Invest Like the Best with Patrick O'Shaughnessy - Ryan Caldbeck – Quant in Private Markets - [Invest Like the Best, EP.110]

Episode Date: October 30, 2018

My guest this week is Ryan Caldbeck, a private equity investor who wants to bring quantitative rigor to the private markets. Ryan is the CEO of Circle Up, which uses a system it calls Helio to identify attractive investments in early stage consumer brands.  While I am of course a fan of quantitative investing, I also know from experience how much harder private markets are than public markets when it comes to the transactions themselves. We discuss this and many other potential roadblocks to bringing models to private markets. Using many individual companies as examples, Ryan explains some of the major predictive factors they’ve uncovered in their research. We also discuss which parts of the private markets might be infiltrated by quant processes first, and which may never be.  I expect many more to go on a journey similar to Ryan’s in the years to come. They serve as an interesting example for ambitious investors out there. Please enjoy our conversation. For more episodes go to InvestorFieldGuide.com/podcast. Sign up for the book club, where you’ll get a full investor curriculum and then 3-4 suggestions every month at InvestorFieldGuide.com/bookclub. Follow Patrick on Twitter at @patrick_oshag   Show Notes 1:39 - (First Question) – Formation of Helio 6:57 – How they handle the relationship building needed to make investments in private markets 10:26 – Why consumer and retail are interesting spaces to apply their quantitative approach in private markets 12:54 – Searching for new relevant data 16:14 – How do they stay ahead of the commoditization of uniqueness             16:21 – Pattern Recognition and Machine Learning             17:24 – Sam Hinkie Podcast Episode 18:00 – Dominant predictive factors in this world 21:05 – Which is more important, relative value or rate of change 21:48 – What does the data say about online sales vs offline (being in a store) 23:30 – Variable that consumer investors think matters but it doesn’t 24:53 – Valuing companies and accounting for mispricing’s             25:36 – Michael Recce Podcast Episode 26:41 – Goes through the process using Liquid Ivy as an example 28:46 – Most interesting sub-categories 29:33 – Future for this model             32:10 – Albert Wenger Podcast Episode 35:19 – Other categories outside consumer and retail interest Ryan 36:28 – Biggest challenges for CircleUp as a business 38:46 – Handicapping their earnings expectations 41:36 – Take on the VC/PE landscape 43:03 – The types of models that are most interesting to the team 45:05 – Quantitative elements of brand that are most interesting 47:30 – Most unique brand and distribution strategy he’s come across 53:27 – Who has influenced Ryan the most 54:37 – His personal values 55:51 – More people who had an influence on Ryan             56:05 – The Innovator's Dilemma: The Revolutionary Book That Will Change the Way You Do Business 57:07 – Thoughts on goal setting at the company 59:29 – Unchangeable factors that shape their long-term vision 1:02:01 – Most interesting individual conversation as part of this journey 1:04:02 – If he could only keep one dataset, what would he keep 1:05:09 – kindest thing anyone has done for Ryan   Learn More For more episodes go to InvestorFieldGuide.com/podcast.  Sign up for the book club, where you’ll get a full investor curriculum and then 3-4 suggestions every month at InvestorFieldGuide.com/bookclub Follow Patrick on twitter at @patrick_oshag

Hello and welcome, everyone. I'm Patrick O'Shaughnessy, and this is Invest like the Best. This show is an open-ended exploration of markets, ideas, methods, stories, and of strategies that will help you better invest both your time and your money. You can learn more and stay up to date at investorfield guide.com. Patrick O'Shaughnessy is the CEO of O'Shaughnessy Asset Management. All opinions expressed by Patrick and podcast guests are solely their own opinions and do not reflect the opinion of O'Shaughnessy Asset Management. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of O'Shaughnessy Asset Management may maintain

positions and the securities discussed in this podcast. My guest this week is Ryan Caldbeck, a private equity investor who wants to bring quantitative rigor to the private markets. Ryan is the CEO of Circle Up, which uses a system it calls Helio to identify attractive investments in early stage consumer brands. While I am, of course, a fan of quantitative investing, I also know from experience how much harder private markets are than public markets when it comes to the transactions themselves. We discuss this and many other potential roadblocks to bringing models to private markets. Using many individual companies as examples, Ryan explains some of the predictive factors they've uncovered in their research. We also discuss which parts of the private markets might be infiltrated

by quant processes first, and which may never be. I expect many more to go on a journey similar to Ryan's in the years to come. His serves as an interesting example for ambitious investors out there. Please enjoy our conversation. So I think a fun device for this conversation would be to treat the core model of your business, Helio, at Circle Up, as almost like a biography of it. So maybe we'll start in its infancy or even the sparkle in your eye, so to speak. Begin by telling us kind of what this is and start at the earliest days. How did you conceive of it and begin building it? Sure.

So Helio is a collection of algorithms and datasets, which go out into the world, first find, companies and then evaluate those companies. We focused specifically on consumer and retail. So the spark, sort of speak, was when I was in consumer-focused private equity, 12, 13 years ago at a business school, I had a job. And the job was someone would hand me a list of companies, four or five hundred companies a week, and say, go through this list, tell me which ones we should reach out to. And we didn't have any data on the company. I would just have to Google that company. And when I would Google the company, I'd find they were in Whole Foods or maybe they weren't sold yet, just rough information about distribution, what I thought of the brand, stuff like that. Candidly, it was really boring.

It did not take a lot of intelligence. And you began to get into a rhythm after you look at a couple thousand of these. You can make a decision in 30 seconds and be 85% right. So after a couple months of that, I kind of began to think, you know, a monkey could do this. A monkey could do this churned into a computer could do this. I had no idea what machine learning was. I was not a CS major in undergrad. It just struck me that this was not what I wanted to be doing with my life.

And this probably could be done by a computer. So fast forward to Circle Up. We started Circle Up six years ago. And we had a online portal that companies would apply to. And they'd apply. They'd give us their financials. And so I wanted to take that application process and streamline it. So we hired a data system.

scientist, who's named Arvin, he's still with us today, doing an amazing job. And we built what we called a classifier, which is exactly what it sounds like, which is people would give us their financials, and then we'd say yes or no to a company. We also then pulled in some external information. But most of the information that we really cared about was the company's financials, revenue, gross margins, stuff like that. And as a private equity investor, I figured that we'd make the decision based solely off of their financials. If the company is declining, we're not going to accept that company. If the company's got 3% gross margin, we're not going to accept that company. When we ran that for nine months or so, we discovered that eight of the 10 pieces of

information that were most predictive of whether or not we would say yes to a company were pieces of information that we didn't need to ask for. Real quick, I want to pause to make sure I understand the original classifier. So the outcome variable, if you will, was an independent, qualitative, subjective yes or no? Exactly. Got it. So you're just trying to, you're trying to basically replicate your own decision-making process. Exactly. The outcome's not an exit or something down the line. It's your own process. Exactly. We'll come back to that later or later on Helio, but you're right.

