Your Pitch Deck Lies

Blue Moon: A Tech-Driven Venture Fund

Jake Aaron Villarreal: I'm Jake Aaron Villarreal, born and raised in Silicon Valley. Here to take you behind the scenes to share what it's like to be a startup founder, the journey they're on, the problems they face, the products I've built in an effort to make our lives better.

I'm excited to have with us today Ben Orthlieb, co-founder and general partner of Blue Moon. Ben, welcome to the show.

Ben Orthlieb: Thank you, Jake.

Jake Aaron Villarreal: Well, Ben, thanks for jumping on here. Before we dive into your story and the company you're building — where are you joining us from today?

Ben Orthlieb: Marin County, North San Francisco.

Jake Aaron Villarreal: Really cool. Love that area, a lot going on. You know, the next couple weeks there's a lot of AI conferences happening, and it seems like it's always happening these days. But really cool.

Well, a little bit more about Blue Moon, which I think is very interesting, because our audience is predominantly VC — really, I would say a lot of startups, but a lot more AI startups now, of course, and in that everyone's at different stages. Some are at seed, some are farther down the path, but Blue Moon is a seed-stage venture fund that's flipping the script on how VCs actually work. Ben spent years at LinkedIn in a leadership position before realizing most VCs run themselves like spreadsheets. So he built something different — a fund that can screen 20,000 companies a year or more, make a dozen investments or more, and asks founders about their life instead of just metrics. The result: a fund that massively outperforms the market.

So, really excited to jump in. If you are a founder, if you are looking at seed investment, this is an episode for you — listen all the way to the end to get all the value you can out of this.

Before we go any deeper, Ben, give us a little background of you. You know, you were at LinkedIn, you spent some time at big corporate companies as well, got into the capital side of things — but what were some of those early experiences that shaped you into doing what you do today?

Ben Orthlieb: Yeah, so I'm originally from France — you'll probably hear that in my accent, a very lovely town in the south. I've done many jobs, and maybe the first experience that shapes what I do today is — so, basically, I went through engineering. I'm a systems engineer by training. I actually was building neural networks before it was really cool, in '99, 2000. And then I switched over, because I got curious about finance and investments, and ultimately went down the path of a mixed path between banking and consulting, where I was always at the cross of finance and strategy in a tech environment.

That led me — fast-forwarding a number of years here — it led me to LinkedIn through various roles. Actually, at LinkedIn, business operations for product strategy, then focusing on the marketing business for LinkedIn, and then finally, because of my experiences as well as my investment and involvement in the tech ecosystem and startup ecosystem, ultimately took over what LinkedIn calls corporate development — which is traditionally M&A but also includes investments and elements of corporate strategy. Which is what I've been doing for the last five-plus years out of my 10 years at LinkedIn.

Jake Aaron Villarreal: Wow, that's great. You know, there's a lot that goes into that when it comes to corporate development and acquisitions, and I know the number is very high even when you invest in companies, that they can get acquired. So we'll maybe touch on that as well, as a strategy, as you grow.

Talk a little bit more about — you're at LinkedIn, you decided to make a change. What was a big breakthrough, or what was a decision in your mind, that you said, "You know what, I think now is the time to do something different and really specifically focus on venture"?

Ben Orthlieb: There's a number of signals along the path where I really got curious, even about startup investing — probably around 2014, 2015. So remember, I come from France originally, I came to the U.S. via banking, I was in consulting, was doing M&A-related and strategy-related stuff. I didn't really get exposed to startup investing, venture, until I stumbled upon it, because I live in San Francisco. Some of my friends became founders, or were founders, I invest on a relationship basis, and then they raise a Series A, a Series B, and you've got to understand — or at least I got curious about understanding — what was going on. That's maybe the first step.

A couple other instances — I obviously started working and met my co-founder, who's also French, also in the venture ecosystem from his job, and got curious about how can we use data to understand better the ecosystem. That, in parallel to starting to work at LinkedIn with the data science team there, we were looking at what are signals of startup success. And there were so many interesting metrics that come up — everything after Series B is sort of an operational nature, so how fast you grow, where you hire from — those types of things are indeed very predictive, and in fact, I think at this point, Series B+ is becoming way more of a science than an art.

