Transcription
Grant Demory is a former US Army officer and the co-founder of One Brief from YC Summer 21 batch. They're creating software to help dramatically speed up military planning.
"What we're doing is on a per dollar basis the most important thing happening in defense tech in the world."
Today, we're going to talk with him about what he's learned selling software to what some might consider the most complicated, hard to crack customer in the world, the United States military. Thanks for joining us today.
Thank you very much, Brad.
So what is One Brief?
So we make software for operational military planning. Every military headquarters around the world, from a giant combatant command like US Indopacific Command down to something small like a battalion. Um, almost everything they do has a has a plan that could be very detailed, like how do we defend Taiwan, or you know, something uh a bit more casual, like what are we going to do tomorrow? And the process, it it takes a long time. It drives almost every military decision, and if we can make it faster and smarter, well, it has a huge effect on how well the military does.
And so you guys have been at this for a few years now and you've started to achieve like pretty epic scale. Um, what can you tell us about like how big the company's got? um valuation, revenue, size of the team.
To put it in perspective for everybody today, our our software is in use in quite a lot of military headquarters worldwide. Um, we're growing usage uh 2.75x quarter over quarter right now. Um, which translates to about a 57x um year-over-year growth rate. And uh, we're at about 144 people um, and we just raised a Series C at a uh $650 million valuation.
How have people in the past or up to this point done military planning from a technical standpoint?
Well, almost all Word and PowerPoint. There's actually been a huge number of attempts to make it not be Word and PowerPoint, but uh, those products are pretty sticky. So, if I'm trying to make like a a a giant military plan, I'll often make hundreds of slides um, and I'll, you know, email them back and forth across every contributor, which, you know, could be 12 different headquarters, and I'll ultimately produce this uh large Word document called an Operations Order, or in in other situations, a base plan.
Tell us a bit about how it works. What's the actual user experience like for people using One Brief?
The overall idea is how fast can a team across a bunch of different military commands um get to a plan, whether short-range or long-range, or a mediating crisis, make it be really fast. We built this system around cards, which is probably really familiar to you. And you're using a bunch of reusable cards where a card might be, you know, a fact about the situation, um, might represent a military unit, might represent a task, and you're using these cards to build all your outputs, like a sync matrix, an Operations Order, which is the the Word document, or maps that present these cards on the map, or all sorts of other outputs that are using the same the same blocks of data. And when you change one, it it updates, you know, all the others. And, you know, you're embedding portions of one document in another.
So, the largest type of military plan is is called an O plan. Um, these are for the the really big ones. Think Defense of Europe. Um, those take about two years to write if everything goes perfectly. You bring One Brief into this equation. What's uh, you know, are we able to compress these times down from two years? Is it a much simpler document? What's the benefit?
The one I'm most proud of, um, there was a course of action that came out of a major plan against a core US adversary, and the, uh, the planners are presenting it to the Chief of Staff at a major One Brief customer. The Chief of Staff says, "No, no, no. This course of action is terrible. It's not what you're going to do. You need to go change it." And that normally would create like several weeks of work. In three hours, they had changed the entire thing. New map overlays, new sync matrix, new written base plan, new slides, presented it, commander signs it, no further changes. That was using One Brief.
So where did One Brief start and what were you doing at the time?
I was a uh US Army officer. Um, had the idea as a very junior lieutenant straight out of West Point. I was in Liberia part of the 2014 Ebola epidemic. This was my first military experience, which actually turned out to be to be really good to like see this with fresh eyes. I was actually working Ebola data out of the Ministry of Health, but it was attached to their J5, which is responsible for military planning. And that meant I I got to see how military planning at a large headquarters worked, which uh, of course, I thought like, ah, this is so cool. I'm I'm really happy to see behind the curtain. But I ultimately felt that it could have been a lot more effective than it was. We're watching people like pass PowerPoints around the headquarters and eventually make decisions months later than we could have made the decisions. So, okay, this is this is really not optimal. I was in another headquarters a couple years later. Um, this time I I think with more informed perspective, having seen this once, I was a little bit more senior. Um, this was CJTF-OIR, the theater headquarters um, out of Baghdad for Iraq and and Syria against ISIS. And I saw other operational planning teams there and said, well, this is this is a pattern. Um, this happens in a lot of places. This is, you know, the core driver of military decisions. And I think fixing military decisions is is the most important thing we can do in defense tech. So, I'm gonna get out of the army and uh try to fix that.
One thing that I think is interesting here is uh for people that maybe have an idea, you have this genesis of an idea of something that I could build and maybe sell and and provide back to the military. I think it's pretty intimidating to a lot of folks. Maybe there's other people out there that have had ideas like this, but they don't even know where to begin. Talk to us a bit about how you understood or mapped the opportunity such that um it wasn't just like selling to a giant monolith. We're going to sell to the army. Um, I think it's actually a lot more nuanced than that and there's many different ways that you can kind of work your way into the into the um the defense ecosystem.