Initially, it was just trying to make our lives easier. Just what is correlated with us saying yes or no, which is clearly biased and everything else, but that was the objective measure. And so that was a bit of an aha. And anyone who saw this data put on the screen, I still remember where I was sitting in our old office, anyone who saw this data would say, oh, if eight of the 10 datasets that are most predictive of whether or not we're going to say yes to a piece of company or things we can get elsewhere, let's just go get that data on a lot of

companies. So then we built another model called The Sourcer, which does exactly what you probably think, which it goes out and finds companies. The Sourcer would go out and find companies, and it started by looking at, we put in the list of maybe 10,000, 20,000 companies we got from trade shows, things like that. But after a couple months, we began to recognize that we could build a new algorithm that would proactively find companies and then pull in the information that we thought was predictive. Over the next couple of years, the more we leaned into that algorithm, the more we got back. Meaning, we began to think, okay, this is helping us find companies that meet this criteria.

Is there other criteria that we would care about? Then we began, and you quickly hit on the concept of the objective measure, we didn't immediately recognize that. But over time, we said, okay, well, is what matters that we care about working with the company or is what matters that the company is going to be successful? Let's go look at what is correlated with success. Now, in the private markets, that's really hard. In the public markets, and you know the super well, it's a bit easier because you've got so much data on pricing.

In the private markets, while you have exits, it's really hard to determine whether or not an exit was successful or not. So we took a long time to pull together a large enough data set that we felt comfortable with that as an objective measure. But then we had a problem because as soon as you start whittling it down to say, okay, what's an exit in food versus personal care versus pet? The data set isn't big enough. So we looked for things that were correlated with exit. We found a few different things, one of which is revenue growth. So revenue growth, which also makes just intuitive sense, revenue growth then became the objective measure for what we're doing. And so we looked for other factors that had high

correlation with revenue growth. And that progressed meaningfully over the last few years. So today, Helio tracks about 1.4 million consumer and retail companies and then evaluates in a series of different dimensions related to brand, product, distribution, and a number of other things. So the interesting insight here is sort of the application of the quant process to private markets, having now had some deep experience in private markets and especially around the transactions buying and selling these things and knowing just how hairy and tricky and qualitative and how relationship driven and emotional these things are versus pushing a button in the public markets. Talk a bit about the importance of that side of things. So ultimately, you're trying to build a

a sourcing engine, something that can basically help you identify through a quant process, good investments in the private markets. But talk about the other side, which is a lot harder than we see in the public side. It's very hard. And it's a lot less efficient than the public markets. And it'll never be close to as efficient as the public markets in terms of getting the deal done. So we try to systematize that process of reaching out, working with an entrepreneur, helping the entrepreneur as much as possible. But at the end of the day, you're absolutely right. It is still a relationship. And we put a lot of weight into the relationships that we have. So the people we hire, while we have a number of quants on the team, we have a fairly large engineering data science team, at least for an asset manager. We also hire folks that are able to build deep relationships with these entrepreneurs. And we've tried to apply some skills that we find in analogous businesses in terms of using the data to reach out methodically, but also layer. on a personal element. So we first start with a funnel, effectively a B-to-B sales funnel. And there's

some private equity firms that do something similar. TA Associates Summit partners that have kind of funnel metrics. They traditionally have relied on the equivalent of, let's say, a phone book or trade show lists, etc. We start with Helio. And then the team has reach-out metrics that we can measure and look at that performance. Clearly, it is dramatically less efficient than the public markets. That's just what we do. And we're able to measure success along each step of the funnel. When we are able to get in touch with the entrepreneur and work with that entrepreneur, it becomes even less efficient. But what we're trying to do is we're trying to bring data to that conversation to transform what has traditionally felt like a used car buying process where no matter what deal you get,

you kind of leave the conversation feeling, did I just get the best deal I could have? Did I just get robbed? How do I feel about this? It's never a great feeling. We're trying to transform that as Tesla has done or as Apple has done in terms of buying a car or a computer. And it's always the same price. You feel a little bit better. So what we're trying to do is just expose the data that we have over time and say, look,

we've worked with tens of thousands of companies over the last six years. we have an understanding of what market is in early stage consumer. Let's expose this data to the world. Let's have them understand why we're pricing their company accordingly. And if they want a valuation that's twice as high, that's going to hurt them in the long run. Let's show them why that is. But this is the valuation range that we're willing to pay and this is the data behind that. It doesn't always work because sometimes the entrepreneur's got a friend in tech who's going to tell them, no, you're worth 20 times revenue. And that's sad, but our quest is trying to systematize this as much as possible and bring transparency and data to the conversation so that they feel fair.

Can you talk a bit about why consumer and retail is such a particularly interesting space for you to apply this kind of quant methodology in private markets? There's a couple reasons. First, it starts with in, let's say, tech, in early stage tech, I think there's too much capital to be Canada through. There's 750 BC firms that chase that industry. In consumer, if you just Google early stage consumer product funds, there aren't 10, 15 in the country. There's a ton of consumer-focused private equity funds that will invest $25 million into a company, but not many that'll invest two or $5 million in a company. But beyond that, I think there's two things that aren't always obvious to folks that don't live and breathe consumer. The first is that in consumer, the business models are

basically the same. Yeah, it's a widget business. It is. And I mean that with a ton of respect, because I think these people are doing really, really important things. But you make a product, you sell it. You don't make the product and give it away for free for five years by putting ads in the front. You don't make a product. You don't make the granola bar and say, actually, Patrick, you can only have the granola bar if you buy a five-year subscription to the granola bar. Unlike tech, where you've got a game for your iPhone, a securities business and a cryptocurrency, wildly different business models. Because it's the same business model, it's the same game of chess over and over and over again, which makes it easier for us to build models on top of it. That's a key component.

The other thing that is not very intuitive to most people is that there's just a ton of data in this industry. So in consumer, if you walk down the street here in New York and go to local bodega and pick up the smallest company you can find, you can find, if you Google that, you can see where that granola bar, let's say, is sold. Not just that they're sold at whole foods, but specifically which whole foods they're sold at. Or Target, Walmart, etc. You can find how many skews they have, meaning do they have one skew or 10 skews? Flavors. sizes, et cetera. You can find information about the pricing, what the end users think of the product. And if you're tracking it, you can see how all those things change every single month and how they

compare to every other company in the category. That data is the building blocks of success for that company. It is almost the entire revenue equation for that consumer print. So we've been pulling that data together for years. Now, it is a really ugly process. It is messy. There's a ton of cleaning of the data, normalizing it, combining it, et cetera. But when you have that data together, it paints a really interesting mosaic about how the business is performing. And there's some interesting correlation between that and the success of the business going forward. So with a standardized sort of, let's call it business model or set of financial statements, more comparability. I'm curious then what the process has been like, maybe thinking back to those