The interesting conclusion for what led to Blue Moon is — before that, basically, seed preceded Series A — the main thing that matters is actually people. LinkedIn will never go down the path of scoring your profile, or my profile. And so that was the limit of the research. But the old saying that basically the only signal that you really have at seed is the founders was holding completely true. And that got me thinking a lot more about how do you evaluate founders first and foremost when you do investments, as I was starting to ramp up my own angel investing.

And then the real catalyst of all this is — in 2020, because of my investments, because of my interest, because of the fact that my day job was to talk to founders all day long, I started a conversation with Emergence Capital, and effectively, eventually, became a scout for them, working closely with Jason Green at the time, one of the co-founders of Emergence. Yeah, that is basically what got us started.

Jake Aaron Villarreal: And as you looked at the market and you said, "I understand what's happening, you're kind of in it, you're doing investments, but you're seeing how investments work and how you evaluate companies" — what was it you thought could be done differently that just isn't happening today? And you talk about looking at building a fund really with more of a DNA around technology, not just a services or marketing arm to going out and getting companies. Walk through that.

Ben Orthlieb: Well, the idea is this — you know, VCs talk a lot about technology. We're back in 2020, 2021, talk a lot about technology at the early stages — most of them use Affinity, which is a spreadsheet with a Gmail connector, effectively. Venture uses very little technology for itself — at the time, such as, and now is true as well — the world doesn't need an extra seed fund that does exactly the same as others. And frankly, our view was: the traditional way of sourcing, relying on your network — I've got a little niche that maybe gives me a few deal flow every year — wasn't going to be enough.

And so we figured, from the get-go, two things: we're going to build as a technology company — we just happen to monetize as a fund, but ultimately we see ourselves as a technology company. And despite doing all that technology in the background, the core of the conversation goes back to what I was saying earlier — it's really just all about the founders. What drives their obsession? What is their resistance to pain? And so we do all this technology so that we can have very human conversations with the founders about their backgrounds, their upbringing, their family. And that's where we take the ultimate decision.

Jake Aaron Villarreal: Yeah. You know, if you're a VC fund or firm, you might be thinking, "Well, we understand the metrics because we've been doing this for years, we understand the data, and we can make decisions based on that that have been successful." Talk about the variable insights you get when you talk to a founder about their upbringing, about their parents, about their childhood, about really the makeup of their character. How can that make a better decision when it comes to investing in a startup?

Ben Orthlieb: Yeah, maybe the best sort of story — parallel story — to illustrate this is the 2012 Olympics in London, which — my daughter wants to be an Olympian, so that's also been forefront — but 2012, the Olympic Games were in London. When London wins the bid, they put together a committee with basically the task of — hey, UK has been underperforming GDP and population results, basically forecast results through the '80s and the '90s. We're going to host the Olympics, how do we not look bad? They put a group together.

Long story short, they changed two main things. One, they went to source much more broadly than usual — so it wasn't just your regional leagues and things like that, they were looking for athletes on a much broader scale. Two — super relevant to the question — it used to be your ranking, your national ranking, was the only thing that sort of mattered in this ability to get selected. They shifted this to say, "Now you need to pass a bar." Right — VO2 max, for example, is a clear metric of your cardio health. You need to have a minimum VO2 max if you're going to play basketball, minimum height, those types of things. But ultimately, that was just a filter — you needed to, quote unquote, pass the bar. Where they really selected the athletes was through a bunch of conversations with psychologists, to figure out who has the grit to go through the training.

You're training to be an Olympian, that takes 10 years, you're training every day. It's very not sexy — you show up and suffer, basically, for 10 years so that maybe you're going to win a medal. That's not normal. What drives that? It's the same with founders. And actually, there are lots of parallels between the profile of a founder and the profile of Olympians. You're becoming a founder, sometimes a repeat founder — you actually know what you're getting into, it's painful for 10 years or more if you're really successful. And that's what we're trying to figure out — what drives that obsession and willingness to go through pain for 10 years so that you're going to be successful.

Jake Aaron Villarreal: Yeah, that's interesting. Yeah, go ahead.

