden you have 100 backend engineers and now there needs to be eight managers because you're fragmenting it across the board. What you have to do before that is fragment the APIs of the organization such that the world team doesn't even know who the scheduling team lead is because it's just APIs and you need it so that's those are two things um the second thing is that you need to pick you really need to pick what you don't care about what does that mean so you need to care about most people start with like 10 cultural things that are really important that you know and you need to get down to two and mostly because in their onboarding you're gonna say like look we're gonna of fire you if you don't do these two things um and for us they're very clear the second thing you need to do is have extremely scalable testing where you're not reliant on a new engineering manager designing a great hiring process for a world engineer the company has spent a lot of time on the world engineering interview test and anyone who passes unless they're trying to be fraudulent is going to be default good but anything other than that is kind of false like you're not going to like everybody you're not gonna um have the same execution culture as long as you have the same pace and the same methodology and the same core product development principles it's going to be fine and then you know the company looks very different every month and that's that's fine you kind of got to really really really pick the sacred cows and most companies have far too many of them or they're not specific enough how do you decide which cows to axe and in your case judgment it's very nuanced some of it's observing like hey we we hired 20 people and they had these six rules and this kind of it wasn't the clarity of the writing wasn't quite there or it actually doesn't matter you know it doesn't matter like um you know there's like luxury beliefs okay like a lot of cultural values and a lot of execution values are luxury beliefs so like what are the two or three that you're really going to care about and are they the most important thing or is that a luxury of being small um and and then you you know it but but it's much more art than science for sure.

43:00Let's say you're like a couple hundred million in revenue right now and you're going to scale very quickly into like the billions, but your pipeline is growing even faster and you're like opportunities are growing even faster. How do you decide which things to go after versus which things to just say no? Mission first.

43:20Chris Power:We have a ton of commercial opportunity that we can't tackle because we don't have the resources. we care a lot about re-industrialization we care a lot about defense point number one point number two is where are the where are the doors that are open and are going to shut and never reopen again so there are some programs that we work with the department of war the primes where it's the first time a customer has ever let a contractor into their division in the last 50 years and they're probably never going to do it again you know and there's 20 billion behind it there's a lot of impact behind it so those are because you're never going to be able to like get in there again as opposed to we're turning down a lot of um factories as a service business for low-cost interceptors makes perfect sense there's a ton of people who've got great designs um they haven't scaled manufacturing these things are not hard to produce you can very imagine in your head a low-cost interceptor factory with a bunch of different designs and it's very efficient for the market it's efficient for us we print money it's much more efficient for the neo primes because it's literally cheaper for them to go with us so we can load balance it then trying okay you know but um there are going to be many low cost interceptor programs so while i hate not being able to partner with those companies right now it's it's below the line of mission or it's never going to happen again and then the fourth thing which is obvious is to deploy a factory maximum lego bricks inside that factory are 50 um we have a long r &d roadmap to do everything some programs our product team will say to deliver this factory it requires um 20 lego bricks 19 of which we have already have okay so we might be five engineers worth of r &d risk some programs we need 20 lego bricks and we have one of them okay but it's not r &d but it's more like company level execution risk that those are the easy decisions to make because it's like how good is your hr software versus the requirements mostly it's mission and then uh like open shut doors if you have like way more demand than you possibly like service today what is the like bottleneck for deciding to grow faster a very careful sense of how over the skis the entire company is does that kind of shift day to day or week to week yep if you had to like think through the past year what have been moments where you felt like you were over your skis versus like completely off on the brakes i think that's slightly the wrong question what's the right one i think as an individual you have a near perfect understanding of where you're at and your own technical capability what i find is difficult is that um my judgment is based off my capability to rapidly hire engineers and build something never been seen before or my 10 best people's capability but we have to realize is that the 10 people running the 10th program you've never met before and you have to put aside your put aside your priors and let the organization show you where reality is at and then judge off the worst case scenario so like imagine if you're recording 20 podcasts and you never wanted to make less than a b plus podcast but you're and you're doing seven a week by yourself and you know when you get to eight the eight one is like a little bit wobbly because you're tired well the 10th podcast interviewer you are is not going to be you um but you're growing so fast that you're kind of like you're not going to show the results of that bad podcast until you see the youtube numbers in month six so how much how much how fucked are you before you can see the results that that is that is uh very intuitive and counterintuitive in your own judgment has there ever been a moment where you felt like you were doing 10 podcasts a week oh yeah i mean uh the first 18 months of the company for sure we raised 1.3 million dollars and decided to build 200 million dollars worth of software and yeah okay and then definitely in the last nine months.

