The biggest mistake defense startups make

Anduril circa 2017 was a highly unsexy company, both for Silicon Valley software engineers put off by its work on the southern border wall and for VCs who saw it as uninvestable.

Now, with the success of Anduril ($625M revenue in 2023), SpaceX ($6B revenue in 2023), and Palantir ($2.23B revenue in 2023), the narrative has completely flipped.

To learn more about the opportunity in defense tech, we teamed up with Sandhill Markets to host a conversation featuring Ross Fubini, founder at XYZ VC and the first check into Anduril, and Scott Sanders, chief growth officer at Forterra and employee #12 at Anduril.

Transcript below.

Our three key takeaways from the conversation:

Adam:

We're telling a bunch of great early Anduril stories, and Anduril—or all of defense tech, is generally having a moment—as anyone who's addicted to their Twitter the way I am, or at least tech Twitter, understands. And it's really fun to hear some of the old stuff. It wasn't—in classic venture storytelling right now, maybe it seems obvious—and it wasn't. It definitely wasn't! from the story that they were telling about where Anduril started, and obviously we have smart people around the table. But we'll get into all that. To kick things off, I’m Adam Hardy, you know me from the Sandhill/Stonks, and from bugging you via email multiple times a week.

We're joined today by Scott Sanders and Ross Fubini. They both know a thing or two about Anduril, both were involved from the very early days. I just got the date a second ago, but Ross called it out like the way I would call out my son's birthday. Very specifically. I think it was 9/26/17? We'll see if he's nodding or shaking his head in the background. But that was around the time that Scott also became aware of Anduril, and what they were building. 

Before we get into that, I'm going to tell you once again who I am. If you haven't been to one of these before, I hope to see you at the next one. If you go to lu.ma/sandhill-markets, we do these on a regular basis. If you’re here, you're probably on the newsletter already. More importantly, Walter from sacra.com, makes us look great, gets me on the phone with way smarter people than I deserve to be on the phone with. I'm always impressed and I'm always thankful. So sacra.com if you haven't been there yet. 

And now, to get into our guests as I kill the first few minutes, and skip some of the intro parts. I put this slide together myself, I don't know if you guys are aware. And most importantly, our first guest is Scott. Now it's not RRAI, it's now Frontera—which we'll talk about, and I put that here, move to Frontera, which was formerly RRAI—Scott was an early employee at Anduril, and saw it from, I think, again, the days of it not being obvious through, and grew in that organization, to being the senior director of counter unmanned systems. Before that, and as a precursor to his experience there, was in the Marine Corps for nine years with five tours in combat zones. So he knows a thing or two about, ultimately, about who these tools are being used by. 

And then Ross, who has a very strong LinkedIn profile picture, he looks very approachable and very VC, is founder of XYZ Venture Capital. He wrote the first check, one of the first checks—we’ll ask him if he was the first check, or one of the first checks—to Anduril in 2017. He was also an advisor to Palantir, and again, investing in defense before it was cool, before you got to call it defense tech investing.

We'll talk with Scott and Ross here for the next hour or so. Ultimately, I have a great list of questions that'll go through everything from how the government has changed their tolerance for US casualties, to go-to market in the defense sector, and ultimately, too, I'm sure we'll get some fun stories from Scott and Ross from the early Anduril days, and the defense sector more generally. With that, I've killed the first five minutes, and I'll get to the more fun part of having Scott and Ross come up. I saw some chuckles back there. Ross, you like that? 

Ross Fubini: I like the venture approachable description.

Adam: Well, yeah, yeah, yeah. I could talk to this guy. He gets it. He gets it.

Ross: I'm a jerk on the inside, so hold tight.

Adam:

Yeah, of course. Fighting for capital, everyone's elbows get a little sharper once you get into the term sheet. That's not the term sheet face that you had on in that LinkedIn picture.

Ross: My partner, Chancy, gave me my literal favorite piece of clothing, that says: ‘allocation is my love language.’ It's my favorite venture swag. I wore it to an event the other day and I had all these people coming up taking photos.

Adam: Good, good, good.

Ross: Chancy nailed it.

Adam: Well, guys, thanks again, and  was gushing a little bit before about Walter from Sacra getting these types of groups together, but I’m super excited to talk to you guys. I gave a little bit of background on each of you, but I'd love to have you both do a better job than I did in sharing your background with Anduril. I’d love to hear about your early connection to Anduril, and also what you're up to nowadays as well, if you're up for it. Scott, are you game to kick us off?

Scott Sanders: Sure. I was getting out of the military in 2017, and wasn't super sure what I was going to do with my life, and through a friend of a friend I heard about Anduril. I didn't know anything about VC-backed companies, and I met the team, and was like, ‘oh, these guys are really cool. I'm not entirely sure what they're doing, but a bunch of guys and gals in a shed with a camera and a stick seems like a better way to do things than what I experienced.’ And decided to take the leap.

Everyone that I knew tried to talk me out of it, and it was probably the best thing that I did when I got out. Getting to work with the founding team at Anduril was awesome. I would definitely not have learned the things I learned along the way if it hadn't been for them. They’re how I got to be here—and now I'm at Forterra, doing autonomous vehicles for defense. I make big giant robots with missiles in the back, it’s the cigarettes of products. It's lethal, it's addictive, and you look cool doing it. So it sells itself.

Adam: Oh, I love it. And as a quick timeline check, how long were you at Anduril? And how long have you been at Forterra now?

Scott: I was at Anduril for just about four years, and then I came to Forterra about two years ago. This is a small engineering company with a 20 year history and a super different approach than what Anduril's doing. I came in at the request of the current CEO, Josh Araujo, who is also a former Marine. We tend to spread like cockroaches, you hire one, and all of a sudden there's like 30 more Marines around the organization. You've got to watch out! He said, ‘hey, there's this great product, but it needs go-to-market, and it needs a way to scale, so will you come build this with me?’ I was like, ‘That's pretty cool. I'm in, let's do this.’

Adam:

Not a bad setup, but now you're in Idaho. We were talking before, you're hanging out at the ranch in Idaho, which is also the testing facility.

Scott:

Yeah, we've got about a 6,000 acre testing facility out here with all the robots. And turns out none of our customers complain about coming to Idaho to check out robots doing robot things. It's like robot farming.

Adam:

It's a strong recruiting copy for the Forterra team. And Ross, I didn't ask you where you were calling in from, but it seems like a mountain-y kind of hangout, maybe California.