Ben Orthlieb: Well, what's interesting is it resonates a lot with founders. They understand that at these early stages, sure, they have a product, they have a little bit of sales, but they know they're going to pivot. What's interesting — you need to be in a good space, but you don't know what your product will be in five years. So all the founders know that intuitively — that it's about them. Honestly, 90% plus of those conversations, they tell us they've never had these kinds of conversations with any VC. They've never shared their relationship with their dad, or what happened to their siblings, or why they're competing with their siblings. All these stories that explain that obsession and that drive — nobody asks them.

Jake Aaron Villarreal: Well, it's one thing to ask the questions, it's another thing to get people to talk.

Ben Orthlieb: Talk about your co-founder. So my co-founder — his first career, he was a diplomat, French diplomat, and he wasn't a spy, but he was trained in intelligence techniques for some of the negotiations he was leading — literally around war and peace treaties. So he's very good at getting people to talk. We have systems that effectively go do a whole bunch of research on people before we talk to the founders, in that second phase of our process. So we show up very prepared — or he shows up very prepared — and then he's looking for patterns in the conversation, how to steer the conversation so that we can get to the heart of the founder's motivation. Sometimes — and it's not one-size-fits-all, it cannot be, right — because those end up being very personal stories. But sometimes it's very stark what's driving them, and sometimes it's less obvious.

Usually we tend to know very quickly after those meetings. We usually take a first-step decision, let's say, after the meeting, and then we sort of revisit the next day, and we're able to take a decision very quickly from there. And, in a few cases, like more recently, it brings up a few more questions that we go and evaluate and make our final decision. But ultimately, by the time we get there, our system found them and we screen them positively based on their background. I've met with them, and I like — with the use of a whole bunch of AI — to give me information on what they're building, their space. So we have a full analysis of what they do.

The round is becoming interesting, right — we're just a small check, $250K. So for us, who else is at the table matters, because adverse selection is clearly a thing — you can see it in the data, it's pretty obvious. But once you've got all of this — those are gates. It's the same thing as the Olympics committee I was talking to you about — it's like, great, you've met the bar, now what are you about, that we get conviction to back you.

Jake Aaron Villarreal: Yeah, that's really cool. I like the motivation aspect, in some ways. I mean, we're in the recruitment business, and a lot of it's very similar. You know, when you're trying to understand somebody, you have to understand their full story. And when you understand their story — what drives them, what motivates them, what's their specialty, and why could they be a fit, what's the upside — you're really trying to drive down into the details, but you have to have the conversations. It's a people business, and just like what you're doing, you're investing in people. The technology is great — what they're building, the innovation — but really, behind the technology is the people. If you get that wrong, you're probably not going in the right direction.

Ben Orthlieb: We spend more on technology than what an associate would probably cost us. But through that, we do so much more. But again, ultimately, it is about having those conversations. The biggest compliment we got, probably three or four weeks ago, was a founder told us, "I feel very seen with you guys." And the specific reaction — you know, we were already 20, 30 minutes into a conversation, so we'd already talked a lot about her past relationship with her mom — actually, that was super important to her. But she had written a short movie 15 years ago, and we'd actually seen it. She was basically flabbergasted, really, that we were like, "Wait, how did you dig that up, and you guys actually went through it and have relevant questions?" It's crazy. But yeah, that's the type of preparation that we go into that part with.

Jake Aaron Villarreal: Talk about the numbers — 20,000's a lot. You use technology to do the tracking, the screening, the qualification. Walk us through the process — what's that look like?

Ben Orthlieb: Yeah, our view — actually, which leads to the numbers, and which numbers are important — our view is there's a couple of things that really matter. One is the coverage, which means how many — what's the full sort of size of the deal flow, or the percentage of deals that you see that end up being done by good funds. At the end of the day, that's how most top funds measure themselves at seed — what's their coverage, at least for the efficacy of their sourcing. And then, ultimately, if you do your job in the middle well — which I'll assume for most people — then it's about winning deals. Those are the key metrics in our model for performance, because if you find great founders but you don't win, then your whole process is not useful.

