Leveraging AI to Predict Venture Success: John Sharp's Journey with Hatcher+

0:02 Cheryl Mack: Founders scale faster on Deel.

0:04 Maxine Minter: Set up payroll for any country in minutes, hire anyone anywhere, get visas handled fast, and get back to building.

0:12 Cheryl Mack: Visit deel.com/dayone. That's d-e-e-l.com/dayone. Okay, 3, 2, 1.

0:26 Maxine Minter: Hey, I'm Cheryl.

0:27 Cheryl Mack: I'm Maxine.

0:28 Maxine Minter: This is First Check, part of Day One, the network dedicated to founders, operators, and investors.

0:32 Cheryl Mack: Investors. If you want to be a better early-stage investor, this is the show for you.

0:35 Maxine Minter: So TL;DR, if you don't want to suck at investing, listen up. So John is an interesting one. He is actually Australian, but he lives in Singapore. I first met John when I was at some tech conference in Singapore, and one of the first things he tells me, he's like, yeah, we've analyzed like 600,000 startup deals. And I'm like, I was like, what? What do you mean analyzed? And he's like, yeah, we're doing this like fund model where we just invest in like the top companies that get analyzed through other programs. And I was like, you need to tell me more.

1:12 Cheryl Mack: I, yeah, I cannot wait for this one. I feel like this one is like a particularly interesting topic to dive into this idea of kind of like algorithmic investing, but into VC, both direct or via funds, just the sheer volume of effort to collect data on 600,000 companies. I'm really interested in, I think for him in particular, right? Was an operator, he has been an investor for a while, and now he is building again, but in this kind of different approach. I'm super jazzed to dive in and learn what has he learned about looking at all of those companies.

1:43 Maxine Minter: Yeah, 100%. His background is super interesting. I think he started in music and engineering, and so he can build, and then shifted into the investor space, and he's been doing this long enough to have seen a few platform shifts. So, I'm sure he can tell us a lot about that.

2:00 Cheryl Mack: Okay, I can't wait. Let's dive in.

2:01 Maxine Minter: Let's do it. John, we are so excited to have you here today. Thanks for joining us.

2:08 John Sharp: Thanks for having me, Cheryl. I'm super excited as well.

2:10 Maxine Minter: Amazing. So I know you've been living in Singapore, uh, so you may not have heard the First Check podcast before, but the first question that we always ask all of our guests when they come on the show is what was the very first thing that you invested in? Anything from, we've heard things about houses and, um, education and even like household clients is. So go for it. Tell us.

2:33 John Sharp: Wow. That's a really interesting thing. Really interesting question. I'd say the first thing I ever invested in was probably an education. And I say that in a very different way than most people would because when I was accepted into Melbourne Uni, my parents actually couldn't afford to pay for it. So I had to defer my then interest in music, I thought, for years until I could kind of save up to go there. And then I kind of decided instead to to buy the textbooks and see what I could make out of that. So I spent $250 on 3 books, um, orchestration by a guy called Walter Piston, uh, the professional arranger composer, um, by a guy called Russell Garcia, and, uh, and a book called orchestration— I forget what it was called— orchestration and voice or something by Nelson Riddle, the guy who used to like do the arrangements for Frank Sinatra. And, uh, and so I bought these 3 books. Oh, and in the 4th book, actually, Dick Grove's, um, book on jazz orchestration. And so I bought these books and I was traveling around Australia, like grape picking, as a firefighter, as a lumberjack, I was doing all these sort of weird jobs. And at the same time, I was going home and reading about, you know, how to orchestrate for the harp.

3:42 Maxine Minter: Yeah.

3:42 John Sharp: With Walter Friston and whatnot. And then I landed in Adelaide and bullshitted my way into a studio orchestration job. And it ended up being ridiculously successful. And so at age 19, I ended up getting smothered in work and I went from earning like $150 a week in some crappy job to earning several thousand dollars a week orchestrating and conducting the Adelaide Symphony Orchestra, which we are not allowed to call it in the recording sessions because they're not supposed to have a side job. But we used to do a lot of film music, commercials, TV stuff, all this sort of fun stuff. So for a $250 investment, that turned into a 10-year career, which is probably one of the funnest things I think anyone ever gets to do in their lives, which is conduct a room full of people, room full of really talented people, I should say, and produce hopefully beautiful music. So that was a really good investment. I have to say that paid off. But here's the interesting thing. That was the first industry, I have to say, that was disintermediated by software. And so That I think was a real turning point for me because at some point I had to make a decision. Do I stay true to the roots of Beethoven or whoever, right? Nelson Riddle. Or do I go off and embrace the emerging world of synthesis, digital waveform technologies, and software? And so at the tender age of, you know, early 20s or mid-20s, I guess by that stage, difficult decision to make as to whether you stay pure to the traditional roots or whether you go off into this unknown new world.

5:22 Maxine Minter: Wow, that's an interesting one.