Ross:

No, I'm in San Francisco. This is just our room that looks like you're in Tahoe, but very little skiing today. Very little skiing today, here in the city. I'll give my background, I can't let Scott talk humbly without mentioning maybe two things. One, I mean to Anduril is transformative. One of the things we'll talk about is the difficulty of procurement in this sector. Scott will show you the wounds of a different sort, getting some of our very, very largest deals done there, and he's modest in that impact. Secondly—I'll talk about my background—but just Scott, I’m always thankful to talk to anybody that served, and that also goes out to anybody in the audience that served. I appreciate that very deeply. Because I did not do that, and I very much appreciate those actions on your behalf. Adam, you want me to give a little background?

Adam:

Yeah, that'd be awesome.

Ross:

It’s just easy to gush about Scott. He's handsome and smart.

Adam:

When it comes to the marine-to-business side, it's hard to compete. It's like, ‘oh yeah, I was just serving for my country and then I fell into this startup, it was called Anduril, so that went pretty well.’

Ross:

Yeah! My background is actually the opposite of that kind of glory. I'll give you two things on the personal side then move to the work side. As it relates to federal, my grandfather was the Assistant Secretary of Defense under President Kennedy and President Johnson. I mentioned that not to attach myself to him, but to make you so aware that I'm just never going to accomplish the things that my grandfather did. So I was like, ‘I'm living in that shadow.’ But I mean, he really was a tremendous human for the DOD, and his father before him, who was a mathematician. There's an idea in math, the regression of the mean, and my whole family is just drifting closer to the bottom of the ocean, in terms of accomplishments. But it made me really quite passionate about the public sector. I always have been.

My own personal background is I'm an engineer by training, and I worked for Netscape, Tellme, and Symantec, and then I started in venture capital about 14 years ago. Related to that, I just had the dumb luck of introducing the first two employees at Palantir, Akash “Aki” Jain and Shyam Sankar together. 

Palantir, of course, is one of the leaders now, in bringing technology and modern software to the DOD. And that relationship actually directly begat my then connection to Anduril, because I got to know Brian Schimpf through them, I was his executive coach years ago at Palantir. 

Then, the founding team, Trae Stephens, Matt Grimm, Palmer Luckey and Brian Schimpf, those were the four people that came together to back Anduril. And I'll tell you, we'll talk about the pitch, but I would've backed Brian building a toaster company. He's so incredibly talented, and we'll talk about the unique skills he has, but he's just an immense human being and great leader.So I was lucky to back that company early in 2017. It was one of the first investments we did here at my firm, XYZ. My firm has a little under a billion under management now. Everything we do is very early, like with Anduril and other companies like Verkada, New Front Insurance, and a bunch of other public sector stuff that we'll talk about that we're just really passionate about. We get involved really early, and then we've had the fortune to then just put a lot of money into these companies, including Anduril, as they've scaled up. So it's great fun to be part of.

Adam:

That’s awesome. The throughline here is this early Anduril connection, and you were both talking a little bit about it beforehand—and Scott has some old photos that I wasn't on the ball enough to get from him before we got on the call—but I'd love to talk a little bit about those early days. I’d love to hear some stories about what it was like in the beginning of that 2017, with a really strong team, but going out into a sector, again, that wasn't very sexy. Now it is, and now it's more commonly talked about. So, one, the early days of Anduril, any quick tidbits or things that you think people would be surprised by, and two, that transition from then to now, and how different it's culturally perceived. So any comments either of you can share on that side would be great.

Scott:

The biggest thing is that there's a little bit of a revisionist history that has happened over the last couple of years, where people were like, ‘oh, defense tech. It was super obvious. There's this big market, and there's really slow sort of entrenched players, and the technology's not very good.’ But back in 2017-2018, defense tech wasn't a word. People didn't say that. You had Palantir, which was my first exposure to defense tech. I was one of the first Marine Corps intelligence officer users of the product of Gotham—which none of us knew the name of the product, by the way. We just called it Palantir. It took a while for stuff to work, and it only worked because Palantir and SpaceX had body blocked their way through the defense industrial base for the previous 10 years.

If it hadn't been for that succession, I don't think there would've been an Anduril. I don't think there would be the defense tech movement, or whatever you want to call it, that exists today. I've heard multiple people, mostly VCs, tell me that, ‘oh, it must've been super easy to raise back then.’ I'm like, ‘well, I wasn't doing any of that. That was Graham and Brian,’ but it didn't seem fun for them at the time. It seemed hard, and it's come a long way. The reality is, the DOD hasn't changed the rules that much. This is a cultural shift that's happening, and there's still a long way to go on the cultural side. It's still a slog trying to get big contracts through. They're hard to build, and they take a long time, and it takes a different type of capital to build these companies,- at least in my opinion.

Ross:

Scott, I'll build on that in two ways. The first is I was able to participate in that first round, because of the connection to the team. And to say that, ‘who knew in 2017, a defense tech company building a prime was the thing to back?’ Even above and beyond what Scott said, it was not a category. More than that, people hated this category. You don't know how to do the procurement, it's multi-year cycles, you're selling hardware. There's been no successes. Palantir was not public at that point. It was this weird thing that was getting protested in 2017.

This was not a compelling opportunity whatsoever and this is before getting to the point that the team barely understood what they were going to build. We had ideas, but this was not a precision focus, let me tell you, on the drones of the future, and that we we're going to make the best counter drone system in the world. The category was hated, and none of the ideas were known. What was a real credit, both to the team at Founder's Fund that was an early backer, and other folks that came in, is really a belief, fundamentally, in probably two ideas. One is just purely the TAM. Yeah, these things are hard to do, but oh my god, it's so much money, there should be a thing here. Then, secondly, it really was just the people around the table—Trey, Matt, Brian and Palmer—who really get the credit for driving this into existence, and then building the team around it. 

Two more things. One, this continued for years. We looked at putting more and more money into this company very early, because we could tell it was working, and we believed in these people. We talked to some of the very smartest investors who called up the former heads of Sec Dev, and basically got told ‘this will never work.’ I mean, 20 times. I mean, over and over, it was like a classic case of the people you wanted to ask were the people that were farthest from the industry. The most generous view is they lived through things, like Scott said, they've lived through failures of trying to drive things through the procurement cycle. This was not obvious from the beginning, and it wasn't even obvious—I don't know. Scott, what do you think?—three years before people started getting it, like it was working?