Coverage — so we use machine learning. I'll start with screening, because it explains how we can scale sourcing. Screening, which is where we started back in 2020, 2021 — we use machine learning to evaluate founders, completely outside — it's not Jedi, it's not "let me teach Claude what I like in a founder" and sort of all the variability that comes from that. It's not a point system, just like, "Oh, you went to Stanford or MIT, five points, you didn't, else." It's pure brute-force machine learning on people's experiences, that if you abstract what that says, it leads to seven or eight pockets of founding teams that tend to be more successful. And so what we get is, effectively, an evaluation of, "Are you close to one of those pockets of teams that tend to be more successful?"

Once you have that, you don't need an army of people to review deal flow. And, by the way, it's way more actually predictive and replicable, because it doesn't change on the mood, or whether you're doing this at 2 in the morning — you just got a dump in your email and you have to check out all these founders, like speed dating, in 30 seconds. Once you have that, you can source from a whole lot of places. So, of course, we have traditional inbounds from network, etc., which leads us to about the same volume as everyone else, which is sort of 1 to 3,000 leads. We get to somewhere between 15 and 20,000 — we're past 8,000 for the year. That comes from about 15 different data sources. Again, we're a little obsessed ourselves, and we think about it as a tech company. So any idea that we've had over time to improve our deal flow, we've tested it and put it into production.

Of course, everything that's public — all the accelerators, sometimes before they publish the batch, all the angel list, traction, Crunchbase — we get all these signals. We have some others — we have a list of over 40,000 people that we follow on a regular basis to check if it seems like they're building a company. Everyone we're second-degree connected on LinkedIn who says they're a founder, we automatically analyze, etc.

This leads — this first pass — leads to 15 to 20,000 teams that look like seed, B2B, North America, which is what we do. And then from there, I guess, our system, our algo, tells us — are the 400, 500 teams that you should probably spend time with this year.

Jake Aaron Villarreal: So it actually is narrowing it down for you. And then talk about the numbers — once you've got that algorithm that's telling you where to focus, what's the next step? I mean, you've got to get on calls, you've got to talk to people — what's that volume look like? How do you manage that?

Ben Orthlieb: Yeah, maybe — well, there's one step, which is — there's about 2,000 seed deals, seed B2B North America, every year. So 20,000 is way more than what actually ends up being done. And so the metric that everybody tracks, for good reasons, is coverage. We, with our 20,000, we get to a coverage of about 70% — meaning, at the end of the year, we've seen 70% of the deals that ended up being done by a top seed fund, top 50 fund, in the year. If you ask traditional funds, good ones with very good brands, they will tell you their coverage is about 30%. So through data, we're able to basically have two and a half times the coverage — not just the pure volume, but the actual coverage of deals that get done from tops.

As you said, the trick then is talk to them. And given we find them, we're mostly an outbound motion — our ability to find you, Jake, and talk to you, Jake — so conversion from finding to talking, to a first meeting, is 70%.

Jake Aaron Villarreal: Yeah, that's very high.

Ben Orthlieb: Most VC outbound is in the 30%, from what friends tell us. What's very interesting that we found along the way — right, I mean, we're iterating all the time and learning all the time — is that being a tech company, quote unquote, is very useful. We reach out with an email that would say, "Jake, I found you with my AI, she's already produced an analysis of your company, here it is, and, oh, by the way, I already have a few questions to ask." In our first meetings, the number of founders who are like, "Oh, this is so cool, those are the questions that I'm asking myself," and, "In this day and age, I'm building with AI, I need more people around me that are doing the same, and you seem like one of them" — and so that cold email conversion is crazy.

And when that doesn't work, we then ask our network for introductions. And the reason people do it — partially just good karma, but also we're very close to a lot of VCs because we have so much deal flow, and we don't lead, we share deal flow with a lot of funds. And so, of course, when I ask for an intro, and you've made some investments or you've seen interesting companies that I shared with you, people are very willing to help. It takes a couple minutes, because on our side it's all automated, so they can get all the forwardables and everything they need.

Jake Aaron Villarreal: So if you invest in a company — $250,000 — what do you give them? You know, companies have multiple options to take funds from lots of organizations. Sounds like you got a good connection with the VC firms — I'm assuming, like, all the big VC firms out there, maybe there's some introductions, maybe there's taking them to that next level. What other value do you give them that other traditional funds might not give them?