early days of Classifier and then starting to build a sourcing engine, about this search for new relevant data. So you just described, obviously, things that are going to be massively impactful on revenue. But what other types of categories are there? And maybe the broader question is, what is your philosophy behind searching for new data sets? You know, it's funny. When we first started the classifier, we had never heard of Kwan hedge funds, had never heard of Renaissance technologies or any of the Kwan hedge funds until one reached out to us and said, we should be doing this with you in the private markets. And when that concept appeared, the entire concept of looking for new data sources and the frameworks with which many quant funds evaluate search for

new data sources came into our view. So early on, we were not sophisticated at all about this. We kind of would stumble across a new data set or someone would get an idea in the shower and we'd try and look at that. Today, we have a pretty robust methodology that the product team has built at Circle Up to find and evaluate new data sources. We care a lot. We care a lot about interpretability, particularly for models that the investors will be using. There's some models that we use that are black boxes that around entity resolution, industry classification that they don't need to interpret. But many of the prediction models are very interpretable. So we look for data sources that are trying to solve core problems for us. So as an example, one of the things

we look at is trying to understand how unique a product is. It was driven, back to the original question, it was driven by a human heuristic that product differentiation in the consumer space matters. If you just have a copycat of vitamin water, you might get shell space for a little bit, but it's not going to end up being successful. We then had to go out and get data on products, packaging, ingredient decks, nutritional profile, etc. To test out this hypothesis. So unlike the packaging, would it be like, are you literally visually breaking it down?

It's exactly what we're doing. So like computer vision stuff and colors and things like that. That's exactly right. And so I'll give you an example in the snack bar category, RX bar, which is a snack bar sold for a ton. Yeah, sold for $600 million. They raised a total of $10,000. That bar, when it came out, it kind of caught investor and consumer's eyes a little bit. Well, more consumers because a lot of investors passed on it. But it captured the consumer eye because there weren't a lot of words on the package. It doesn't sound like a big deal, but it stuck out in terms of how clean it was. So one of the things we looked at with. computer vision is we took apart the package of that and I think we tracked about two or three thousand other snack bar companies and money more than that products. And we just looked at the number of words on the front of the package and looked at a correlation between that and revenue growth. And it turns out that there is an inverse correlation, meaning more words,

the less the company grows. That is an example of something that we would never invest solely based off of that metric. That would be a very bad investing strategy. But we try to combine that with many, many other factors to look for things that make intuitive sense to us, are interpretable, and are correlated with future success. So this next question is about this exact, you know, snack bar idea, but it's a broader one too, which is I'm obsessed with this William Gibson book pattern recognition. And there's a little quote in there that's something like commodification will follow identification. So let's, so you've identified a trend here. So simpler, fewer words, more distinct catches the consumer's eye. Rxbar gets sold for 600 million. Probably wouldn't surprise me

if now all the marginal new stack bar companies, you know, mimic them. And the, what made them unique is now a commodity and no longer unique. How do you stay ahead of that kind of thing? Is it, is the evaluation of, let's say, uniqueness as a trait that's related to revenue growth? Is the half-life of these models very short? So to date, we have found that they are not very short. Now, the life cycle may shorten over time as information about what breed success increases as more readily available to entrepreneurs. But I, think like any industry, just because one company had success with a strategy, doesn't mean that everyone copies it. I mean, we see that even in the public markets where something that may work

for someone, and I've heard this with some of your guests, other folks may not copy as readily as perhaps that they should. Sure, good point. Or when Sam Hinky was on talking about what he was doing in basketball, not every NBA team wants to copy that for some reason or another. In this case, I think that kind of copycat mentality may occur over time. I think it is a long period. And I think we keep going back to the general thesis, which we've been able to prove out, that differentiation matters. Uniqueness matters.

So if everyone moves to clean label, the clean front of the package and fewer number of words in that small, silly example, then something else will matter. Some other type of uniqueness will matter. And it'll be our job to identify that. One of the things that we find in public market quant research is it's just very difficult to find something marginal that's useful that improves your core model. And inevitably, there are, I think in most predictive models that I've worked with, a pretty small number of them them that really drive the vast majority of the predictive power.

So, you know, everyone is familiar with what I call the commodity factors in public markets, value, momentum, quality, volatility, etc. and finding stuff on top of those, it's just really hard. You can improve them, and we believe you can improve them a lot, but the fact is they drive a huge part of the forward outcomes. So you mentioned differentiation or uniqueness, maybe as one of those factor themes. Talk about maybe the handful of others, broadly speaking, that you think most drive the outcome that you're after. Like, what are the dominant predictive factors in this world? Yeah, I will. And I'll also give one that most consumer investors think is really important.

and we have found is not. Yeah, those are my favorite. So we have looked at a lot of factors, hundreds and hundreds. The things that matter tend to fall into buckets. So distribution, brand, product uniqueness are the buckets that matter the most. Distribution, so online or offline distribution, if it's offline, it is basically breadth and quality of doors. Doors meaning are you in one whole foods or 400 whole foods? that data is outrageously valuable and it's really hard to pull together. It's really messy.

There are data sources like retail level sales providers, IRA, Nielsenor 2 that provide it for a small number of the largest stores. They focus on grocery only and it's usually the top 50 or so retailers. They don't track the long tail. And that's what's so critical in this space to identify innovation because innovation tends to not start at Kroger or, say, Air Albertsons, it tends to start at one of the smaller chains that isn't tracked by those providers. So we have to find that data. So breadth and quality of distribution is one. Another is brand intensity. So brand intensity, what I mean is if you think back 12, 15 years, vitamin water had an intense positive relationship with the consumer. Today, it's owned by Coke. People don't have their relationship anymore with vitamin water. But at the time, people loved it. They tell their

friends about it. Isn't it neat how they have these little sayings on the side of the bottle, et cetera? And that brand relationship with the consumer is expressed in a lot of different ways. This is one of the beautiful things about consumer. In, let's say, enterprise software, no one's talking about that business publicly. In Vitamo Waters case, or RX bar, halo top, those are brands that people talk about. They talk about whether or not they like them. So they're talking about it on social, on reviews, et cetera, blogs, et cetera, all that information is out there. And so we look for positive and negative sentiment.

and it turns out that those kind of signals are very correlated with success of the business. Success going back to the objective measure of revenue growth. And third is product uniqueness. But distribution and brand are two key ones. How much of this is the absolute level versus the rate of change? So if you're predicting revenue growth, does it matter more that accompanies in 400 Whole Foods or that it's moved from 200 to 400? It's a complicated answer that we haven't nailed yet. And it's in part because.

because it varies based on the size of the company. We focus on one to $15 million revenue companies, but even within that, it varies a little bit, and the category itself. So we're talking about consumer, this massive industry, as if it's all one thing. But there is still personal care, which is sold through, let's say, Sephora versus food sold through Costco, etc. So it depends, is the short answer. But to sum it up, we have found a better correlation with dollar revenue growth than percent revenue growth. What about the difference between, you mentioned online and offline, sales. Any interesting insights that you've discovered there, whether or not the kind of direct to consumer online sales tells you something interesting or matters more or less than just being

in a normal grocery store? My VC friends in Silicon Valley will not like this. Offline matters a lot more than they think it does. I mean, it matters a tremendous amount. And I'm not going to say that DDC businesses, there are some DDC businesses that are being built that are really interesting businesses. A fun game to play is to make a list of all the DTC companies that have ever existed that have had successful exits and raised more than $50 million. That list is very, very short. There's a lot of DDC companies that have raised a lot of money. Very few have ever had successful.