Scott:

Probably three years in it was like,’oh, this is working. We're making progress here.’ We were just white-knuckling it every day. Like, ‘oh, I don't know.’

Adam: Meanwhile, Ross is piling on money, which is good on him. Two questions that come out of that for me. One, was there a moment that you can think back to where the white-knuckle went away a little bit? Was there a contract? Or, even internally? And that’s the second question, what was the internal culture and the investor culture like over that three year period? Were you wondering… are we doing the right thing? Are we going to make a lot of money? Was there a moment that you can think back to where either of you were like, ‘oh, I knew we were onto something, but maybe this is that inflection point, where it becomes a little bit more obvious?’

Scott: I think the three obvious ones are the first time we actually got paid for something. Someone was like, ‘I'll actually give you money in exchange for goods and services, with the power program, with CVP.’ Customers seeing it work was pretty impressive, and for an early stage product, it was pretty good. The second one was the $12.5 million dollar contract we got in the Marine Corps, which was a FAR Part 12 sole source. You’ll see a lot of companies today hammering away on SBIRs, and doing CRADAs (Cooperative Research and Development Agreements), and honestly, doing things that just have no path to scale. The odds of scaling a SBIR are like the odds of becoming an astronaut, they're pretty low.

Adam: What's a SBIR? 

Scott: It’s a Small Business Innovation Research Grant. It's like an $8 billion fund per year that's generally designed to enable small businesses to sell to the DOD. There's another version called STTR, and I'll be honest, I don't even know what the name stands for. Neither one of them are good pathways. What we did is this FAR Part 12 contract, which is a traditional acquisition mechanism, like, FAR 12 is how you buy commercial products. It was relatively quick, and we closed that right before the series B. I don't know, time blurs, but those were two big ones. Then we were like, ‘okay, we're actually selling something here.’ The customer wants it, they want more of it. The next one, at least for me, was the billion dollar Systems Integration Partner contract, which was our first ACAT I, the biggest program you could do when it comes to acquisition authority. It was really complicated. It took 18 months, it took going to Iraq and Syria and deploying technology down range, while trying to win a competitive competition. The way the engineering team came together to make that happen was like watching magic. 

The senior engineer for that project, Gokul Subramanian, is still at Anduril, he had this band of 20 people, and it was day and night testing, nonstop, sleeping in the trailers. It felt more like being in a special operations team than actually being in a special operations team felt like sometimes. Closing that deal was like, ‘okay, we can do this. We can compete with more entrenched companies, we can compete against the big defense primes, so we can out cycle them and out hustle them, and build better technology.’

Ross: Dude, we're all going to work for Gokul one day. That guy's a monster.

Scott: I've been trying to do it for years. I just can't get him to do anything.

Ross: You and me both, brother. There are two things I want to add to that. The first is, with that primer, you now know more about defense procurement than basically everyone in venture capital. This whole category is super hot, and Scott will tell you, I mean, God bless them, there's very few people that really grok the complexities of the procurement space. That's no shame on any individual, that's just the category has gotten hot, and people are still learning about it. I mean that truthfully, those of you that are entrepreneurs, the next time you're pitching, find an awkward way to ask that venture person, ‘hey, what does a SBIR stand for?’ And the glaze over…

Adam: They won't be as honest as I was! I was this close to nodding through it too, as opposed to being like…

Rossi: The second thing I would say is, it's really wonderful to give you the venture side—and this really is the truth—of Scott living and breathing and operating the show, and then me being, happily, very close to this team. But the venture people were always sort of inside/outside the tent for  the day-to-day. My equivalent to what Scott said, and it really is the other view of the exact same facts, is what I could see, at the time, was one of the things that people don't want to appreciate about Anduril now. Which is, in my view, what they are incredible at is there's three different stages that were going on at Anduril. 

One, was crazy shit that we could go do. Firefighting humvee, breach device. Scott, you remember all the weird hardware things that Palmer was like, "Wouldn't it be cool if..." and we were like, ‘I don't know, man.’ Then, we got really good at selling stuff in the range. Autonomous helicopter, no idea what it's for, but a three star general is, like ‘Yes, we need $2 million to do that.’ Then, the last phase is things that are scaling, and this is the sensor towers, and then anti-drone, and other products. That phase, if you think about a company that's in a market this big, you could create stuff rapidly prototyping. You could move things to the range, and get paid for them, or do some kind of validation, and then you could really scale up, and now to the tune of hundreds, now thousands of units.

The ability to do that, and the speed with which you could see Anduril was doing that, I thought was quite shocking. That's what you could tell, and then the contracts came behind that. Just to give one other example that's in that exact same time period, I just had the dumb luck as I went out towards one of the range areas. The first day they were doing... Scott, what's the name for the anti-drone stuff? The kinetic attack?

Scott: Anvil. Ross, I took you to that range. I was your handler for that day.

Ross: Oh, shit.

Scott:

Matt Grim dropped you off at my desk, and was like, ‘Attend to this man.’

Adam: The VC's in town. The VC's in town.

Ross: Oh, dude. Okay, well, I want to tell the story. I completely forgot you were there. We went out, and they were literally flying around drones. One thing I know definitively, is they were recording it. So they were basically like ‘hey, fly a drone, smash it in another drone.’ Kinetic attack, fancy word. It's like that was a thing we'd done, it's a very hard problem. That day on the range, they happened to be filming and it worked, and it ended up becoming the video that was on the website, right?

Adam: Just recently?

Scott: No, this is 2018.

Adam: But the release of it?

Ross: This was like an idea. This is pre-Ukrainian war.

Adam: Okay.

Scott: Now, I've got to tell you the story of what really happened.

Ross: Yeah, give it to me.

Scott: Ok, so that did happen, but here's the backstory. It was Palmer, Jason Levin and I, and we were sitting around, talking about this counter US thing, because it was starting to get some traction and people were talking about it. Palmer was like, ‘I bet you could hit another drone with a racing drone.’ Turns out that's really hard to do, even with a professional racing pilot, like an FPV pilot. We went and tried that first, and that didn’t work. Then Jason was like, ‘I bet I could put a small seeker, like a Raspberry Pi on a racing drone, and have it hit another drone.’ And I was like, ‘I double dog dare you. There's no way. That's a legal dare. There's no shot you can make that work. You're not going to have the compute power. There's no way.’