5:23 Cheryl Mack: Yeah, we've never had anything even close to that. When you started first talking about orchestration, I was like, oh, like tech in software orchestration. And you're like, no, no, true music.

5:34 Maxine Minter: I thought like maybe he's saying origination wrong, like deal origination is what he meant.

5:40 John Sharp: Yes, exactly. That's what I do now.

5:42 Cheryl Mack: You have an investor hammer, so everything looks like a nail.

5:46 John Sharp: But yeah, it was so, the super interesting part about that I think is the part where When I was renting out the Opera House in Sydney and charging $30,000 for a commercial, there was some group of guys down the street that had just bought a Fairlight synthesizer and charging $300 to go, "Da-da-da-da," you know, on their synthesizer. And so there was this giant change that happened economically in the business where the typical film score went from being like a $100,000 gig, for example, to being a $5,000 or $10,000 gig. So everyone that has sort of based their lives like I did on that decision to do a lot of learning and a lot of, you know, a lot of training to get to that point suddenly was left in a situation where automation and software and dare we say AI had come in and completely transformed the industry.

6:36 Cheryl Mack: Wow, that's so interesting. Do you remember the first time that it dawned on you that software was gonna come and disrupt your industry, that it was gonna come and eat your lunch? And was it consensus at the time?

6:46 John Sharp: I lost a gig to one of these guys and the resulting commercial sounded atrocious, at least to my ears, because it was very early days for synthesizers. And I called up this, the guy who was the concertmaster at the Sydney Symphony, because he wondered why we lost the gig. I said, well, we lost it to these guys out in Manly that have this like Fairlight synthesizer and whatnot. And he said, what's a Fairlight synthesizer? And I said, it's this thing that replicates, you know, digitally the sound of an orchestra. And he said, "Oh, we gotta fight these guys." And in that moment I thought, here's this like 50-year-old dude telling me that we gotta fight these guys. And I thought, I've seen Terminator, I know how this is gonna go down, and I've gotta go with the software guys. So then I did the same thing I did like years ago, and I went out and bought books on digital waveform analysis and just started learning a lot more about— Yeah. That. And that became really fascinating to me, actually. It really roped me in. And that was sort of the lean into software and networking and moving stuff around networks. And that led down an entirely different road. And that road included scalability because suddenly if you're doing those kind of things on a network, you can do work for Saudi Arabia or Singapore that you're not just locked to Sydney anymore, or Los Angeles, as we ended up doing stuff for. So you end up doing a lot more stuff when you broaden yourself out and you use software and you use networks. And so that was a real eye-opener for me and I've never gone back, obviously.

8:13 Cheryl Mack: Yeah.

8:14 John Sharp: If you put me in front of an orchestra right now, it would genuinely sound like crap.

8:18 Maxine Minter: I always thought that those people with the sticks didn't really do much anyway, so I'm sure you'd be fine.

8:24 John Sharp: The guys with the sticks don't really do much, but it is an interesting management exercise, right? Because you're standing in front of 60 people with big egos massive amounts of talent, and the guy with the stick has to basically— let's call it the guy with the baton, that sounds a little more—

8:39 Maxine Minter: Sorry.

8:39 John Sharp: Yeah, a little more sort of like artistic. The guy with the baton has to basically stand out in front of them and get them all to play at the same volume and the same tempo and with the same kind of dynamics and emotional kind of context. And so it really is probably the most amazing management exercise I think that any manager can go through. 'Cause we all have to deal with incredibly talented people, and getting them to play together is one of the hardest things in management. Just leads to so many political fights and so much drama. And so it's a really good training ground for all that, because yes, orchestras have a lot of drama and a lot of prima donnas and a lot of other things going on inside of there that you can only imagine. I came away from conducting orchestras with the belief that a lot of what goes on in orchestras is much worse than what goes on in rock bands. Those guys are like— Yeah. Complete pussies compared to what's going on inside your average orchestra. So, a lot of respect also for those guys because they played hard, they partied hard, and just enormously talented people. But yeah, the guy out front has to respect those people, but they have to kind of respect the output that's coming out there. So, yeah, very interesting management exercise.

9:50 Cheryl Mack: Oh, yeah. I mean, we, from the executive coaching side, a big part of helping new leaders, especially if they haven't kind of experienced it firsthand before, new leaders try to access somatically or experientially circumstances when groups of people are doing hard things together. And a lot of folks, you know, have never played sport, they haven't spent a lot of time around sport. And so kind of orchestras or bands playing together is another entry point that people have seen, like the reality of building together, building trust, you know, that smooth handoff that happens when one player kind of hands off to the next. It's a very beautiful thing. I've actually never spent any time conducting, but it can be a wonderful entry point for a lot of new time managers to learn how to do that. So that's cool that that was your actual entry point, that you were actually conducting these large groups of people. Mm-hmm. I think that's a really interesting experience to have had relative to today's moment in AI prediction, just the kind of general disruption and widespread disruption that's happening. At the moment. I'd love maybe if you have some thoughts, maybe some pearls of wisdom for our audience. What did you learn going through that moment of disruption that is useful for today and to help people navigate this particular moment of disruption?