47:53Chris Power:But I think I'm much better at it now. And I can self judge like, wow, you know, that team is actually doing really good. And I just fucking blew them up. Like, I'm pretty sure I haven't slept, you know, it's like drinking too many beers, right? By the time you had the six beer, you're not gonna be able to judge whether it's a great idea to like leave your keys at home or not. The only way to do it is, um, write down a rule that is once I have my third beer, no matter how confident I'm feeling, the keys are staying, you know, put aside because when you're tired or when you're manic or when you're winning or losing or whatever, you actually degrade your judgment so much.

48:36Chris Power:So you have to write down a series of what I call Ulysses contracts with yourself. Because when you're in this state of sleep deprivation or hype, and actually, actually, I think it's more hype is what kills people. A great phrase a big hedge fund investor told me is, and this is how I think about it is, why do real estate cycles exist? Real estate cycles exist because all new real estate investors are 25 years old, and have not felt the physical pain of losing their shirt on real estate bets. And therefore the cycle continues because they literally don't emotionally feel like they're in a real estate bubble.

49:18Chris Power:Which is obvious because why is the market growing? Because the market is a poor proxy for mania and depression. And the new wave of real estate investors are going like this. And it's not the fact that there's a risk book that says you're over leveraged. It's that you have not felt the gut punch of losing your shirt okay so if you're really serious about it then there's a couple of things there's a couple of rules you don't break and the point of it is you have to write them when you're clear-headed so that when you're tired or over your skis or whatever you don't you don't suddenly decide that the rules is no longer valid because you're incapable of nurses do this very well you know earlier today you mentioned that when you started the company or even three years ago you just didn't know almost anything especially on like the government sales side and how that worked but then you also had this thing where you were like but you just fucking do it sure and how do you kind of in your mind decide what is worth just fucking doing versus like oh i don't know that and i'm probably not going to know that well i think if you are starting the company and let's say that i think a common pattern is you're you're great you're a great engineer but you're terrible at recruiting and the number one thing you need to do as an engineering ceo is recruit but great people and let's say you've never recruited anyone in your life well most people are wrong um you can ask people for advice but the best way to get good at weight lifting is just lift weights okay so so what i did literally in the first 12 weeks of the company and i've never really recruited before is yeah i read a couple of blog posts and then i was just on 20 screening calls a day i was doing outbound i was screening people i was interviewing people blah blah blah and very quickly after eight weeks you pretty much get it um what most people who prevent themselves from doing is being completely okay with it's not failure you're just learning really fast but it's like bodybuilding like if you've never lifted if you've never done a bench press before and you can't lift 200 pounds that's fine you just need to start with 20 and ramp up but for some reason in business whether it's product development or engineering or manufacturing or go to market or recruiting or finance all this stuff is super learnable but people get very afraid that like, I'm not great at this in two weeks and suddenly their whole construct of, are they a good human being or not collapses?

51:55Chris Power:That's point number one. Point number two is if you're a really good CEO, apart from the basics, which I struggle at every day, I think if you're running a complicated company, you have to be the second best person in the company at every function. The reason why I say second best is you want your CFO to be the best accounting leader capital markets you want your VP of talent to be the best recruiting leader etc but if you want to hire the best people the best people report to the best CEOs which means you've got to have empathy for the function and you've got to be able to partner with them on being world class I'm not the best engineer at the company by a long shot but I am good enough that I'm probably one of the best engineering CEOs that people have worked for which means I I can make real trade-offs with staff engineers in real time and make decisions instead of being like, oh, write me a spreadsheet.