Like Dollar Shave Club is one, but then the list gets really, really short. Yeah, it's also interesting how they all, a lot of them are opening stores now, right? Like the Warby Parker phenomenon that I've talked about before with some guests. I think DDC is an amazing channel to iterate on a product, to basically A, B, test a product, because it's hard to do that with a physical product on a shelf. if you have to pay slotting and change it out, et cetera. But to scale a business, the customer acquisition costs online is just brutal for these companies.

And tech vCs tend to think, gosh, because I can strip out the costs of going to a store, this thing is going to be a higher margin. They don't take a step back and say, RXBarr sold for 600 and raised a total of $10,000. Halotop just had a billion dollar valuation and raised a total of $2 million. But my company is raising $100 million. Why is that? It's because these things aren't profitable. selling only online is a really difficult way to make money. So the thing that we've learned,

which is intuitive to us, but we've been able to show it, offline matters a lot still. I don't want to forget the, you mentioned there was one variable that you said consumer investors think matter, but doesn't. If you ask any consumer investor or just someone that's lived and breathed the space, they would say that velocity is, if not the most important metric, one of the top two or three most important metrics. Meaning like how quickly something sells through the store. Exactly. Exactly. So units per store per week for a given skew. So if I've got two granola bars selling at the same store, one sells five units per store per week, the other sells three, the one that sells five must be better. Across tens of thousands of companies, we have not been able to show any relationship with success. And that has been just mind-blowing. It's been something that we haven't really talked about publicly before because I can't give any economic intuition for why this is true. Other than this, we think the only possible explanation could, could be that buyers, they are the people that make decisions at a given retailer. So I decide what

goes into my store may value uniqueness, a different sale to a new customer more than just whatever is the top performing skew. So that might be the logic instead of having Budweiser, Miller, Coors to have Budweiser Miller and the craft brew. Even if the craft brew doesn't sell as much, it brings in a new consumer into the case. Yeah, really interesting. But it is it is the metric that every consumer investor pays a lot of money for, and we have not found a relationship with success. All right. So the last part of the answer there brings us to the all critical price component in all this. So I think my general belief in the early stage investing world is that the understanding of mispricing just isn't baked into the process well enough. So maybe they're good

of identifying great future brands or revenue growth, but any good revenue growth could be destroyed by too high evaluation. So talk about the discipline behind. valuation, how do you think about it, what are the trends that you find interesting and maybe any data-based works on valuation? We've talked pretty much exclusively about the E side of the equation if you think about it like a price to earnings ratio or price to revenue ratio or something. So talk about price discipline, valuation, and mispricings. I remember you had Michael Ricci from Newberger Berman on a while ago and he talked about pricing in the public markets and evaluating the strength of the company, the operating company itself versus the price, which might be impacted

by the behavior of the public markets. In our case, I think our problem is a little bit easier from that regard in that we don't necessarily have to predict how the price will respond to the public markets. We don't have to do that. Pricing for us, it is a revenue multiple. And we have a ton of data about what fair is, about what market is. We use that data and then a rough sense of how that historical multiple has compared to revenue growth and apply it to a given company. So if the average in a category is 4x revenue and the average revenue growth for that data set is 200% and this company is growing at 150% that'll cause us to dial back that multiple a little bit. It is candidly still a bit human heuristic driven.

We're working through building a rules-based system that we haven't finished yet where we would give assign points effectively that lead into evaluation process, but we haven't done that yet. I wonder if you could tell a story of a brand and how it kind of moved its way through your process, both through Helio, but also through Circle Up as an investment firm, just to highlight some of the process through a story instead of just through the framework. There's a company called Liquid IV. Liquid IV is a hydration company. So think of it as a powder that you pour into water. And 75% of Americans have a hydration problem. I mean, they don't get enough water. They don't get enough hydration. And they're dehydrated. So doctors tell

you to drink 10 glasses of water a day. No one drinks 10 glasses of water a day because it's really hard to drink 10 glasses of water today. With liquid IV, if you pour it into water, it's the equivalent of having three glasses of water. One glass equals three glasses. And this company, we found solely through Helio. I'd never heard of the company. Kindly had never even really looked at the category as an investor. It wasn't found in any of the typical ways an investor in this space would find companies, which is, and this sounds like a joke, it's not. Typically investors just go to trade shows. That's all they do. This company doesn't go to a lot of trade shows. We found it through Helio. So we were looking in Helio for factors that we have seen to be predictive of success, one related to distribution, another one related to brand in this case. This company spiked.

Reached out to them, developed a relationship of a couple months, and tried to invest. When we first started talking to them, they were a couple million dollars in revenue. And we closed the round a few months later. They had a term sheet on the table for evaluation that was about 50% higher than ours. They took hours because they wanted access to Helio. They wanted access to the technology. That's one of the interesting things about the space. There's such a lack of data and information in this industry. People are hungry just for transparency about how do companies grow? What leads to success?

What can we do to help ourselves succeed? So this company took a meaningful haircut evaluation to work with us because of Helio. A year after we invested, the company had grown almost by 8X in a year. So that's over the last 12 months, still relatively early in the process, but we're thrilled with how the company's doing right now. Really interesting. What are the most interesting, I guess, subcategories to you? And what are the structural reasons why that might be the case? We've mentioned a bunch of examples, hydration, snack bars, personal care stuff. Are there sub industries that you find most appealing? And if so, why? We have a model on category that looks at things like category momentum, also the proportion of a given category that is, dominated by stale incumbents. So categories where incumbents dominate a high percentage of the category historically have been categories that are ripe for disruption. We've been able to show that

with data. Those are the categories that I like. To be frank with you, I don't spend a lot of time myself as the investor picking the categories. We more look at it through that lens of what's working in the data. The future for this model. So you're focused on consumer and retail now. How much bigger is the umbrella here? So I'd like to talk now more generally about your take on applying quant to private markets. What's the timeline for that? Certainly some big firms, Two Sigma being one has begun talking about this or experimenting in this world. What's the timeline? What's your take on this as someone that's deeply in it? And how quickly will it expand beyond business model like consumer that's that's so repeatable? We think quant in the private markets

is going to be a meaningful part of the future. I think it'll be here sooner than, and a lot of people gave it credit for. In the public markets, it took years and years, even with the advantage of being able to backtest models fairly efficiently. I think, though, that that has helped pave the way. And I think there's a number of different dynamics that lead quant in the private markets to be something that will catch on much more quickly. First, there are LPs and asset managers alike who are desperate for access to the private markets, but want a scalable, repeatable investment strategy. The problem with VC or private equity is it is not scalable or repeatable, if done the traditional way. You basically hire a team of star

analyst, star private equity or VC pros, and they source companies the same way today that they did 30 years ago, which is they go to a trade show, they go to a cocktail party, whatever, but it is manual. And you can't figure out however someone found Uber or Snapchat, they can't. They can't repeat that into other things. That's a problem for large asset managers and large LPs who want to believe in something. So just to push back in that a little bit. So, and I don't know this data intimately, so I don't want to hold it out too strongly. But there certainly seemed to be a top flight of VCs that have delivered pretty consistently the best returns of the 700 plus VC firms. And that sure seems like repeatability to me. That's a great point. So how do you think about that? Yeah, that's a great point.