Ben Orthlieb: Yeah, it's very interesting, because our win rate is basically 95% plus. In the first fund, we've made 50 investments, we've been told no twice — there's a company that I'm still waiting to hear, so I don't know if it's a no or not — but it's basically two or three nos in the span of five to six years. So it is an incredible win rate.

Where we started was traditional — and so we pay for some services outsourced, so you get an exit coach, plus, three exits, you get a program for engineers, mentorship, sales coaching, a few other things. A key piece, as you've mentioned, is the intros, because we're a small fund but we're very highly connected — we're able to make lots of intros. So we tend to talk to people early, they don't necessarily have a lead — I've got a whole system that helps me reach out to potential leads or co-investors, which means last year we've made 278 intros between founders and VCs. And so I reach out to you, we have a first chat — of course you don't really know Blue Moon, but if I like what you do, within a week you get five or six intros to top VCs. All of a sudden you're like, "Wow, Ben's really effective" — that will be helpful for my Series A and my Series B, which is also where we help. So that's the second sort of place where we help.

And the third one is — I think you've mentioned this a couple times — 92% of exits — actually, that used to be my day job, so I know exactly how the tactics happen, how the workflow happens, not just purely like, "Hey, talk to a banker," or anything like that — I get into the nitty-gritty with founders, from, "Hey, I got an inbound from a company," to helping them through negotiation. It turns out those are nice, but that's not why we win. Ultimately why we win is much more emotional, which we found, again, along the way — it wasn't by design, but they love the fact that we were effectively builders, they love the tech that we show them, because they haven't seen this from a VC — forget in 2021, but even still in 2026, they're like, "Wow, this is sort of crazy good." In fact, some VCs are actually getting it from us — they're getting technology from us — very well-known VCs.

Two, they absolutely love the conversation about them. Of course, the context is a $250K check, so I'm not here to win the lead position, which would be crazy for a whole sort of reason — I'm here to win an allocation in the 50% that the leads usually don't take. Once you've had this conversation where they tell us what happened to them in their childhood, or versions of that, you became pretty close, and they absolutely love that. And so this whole package — with the intros, the demonstration, the value-add, plus the personal connection — when a couple days later we say we want to invest, it's never really a question.

Jake Aaron Villarreal: Yeah, I like that. Well, you know, you're good at getting companies and having the conversations, success rates in converting that — how hard is it to stay disciplined and not want to go upstream and start cutting bigger checks, if you really feel good about an opportunity? Like, is there a challenge there for you? Because it looks like you got the volume, it looks like you're hitting the target with the right kind of companies, the right kind of founders, you're at $250K — do you have an interest in going higher than that?

Ben Orthlieb: It's a great business model question. If I go back to the things that matter — coverage, we're always doing better, but we've sort of solved the problem. We can always improve, and we're always looking at it, but, okay, problem solved. I don't want to compromise the win rate. And there — it's the other thing in the middle — Blue Moon, as much as I love it, is not a VC brand yet. If I were to compete against Sequoia to deploy a $2 million check in your seed, why would a founder take me? And it actually would be — honestly, I'd be concerned if I was to win, because maybe I have differentiated deal flow — is that really 12 investments a year where you're like, "I see something that none of the top funds can see"? I don't actually believe that. Collectively, the top funds see everything, and they're actually very good pickers. I mean, I track that data very closely, to even know who I want to co-invest with — there is something there, it might just be the brand that's become self-fulfilling, but the top funds are top funds for a reason.

So me competing would be crazy, because I don't have a brand — versus me co-investing with your Sequoia, your NFX, your First Round — is actually cherry-picking from their collective portfolio. And it turns out our metrics — in particular, graduation rate — is better than theirs. So if you think about this, for a whole bunch of reasons, top funds do perform better, and then, with our process, we're able to overperform from a graduation rate standpoint, and frankly, from another metric standpoint — so far, in the first fund, TVPI, DPI, we're able to overperform. And so that's the key to your question — like, how much can you invest? My answer is: as much as I can without losing my ability to win and be in the right deals.