52:50Chris Power:Same thing with finance, same thing with M &A. That doesn't mean you're doing it yourself, but you have to be good enough to be able to judge what is a good executive or IC or non. And the only way I've found to do that is, you know, you don't have to be the best engineer, but you better fucking be able to write some level of code to be able to judge what good or bad looks like. it's like who's the best jazz musician like i don't know like not me but i know enough to be like that's world class versus that's like a poor rembrandt you know if you think through like the competitive advantages of a business like this i think one of them is the finance side and being able to just raise unimaginable amounts of money and on good terms how did you kind of think through building out the finance function so that you guys can raise huge amounts of debt and all these weird wacky structures in order to pay for all this build out so um first principles and the first principles of capex heavy businesses is there are only two successful capex heavy businesses um number one is you happen to be uh in a market like early spacex where the customer set is already pre-paying for your build out so nasa pre-paid all their contracts the pentagon doesn't do this retail customers of base power or solar city don't do this it's very rare data centers don't do this if that's not true then you have to architect your business models such that you can access long duration infrastructure credit at some point at some point and you have to be able to be smart enough for predicting that over a four-year time horizons that you can tell your investors where's my cost of capital and duration curve coming down because one of the things you have to subsidize so all right what is engineering engineering is a subsidy of gross margin because you're saying like hey i'm inventing this new um i'm not picking on jordan wire harness business and at some point we're going to be faster and cheaper there is at least a year probably a two year period but that's a lie so but everyone's used to the fact that hey hire software engineers and at some point it's going to be more efficient because you're building more software okay that is a subsidy of your gross margin now as a good ceo you don't want to be more than six months off that because if you say it's two years and ends up being six then you know you're there in us it's the same with capex capex duration and capex cost um so we spent a lot of time saying what are the business metrics have to be and what is the business stability contract duration capex useful life capex uptime such that we would be able to scale factories using low cost of real estate or infrastructure credit and every round you know we started off with venture debt and then equipment leasing and then you know a couple of rungs down but we were underwater on our financing costs until this late series b and the only reason why people let us do that is because we were good enough financial engineers to be able to paint that paint that picture and paint it accurately so if you called someone that may fund you in year four and they said hey if there's a sort of business that looks like this and they're doing 100 million in revenue and the durations look like this you know is this financeable at this rate people that we never met before and a venture investor calls you know someone in a big credit fund they say yeah then there's trust that you'll get down that curve but you have to run the curve like you're running a software engineering curve or a cost of customer acquisition curve you know but it's first principles because it has it has to happen otherwise it's not a venture scale business and eventually the game doesn't work was there some process uh that you had to go through where you were like educating venture investors yeah every round what is that like a story or narrative been like throughout the different rounds well there's like big demand signal what's changed etc the normal venture stuff the financial side was mostly um i would argue until us there is no So I think every venture capitalist that understands manufacturing or CapEx heavy business is invested in Hadrian.

57:20Chris Power:And every round we have to teach people that our way of cost accounting is correct. What is your way of cost accounting? It's nothing fancy. It's just that very few growth funds or venture capital funds can accurately understand a manufacturing balance sheet and how we think about it, which is very different. Not like we're making up our own accounting rules, but we do think about it very... It's a first principles way of thinking about where costs are in manufacturing, where they aren't. And it's correct. But, you know, you're tapping a venture market that's used to investing in product companies like Anduril or SaaS companies like Databricks, both of which carry almost no CapEx.