11:10 John Sharp: That's a really great question because I did live it and this is years and years ago, decades ago when this happened and it had a profound effect on our industry. I remember calling one of the very, very best people in our industry who'd moved to LA and finding him on top of his roof on his cellular phone doing the roofing himself. And I said, "Dude, you know there's people you can hire for that?" He said, "Not on the budgets I'm getting now." And he'd gone from earning like literally hundreds of thousands of dollars a quarter to earning not very much money as a result of that because he never really did make that move. But if you speak to your average musician today, they're completely across all the digital tools and everything. I went and spent some time with the drummer from my rock band that I joined in my mid-20s in Cincinnati recently, and he had just the most amazing digital ensemble of tools that created the most incredible abilities in the context of his living room that it would have been impossible to create 20 years before. So, I think there's two paths. There's the path whereby you don't embrace it, you just ignore it, and you hope that by ignoring it, you'll get to the end of your useful career without too much disruption. But what happened to us back in that time happened so incredibly quickly that you had two paths. You either embraced or you went, you know, you went away. So I think the most important thing here is to embrace change and understand it. And if you don't understand AI, figure it out as quickly as you can, because if you embrace the AI options that are available, you'll become expert at them. I mean, the nature of the way these tools are being presented to us. I just tried NotepadLM a few days ago for the first time, which is just absolutely insane. You insert a document, it outputs a podcast. I mean, it's just madness, you know, but it's so incredibly easy to use that there's no excuse not to use it or to become familiar with it. Now, is it going to be a podcast as, as good as Maxine, Cheryl, and John? Absolutely no way. Does it sound a bit like us talking? Yes, it does. Yeah. So you can see where that's going to evolve too. Yeah. And if you feed it enough podcasts, then it's going to be interesting, right?

13:21 Maxine Minter: Absolutely wouldn't sound as like us. It would be a creepy, creepy version.

13:26 John Sharp: Yeah. Even your industry now is getting transformed, right? Yeah. These— NotepadLM is your moment here that you'll have to adjust to. And if it's going to flood the internet with, you know, 20,000 fake podcasts a day that are all fairly listenable and you know, with feature chirpy voices and supposedly intelligent banter, that this is a problem. And so, you have to figure out how you leverage that to make that work in your favor. But yeah, it's gonna affect all of us and it's gonna affect all of us in the areas that we totally didn't expect it to.

14:00 Cheryl Mack: I think that's the bit. Like, it's gonna affect you in the areas you totally didn't expect. For example, if Notepad manages to replicate Sheryl's shit-talking, like, it can have it.

14:13 John Sharp: No one can replicate Sheryl's shit-talking.

14:16 Maxine Minter: Not a chance. Not a chance.

14:18 Cheryl Mack: But I do, I feel like the areas that you are not expecting the innovation to come in, I think those are the really interesting ones. I'm in San Francisco at the moment and it's Demo Day season. Everyone, you know, like YC just had theirs, EF just had theirs, 500 is having theirs, Pears was today. Like it's all on and there's lots of the kind of traditional ones that you would expect, lots of innovation happening in kind of like influencer spaces, marketing spaces, et cetera. And this is just like completely out of left field, like innovations in places I have never thought about as industries or the picks and shovels, seeing a lot of that kind of LLM or like GenAI optimization tools that are just like, wow, I just hadn't thought about that need, but it's so clear now that you talk about it. And then the whole bucket that's like, I've never even thought about, I have no idea how that would work. I'm going to go ahead and pass on that because I just don't even have an access point. But I do, I think that the pace of change that's happening at the moment is— unheard of, but it will be, you know, in 30 years we'll have another version of this conversation that will be considered a slow pace.