So I wouldn't say it's scalable, though. So most of those firms, so think of Sequoia's main funds. Sure. So it came to mind. Yeah. Sequoia's main fund or benchmarks main fund or Union Square. Most of the funds that we'd list at the top five that have stayed in the top five have remained relatively consistent terms of the size of their funds. Right.

They're not huge. It's right. And so if a big endowment comes to them and says, we want to give you a $3 billion, it's really hard to do that. Whereas in the public markets, not all strategies. but some strategies can scale much more easily than they can in the private markets. I can't take Bill Gurley at Benchmark or Fred Wilson or Andy Weissman at Union Square. I know you had Albert on the show and replicate them infinitely. So what their brands do and they do is a really good job of maintaining their position with what they're doing,

but they can't scale the fund infinitely. We think that you can scale a quant VC fund much more easily because you're not relying on just Andy at Union Square or Friday Union Square to make decisions themselves, you're relying on the models. And that, I think, is going to be extremely attractive, the ability to know why I found a company and be able to replicate that over and over again. Now, I don't think this works in every market to answer your other question. I think that there are some markets where this will be really tough. Tech is actually an example. So in tech, let's go back to the framework that we had before about why this works in consumer. Same business models, a ton of data that's out there.

In tech, there aren't the same business models. But on top of the not being the same business models, there usually aren't prior examples of success. So when Uber hit, there weren't a hundred other Ubers. In the case of RX bar, a snack bar that hit, there's a lot of other food companies that looked really similar in terms of those hitting. That's an issue in terms of building the training data to know what success looks like. So now I see analysis, especially on Twitter, of the scooter companies, bird or lime, and they compare them to the ride-sharing companies of Uber or Lyft, you're comparing it to an N of two. And yes, the growth patterns look similar, but we've got two companies we're comparing to two companies. That doesn't pass muster with a lot

of data scientists. That worries me about tech. The other thing that worries me of tech is the lack of data. So we see some VC firms that are trying to be data-driven. They still have human, heuristic, driven investment committees. There's a team of folks sit around and make a decision, yes or no. It's not a systematic fund. It's just a data-driven VC fund, which sounds novel in VC because VC's traditionally haven't used a lot of data. But in the case that they're pulling data, they might have one engineer in house. The data that they're pulling is almost always LinkedIn data. It's data on where engineers are moving. I view that as a derivative metric of success. So just because a company raised a lot of money and can hire engineers does not mean that that is a

good company. It may end up being a good company, but there's a lot of amazing. tech companies that raised very little money, or there's a lot of bad tech companies that did hire a ton of engineers and didn't work out. We think, though, that the other problem in this space is if I build an algorithm and a series of datasets, they're able to accomplish and get over the problems that I just mentioned. You still have the issue of competitiveness in that market. So if I build the helio of tech and I find the next Uber, the problem is that that company is going to call Sequoia and Union Square and benchmark. And now I have to beat them. And I don't believe that they're going to be beaten. That doesn't make sense to me. Maybe I can do it in a couple

cases, but to build a sustainable, repeatable, last manager, I really struggle to believe that. That's very different than some other industries, consumer being one, where there aren't a ton of venture style investors in that space. What are the next adjacent category? So if tech is the other end of the extreme that's hard to introduce this kind of process to, what other than consumer and retail are most interesting to you? We look for categories that have those characteristics of repeatable business models, a lot of data. So real estate is one that makes a lot of sense to me. And there are some folks that are kind of doing this either publicly or privately trying to do this in some way or another. I think that that's a space that is ripe for this. Now, there's also a

problem of, I'm not sure it's a very inefficient industry, but there still may be an edge that you can gain in real estate. I've heard some folks lean in on parts of the media industry. So whether it is music rights or certain areas in movies. So if you think of like the long tail in both music and movies, could we capture that long tail more effectively with data, with some sort of quant strategy? Not to say that you'd use it to base the next $400 million budget film on, but to do the next $10 million horror movie, it may make a lot of sense. industrial categories where there's a lot of data. It's kind of repeatable business models over and over again. What have been the biggest challenges for Circle Up the business? So this is something new, which everyone

says they want. And then oftentimes don't want to be first. I'm thinking here about LPs specifically. Talk about any challenges or interesting things that you've learned on the business side. So leaving Helio now for a moment, talking more about Circle Up. There's a lot. The two biggest challenges and I think risks to what we're doing or if anyone else does this as well, One is attracting and retaining talent. Look, I think any CEO would probably say some flavor of that. I believe we have a pretty good argument on why it's trickier for us. Our team is basically two sets of people, business folks and technologists, engineers,

data scientists. The business folks often come from finance. Well, that's a world where there's a lot of money floating around. And if you believe for a second that private investing will move more towards a quant space, or at least quant will become more of a thing in private market. markets, those firms, whoever is going to do that, is going to look at circle up and say, why don't we poach some of their people? Why don't we just poach some of them and try and pay them more? That's an issue. On the engineering and data science side, it is a bloodbath. I mean, nothing will convince someone more that we need to change our educational system in the U.S. than trying to hire engineers and data scientists right now. The disconnect that supply, demand, and balance is just absurd. And so trying to attract and retain that talent is difficult. I think that's a core risk. The other risk relates to what you mentioned around LPs. It is raising capital for the funds. In the public markets, when you've got a quant strategy that works, and you could tell me if you disagree on this, but you can often backtest it and show those results to LPs. Not all LPs will believe it. You can poke holes in back tests, etc. I understand that. But there's still something to show fairly robustly. The private markets, it's much harder to demonstrate success. The feedback loops are longer. We can do the back test, but it's certainly not. on as many companies or as long of a time period. And so being able to get LPs to believe in our

vision is a very big challenge. And that doesn't just stop then with that technology. In our case, we also have to demonstrate that we can build an operations machine, a well-oiled machine that can reach out to the companies, convince them to work with us, help them post-claws, lead them to exit, or is in the public markets that isn't needed. Sure. It's obviously a really hard string of things that you have to get right. I'm curious when investors ask LPs, let's say, why would I bother with this? So there's lots of private equity funds. There's lots of venture capital funds. Are your perspective returns better? Like, how do you handicap what kind of returns you're trying to earn through this process? So a couple things. First, it starts with exposure to an asset class