Jake Aaron Villarreal: Yeah. Well, you talked about you're building a technology company and really funded by your investments — so I guess the question becomes, how quickly do those investments have to start paying off for the model to work?

Ben Orthlieb: It's a very good question, no one asked that before. Well, there's a couple things. One is, does it pay off operationally? And the reality is the volume of deal flow, conversations, analysis that we're able to do is incredibly high for a fund as small as ours. I honestly don't understand how most solo GPs even run it — and it's a compliment to them, like, there's so much going on in the background. If anything, admin is crazy, which we're automating most of it — but even just the time you have to spend on admin, working with your Carta or your Angelist, it's crazy. So we've got incredible leverage.

No fund can do seed and show up to the founders the first time with an analysis of their company — it is just not scalable. We send them an analysis of their company before we even show up, which means our systems run analysis on 500 to a thousand companies every year before we even reach out to them. That is just not doable. And those analyses we've built — just to be clear, this is not ChatGPT or Claude or whatever, like, we've built a RAG for the last three years, we scrape 150 sources curated about early-stage venture in B2B, which is what we do. The quality of this analysis is honestly mind-boggling, even though I know exactly how it works.

And the same thing for every conversation — I get a five-page memo based on the conversation, after a conversation with founders. This is, again, something that is not basically doable if you're going to use mostly humans to do this. And the quality of just shipping it to ChatGPT or Claude is not good enough at this point. And so, is it worth it? Well, we do all these things — the volume, the analysis — more than, honestly, your top seed investors. You could say we could cut the spend, but then we'd be the same as every other fund.

Yeah, and so — but the real payoff, sorry, that goes a long way to actually answer your question — the real payoff is in the returns. The first fund is doing very, very well according to Carta — we're in the 0.1% funds for our vintage, in 2021, which is a really bad vintage for venture in general. But we have about returned the money already, 1x DPI, to all our initial investors, and the paper markup is close to 5x. So, all of this to say, it will take time, but the payoff is — as with any fund who have been investing in AI in the last year or two — it needs to show up in the returns.

Jake Aaron Villarreal: Yeah. Can you talk about some of the companies you've invested in that the listeners would know about?

Ben Orthlieb: The most famous one is probably Mercor, and we invested at their seed, and they publicly raised at $10 billion in October. So that is, of course, a very, very high return. And what's interesting is it's textbook our model, how we got there.

We found them because they were trending on Crunchbase — so we track Crunchbase not just for Crunchbase, but they actually have sort of pretty obscure traction metrics in the background. We found them — actually, an at Village Global made the intro on that one, because I saw it in my CRM in the morning, I was having lunch with her, she knew of them, she made an intro. We talked to them, which is interesting, because they weren't actually at the time doing exactly what they're doing now — they were more going after the model of, "We'll find teams and build teams of engineers for you," sort of on a contract basis, when they quickly evolved into reinforcement learning, which is their business.

The space was interesting, right — what was very interesting in the first conversation is our systems screened them positively. At the time, there were a bunch of — like, there are three 20-year-old dropouts from Georgetown, and this is B2B, right — it's not, "Oh, it's a good sign maybe to have dropouts in B2C because they've seen a trend" — they are in the — no, this is finding engineers, finding people in a B2B context, our system saw them and scored them very highly. A lot of people — just because they're not the right background for traditional these kind of jobs — so we found them, our system screens them positively, we talk to them, like, what they're doing comes the personal conversation — they are exceptional. One of the easiest conversations or decisions we've had to make.

Brendan himself basically started doing AWS consulting on nights and weekends when he was 13. He's not the only founder that we've backed that started at 13 doing sort of consulting or companies, but I mean, it's pretty exciting.

Jake Aaron Villarreal: That's young.

Ben Orthlieb: Yeah. I mean, I wasn't doing that when I was 13. And then a whole bunch of stories — he and his co-founder — so Brendan is dyslexic, he's talked about this, to prove that this wasn't an impediment. He and his co-founders signed up for debate — debate competitions — and won the state championship three years in a row. Nobody's done that before, nobody's done that since. And a whole bunch of stories like that. And honestly, you meet people like this and you're like, "Take my money." You can't give them to — you know, to partially answer your question earlier — you have to stay disciplined, you can't just make one bet, you have to make — for the numbers to play out at seed, you have to do actually quite a few bets. But once you get past the, "Oh, they don't have the traditional background," and, "Is what they're doing interesting or not" — they were an obvious investment. Level of drive and obsession, it was just off the charts, and that's ultimately what we were looking for.