15:25 John Sharp: So we just made an announcement, a soft announcement I'd say, in Sydney. Actually, the last time I saw Sheryl was a few weeks ago, and we just slipped into the audience some findings that we'd had recently with using our AI for venture. And so we've been building a venture selection model using AI for close to 8 years now. Dan, my— Hooked Up, my partner and I. Dan's amazing. Dan has been working in AI for 30 years. He's had numerous AI startups. I think the last one before us, the last startup he worked at before us, he exited for a couple billion dollars. He's done really well in that space. And the algorithms and models that he's built for Hatchr are starting to really bear fruit. So we're even seeing within venture, no one would have predicted that it was possible to score a company and then 2 years later have that score be predictive of the valuation uplift that we're seeing there. But we're now seeing that happening with the Hatchers score. We can actually correlate a high Hatchers score now with a higher valuation markup or a low Hatchers score with a lower valuation— with a low valuation score 2 years later. And we've been— yeah, we're pretty excited about this because this is a first in venture capital. We're still trying to get our heads around how we actually release this information. Because it's a pretty bold statement to come out and say, well, we're able now to predict venture returns. But it's the old story. If you can— if you have an algorithm on a hedge fund and it works 51% of the time, you're going to make money. And so we have an algorithm that appears to work more than 51% of the time for venture capital. And so we previously had focused on the— as I think you guys are aware, we focused on how you construct a good-looking portfolio in venture. Mm-hmm. And what the modality and what the shape of that looks like. And now what we're able to do is say, here's how you might want to think about selecting for that portfolio. So these are now— what we're getting now is to the point where for an average GP working in venture capital, we can do a lot to assist them in the selection process and make sure that their portfolio is as good as it can possibly be. And that's based on a pretty rigorous approach that we've taken over the last few years. To get to the point where we're able to know that that model works. So, from AI, here's an example, to your point, of AI affecting something no one ever predicted that it would, and everyone told us it was completely stupid to even focus on doing predictive analysis in venture because it wasn't possible. Well, guess what? It is possible. Just like all the things that people say AI can't do, this ends up being another, yet another thing where AI can very successfully throw its hat in the ring and say, here's some— we've got a surprise for you. We can predict what kind of startups are going to be successful.

18:14 Maxine Minter: Yeah. So you and I, I mean, we hadn't caught up in a number of years, but we met up again for the first time probably in, what, 6 or 7 years at a conference that you were at here in Australia. And you showed me some of those models and the scores. And I was really impressed with how you were able to track the valuation in that range and where it was going. It might be helpful for the audience to just start at the beginning a little bit and explain exactly what Hatcher is and what you've learned doing some of this analysis.

18:45 John Sharp: Yeah. So, the beginning of this is I had an exit in 2011, 2012, moved back to Singapore and started investing in startups. And that led to being engaged with a group of about 40 other angel investors. And we made like 20 investments and we put about $20 million in the pot and invested that. I put in like about 10% of it and my buddies put in the rest. And then we watched as over the next 4 years, this did absolutely nothing positive. It stayed at $20 million. And so between 2012, 2016, I'm sitting there looking at this kind of like line and it hasn't moved. And I'm like, this is super annoying. So basically, what we started doing is looking at why this number had stayed the same. Why was our portfolio still worth $20 million in 2016 versus $20 million in 2012? And I didn't understand J-curves. I didn't understand very much at all about venture capital back then. So I started doing some analysis of what kind of portfolios might be predictable. And I did that with Wassim Ateqi, our other co-founder at the time. And, uh, and we seemed to get some good results when we loaded up a portfolio with a lot of startups, like, yeah, 500 or 1,000. Turns out that Dave McClure is absolutely spot on with 500 startups, by the way. Uh, because that's, that based on our later analysis seems to be really the, the cutting point where you can actually get a predictable return and his returns would seem to have indicated or validated his model. We did a lot of analysis around that. And I remember calling up and I had a buddy, Dan Hookter, in Florida who was the smartest guy by a long shot that I knew in data. And so I called up Dan and I said, "Hey, Dan, I've been doing this analysis of a venture fund portfolio. And it appears to me that if you get a lot of investments and you put them in a portfolio, that this could really create a much more predictable return." And he said, "Basically, you're dreaming. Forget about it. It's not going to work." many people have tried, no real value in doing it. And I said, well, look, would you mind just having a look at the numbers? And he said, sure, I'll have a look at the numbers, but don't expect anything great. So the thing I love about Dan is Dan is full scientific method. If he makes a statement and then the data says that he should revise that statement, he will, which is a wonderful thing to have in a partner, by the way, just amazing. So he called me back about 10 days later. He said, look, I've looked at these numbers. He said, could you get me some more data? And I said, oh, it sounds like you're interested. He said, yeah. He said, "I think there's something here." And so that led eventually to about 600,000 rounds of data being analyzed. And now we are able to absolutely show that the right kind of portfolio is not 20 companies because that's just an absolute crapshot. That is all probability-based. The right kind of portfolio is actually, you know, 100, 200, preferably 500 or 1,000 companies in a portfolio. that will return essentially an index, for want of a better word, of venture capital returns. Anything less than that, like anything in the sort of 10, 20 company area, like our first fund, you're playing with fire. You are making a statement that somehow you're a magician and you can take this incredibly small data set and make it work against the odds of that, that are literally 1 in 10,000. The actual odds are Only 1 out of every 100 business plans get selected, and only 1 out of every 100 of those goes on to be a unicorn or, or an absolute portfolio maker. So it's a 1 in 10,000 bet, right? So if you're creating a 20-company portfolio, you've got 1 in 5 chance of having a unicorn in your, your thing and having an amazing portfolio. So that's why our first fund was, was basically where it was at. And the second fund we created around this model. Now, the interesting thing about the first fund is we're now 3x, 20% IRR, everything's great, right? But that is pure probability. We ended up having 2 or 3 really decent bets in that portfolio that have gone on to create good value for those shareholders. But that's just probability. We were one of the very lucky funds that has 20 companies in its portfolio that does 3x because 75% of those funds do not do 3x. That's the odds. So, we all hear about the ones that do the 3x or 4x or 10x, right? We never hear about the funds that don't return the money or do less than that because they just quietly go away or become zombies. So, I'm very happy that our first fund is returning, is putting up a nice multiple, but I would never go to investors and say that's because of our genius. I would say that is purely probability-based because that's the way venture capital works. And if you want it to work for you, you need to recognize that just like the gamblers did in the 17th and 18th centuries and go to the casino knowing the odds and knowing what kind of strategies you should deploy with respect to those odds. And if you work the system that way, you'll do really well. If you pretend that this is all based on genius, you're making a fundamental error. Of forgetting the fact that everyone in this industry is pretty smart. Right? And so, if you're setting yourself up to be a little bit smarter than that, well, there's not a multiple involved there. Right? So, I don't know. Some people are pure geniuses and they could probably make that claim, but not the entire industry. So, if you want to succeed in venture, you've got to truly play the odds in a very intelligent way.