It was probably like a week or two later, and Jason's like, ‘I think I've got it working.’ So we’re like, ‘Okay, let's go to the range.’ And we had a DJI Mavic. Anduril wasn’t a successful company at the time, so we didn't just throw money away. But we had this Mavick, and I was like, ‘Jason, you should knock the Mavick down with the drone.’ And he's like, ‘Well, who's going to tell Matt and Grim that we smashed this $1,200 drone?’ I was like, ‘I will die on that sword for you.’ And yeah, it worked.

Ross: And Ross was in the background!

Scott: Ross was sitting there, we've got an engineer holding a balloon in the air—

Ross: It was somebody who was happily videoing, and my memory is they randomly were pointing it in the right place to set up a shot, and then like, boom, it hit it, and it looked cool as shit.

Adam: Ross is like, ‘so, do you need more money?’ Scott was worried about the $1,200 drone he just blew up. Ross is like—

Scott: No, I wasn't worried about the $1,200 drone, I was worried about going back to Matt Grimm, and being like, ‘So we need a new drone.’

Ross: Grimm's a cheap bastard. Still now, I mean, it's a big office, but every pencil gets counted there. Grimm is the COO at Anduril. To Scott's points, this is hard to put in time, because obviously anti-drone ended up just being so critical to so much of the US/Ukrainian conflict and any warfront now. But this was an example back then, or my experience was, it was similar to what Scott's articulating, the ability to be like, ‘Hey, this could be cool, find, in that case, a solution.’

The point Scott and I were talking about is a little bit hop forward, but actually relatively, in my memory, quickly, we got people to pay and be interested at the range. We went out and won some technical proof of concepts, and then, of course, it went into scale mode quickly, because there was demand. But that ability—we have a couple companies that are able to do that—but that is very, very, very rare. Even if you put your brain on, and think about the number of companies, in the world, that generate multiple products. It's like Apple… maybe end of list? It's very, very, very, very short. 

But you could see Anduril was starting to create this motion very, very, very rapidly. That's what got me excited, and then largely got shot down by still other very traditional investors, who were like, ‘what are you talking about? These things trade at two x revenue, dude.’

Adam: Because the flip side of this, is that, again, defense was not sexy at the time—and not just from a moral standpoint, because there is that moral conversation with some VCs who are like, ‘oh, we don't weaponize tech, we don't do that.’ But then, there was also, from a dollars and cents perspective, this procurement process that Scott touched on. Just from a purely, ‘how do you make money here?’ perspective, was that the bigger blocker? Scott, this is what you did a lot on, and it sounds like Ross is crediting you with a lot of work in that sector. Taking away the moral hazard area, and getting into the actually making money part of it, was that the blocker? How do you get through this convoluted government process where they're used to talking with a small list of people?

Scott: The DOD has always relied on a lot of very small vendors for small component parts, not typically big programs. There's 30,000+ companies that sell to the DOD, so that part's not new. Initially, it was like, ‘okay, how do you build a product that people want?’ That's still a thing that I see in this space, where it's like, ‘oh, we're going to go do this thing over here. It's a really cool tech idea.’ And it’s like, ‘that doesn't contribute to the things that DOD directly needs, so that's probably not a great bet.’

But the way Anduril’s founders approached this—and then especially bringing in Matt Steckman into the company, it was like mid '18, mid late '18—it was about having the framework for how you attack these bigger problems, because those little contracts can be really meaningful. You're not going to go get a $10 million contract as your first contract. You're going to get a $100,000 contract, and you're going to build trust with the customer, then you're going to do two and a half, then you do five, then you do 10. Then you can do a hundred, then you can do a billion—there's progressive steps. The DOD, and the way the acquisition core works, they're just not going to go YOLO a big chunk of their budget on a thing that they don't know, with people they don't trust.

It looks a lot more like enterprise sales than it does, really, anything else, if you're in this space. You have to have a couple of key things to make it work. If you want to build programmatic scale, you need tech that is highly differentiated. If you build a 20% better tool, the acquisition friction is just not worth it. It takes so much time and so much money to get through the acquisition process. Why would you? It's 20% better, who cares? You need to be a step change different, you need to be half the cost, or twice as good, or whatever metric you want to use. 

When you look at counter drone stuff, it was a pretty open space. You've got to have that. You've got to have someone on the inside, not a general officer, not the Lance Corporal. You need the program manager, the Lieutenant Colonel, the Major, the person who's going to be in that seat for a while, to really buy into what you're doing, and buy into you. It's as much about the trust that you build as it is the tech. Because I hate to break it to everyone, highly differentiated tech doesn't actually move the needle in defense. It's how you go capture it.

You could build the best thing in the world. The joke we used to tell was you could build a six generation fighter, that’s a quarter of the cost of the F35, park it in front of the Pentagon, it would still take them two years to buy the first prototype. That's just how long it takes. You also have to be building a piece of tech that directly contributes to the national mission. So you look at what is the priority right now, it's denying an invasion in Taiwan, it's defending the Asian Pacific region, and it's bolstering NATO's Eastern flank.

Well, a lot of stuff that people go build doesn't directly do that. So it might find a small home, but it's not going to go find a big programmatic home. That's why we are so bullish on our approach to defense tech over here. What you want is a lot of robots, because I don't want to go back into combat, and no one I know who's served wants to go back to that either. I want a robot to go shoot the missile, because people shoot back. I don't want to walk 50 miles to put a sensor on top of a mountain, I want a robot to go do that, or a drone. That's the only way we can fight with scale.

If you don't have all these things lined up—and even if you do, even if you lay out this plan perfectly—you've got to deal with Congress, and you've got to deal with… how is the FAR going to buy your thing? Is it a true commercial product? Is it a non-developmental product? The list goes on. Without the experience that Trey and Matt and Brian and Steck[man]—and then when Brose, he came on—had, you wouldn't have been able to do that. You need a bunch of cycles, to understand how this process works. It's not something you can pick up in two months. It's something you pick up in two years, or four years. and it's hard. It's a hard space to work in.

Ross: Building on that, and you can hear it in the complexity, the two problems viewed on the venture level—I still, by the way, can't get over the fact that Scott was the guy at the range. I'd totally forgotten that. It's a big part of my mental experience.

Adam: What a great story. Yeah, what a good story.