Yeah. So uh, basically every military, and especially the US military, is a huge number of different organizations that control their own budgets and and these are often like, you know, eight or nine, and sometimes 10-figure uh, you know, budgets for these individual headquarters. And in the US, there's there's about 1,200 commands that that need our product. Basically, what this means is that you don't, at least initially, have to sell to DoD as one giant customer. You can actually sell to the individual command that most needs your product, which is so much faster and easier and keeps you a lot closer to the end user.
How many people at a given headquarters do you usually have to actually present to or convince?
Well, our first deal is one. I now these are, you know, pretty standard enterprise sales where there's, you know, a committee of people who need to, you know, be persuaded and stakeholders, but I don't think that's helpful to say a very early stage founder doing defense tech. We were able to get the initial sale on a government purchase card. So there's these basically credit card, I mean, they are credit cards that uh a lot of military officers and NCOs have um if their organization has something of a slush fund, and they can buy things up to, depending on the place, sometimes it's up to $25,000 and just buy it. Very little process, very little contracting. And it's a bit of a hack to avoid the contracting process because if you're coming in and and don't know how to do contracting, um, you can just get in the headquarters. So it it actually helped us a lot by closing a small sale in, you know, relatively little time, just to get physically in the headquarters and be exposed to users.
Yeah, that's a great hack. We're always telling YC founders um to find some simpler way to validate that people will actually pay for the product that they're building. And so instead of going through a multi-month procurement, you're able to get one person to make a decision and put it on a card. When you first set out to start working on this, you decided, okay, I'm going to go build a company. What was the first thing that you did to kind of figure out what exactly you needed to build and what the customers actually wanted from you?
We started with the with the the wrong answer, which is started just building. I don't totally regret it because it's how I got my co-founder, who turned out to be like truly excellent. I started making these like clickable mockups of here's what I think the product needs to look like, and then actually like wrote an early version of the product in MATLAB because it's the only thing I knew, and it was something of a show of seriousness of like, hey, we we really want to build this thing that's going to fix uh, you know, military planning.
Talk to me about the user interviews for a minute. Like who are you reaching out to?
We basically reached out to every stakeholder in military operational planning. It was about 200 interviews and uh, well, it was it was really, really good, and uh, we were pretty wrong to think we didn't need to do that.
How did you get early users for this thing? I know you guys did a bunch of like things that don't scale to get people into your early software. Tell us about some of those.
So the problem we had to solve is we couldn't get access to real users. This is before I had any contract. This is before we were in the room with in a military headquarters. There were just too many barriers to get any usage and see how this worked.
This is like the first year or so after you started working on this.
That's right. What we ended up doing is cold LinkedIn messaging a bunch of uh military planners, usually majors through colonels, and said, "Come to my all-day Saturday exercise. We'll pay you, but you need to be here for eight hours." Okay? And we put together this like pretty robust military scenario. Like I'd spend like almost a hundred hours writing these scenarios. And I'd put them into two teams and say, "You six are going to solve this military scenario in the old way with Word and PowerPoint, and you six are a team and you're going to solve it in this like proto version of One Brief, and we're all going to watch you and we're going to make the product better."
What kind of feedback were you getting or what sort of lessons were you learning from these exercises? I imagine throwing these people at your product in its early stages was maybe a little scary and a little sobering at times.
So the response was a lot better than we thought. So we tried like a whole bunch of go-to-market things already that were they didn't work that well, and this was the first time where the response was genuinely great, where people were like, "I am very excited. Can I introduce you to people in my headquarters? Can I help you?" Okay? So you did these 200 customer calls, you're doing these exercises with planners, how did you go from that to actually getting your first proper contract?
I had a phone call via a warm intro um with this gentleman, Captain Dave Fields, at US Pacific Fleet. And I actually didn't know he was as big of a deal as he was when I had the the phone call. So I I told him what we were up to, and the first thing he says is, "I've actually evaluated at least 10 attempts to solve this problem. They've all been terrible." And I could tell this was like a, you know, a sharp person who really cared about it. And we ended up chatting on like this first discovery call for about two hours on the phone. I think the the level of honesty about like what we did and did not know was pretty helpful. um in like building some some real trust. But he came out of it saying, "This is this is promising." And I said, "Great. Can I embed in your headquarters? I didn't want to just stay there. I want to stay there for three months. I want to try a proto version of our product on some of your operational plans." And I was like, "Just put in a government purchase card for some like nominal amount of money." And he was just like, "Okay, let's do it."