24:24 Cheryl Mack: Right. And I mean, I think the reality, if you've spent any time around venture, you spend any time You know, with funds, a recognition that the fund performance is distributed along the power law in the same way that the underlying asset is distributed along the power law. And I will say it's very common when people first start investing in this class, 'cause the power law doesn't pop up in a lot of other asset classes. So, you know, it pops up in maybe the creator economy and content, but it all, but it pops up in venture. And so it is something that—

24:54 Maxine Minter: But not bonds or— No, not bonds. Or, you know, treasury, treasury bonds or houses.

25:00 John Sharp: No. But probably there's someone out there that's shouting at the screen right now saying, "Yes, it does to this very small extent." But I think the best analogy is sports, right? Moneyball was such a great movie and book in respect because it just shone a light on the fact that people have been selecting for prettiness when they should have been selecting for talent. Billy, what's his name? God, I've seen your moment here. But the guy who's the manager with Oakland A's, When he set about looking purely for talent, the guy played by Brad Pitt in the movie, when he set out looking for pure talent and based his whole thing on statistics and runs batted in and all the useful stuff, he did what no one had really done before. They just had a bunch of scouts going down there saying, "Oh, that guy looks like he can hit." They hadn't spent enough time looking at the raw statistics of what win games. Now, the difficulty occurs, as it did in Moneyball, is when people start replicating the strategy. Everyone's on the playing field now with the same strategy. And so that's what's also gonna happen with venture. Like we're one of the top 20 global data-driven VCs according to this report that came out last year. That's great 'cause there's only really 20 people working hard in this space at the moment. When that goes to 200, right? Or 2,000.

26:14 Maxine Minter: It's gonna be a bit harder.

26:15 John Sharp: Yeah, when it's 2,000 and it's the entire industry and it's every fund and every fund is doing what we are doing, which is using large language models and algorithmic algorithms to generate returns, then it's a completely different playing field. But remember, this is what happened in the '90s with public equities, right? We went from a bunch of like manual processes at Solomon Smith Barney to multibillion-dollar dark pools at Goldman Sachs, all trading algorithmically amongst each other, right? Right.

26:44 Maxine Minter: It's all trades are done by bots now. So humans aren't even involved.

26:49 Cheryl Mack: Right. Yeah.

26:50 John Sharp: So we would be mad not to expect the same thing to happen within venture. And my prediction is within 10 years,— you'll have some pretty aggressively algorithm-driven— hopefully, Hatcher will be one of them— very aggressive algorithm-driven funds that are using their fundamentals and analytics that they're able to derive from these companies to build portfolios. And the whole thing will completely change. It'll be a lot more like public equities at the end of the '90s and a lot less like the manual sort of arcade processes that we see now.

27:22 Cheryl Mack: Yeah. I mean, I think that the key thing though is I would imagine there's an enormous amount of infrastructure that has to be built before you're ready to like purely algorithmically trade. And in venture, if I think about the earlier stages, right? If I think about pre-seed where we play, even seed, increasingly today, Series A, like there isn't actually that many data points that you can access. Without a human being an intermediary.

27:52 John Sharp: So we hear that a lot.

27:53 Cheryl Mack: Yeah, what kind of data points do you grab?