Ross: Yeah, it really was. One is the complexity of procurement, and this is still the thing today. A hundred percent, it's the biggest challenge, by such orders of magnitude of every single business. And Anduril's done it, Forterra's doing it, a small number of businesses are able to do it, but this procurement process is just such an incredible hurdle. Then that, combined with the fact that there weren't instances of success that were really visible. Again, Palantir wasn't public, SpaceX, it was known it was making money, it was well-marked, but it was not as visible, and the DOD line wasn't being as drawn as clearly. So there weren't those midpoint examples.

Then everything at scale had been consolidated down, and it was time and material structure. There's a whole rabbit hole we can talk about, about the wonderful complexities of procurement and about the complexities of multiples. One of the reasons those multiples are so low, for example, is because the government is funding a lot of the development process for those projects. That is, someone shows up with a cool prototype, the government signs up, says, ‘Great, we'll give you a half a billion dollars to make this real.’ That's why the time and material structure came in, and time and material, ten percent, that's why the small number of players in this market trade that way. 

Versus a Forterra and Anduril, Cape, and other new solutions that are coming online, when they are that—not 10% but truly 20 times better solutions—robots, not humans in the field, a true security innovation that just does not exist, etc. Those things show up as products, and therefore, they're able to get higher margins, because they're not being funded by the government, they're buying them. But again, all that's incredibly hard. For example, the government's terrible about buying software. It's awful at it, but it's really good at buying hardware. So you have to do all this work so that you can package things, so that hardware can get bought as software. And then, don't even get me started once you start working at scale, talking like, ‘well, which congressional districts are you actually in?’ The rabbit hole gets deeper, and Benjamin's like, ‘no, thank you. You have a data warehouse that I could fund. I hear this Snowflake is a big deal.’ Why would you bother?

Adam: Yeah, yeah, yeah. No.

Ross: One thing about this that really changed, when it comes to why we bother now is the situation with Ukraine. That radically changed things, that, and all the terrible things that have happened since then. That made the mission need sharper. Engineers were drawn to the challenge, and it wasn't just folks that served, like Scott, it's hardcore nerds I know, that are now, ‘I want to work at Anduril.’ Because they feel the mission component. And the government is still crappy at buying stuff, but it got a little bit better. A little bit. It still needs to be writing checks in the orders of hundreds of millions of billions, and it should be making Scott's job easier, but dead serious, it should be making everyone that sells innovation technology’s job easier. But it's still way too hard.

But all those things got better. Anduril's success, Palantir are going public, a few other things, and then suddenly, people believe that you can make money, engineers want to work on it, and the buyers are more forward-leaning. Then you get the hype cycle we're in now, you get the enthusiasm that people have right now in the market. And those things all change, at least that's what I see. I don't know, Scott, if you'd agree with those shifts, or others.

Scott: No, I think that's super accurate. I mean, it's good for people like me who can't code that it's really hard, because we have jobs. If not, I'm just going to open up a barbershop in Ketchum, Idaho, and cut hair or something cool, be a ski bum, or whatever it is my calling in life is, besides banging my head against my computer every day, going through the acquisition process with our team here.

I think the other thing that a lot of companies probably didn't understand, was that if you try to take a playbook from a Raytheon, a Lockheed, an SEIC, a traditional defense company, you can't run that playbook if you're making these big bets—and they're all big bets. If the sentry towers didn't work for Anduril, that would've been pretty bad. You need to have a core anchor product that you can prove out to the market. Yes, you know how to do this. You know how to scale it, you know how to build it. But by the way, the DOD is really bad at buying software. 

But they're really good at buying hardware. They also need a lot of hardware, because at the end of the day, what is the military designed to do? It's designed to go fight and win the nation's wars, and that means you've got to go somewhere and do something. Now, there's a lot of things that happen ahead of time, that you can do a lot with using software, space-based sensing, and all kinds of cool stuff that's really important. But when you think about the end of the day, someone's doing a frontal assault somewhere every day. That's humans for you, we're fighting each other everywhere, kind of all the time. You fight with hardware. So you've got to make some pretty expensive bets on what your technology is going to do.

And to Ross' point, that gives you an advantage when you're going to sell it, because you're not going to the government and saying, ‘hey, fund all this.’ But, there's a reason those margin profiles and those multiples look the way they do, because you're not taking risk if the government's funding it. Well, if you're not taking a risk, last time I checked, we live in a somewhat capitalist society, you don't get paid. The downside is, if you go use your own development dollars to build something out and no one buys it, you're up a creek. It's going to be really hard to recover from that.

So you have to be really smart about where you are going to go make bets. You can't make too many, you've got to make more than one. What does that look like? Know when to pull the plug on products, which is hard, because they become emotional things that you build from the ground up. And when you finally decide, ‘you know what? This just isn't going to scale.’ You've got to know how to pull the plug on it, or else you're just going to be hemorrhaging cash on things that are never going to generate a return. That level of discipline is, again, back to the founding team that put it together, really important, and not a commonly understood concept throughout the market. At least, from what I see today.

Ross: Yeah, I agree so wholeheartedly, it's one of the reasons that I just don't think there's another Anduril to build. I think there's a Forterra to build. I think there are companies we backed, like Saildrone, which is a great business, from what I've seen. I think Cape, Turbine One—those companies we backed, because you love your own children—we think they're good businesses. But they're not trying to build the same prime in the way that Anduril was. They're running different playbooks.

They're starting with different customer bases, the same way Anduril's first customer, Scott, correct me if this is wrong, but it was Homeland Security, right? It wasn't DOD for many years. That's just an example, where it wasn't the full-frontal assault, so to speak, against a prime, that you've got to find your own unique path. We see a lot of people that are now pitching themselves as, ‘we're going to be the next Anduril.’ I'm like, ‘you don't realize how hard that was, how hard that is to do.’

Scott: Yeah, and how big of a moat you now have to go through because they're there. You're not going to compete with that mass, and that user uptake, and that's really hard to do.

Ross: We do a lot in this space. It doesn't make me negative on this space, but it means that your next thing, it maybe rhymes with Anduril. The next success maybe rhymes with Anduril, like Anduril rhymed with SpaceX or Palantir, has some of the same people, attributes, and skill sets, but it's going to be different. There are going to be more, because for our nation, we need more solutions, but they're going to be different.

Adam:

Yeah, let’s jump into that rabbit hole a little bit. Because there's this moment, and you mentioned some of the hype cycles around this. In El Segundo there was the Gundo goons who are getting a lot of attention. The Washington Post, I don't know—Ross, you're looking at me like you've never heard of this. You've never heard of this?