Jake Aaron Villarreal: Yeah. Well, that's not going to show up on a resume, so you got to sit down, you have to have that conversation.

Ben Orthlieb: Yeah, 100%. Because sometimes those conversations — again, sometimes those conversations are flat with people that have excellent resumes, and it just doesn't work out for us.

Jake Aaron Villarreal: Yeah, that makes a ton of sense. Talk about your company — it's Kubi(?), you've got a lot of technology running a whole bunch of agents. Where has it gone since when you started? Are you hitting breaking points yet within technology? Are the token spends — like, you're seeing significant increases in spend? I mean, I think the big question we get a lot of right now is, "I'm building agents, I've got tokens running, I hope there's an output that's actually, you know, an ROI behind what we're trying to get done within our company" — but how to manage it, how to make sure that we're not overspending in areas that don't make sense. Like, what's working for you?

Ben Orthlieb: What I think is different, and we help a lot of bigger funds on that journey — again, some of them actually buy or get our tech analysis on companies, a few companies have that through APIs, things like that. Other funds — what I think has been key for us, and makes us different, I think, sustainably, is the people who are building are the people who are operating the fund. And so we can very quickly test things and evaluate, "Is this worth the output? Is this worth the tokens?" Because ultimately, the tokens spend is money we cannot pay ourselves, and, trust me, it's tight. And so we only do things if it makes sense, we test a lot of things.

We have a vision of where we want to go — we had a vision of what we wanted to build, you know, in the last two, three years, completed — I would say that first sort of vision, which is, at this point, everything from sourcing, we've talked about, portfolio management, has technology, and, at this point, AI, that gives us incredible leverage. But the building and testing and dogfooding all the time is crazy important for us. And then, frankly, once we get to a system that works — like I was talking about, our RAG — very complicated, actually tricky to build something that performs very well for what we're doing — and then you can replace the models, because once you've built the infrastructure, changing models is actually super easy. And so every time there's a new model, we can test — okay, does this become significantly better from a whole bunch of perspectives, versus how much we use? And, in fact, what we found is once we like something, and it's close enough, or it's not crazy on a cost amount — because the cost goes down so quickly — it's actually oftentimes worth it to keep going.

And so, yeah, we're a small fund, we probably spend the equivalent of an associate salary on all our tech, and we have incredible leverage. So, in the aggregate, it's totally worth it.

Jake Aaron Villarreal: Yeah, but I think it is the question that you were asking, like — are you building for the sake of building, or are you seeing a direct impact on your investment, which is the core of what we do, that influences what you draw, what you try?

Ben Orthlieb: And we've seen a lot of funds having issues where — they hire a developer, there's probably adverse selection already there, because working for a fund as a developer is not necessarily the sexiest thing you could do in the world, but some people have passion — but the developer is not in the workflow of the investments. The other challenge is most firms, because they've developed over time, each GP is his or her own island. And so a lot of times we're seeing an issue where it's like, "I've got four GPs, they all want something very different." And so either I build the minimum that can help them, but it's actually not that super helpful, or I need to convince them to start standardizing the way they do things, which is very hard, because successful funds have been successful, and so people don't really see the impetus to change — again, beyond, "Oh, let me be a little bit more operationally efficient." So a little bit of sourcing, a little bit of, "I'm going to teach Claude my scoring system so it can screen the exact same way I'm doing," as opposed to, "Can I use AI to actually improve my results?" — things like that.

So we're in that phase where people are trying to figure out what they can do with AI. Some of it is just a little bit of feel-good and marketing versus — do you go all the way to rearchitecting your company to perform better? The latter is very, very hard and costly.

Jake Aaron Villarreal: I mean, yeah, we're in the middle of it right now, which is why I asked the questions. Yeah, yeah, yeah. Really cool. Well, exciting about what you're creating. I'm assuming that if you've got systems that go out there and can track really a lot of companies that are in that pre-seed stage and beyond, or earlier — what should founders out there be doing that will show up in your algorithm? Where do they need to be so you might see them? And if not, how do they find you, or how do they find Blue Moon?