What else do you think might have been the things that kind of persuaded him to give you some extra time and to really bet on you guys for this?
I think there were some creative parts of our approach that weren't take the existing government process and put it directly into a product because we had this presumption the existing process was was bad and needed to change pretty dramatically. Um, and we would build a product around the future instead of the past. You know, sometimes people can just like detect a certain level of like energy and determination where like, oh, this is this is real. I think the ask of, "Can I embed and just get it done?" is a more earnest ask than, "Can you give me money?"
How did that lead into an actual contract eventually?
So, I was away for it was a little over 250 nights that year, most of them to Pearl Harbor, um, Hawaii. And we embedded around around two plans. One was this uh large exercise called Pacific Fury. It's like one of the largest exercises in the Navy. It's really important. Our product was used for like a fraction of 1% of it, but it was still real, and it was still the first time we've been used for for anything real. And there was actually like a pretty interesting barrier to entry, which is a a theme in defense. So, they can't use software that doesn't have an Authority to Operate, especially not for classified work because like these are are secret and top secret.
What does Authority to Operate mean?
Uh, Authority to Operate is the accreditation that like lets you put software on their internet.
I see. How does one get this?
The first thing you need is the contract. And the second thing you need is to work through some administrative steps to get approved. There's actually like a lot of art to it. Like there's ways you can do it that will cost you many, many millions and take you many years. There's actually a lot of defense tech companies that don't have one after being in business for more than five years. What we did initially was just circumvent the problem. We knew too little. We were too early to to figure out ATO. And I believe if we'd tried at that stage, it would have been this all-consuming problem that uh would have meant that we never found product market fit. So what we did to get around the problem is um deploy our software on one computer, my computer. All right. And went to their their N6, which is their word for IT department, and said, "I want to bring this one computer into your skiff," which is their word for secure facility that handles like top secret and above information. And we actually got to yes on that. But we got to yes by agreeing that the Wi-Fi was disabled, Bluetooth was disabled, and the laptop would become Navy property to dispose of and destroy when we were done.
That's amazing. So So let us do it. And I was the only user. So we we weren't at the point that any military planner like wanted to change their workflow and like use this thing, which would just be like a totally worse experience anyway because it's not connected to the network. You can't even like email your results. You can't share your results. It's collaborative software with no collaborators, which sounds terrible, but they like allowed me to to be a planner. And we would like hook the laptop via HDMI cable up to their screens, and they would have these video teleconferences with all their collaborating units and would share my screen that way. It's like a it's a pile of hacks all put together to to get you guys into business with these folks, and it it ultimately like got us there. In a sense, this was this is very unimpressive because the revenue numbers were like about $15,000 ARR. So, we didn't have much real revenue. We didn't have any like real accreditation to use the software, but we actually got to um both an exercise complete and a major real-world con plan, which is basically a large military plan against a big adversary. These are hundreds of pages written entirely in like on my computer in the proto version of One Brief, which was like a a huge feather in our cap, like, hey, somebody got through a whole con plan in your product.
So now you've got your first customer, you've got some money in your pocket. What was like the next big barrier that you guys had to figure out after that point before you could start really going after all these customers and being aggressive on the sales side?
So we got a verbal on it shortly before YC, and it was signed during um was our first real contract um, and this was with US Pacific Fleet. And to us, it was a huge contract. It was $322,000, and that was all like ARR subscription revenue. We were so proud of it. So the contracts actually came with something that was way more important than the money. Um, which was they sponsored us onto JWIX, which is the network that the military uses for secret and top-secret data. It's basically an internet, and having like a real sponsorship on this was the key to getting our software deployed in places that users could like access it and use it and collaborate. You know, we wanted, you know, collaborative software that would eventually become the way you run every military staff. And you you got to be on your networks to to do that.
You get this first deal, you're on the network. What are the next steps then? Do you go and just start doing a full-court press to the other headquarters? Were there a few specific ones that you went after?
We understood that this was make or break, that we would have no real usage until we were live, that everything we wanted to do was post going live. So, we were like pretty motivated. There's an art to helping the bureaucracy move fast at something it already wants to do, and I think we like availed ourselves of that art.
Was there a moment at which you felt like you'd finally kind of broken through and you you were in and able to just kind of run this like a more traditional software company and not have all these barriers in front of you?
Not really. Even today, the major barriers to entry of um, you know, DoD software, and like there's there's really only three barriers. So if you if you beat all three, then then it's just like a commercial business. Um, and the three barriers are facilities clearance, um, Authority to Operate, and contracting. And they never really go away because, you know, now that we're at at this certain scale, we we've like pretty um grossly exceeded the scale that our original infrastructure was was intended for, and now we're, you know, we're doing another one. And we actually have yet another ATO process for like our future infrastructure planned after that. Being good at the art of contracting quickly uh really helps you in defense. Facilities clearance kind of went away. Once you have it, you have it. Um, but I I guess what I mean there is if you know what the three barriers are in defense and you get good at overcoming the barriers, then it's a little closer to a commercial business.