that the LPs effectively can't get anywhere else. So they can get late stage consumer, but they can't get early consumer. And that's an important difference. In consumer, just to go on a little bit of a tangent for a second to explain. In consumer large brands are losing market share to small brands. They are getting slaughtered by small brands. If you see that trend, you can play it in one of two ways. You can short the large companies in the public markets or you can go along with small companies. But private equity is stuck in the middle. Private equity, which is, there's $80 billion of consumer-focused private equity in the U.S. That is a space that is kind of, they're investing into kind of the mid-tier brands that are doing fine, but they're not playing that

trend, that massive trend that everyone consumer talks about. So the only way to play this trend is to work with Circle Up. There aren't 10 other firms in the country that do this reliably. You might have a family office that'll invest in a one food company every two years, but it's not a reliable thing. So that asset class, that long-term trend, has resonated a lot with LPs. Then to demonstrate the performance of what we've done, we've worked with some amazing brands like Halo Top or Beyond Meat, which just said they're going to go public. And to show them that the models work, that this is a scalable, repeatable process for building an investment firm. We look to work with LPs who have a longer-term vision for what we're trying to build here.

And we've been really fortunate to work with some absolutely amazing LPs so far in our current funds. But the LPs see that disconnect in the consumer market. They also see a broader trend, which is if we're investing at the early stage, let's say one to 15 million in revenue. And there's this $80 billion of consumer-focused private equity comes after us. We have two options. One, we could sell into them, right? It's a very robust market where they're bidding up the price. Or two, we can just ride the winners. Why do we need to hand it off to them? We could just either give it to our LPs as co-invest or raise a fall in growth fund. That attracts them. To your original question, the returns that we're targeting for this

systematic fund that we'll do are about 3x, which is in line. with, I think, what most VCs and private equity firms would say they're target. I think if you look at TechVC, it tends to be dramatically lower than that in reality, but they'd say they target that. We just think we have a process and a technology that can deliver those returns in a more repeatable and predictable way. But we need to prove that. Chris, to get your take then on the broader VCEPE landscape, what's working? What isn't? What should investors be concerned about? Obviously, you're a member of the early stage investment community. So what do you see there that you think people should take notice? I'm seeing a huge bubble

in tech investing to be candid with you. And there's a lot of folks in the city I live that are not going to be happy with me saying that. But tech VC, you kind of look at the things that are getting funded, the valuations that they're getting funded on and the what you need to believe in order to make money on this. And it just terrifies me. And we see that in tech VC. Then we see tech VCs also trying to expand other categories because tech has become so crowded. So we see tech vCs investing into industries where they're not as familiar. I'd invest into a food company and give the food company $75 million. That food company shouldn't raise $7 million.

And that's an issue. In private equity, kind of later stage private equity, we're seeing private equity firms really begin to lean in to data. Now, they're trying to figure out how to do it. So some are partnering with quant hedge funds in the public markets and saying, gosh, could we have some mutual information sharing? here in a partnership. Some are trying to hire data scientists in-house to leverage the broad concept of more sophisticated data sources, application of data science to what they're doing.

We're seeing that almost universally at funds above a billion dollars. The other curious thing that I think is really important that we spend a couple minutes on is the type of models and how you think about it. So there's a couple terms that you use that are good, like brainy versus brawny, diagnostic versus the other categories. Talk a little bit about those differences and where the team is most focused. We use the term brawny models to talk about non-interpretable models, often machine learning models where the end user doesn't really need to know why that prediction was made. Brainy models, in the other hand, are very interpretable.

They might be regressions. They might be other types of models that are very interpretable. The brawny models we will use for something like industry class. classification. So we need to, when we identify a company or a product from afar, we need to say it is a popcorn company, not a pair of shoes. Because if we're comparing popcorn companies against pairs of shoes, that doesn't work. So when that model, the industry classification model, or another example is entity resolution, when that model is looking at a popcorn company and classifying it, the end user then looks at the results. The end user doesn't need to know why the prediction was made.

It's either right or it's not. And sometimes it is wrong. We get it right. Actually, we've been able to show that we get it right more often than humans can. But there's still, there's not that need for interpretability. When the user cares about the prediction, about why the prediction was made, interpretability is vital to us. So the models we've talked about before are the kind of factors that we talk about are related to brand, distribution, product, and then predicting the future revenue, predicting the future distribution. Those things need to be interpretable, think. And so we are building models that they're able to dig into. we call those brainy models, that they're able to understand, well, why do we think that this

company will go from 100 doors to 1,000 doors over the next years? That's really important to us. Most of the team is more focused on the brainy models, but there are a couple vital brawny models that we work on, particularly related to entity resolution. Everyone is thinking about those brainy models. Everyone's so interested in brand now, especially around social. You mentioned before sort of some of the dimensions that matter about brand. But maybe you could talk a little bit in more detail, because for anyone building a business out there or investing in these kinds of businesses, it's something that is kind of fuzzy. You know a good brand when you see it or something. But what are the quantitative elements of brand that you found to be most interesting? It is very fuzzy.

And it's the type of factor that if you talk to consumer investors, they will all agree that brand is important, but that it takes 10 years of experience to tell you what a good brand is. We took the viewpoint that we can't replicate what the investor sees necessarily, but we can replicate or we can evaluate what the consumer is saying. So we're kind of skipping that middleman of the investor and just going directly to the consumer. And we're looking at what they're talking about and how they're interacting with a brand. And we find that what we see people say in reviews, on social, how they interact, there's a high correlation. between that and the performance of these consumer companies. So I'll give you some examples. When a company tweets, send me a direct message, that company is three times more likely to not end up

being successful than when it doesn't tweet that. So if you think about why that is, why would a company send someone a message which says, send me a direct message on Twitter? The answer is because they're responding to some negative comment on Twitter where the person's complaining about a product and they want to take it offline so that no one else can see that conversation. That's another example of like it's a really small detail that we look for other things like that that have high correlations with success that also make intuitive sense. We've also found, and this isn't particularly proprietary to think other folks have found this too, that number of stars, for example, doesn't really matter that much.

It's a lot more about the volume of engagement with a brand. So we look at quality of engagement. We look at amount of engagement. We look at the rate of growth of engagement for a brand. We look at a number of those factors to try to understand how that brand is connecting with the consumer. We have done a little bit of work also going back to the product. The product is often the best brand billboard, and we've done a little bit of work to understand how that brand billboard relates to the performance of the product from a brand standpoint.

Haven't found enough of a correlation there that we get excited about yet. Those are the things we look at. So I'll take one of your three categories and apply to the other two. So uniqueness matters. What is the most unique brand that you've seen? And what is the most unique distribution strategy that you've seen? I think the most unique brand we've seen is Halo Top. Halo Top was a company that was doing well, 2013 to 2015, doing very well.