Ross: I literally don't use Twitter. Well, it didn't matter before. I literally don't. I literally don't.

Adam: Now, Twitter loves your stuff. You'd be super popular on Twitter now. Everyone who would have given a shit, like recently, but now—

Adam: But there's this cultural moment happening around defense—and I think that you mentioned, in something that's been building for a long time—engineering talent, that's really interested in solving hard problems, and attracted to rocket launchers a little bit more than they're attracted to B2B SaaS. Which seems obvious now, but actually wasn't obvious, not that long ago.

Ross:

A hundred percent.

Adam: And there’s a lot of VC money now flowing in also. LPs want it, it's something you can do, like, ‘okay, there's now wars.’ There's a quote in a video clip from a John Cogan video that he did about Anduril—which people in the audience, if you haven't seen it, it’s really great. But Palmer was sitting at a web summit or something like that, and talking about how there was a period, not that long ago, where people thought that large scale conflict was over. Like ‘oh, everyone's too intertwined to go to war.’ And clearly not true, now unfortunately. Scott touched on it, as part of the nature of humans, blah, blah blah, that’s longer conversation. But, there's clearly a need for these products. There's a cultural tailwind, but you both mentioned a lot of things that people are doing wrong.

What do you see as a really common misunderstanding or a really common mistake? Again, El Segundo had that hackathon for products, and it's like, ‘seems fun.’ But then you talk about procurement and I'm like, ‘this is so far away from anything that actually gets used by someone on a front line, because of this cycle.’ So, what do you think of common mistakes, on both the operation and the VC side?

Scott: I got an easy one, and hopefully Ross agrees with me. One is that the DOD is not a unified market. It's not. You can have two program managers across the hall from each other, and they have completely different acquisition strategies about complete different things, using different colors of money. It's a conglomeration of micro markets, and they're all very different.

The second part of that is understanding the DOD budget. They're called J books. They're public, office of the comptroller, look them up. 36,000 pages of PDF, please, someone, build an LLM that discombobulates them into something useful. It's been tried. There's a lot of people trying, it's not accurate, just make it accurate, but give me an Excel. I'll CTRL-F my way through it, instead of the time honored tradition of when those books come out, around this time every year, and you have your analysts sit there for three hours, just downloading PDFs and smashing them into one gigabyte sized PDF. Also Adobe, why is that the limit? Why does it work?

Adam: These are good requests for product.

Scott: I've got some requests for products here. But understanding where that money is, and how big it scales, you hear a lot of folks saying the most wrong answer—$800 billion defense budget. Nope, that's not how it works. You have roughly $150 billion that goes into procurement funds. You get about $140-150 billion that goes into RDT&E, and you've got some O&M stuff over there, and then there's MILCON and MILPERS, but none of that matters.

So you're actually looking at two types of funding markets inside that market. About 75% of that goes to traditional, large acquisition programs: building ships, building missiles, building trucks, things like that. So actually, your TAM is still big, it's just not as big as you thought. And if you make a single threaded product—I'll use my own example here. We build ground autonomy systems, robots. We put them on vehicles so that you can have a large robot and a small robot go do things that you know you don't want a human driver to do.

But when you start boiling it down, and actually looking at those program execution lines, the PE lines inside the J books, you can very easily come up with a reasonable, addressable market. So for us, looking forward to the 2025-2026 forecasts that are in the books, it's roughly $4 billion. That's it, right now. It might grow, it might go down, err is about six to seven, six to eight, and 95% of that goes to UAS, that are group four, group five or black programs.

If you make a small drone, your addressable market inside the DOD today is in the hundreds of millions. Not billions, not trillions. And the DOD doesn't want to get vendor-locked in a single solution. You're going to have competition, and that means you need a pathway to scale. There's two pathways you can scale. You can either go be a multi-product company, which is extremely hard, extremely risky, and requires an incredible amount of capital, and, like, magicians of a founding team to go put together.

Or, you need to have a thing you can sell commercially, and go tap a bigger commercial market. That's kind of it. If you're just building a product that fits in a very small niche part of the defense budget, and you don't have a plan to go either direction, the size that you can become is very, very small. And if you're raising at venture multiples, you're killing your ability to get acquired at some point, because those companies that are valued at 2x revenue are not going to come buy you for 50x revenue. So what's the exit plan?

Ross: So it's Scott Sanders, S-A-N-D-E-R-S. I shit you not, this is amazing.

Adam:

This is a masterclass, yeah.

Ross:

This is a TAM breakdown. Scott can do the same thing on procurement so articulately. This is the challenge, as you laid out so well, in basically building a single product, as an example of a challenge that exists there. And again, lots of venture people know this, lots of founders know this. Maybe it's not widely known, or it's something to be aware of and be navigated, and have a plan around, like you said, and have a platform approach. I have to caveat, because I want all these companies to succeed, I want founders to succeed, and I want there to be new capacities and capabilities. But I think it's a great example of the true TAM of something. 

Something, Scott, you just mentioned, we're big believers in dual use. One of our companies, Airspace Intelligence is one of the dominant providers, emerging dominant providers of technology for large airlines. Airlines are a really interesting market. They've solved a really hard problem. But the market expansion of going out and doing things for the FAA, for air command, for our allies, and basically addressing logistics and movement and observation of our skies, is an opportunity that group can go after. On the more traditional security side, we have another company called Manifest that provides software bombs and building materials. They sell all over the government, including to the DOD, and they also sell that same functionality to government contractors, but also to consumer businesses and enterprise businesses.

Dual use is an obvious and big one, and I would add real weight here in my comments, because again, this is something that's always been true. Federal government is one of Cisco's biggest sales channels. But, historically, when you're setting up a new company, you're building a new Splunk. Let's pick a real idea. The government would not be on your list until year four, or year 40, of the business. And now, I think the possibility of going and getting that half a million dollars SIBR, $2 million SIBR, can work, Scott. Gosh, I hope so. It becomes a $12 million stratify, or you're finding another program, or some equivalent acronym where you find an entry point, you can build a dual motion. This is hard to do.

That dual motion, it actually gives you, usually the dollar slopes show up slower but larger, initially. So it can be both a good go-to-market pairing, but it's hard. Because you now need a Scott, and you need someone that does product-led growth, or core enterprises, or you need Scott to do both of those things. Again, it's not easy, but we see a lot of success doing dual use, where it's applicable. Some things aren't applicable. We should talk about who in the private sector can buy robots with missiles.