27:55 John Sharp: Yeah, exactly. We hear that a lot. There's not many data points, but then if you go to your average VC conference, right? And then Cheryl, this is, I've heard you speak on this as well. Yeah, what do everyone talk about? Everyone talks about the team, right? So there's not a whole lot that changes with Maxine or Cheryl or John when we go become a CEO of a pre-seed company and now we're a CEO of a Series B company. All that's changed on our CV really is 3 years or 4 years of working at this startup, right? But our CV is effectively unchanged, where we went to school, maybe we did some course in the middle, but I doubt it because we're working 18 hours a day on our startup. So, there's a lot known about the team at the beginning and as that kind of comes in, there's a lot known about the market that you're going after, the problem you're trying to solve. And so, for A company like us that does a really brilliant job of collecting all the data that's available on a particular thing, either by iterating with the founder over a series of— like we do this thing, like we do pitch deck analytics, we do company description analytics. And so, if we score a company description 17 out of 50, the message we're giving the founder is, hey guys, there's not enough data here for investors to make a decision. And so, we're going to keep prompting them and pushing them to fill up as many of those silos as we can till we get to a point where they have enough data to give us for us to start making an assessment. Now, that's just on the stuff that they know, but there's a lot of stuff out there that the world knows about the problem that they're solving, whether it's breast cancer or climate change or rocketry or whatever that problem is that they're solving. You can very quickly find out who all the competitors are. Which we have a database of 600,000 companies. We know who your top 10 companies competitors are when you pitch us. We just go bang and we devolve your application into math and we go out there and we find the top 10 guys. We actually find the top 250 competitors. So, we know what the competitive matrix is, we know what the team buildup is, we know the problem, we know the market size intuitively within 15 seconds. Yeah. If you submitting your business plan. So the data points are actually out there. It's how you gather them, it's how you make sense of them, and how you then construct a scoring methodology around that. We spent 8 years constructing our scoring methodology. Takes a heck of a lot of work to get that right, but we're now at a point where that's a fairly reliable methodology. And so I would say that we are at the beginnings of figuring out the selection piece of this. And then the next phase for this is how you go through the back office piece of understanding what's going on at that company and how well they're executing. Are they on track? What milestones should be in place for that company that every other company has? And we've started doing that with our valuation milestone, um, analytics where we actually target what the valuation journey should look like for that company, which is pretty cool. And so these are all things that we're gradually building as capabilities into the FAST platform so that people can start moving into that journey of a data-driven or data-assisted venture selection and venture building process.

31:14 Cheryl Mack: So, it sounds like at the earlier stages, the data points that are really valuable for— that your system kind of orients on is team and team CV, market dynamics, competitor dynamics, And then as you are progressing them along that journey, kind of traction dynamics and execution dynamics.

31:36 John Sharp: Yeah. Also we, we look at comparative valuation expectations, which are indicative of, of, you know, whether or not we're gonna be able to make money out of this as sort of venture investors. We're looking for market size.

31:49 Maxine Minter: Do you do that based on location?

31:50 John Sharp: Yeah, we look at, we, we've weighted for location, for sector. For stage. Stage is a really tricky one because a Series A in Singapore might be a seed in San Francisco. And so you've got to have these kinds of adjustments that you make. So we actually use round numbers. Yeah. So we'll, we'll refer to round 3 rather than calling it seed or Series A.

32:10 Maxine Minter: The naming is so messed up anyway. It may as well just be like fruits at this stage because it's been changing.

32:16 Cheryl Mack: Yeah.

32:17 Maxine Minter: Right. Like I'm doing my mango round.

32:18 Cheryl Mack: I'm going banana. What's the traction they need? And banana. That's really funny. But I think, I mean, it's constantly moving, right? But as you know, that has been the case for a really long time, right? Seed never used to be a thing. It used to be like VC and they were all coming in at Series A. In fact, there's a— Bessemer open sources their investment notes on certain deals and their one on Shopify actually makes me want to cry. So they, it's like them agonizing over the decision to invest in Shopify. And they're like asking themselves the question, like, can you even build a good software business in Ontario, Canada? Shopify at the time had $5 million in revenue, sorry, $5 million in profit. And it was, the valuation was $5 million. And they were like, do we do the deal? Dunno. I was just like, oh my goodness.

33:11 Maxine Minter: Wow. What an indictment on Toronto.

33:14 John Sharp: But yeah, we all hear about these crazy deals that are like that, right? By the way, Shopify was the first time I ever saw an enterprise UI that I just thought was wonderful.

33:24 Cheryl Mack: Really?

33:25 John Sharp: Like, one. Yeah, I just thought it was just so incredibly well designed and well thought out. It made me weep for all the other UIs and UXs out there because I just thought it was just stellar. And their integration with PayPal and all these other— it was just Brilliant.

33:41 Cheryl Mack: It's just been beauty. Anyway, I mean, I think like obviously the game has changed so much since then, right? To build a business when Shopify first started, to be fair to Besma, so different than to build a business as it is today, right? And I think what, that's what's really interesting given the nature of what Hatch+ is trying to do is kind of pull these data points and especially historical data points and make predictions about kind of future valuation dynamics and stuff. And it just sucks. Seems like such a dynamic system to be using historicals for prediction. Now, that's what humans are doing, to be fair.