It's ice cream, right? Ice cream company, excuse me. Yeah. So Halotop is an ice cream company, independent company, now sells more pints of ice cream than Ben and Jerry's and Hogendos. And 2014, 2015, the company was doing well. And we worked with them. If you look at the package, though, the package had nine or ten different claims. So I don't remember all the claims, but let's say it was gluten-free, sugar-free, everything-free.

And it was overwhelming to the consumer. You looked at it and you just got dizzy. They changed the package sometime around mid-2015. And we can demonstrate we've done this in a blog the growth of the company before and after that point. They changed the package to strip out all the ancillary claims and just focused on one. thing. They had an insight that when you eat a pint of ice cream, you either have two bites and don't feel satiated, you don't feel full, satisfied, or you eat the entire pint and you feel guilty by yourself. But either way, it's a bad experience when you're done eating the ice cream from the

pint. They had the insight that if you could give the consumer permission to eat the whole pint, that they would love that as a different quality dimension. See, ice cream companies historically have just focused on one quality dimension, which is taste, which what is the best tasting? It's really hard to beat Ben and Jerry's on taste because it's not very good for you. And I'm from Vermont. And in the case of Halotov, they said, look, our ice cream is going to taste good enough, but we're going to destroy everyone else on another dimension, which is number of calories in the bind. So they took out all those claims.

And then in the front of the package, they just put one thing, which was the whole thing is 300 calories. When they did that, sales took off like no one has ever seen before in the consumer space. It is probably the most successful consumer company in the last 15 years. And it was funny, it took off and we can show this in a graph, about a year later, after it began taking off, I got calls from 18 of the 20 best consumer private equity firms in the country, all in a two-week period. I mean, it was an absolute clown show. they are all reaching out one after another and saying, hey, can you make an introduction to Halotox? They knew we had worked with them. What was happening is they all got the exact same data dump from the exact same data provider,

which is a retail level sales provider. And then they all saw the same thing, which is this company is taking off. They wanted to then trade on that, but that's the problem with everyone using the same source of commoditized data, that they were focused, that they were all trying to get in. In that point, the company had already passed. So that, to me, is the most interesting, unique brand. they were able to hone in on that value proposition. Before we get to distribution, I don't want to forget that because I'm fascinated by

distribution strategies and channels. When you see you worked with Halo Top, like what specifically does that mean? So we used to have a marketplace that companies would come on to raise money from other investors. This was before we had a fund. The marketplace, and the investors might be family offices, might be individuals, might be really small funds. Like almost like an angelist type set up.

Yeah, exactly. And that marketplace was where Halotop raised money twice. So anyone could have invested into HALA. hillotop, and several folks did, but other folks could have and passed. Sounds good. Okay. So back to distribution. So most unique, interesting distribution strategy that you've seen. And this doesn't need to be quantitative. It could be qualitative. I think, you know, we talked before about D to C, direct to consumer, and how I think DDC is an amazing way to test a product, a very difficult way to scale a product. The exception to that has been

probably the most interesting strategy, distribution strategy that I've seen in the last 10 or 15 years, which is Dollar Shave Club. Dollar Shave Club was able to build a pretty sizable business before it sold for a billion or so to Unilever, basically solely through D2C. And when you ask about distribution, there were two things that I thought were interesting there. One was the distribution of their brand. They recognized, and unclear how much of this was fully thought out, but I think a large portion of was. They recognized that we were moving from an era where you paid kind of fixed fees for marketing, buying an ad in Us Weekly or whatever it is, to social, where it was effectively variable cost marketing. And they put $50,000 or so into a minute long ad on YouTube.

It's an ad that's been seen between $25 million. I could picture it. Yeah. I remember it very well. And look, it's no one can ever build an ad that they know will go viral. That thing happened to go viral. But they relied on that for a couple of years. And what I was impressed by was that they did not take another road, which was, okay, let's now go raise money to put into the Us Weekly ads or do the traditional. They relied on that channel. They also then stuck to the DDC channel, I think, in a way that gave them the opportunity to get bought by Unilever. And what I mean by that is, by all accounts, the metrics on that business would not have justified a billion dollars. Unilever was buying capability. They weren't buying Dollar Shave Club. The capability that they were buying was the ability

to sell products online in a DDC fashion. And the hope was that Michael, the CEO, that the CEO and the team there could help teach Unilever how to sell other products online. Similar analogy was with Jet and Walmart. Walmart wasn't necessarily buying Jet. They were buying e-commerce capabilities. Sure. And the CEO, I think, who had done that a few times before, like we built these fast companies. It's pretty amazing. Since you're doing something unique in this world, I'm always curious to hear who has influenced you the most, maybe in your past or actively today. Who are some mentors, people that you think are doing, pushing kind of boundaries, doing interesting things across the investing ecosystem, not necessarily just in your corner of it. There's a lot of folks. So personally, my parents have influenced me a ton, mostly just in terms of work ethic, value. and helping me to believe in myself. But in terms of professionally, a number of folks, some of which

you've actually had here on the podcast that have influenced me longer term and more recently. Jeff Jordan at Andreessen Horowitz, I've learned a lot from him just in terms of VC tech. That market historically has not been a market where there's a lot of good actors, where people are able to maintain their value system. Jeff has done that in just an amazing way. He is a really high quality guy who also has built some amazing businesses and also been a phenomenal investor. And his ability to maintain both a set of values, be a good person and be successful in that world has been just truly inspiring. Just to pause there. So if you had to sum up kind of your personal values, what are they? What are the big ones? Integrity. I care a lot of integrity

from the people around me, from myself, perseverance, being able to run through difficult things. I think as an entrepreneur or my personal life, being able to get over challenges has been a really critical thing for me. And I look for that from other folks. Another that we have as a company value is trying to just be a solution. I really don't like spending time with people who just complain, which is part of the reason I'm sometimes hesitant to read. Twitter. I post a lot on Twitter, but I'm hesitant to read it because there's just a lot of complaining there. I like folks that are able to offer solutions and come to a conversation with the opportunity to build, not just tear down. So those are three, integrity, perseverance,

and being a solution. I came across a great quote in a book written by a guy I met with yesterday in the kind of family office, private wealth management world. The quote was from 100 years ago or something from the founder of the carnation kind of milk company or that product, which is at the best, the ultimate measure of success is to be useful is how useful you are. I thought that was pretty unique and really resonated with me. I love that. So I'm sorry, I interrupted you on people that have had a large influence. So Jeff, another person that has had a lot of influence of me in the last five or six years is Matt Christensen, who is a son of Clayton Christensen, or wrote a book called the Invitational Lema. Matt was one of our first investors. He's had a big

influence in me for a couple reasons. One is just able to see not the five-year vision, which you kind of get sucked into a bit as an entrepreneur, particularly because investors talk about it, the team talks about, et cetera. He's really pushed me and us to focus on the 20-year vision. This can change industries over 20 years. And it is particularly refreshing, inspiring to hear that from an investor. That's been amazing. Another person is Sam Hinky. I know you had on this podcast. I love analogies from other categories. And he is as good as anyone I've ever seen at being able to pull obscure analogies from categories or situations that I've never heard of and learn a ton from. So he, he's a friend for the past 10 years or so. But in particularly as an entrepreneur, I've learned a lot from him.