Scott: Ross, I'll sell you one.

Adam: I know there's plenty of guys in ranches, not so far from Scott right now, that would love to have a rocket launcher on their property.

Scott:

You can't buy the missiles, but you can buy the truck. I mean, that's our approach.

Ross: Can I buy the missiles separately? Is this like I could buy the pistol with the other—

Adam: It'll be like a Nerf missile, they'll put like a Nerf missile on it, for the dude ranch version.

Scott: No, but that's our approach. You have a single unified hardware and software product, and whether it's on a JLTV or an FMTV, or some acronym, military vehicle, monster truck thing, or it's on a Mack truck, or a yard truck, or something like that. For us, it's all the exact same. Whether you're backing up a missile trailer, or you're backing up a 40-foot shipping container, it's all the same stuff.

There are very few markets that can truly do dual use, where you can lead a defense first one. I think cyber, situational awareness, maybe hypersonics, ground autonomy is in this, but I don't think there's a ton. If you're building small drones, or hyper-specialized products, you might be making a gamble on dual use. When it works, it works. But there are very few really good examples of it. Aerospace Intelligence is one of those, that it works.

Ross: Airspace is years in on doing it, they work with airlines and then came over to the defense side. We’re seeing other people that are doing it. They're just less defense companies, as you said. They're like, ‘we make a better database, we make a better…’—I'm trying to think of ones we can talk about publicly— ‘we make better products.’ Yurts would be a great example. Yurts is a company that basically makes it so that people in service can use tools like OpenAI. They can use modeling. That's an example. I'm always careful, you don't want to talk about your book too much, whatever. I'm sure there are other great companies, but these are the ones I know. They've done a really good job in the public sector, getting pulled into defense. But to your point, it's software that's applicable across both use cases.

Then, there's a nervousness there too. If you've done a good job at your defense use case, are you the best at the enterprise use case? And that's a tension, but I think it's navigable. And again, I just think it's a new thing. Literally, this did not exist, in my experience, before Anduril, this awareness. A lot of it is because great people are coming out of service, and I know I keep shining on Scott, but quite seriously, people like Scott are coming out, and they can operate in this way.

You cannot understate, in my opinion, the importance of Palantir and the four deployed engineers. The Deltas and Echos, is our vernacular. That skillset, we've plugged into many, many, many companies, and they have the ability to then help the aerospace companies of the world do dual use.

If I could say one other thing, in case there are actual entrepreneurs also listening, and a really important thing is, venture people really and truly do not understand this stuff. It's because it's hard. 

As Scott mentioned before, we spend a lot of time thinking about how to present government contracts in a way that people can get comfortable with. I'm going to run 10 small contracts for $2 million each. We're going to win them all, or we're going to work hard, down to the coverage, we're trying to get 20, and we're going to get eight, actually, not 10. And a year from now it’s going to become four to five $15 million contracts. And then another year or two years out, a couple of those are going to be systems of record, it could be $30 million.

I'm just making it up. But that's very different from ‘we get 50k, we're going to get $250K and onto a million." It's a different articulation of the market. Scott, very eloquently put earlier, you're trying to explain how the government buys, in a way that maps to how investors think about it, which is, you're mitigating risk at every step. But it's different than just calling up a customer and saying, ‘hey Adam, you spend $15K on your podcasting software, will you spend $50K next year?’

That's a different process than the budgeting process, and entrepreneurs need to get better at articulating that, and venture people need to get better at understanding it. I'm still in the JV lanes of it, and I've been doing this for a whole bunch of years in this little category.

Scott: The sales cycles are very different. Maybe you can look like a SaaS sort of revenue profile at some point, when you're in a major program, but the fastest you're going to get to a major program is probably 18 months to two years. They take a lot, and I think the risk, to go back to the SIBR, stratify, TAFI, and small business pathways—and look, there are great products that have come out through the SIBR program—but just look at the volume of inputs to give you the outputs. The odds are low.

The SIBR programs specifically, if they're not done right, they can take a lot of your manpower, or people power, sorry. If you're focused on, ‘I need to go do this motion and build this thing,’ and the SIBR is pulling you this direction, because you signed up for this $1.8 million thing that you need, because you're a year-old company, and you've got to show revenue traction, you don't really have the ability to just be like, ‘well, I don't want to do that.’ 

You lose the ability to say, ‘well, I want to build this great product, because I think it solves this problem.’ You're getting contractually locked into something you can't walk away from. And to spin those down, to focus on a product-based approach, is really, really hard, and you're going to tear engineering apart doing it. And if your business teams aren't partnered at the hip with engineers, then this doesn't work.

This is really a Palantir-ism that's echoed through the network, of just how you go create the programs, and how you partner really good engineering with the capture of business development, growth, and whatever non-offensive sales term you want to use, to call your salespeople, to go get these products in those spots. You also need salespeople that are technical enough to understand the product, because you're selling to a relatively technical buyer.

When I buy Salesforce, I don't sit there and ask them how they're using Kubernetes, or something in the background to make my dashboard work. I don't care. It's, ‘make the product work, and please stop charging me an arm and a leg.’ That's not the way it works in defense. You've got to be able to understand the basics of how does a radar work, how does the LIDAR work, how do drones work, how does airspace work. Oh, and by the way, you've got to go put these things in an operational environment, for which there is no regulatory framework in a lot of cases. And you've got to do it safely.

The DOD’s got its own way of doing safety things, like DO-178 on the aircraft side, or 882 Echo on ground systems, that are huge bodies of information that you need to become a mild subject matter expert on. You've got to hire the right people on the right team that can do that, because it's not just a one-person show. I was really lucky, and got to hire incredibly talented people here at Forterra that are technically-minded, knew a lot about defense acquisitions, came from places like Vannevar or Palantir, and then put together a playbook, as a team, that you can run. Because the playbook for every company is going to be different. You can't just go lift and rip somebody else's.

Adam: Sounds hard. Sounds hard. My biggest takeaway, I think, so far. Culturally, it's awesome to see more attention, but it seems to be very much outside of the classic venture scope. I invest in YC pretty aggressively every batch, and I just went through that batch of founders, and these two person teams, and some are really young, coming out of whatever, doing new software things. 