58:15Chris Power:And manufacturing accounting and gap accounting and revenue recognition is extremely different than any other business model. And it's not that venture investors are dumb. It's just that they've never seen it before. so you have to be better than their lps at being so accurate about your numbers and how you present them and how you think it through they can trust you enough to over time outsource outsource their critical thinking on that domain to you which is a big leap but now when we explain that we're going to scorecard a business unit like x it's largely understood that we've thought it through in the level of detail that if it was a public company anyone would look at it and go like actually that's not inventive but that is the correct way to outside and think about this company very high bar but that's the only way to do it when you think about the imagine competitive advantages are like vectors and each one you know it could be like you get a smaller lead time or like better financing options and stuff for customers or you might be able to get raise capital cheaper or on a different timeline than other competitors if you think about like your competitive advantages in this business and the other ones that could exist but you've decided like these are not the vectors that we're going to optimize for what are those so i'll say two things is lower cost of capital or large quantities of capital are a derivative of being better at executing than anybody else i don't think they're actually good fundraisers or bad fundraisers there are better storytellers but i think people that actually end up with large amounts of capital have better businesses in terms of the things to not optimize for and one of the things that i was wrong about is speed so we started off with the theory that so there are um there are two markets in american manufacturing very broadly at least on the tactical side and one is print no print um and our good friend jim belowzik runs an incredible no print business and there's nothing wrong with it it's amazing we were always in the print business technology's harder it's higher compliance um and we picked our first customer which is a very large rocket company with the theory that they're the toughest customer and if we built a thing they were vaguely happy with, we could sell to the primes, which is a pretty standard design venture pattern.

1:00:49Chris Power:Anyway, it turns out that when we expanded to other customers, the theory being that there were many other growth companies and they would pay for speed at this complexity, high tolerance print level. Anyway, it turns out that two things were wrong. Thing number one that was wrong was there were no other companies in that rocket company of any sort of scale. at least at that time so there were no other startup customers and that everyone else's business model um they didn't care about speed at all and this particular rocket company i think is best summed at is they're the only space or defense prime that can convert speed to value so uh now there's there's an outer limit to this statement which is like you can't be slow but the engineering or manufacturing cost and complexity of being everyone else is at 12 weeks and you're at eight and you're cheaper versus you're at two and everyone's at 12 weeks is odd as a magnitude so most of the market just doesn't value speed the market very much values flexibility scale total program cost um a lot of other things but that was one thing i got wrong in the early days that we pivoted into.

1:02:08When you think about the government side, one of the main inputs to having the contract in two years is having people that trust you today.

1:02:15Chris Power:Yes. How did you learn to build the relationships where the government wants you to win and is like incentivized for you to win? I will tell you how we built trust, but it is an inverse of what you just said. Because what you just said was, we want government contract, therefore I would want to build trust. And I have found in my personal relationships, venture investors, certainly the government, certainly large enterprise customers, that if you are serious about the mission and you are serious about creating value, the first thing you should do is create value and then be so trusted that people come to you with your hardest problems, that if you design a solution that can credibly execute them, you will receive a contract instead of trying to fake trust or credibility in order to win a contract.

1:03:23Chris Power:The tactical answer is we spent two and a half years telling the government in Biden one, we're not ready to tackle any of these problems but here is our advice on how we think you should think about policy and manufacturing to the point of near frustration and it was real because we could have probably won contracts earlier but there was no way we're even close to the line of execution or scale there's a lot of stuff that had to be worked out right and i think when you do that honestly not as a sales tactic because people do this as oh shucks we're not really ready for this you know if you do it honestly it builds so much credibility over the time that when you do say we are ready for this people do trust you and equally we have built a reputation of saying no to many contracts because we don't think they're going to create value to the customer and that is i think a slow way to build a billion dollar company but it is a very fast way to build a hundred billion dollar company like do you want to go on a date and kiss the girl by like acting like you're rich and taking photos with your Porsche on TikTok or are you looking for a wife and we are just the type of people that we're not happy to wait but we are honest actors and by being honest and execution driven you build credibility versus I want credibility and therefore I'm like marketing to credibility.