34:17 John Sharp: It's in that category of business where you've got to go big or go home. So what we're really building at Hatcher is an operating system for venture or for family offices or for an emerging class of alternative investments. Let's just call it that. And we have to get We have to get AI across the entire infrastructure here for it to make sense. As I mentioned before, it's not just about the initial selection, it's about following through on that, which means this permeates your fund admin, permeates your managers, it permeates your reporting and what you put in the reports. And so it's a massive, massive, um, exercise as it turns out. We've spent literally tens of millions of dollars so far getting to this point. Um, because we want to win. Yeah, we, we want to be the guys in this space that have the integrated approach. We're kind of building an iPhone approach, by the way. So, we're not betting only on our own technology. Like, if we license our technology, for example, to Microsoft, which we just did for Venture, and they had some technology that was really cool, we'd license it back. We'd bring that back in and we'd say, "Okay, we're going to allow that to sit next to our stuff and may the best technology sort of win." Because we think that approach is a lot better. If I can offer— Yeah. 2 AI-powered cap table management solutions or 5 instead of 1 and allow you to choose the one that works best for you or the one that you've been using, then you don't have the switching cost. You can come right on the platform. So our whole approach here is to create a platform that where AI starts here and, and goes right throughout the platform and the process automation sits on-chain and supports all of those decisions that are made up here that are assisted by AI. Mm-hmm. So, it's a very— what we're building here is a really massive bet that people are going to move like they did in the '90s from completely manual or very, very manual systems to completely automated systems. And the reason we know that's going to happen is that the timeframes, there's so much capital sitting out there that if I can present you with a system where you can deploy that capital with a really good feeling that you're going to make returns from that, and allow you to operate fairly seamlessly, consistently with the same best practices in play at one of the top Valley firms, then I know that I can make that bet on that technology offering because there's so many people that would like to be able to do that, but they don't have the analysts, they don't have the ability to create those giant people-based structures, but they do have the ability to invest a few tens of thousands of dollars in annual licenses of a patch of software. And so, that's the business that we're in is trying to transform ahead of time what we see as this massive opportunity in venture, but make it digital and make it accessible to everybody that wants to play in that alternative space. Because let's face it, I mean, with— Yeah. Yeah, I don't know what the Dow is today, but it's pretty toffee. Bonds, not that interesting. Real estate, extremely well known and well understood throughout the world, getting much harder to make a buck there. And so alternatives and playing in that space intelligently is an area that everyone is rapidly moving towards. Private credit, private markets, these are crypto. These are all areas of extreme interest to people, to investors now. And we're in the front driving seat of leveraging AI in those areas to help people make those investments, but make money from those investments.

37:41 Cheryl Mack: Super interesting. So of those, I think you said 600,000 data points. That you've analyzed and I think you're kind of active.

37:48 John Sharp: Well, 600,000 companies actually.

37:50 Cheryl Mack: 600,000 companies that you have analyzed. What have you learned? What correlates to a successful business and what doesn't?

37:59 John Sharp: What have we learned? That's a really, really great question. We've learned that the best qualified founders can sometimes be pretty flaky and run away at the first sign of trouble. That's nothing new to anyone who's listening to that podcast and has invested before. You are looking for resilience at the end of the day. That's really the goal of all of us as investors. I think we've learned that when you go into a crowded space in a developed market, that can be really difficult. But if you go into an underdeveloped market with a cloned offering, you know, you can do really well out of that. So there's a few sort of basic rules that emerge, but I think these are already rules that are known to all the investors in this space as well. I think what we've learned from the AI is nothing that would be new to the ears of people listening to this podcast. There's no aha moment. What we've really learned is, goes back to what our initial portfolio level analysis. What we've learned is that if you're making 20 investments or 2 investments, we've all met the guy that's made 3 investments and they're all unicorns. We've all been to dinner with that person and he's a pain in the ass.

39:06 Maxine Minter: Those ones are the worst.

39:08 John Sharp: Yeah, he is just the worst. You know, just don't want to ever have dinner with that person. They're statistically out there, right? You know, 1 in 10,000 of those guys exist as well. And it's super annoying because it's just not what happens. Most people make 3 investments and they absolutely plummet in value and they have a hard time figuring out what went wrong. What went wrong is they just didn't make enough investments. Statistically, 1 in 100 of those is going to pay off big time, like a billion-dollar exit. And maybe 10 out of 100 are going to pay off really, really well, but not billion-dollar exits. And maybe 30 out of 100, according to Fred Wilson, are going to do okay. And the rest you can just write off. And so if you're writing off 66% or 70% of your investments, you know, you've got to be making more than 3 because statistically you just need to, right?

40:00 Maxine Minter: Power law. The power law is everything. That's the one thing, like you said, Maxine, it's not something that doesn't, that comes up. An other type of investment, but you enter the venture world and like, I can understand why it's confusing because you enter the venture world and you're like, why wouldn't I apply the same investing logic that I've, uh, that has worked for me in other assets.