And I wonder too, and you mentioned the Anduril, rhyming with Anduril, and it sounds like SpaceX, Palantir, Anduril have, you mentioned, kind of elevated and made it possible in a lot of ways, but that team has so much success in their past already also. Like Palmer and everyone else around that table were serious building people, with track records of doing things, and I wonder about the ability for small, scrappy teams to the product, to draw attention. It seems like it's more like super team biased. It’s more like lots of people coming together to get something across the finish line, rather than two people in a garage tinkering to make something work.

Scott: You're not going to get a WhatsApp-like exit in defense. You're not going to get seven people in a garage, build the thing, kick it over the edge, and all of a sudden you have a couple million monthly active users. The superteam thing is right. I know folks like Max Wang are listening right now, he's one of my favorite early software engineers at Anduril. Everyone should ask him how he tried to cut his own hair one time. It's a great story. DM me for photos.

But you've got to have this team of super driven people, and you've got to create a culture that says, ‘we are going to hit the gas until the mission's done.’ That might take five years, that might take 10 years, and it's hard. We've got a test range out here in Sun Valley, Idaho, where there's 20 engineers, techs, random people down there today, testing robots. It's 20 degrees outside with snow. They've been out there since 6 AM, they were out there this past weekend. They're going to do that, day in and day out, until this next batch of software is released. That's a different breed of person than, ‘I sit behind my quad screen monitors. If my mouse doesn't work today, I quit and go home.’ I know it's a vast generalization, but it's just a different thing. You've got to want to do it, because it is hard, and it's physically hard sometimes. It's definitely emotionally hard.

Ross: I’d like to put two things to that. One is, I've got a sad update for everybody that's a company builder, which is like, it's all goddamn hard. We've got a procurement company that's killing it, a legal tech that's killing it—and that's all hard, and that's going to take 10-15 years to matter. Don't tell my LPs, because the DPI is a ways out. It's all hard. Now, some things are more physically hard, rigorous, out in the snow is different than being behind the keyboard. But I think the one thing is, the awareness here is, it is just the difficulty of being a founder, universally. 

And the one thing I was going to say—and this is because I want to have skepticism, but I don't want to beat the joy and belief out of people—is I do think it takes the super team. That's a great phrase, but you can go build it. There are people that have savvy, and actually, we talked about this. Palantir is full of young people that are learning this skill set, or even not that young, or maybe had that as their first job, that are being trained up. Palantir famously has a lot of very young people that come through, do great work, and are then able to understand procurement. I'm sure SpaceX has the same, I'm sure Saildrone and post-IPO, Forterra will have that.

There are people that are getting trained this way, and then those people can build their super team. That possibility exists. I do think there's naivete, and then there's just awareness of ‘we're in here for the grind,’ and you can get good outcomes. The last thing, just because we’re talking about defense, is unlike things that end up economically, like, ‘yeah, we're public, we're a great MarTech,’ and those are terrific and we'll back those. Max, it's [email protected], when you're ready to start your company. 

But when you're really on the mission, there is a lot of joy in that. We have 107 companies in the portfolio, I love them all, but the ones that are doing stuff, that are keeping rangers safe, keeping our enemies at bay, those are my favorite, because they're mission-aligned, and they matter. They matter a lot. The messaging ones are fun, but they matter a lot to me.

Adam: We're coming on time, I don't want to keep you guys too late—I know that Scott needs to go do some rockets, and Ross has to look at some pitch decks. I've had conversations, recently, with some young engineers. People who are on their way out of school, like super, super bright kids that are interested in working hard, are interested in these spaces. And again, because of this cultural moment that's happening, one of the most exciting parts of it is giving young engineering talent a new perspective. They're opening their eyes to ‘oh, I could go work on something that’s keeping another human safe, who's fighting those types of missions and stuff,’ like you said, Ross. That’s super exciting to young people and young engineers.

The thing that’s different, though, in this, to paraphrase a little bit, in software there’s this ‘drop out, just start something, start screwing around, get out there.’ The YC mantra of talking to customers and going for it. Whereas what you guys are saying a little bit more is that in defense, or maybe in some of this hard tech stuff that needs some scale, that needs some super team, that there's a path of going to a place, like a Forterra, like a SpaceX, like a Palantir, like an Anduril, to get trained up in this space that is very, very deep and very, very different than other places. It sounds like, generally, it’s not as open as just internet software. Maybe that's an obvious statement, but there's this path of there, that’s hard, but you can learn it, and you can learn it within companies that blaze the trail, and are looking for exciting young talent to come learn, and are supportive afterwards.

Scott: I mean, I got an economics degree from a middling state school at best, so if I can figure this out, I think literally anyone can. You just have to love pain.

Adam: Easy for the Marine, you’ve got to love pain. But again, there's a lot of people that I think are drawn to that. On the software side, there's a lot of software development that is pain.

Ross: Anybody that's a founder, we all know this, or just on an executive early stage team. It’s all a force of will. It’s all pain. I mean, it's not literal. Some people live through literal pain, but it’s just like you're grinding the venture. Whoever doesn't like you, the customer doesn't want it, the fucking engineer that quit. It's all hard.

Scott: It's all hard.

Adam: That's a good takeaway. Defense is hard, but it's actually, it's all hard.

Ross: It's all hard.

Ross: But the other thing is really boring.

Scott: Yeah. It is.

Ross:

This is why Scott isn't a ski bum. You should become a ski bum, that'd be awesome.

Adam: Well, I don't want to keep you too much longer. I know that we're over the 10 minute mark. This was really, really awesome. I really appreciate you guys, you jumping on. For everyone who was here live, I know for a fact everyone's been enjoying it. I've already got some texts from people asking how I got ahold of you guys, which is thanks to Walter, and some super fun old stories from the Anduril stuff. I'm glad that Scott, you were able to tell Ross the background on his moment, on his like, ‘holy shit, we've got to get more, we're putting more money in.’

Scott: I'm going to get yelled at. This is a great one.

Adam: Uh oh. But, yeah. This was really, really, really exciting and awesome. And again, we'll cut this up and make everyone sound smart, and pull the key takeaways out. And thanks again, and I'm sure I'll bug you guys soon for something. Maybe get some old stories again. Maybe get some of those photos, Scott.

Scott: Oh, I've got some great ones. As long as you can fix this and post, we'll be good.

Adam: Yeah, yeah, yeah. The handsome Marine. Yeah, the handsome Marine. Well, awesome guys, thank you. Jesse's pulling us off. Thanks again. I'll send stuff over email, and we'll talk soon, I'm sure.