How did you figure all this stuff out? Like you didn't know about any of this stuff five years ago and now you are the expert in it. Like what how did you go about learning all this?
It it almost helped me not to know how difficult some of those things were going to be because I couldn't be distracted from them. I think we were properly focused on just get to product market fit. If we can get a product that every military staff really wants to use, um, we'll be able to figure out the rest.
How, if at all, are you guys today at One Brief thinking about AI? Are you using it personally to be more productive? Are you guys using it internally? Are you thinking about adding it into the product in some way?
Inside the product, basically two buckets. Uh, bucket one, we call the mundane utility bucket, and that's the stuff that's live today. So the goal of the product is make military staffs fast and efficient. If you can get a military command to a a decision faster, you basically help them win the war. Um, so any little automation of mundane parts of the military operational planning process, I think there's huge numbers of automations available. The second bucket is I think more exciting. Um, so the second bucket is very special military, and the second bucket is about automated course of action. People have been trying really hard for a long time to automate the course of action, which is basically the meat of a plan, the like how do you defend Taiwan? How do you do the specifics of operation X? And we want to automate that, and we think we're finally at a point where we can ultimately where we want to get people is um, you know, AI generate a course of action that are faster and smarter than any human could come up with.
You're obviously talking about AI with folks on the other side with your customers. What what's your sense of how the military is thinking about AI and using it and like uh anxieties, excitements? What's your view on how they're thinking about applying this?
I mean, I think there's a lot more excitement than anxiety. I'm very careful to have every conversation like oriented around outcomes and what the product will do for you, and like make sure that the value pitch doesn't have the word AI in it. I think that's important because if you get AI in the pitch too early, it becomes "look at my cool technology" rather than "look at my business outcome." I think the the military is very excited about AI but a little bit skeptical of being sold snake oil because they probably are sold snake oil like pretty pretty often. So I I try to, you know, establish trust around the core product, around what we've already accomplished, and then around like the uses of AI that are currently live and working and delivering value.
You've hired dozens, dozens and dozens of people. You've expanded. You've signed many more contracts. What advice, if any, would you want to give to yourself when you were setting out to do this?
We could have moved a lot faster if we had known which things were the real bottlenecks. So, at no time was I working on a thing where I was like, "I don't think this is the most important thing to work on right now." I was just wrong about what the most important was. If I'd known that the exercise thing was going to work, we could have done it with a product that we'd built after one month. And in reality, we did it with a product that we built for two years. Same for getting on uh Cypernet and JWIX. That one's a little harder because I don't know how like my past self would have figured that out without a lot of trial and error. But just knowing this is a really big deal. You're going to have to find hacky ways to get to it. And the sooner you get there, like there is no product market fit before ATO in the military. So if you don't work ATO fast enough, you're you're in trouble.
What's next for One Brief?
Ultimately, where we want to end up is to be the way that you run a military staff from battalions up to the Joint Staff for the the US and our allies around the world. We want to make these staffs lightning fast. Something changes in the world and you know, you you're able to immediately react. You have an instant plan. We want to make them require fewer people. I think it's kind of insane to have a headquarters of a thousand people. I think you should have much smaller headquarters, you know, heavily, you know, armed with the the right tech tools to get there. We're well investing pretty heavily in core product um leveraging a network effect and use that to get usage to grow quickly and use that to get everybody to use it.
So founder motivation is the single biggest determinant of success we've seen here at YC. We can control for all sorts of things, but if a founder is not really in it to win it, it's not going to happen. Why are you so keyed up and passionate about One Brief?
I think what we're doing is on a per dollar basis the most important thing happening in defense tech in the world. What I mean here is it's very expensive to scale hardware across the military. It solves a real problem. We do need drones, definitely need submarines, but the military is often bottlenecked by decisions. Um, I think that was a huge cause of what happened in the Global War on Terror. There's a giant gain, like almost a win-the-war level gain in, you know, Taiwan, South China Sea, Korea, Europe, Iran over having much faster, like superhuman smart military decisions. For one thing, we're we're doing that. And uh, for another thing, we think we can do it very, very cheaply relative to say submarines, where once you've built software at once, you can scale it to every battalion, every brigade, all the way up to all the joint staffs for allies around the world, and effectively have, you know, one piece of software that drives almost every military operational decision worldwide.
Well, Grant, thank you so much for coming in today and talking about all this with us. It's been great.
Thanks, Brad.
[Music]