40:18 John Sharp: You're so right. And like I said, I spent 4 years in venture not understanding the power law or the J-curve or some really fundamental things, um, about venture. And, and so yeah, I think there's some, some real basic good learning that you can do. But I'd say the basic thing to Maxine's question, the basic thing I think I've learned is the importance of doing a lot of investing in venture, taking an index approach as opposed to trying to pick the winners because the wrong founders will seduce you. The right founders will stammer and not say the right things at the wrong— or they'll say the wrong things at the right time. And it's just, it can be confounding to try and pick winners in this space. One of the companies that I thought had the least chance for success is now one of our top performers in our portfolio. And one of the companies I thought had the greatest shot of success is right at the bottom and is out of business.

41:13 Maxine Minter: It's always like that, isn't it?

41:15 John Sharp: Yeah, it's super annoying, but that's the experience of every investor. Yeah, like the best company that was ever put in front of me, I passed on because I just thought the guy had a really limited understanding of his business. And and just a tiny, tiny set of assets relative to his valuation at the time. That company is now an $8 billion company. And I was offered at a $10 million valuation. So that hurts. But on the available information at the time, there was absolutely nothing that indicated this was going to be an $8 billion company.

41:47 Cheryl Mack: I think that's actually something you just have to truly internalize. To do this well, you have to worship at the altar of the power law. Like, it dictates everything that you do. You really have to recognize how important that is. It is unavoidable and dictates the outcomes. And I think diversity— yeah.

42:07 Maxine Minter: Just show up every day, pray to the altar, power law. Oh, please, please, power law, grant me successes.

42:14 Cheryl Mack: Grant me successes. It's so true. And I think, like, you know, when you are especially picking at the earlier stages— and we spend a lot of time thinking about this as well, which is there are companies that I know I have passed on because there wasn't sufficient data for me to be able to make any reasonable prediction that it was possible that it would be successful. And they might be hugely successful, right? I think that they're, you know, especially in the Australian ecosystem, we throw a lot of shade on the people who passed on Canva when it was, you know, a yearbook automation business. But like, if we take you back to that decision and the data that you had, like what is the probability that you could have reasonably predicted that other than if you were doing founder investing, right? The market analysis would just not have got you to that outcome. Maybe the founder analysis, but the market—

43:03 John Sharp: Well, also market analysis, how do you predict a pivot? Cannot.

43:07 Cheryl Mack: You can't.

43:07 John Sharp: Every one of these companies, there's been like Canva, you just mentioned that had a consequential pivot. Airbnb had a massive consequential pivot. I think it was the drummer of—

43:17 Cheryl Mack: From cereal.

43:18 John Sharp: When they went to like whole houses and— Instead of just mattresses.

43:22 Maxine Minter: Mattresses on people's living room floors.

43:23 John Sharp: Yeah, exactly. Mattresses on— every one of these companies at some point has a pivot and you're betting pre-pivot in so many of these cases and you're hoping that the founder has enough clue and support from his board and all the rest of it to be able to continuously pivot till they get it right. And the other thing is you're hoping that they have a board that understands that they're in the long term. Pharmaceutical business. They're not in the short-term, sell something for profit business. And this is another huge problem that happens in these startups is you have the wrong people backing the right kind of model. And everyone knows that for biotech or whatever, it's all about getting those FDA approvals and everything else. And you're not going to make a dime until that happens. But when it happens, your valuation could jump 40x, you know, like the next day when that stuff comes out. So, if you have people on your board that don't understand those dynamics, you're screwed. You might as well give up now if that's the majority of your board. And so, that's another thing I think it's so important. You have to have the right kind of backing for the right kind of model. There's so many things that have to go right for a startup to succeed in this world. And that's San Francisco, Singapore, Australia, Sydney, doesn't matter. You've got to have all these things aligned, which is why the CEO needs to keep all this stuff in mind, I think, when they're building their businesses. But look, it's a fascinating world that we inhabit, which is why we do what we do, right?

44:54 Cheryl Mack: 100%. Absolutely. I feel like we can keep having this conversation forever, and sadly we're coming close to the end of our time. So the last question we like to ask everyone who joins us is what is your biggest big kindness moment a moment where you felt really brave?

45:09 John Sharp: My big cajones moment was when I was 19 and I bullshitted my way into conducting the Adelaide Symphony Orchestra for a commercial for Australia's biggest department store on the basis of almost zero experience and a modicum of talent, and I managed to pull it off. And standing in front of those 60 people, never having conducted before, never having actually written live for strings before or anything like that, and pulling it off was certainly my big kahunas moment. And it took me from a tiny, tiny salary and no prospects and being dirt poor to being in a really great place in the world. So sometimes you got to have those big kahunas moments. But that was very early in my life. And there's been a couple since, but that was a big one for me and proof that if you have confidence in yourself and you know your stuff, sometimes you just got to put yourself out there. And, you know, follow your gut that you know how to do this.

46:04 Cheryl Mack: I love that. So good. Thanks so much for joining us. Thanks, John.

46:09 John Sharp: It's super fun, guys. Let's do this again.

46:11 Cheryl Mack: Love that. Love it.