Ben Orthlieb: I mean, finding Blue Moon is easy, it's blue-moon.vc, so that's easy. We're getting more inbounds because we're in the sort of Crunchbase list of active investors, etc. — that discovery, that way, is easy.

I was thinking about this this morning — the question is, do you put stealth, or do you actually just say what you're working on? I capture people both ways, but I find, as an investor using a lot of AI, more useful when the screening will be the same, because it's independent of what you do — so it's not why I was screening you positively or not. But it actually, honestly, I think saves founders time if they say what they're doing, because then VCs will not reach out if they're obviously not a fit — which may not be obvious if you're just saying you're in stealth, and stealth is still exciting for VCs for a lot of reasons. And so I know people put stealth oftentimes not to be bothered, but I actually think people who quietly say, "Here's my company, and here's what it does," but without making a big fanfare of it, they probably get much more qualified inbound than if you just say yourself, and you are coming from a sexy company or sexy sort of university background.

Jake Aaron Villarreal: Yeah, makes sense, I agree with putting it out there — these people know what you do and they can help you. Last question for you — if you were to restart this company today, what would you have done differently, if anything?

Ben Orthlieb: Interesting. Well, there's things we've learned along the way, and so maybe you can argue some of that learning could have come faster. So year two is when we really started looking at what we call our anti-portfolio system — so we track all the companies we've evaluated, we've missed, why we passed, all these things. But year two is probably when we realized we were missing companies because our deal flow was not big enough, and that's where we got pretty obsessed about this question. At the time we were probably 3,000 companies every year, we're now five, six-x that — but that could have happened earlier, and there's a few companies that we missed because of that.

Another learning that — you know, is part of the job, but will probably haunt me for a long time — we passed on Perplexity's first round because of pricing. There's good reasons, right, 2022, '23, their round was at $90 million, which was not at all usual — let's put it this way, not that it's unusual today, it's overblown, but as a percentage of sort of rounds, that is in that zip code. So we could have gotten in, we passed purely on pricing — obviously they proved us wrong. But that's a little bit of our learning — which was more, not that pricing doesn't matter, but you have to think about a little bit more flexibility, basically.

And frankly, the founder's obsession — it's almost like it was in front of us, because the machine learning, for example, is all about the founders, but "let's go much deeper" — and we have a secret weapon in my co-founder's sort of background and ability to do this. And so it's almost like it was in front of us the whole time, but it took us a while to really go deep in there, and then realize that it was actually very helpful for our investment performance. So you can definitely see before and after — two, that sort of reflexive thing, that this makes us win more. Not that we had a win problem, to be honest, but the check was smaller and we could have maybe scaled faster. So those are the sort of learnings along the way.

I would probably have picked a better brand from the get-go, also — which — we used to have a complicated name, and then we're very close to James Currier at NFX, and one day he sends an email and he's like, "You got to change your name." And so he gave us a masterclass on venture branding, and then from there came Blue Moon. But that also, you know, things you learn along the way.

Jake Aaron Villarreal: Yeah, I got it. Well, really cool, really exciting to have you on the show, and it's going to be very valuable for a lot of founders that are looking to get funded at some point. So thanks for joining, Ben, appreciate your time.

And for the listeners, for listening, it means a lot to me. If you're out there and you've listened, do me a favor, hit subscribe, it's free, it helps us — we've actually grown like 1.02% in the last 45 days of subscribers. So continue to do that, it's happening for a reason — content has gotten better, we've done a better job, we're accelerating how we operate with AI, which has been helpful. And ultimately, we're just trying to bring good stories that help everyone out there find their dream and do it with the right companies at the right times.

I'm your host, Jake Aaron Villarreal, signing off for now, but can't wait to catch up with you all in the next episode. Until then, Ben, the world — take care. If you like what we're doing, don't forget to subscribe, leave a review on Apple Podcast or wherever you listen, and follow us on YouTube, where we go behind the scenes to learn what it takes to be a startup founder.

‍