So Sam's a good excuse to ask about long-term thinking and specifically goals. So and the idea of having a 20-year vision instead of a couple-year vision. So just like your philosophical take on goals, how useful they are, whether or not you use them at the company. What are the pros and cons in your minds of goal setting? We use them at the company and I use them individually. I think it's interesting the way you ask that question because I do think that there are pros and cons. And I'm not sure we've totally figured out how to balance the pros and cons. But as a company, we have, we called MoKRs, objective and key results.

We took it from there's a great. John Doer. Yeah, John Doer, and there's a great Google Ventures video about it. We set them quarterly and we have year-long goals. The goals that are for the company, then each of the teams has goals, and individuals can also have their own goals. We have found them to be an interesting framework to align on where we're rowing as a company. we also find it to be constructive to help prioritize on their own.

And those frameworks, we think, help to not prevent, but limit micromanaging. Meaning if you and I agree on what your OKRs are this quarter, go run, do what you need to do to get those things done. But if we agree that your OKRs are X and Y, and you end up working on A and B, that's a problem. We need to have that conversation. So we find it really useful to align and prioritize. Now, we also struggle with a short and long-term balance. So you talked before about longer-term vision.

There are things that we know we should do now to make us more successful in a decade. But if we only focus on those things, we won't be here in a decade because we won't have a business in three years. And that balance is really, really hard. And so then you get into, well, maybe I should set five-year goals and one-year goals and six-month goals. And at some point, when you add so many goals on, they become, to me, a bit worthless. It's almost like you see some companies have 10, 12 values. You ask someone that works at that company, what your company values are.

They can't name one. So we try to limit the number of goals to three to five so that they're actionable and memorable, but that long and short-term balance is very difficult. So probably the best, I think, thinking on. on this notion of long-term thinking comes from Bezos, which is like, we only care about the things we know will still be true in 10 or 20 years. And we can optimize around those, but predicting stuff is really hard. And so for them, it's, you know, the customer will want stuff, you know, better, cheaper, faster, right? Like, that's pretty high degree of confidence that that's going to be

true in 20 years. Do you have any sort of similar unchangeable, things that you just know or are very confident will still be true of your clients and what they want in 20 years? that helps you determine that long-term vision? I think so. I don't know if it's as crisp as what Jeff Bezos would say. But a couple things that the team that we have all aligned on as a team. One is that our mission is not changing. Our mission is to help entrepreneurs to thrive by giving the capital and resources that they need. So what does that mean? A real example is, and this is going to sound like a little bit of a humble brag, and I really don't mean it to be. We've had 30 public consumer companies reach out to us in last year and said, we want to give you

money for Helio. We want to use Helio to help us. And when you kind of scratch the service on what they want to do, well, they want to help grow their $5 million yogurt line, $5 billion yogurt line, rather. That's not consistent with our mission, right? And that's not going to change. We don't want to lean in on that. It could be short-term revenue. It could be amazing short-term revenue.

But it's not going to be consistent with our mission in 2018 or 2028. That has acted as North Star. Our vision, our mission, and our values are things that, don't change. And that helps to set an interesting balance for us. Our values are do it right, be brave and be a solution. And so, you know, there's also been cases where we will get a business opportunity or an opportunity to do something with a product that may not be consistent with one of those values. That's happened in the last six months. And we, someone on the team will raise their hand and say, hey, a little bit of a flag here, this isn't consistent on our values.

those things help. I think in terms of the entrepreneurs that we're serving, we don't have any plans to move outside of consumer. So that's another way that we focus. We get opportunities and asks to, can you move this into health care or real estate, as we talked about before, education, whatever it is. That's not something that's interesting to us. So those are some of the building blocks or I guess constraints that we have to help us focus. But it is still a challenge. And that long versus short term tradeoff is still a very difficult. difficult one to make. What is the most interesting individual conversation that you've had across all this journey? The second conversation that I had with one of our investors, Andy Weissman, he pushed us to lower our pricing. It was then a marketplace, right? And we had a pricing strategy. And the pricing strategy was working really well to be candid with you, which is, I think, one of the reasons that they were interesting. And he pushed us to lower our pricing. It doesn't sound like a big deal, but from the world that I came from, I used to work in consumer-focused private equity. And the concept of trying to lower price when things were working was just totally antithetical to what I believed to be right in the world.

And so I remember him asking me, why don't you drop price as just a pivot point in my career where, and I remember where I was sitting at the time, with my co-founder Rory, I remember it being so important because I immediately realized, A, I don't have an answer for that question because I've never thought about that question before because it's totally different than anything I could imagine. And B, this is a new world. What I'm doing as an entrepreneur is very different than what I was building or what I was doing as a private equity investor. And it began to help inform and reshape how I, how I appreciate how I approached a lot of different conversations. It gave me a thirst, candidly, that I did not previously have to talk with other entrepreneurs about the problems that they were solving, talk to other

investors about our problems, to poke holes in it. It gave me a thirst to learn from different types of people. I'd always have a thirst to grow and to learn, but it opened my mind to new frameworks, new mindsets that I hadn't yet had. It was a small question. I'm sure he doesn't have remember it, but it transformed how I thought about things. So two closing questions for you. The first is if you could only keep one of your data sets, what would you keep? Oh, gosh. I'm not sure any of the factors themselves would be valuable without the training data. And the training data that we have is by far the most important data we have. So the training data just give a little bit of background. We've got factors that I mentioned before, like distribution, brand, product, et cetera, that we seem to

be correlated with future success. The only way we're able to see that is because we have private financials on thousands and thousands of consumer companies. We have by far the largest data set in the world of financials for consumer companies. And that's because they gave them to you. They've given it to us for six years. So it started because we had a marketplace and then because we have credit and equity funds. And they were willing to give it because they needed the capital and resources to thrive. It's unlike if you hang a shingle in tech, I've got 700 other options I can go to. And consumer, They don't, which is why amazing companies like Halotop or Beyond Meat, Rillen, to work with us. That training data set is core to everything we do.

So that's the one data set that I would keep. That's a great answer. So my closing question for everybody is for the kindest thing that anyone's done for you. I love this question. I grew up in a small town in Vermont called Shelburne. And both of my parents were small town lawyers. They worked together in a private practice about 100 yards away from where our house was. And I started playing basketball relatively late in life around seventh grade or so.

And by the time I was a sophomore, I knew I wanted to, sophomore in high school, I knew I wanted to try and play in college. And so my dad used to leave work every day to go rebound for me. So he'd leave work at around 3 o'clock. And we were middle class for him to take off a couple hours in the afternoon every day. He was for going income. He didn't work as part of a larger firm that would still pay him a salary. When he didn't work, he didn't get paid. And so he did this every day for three years. He'd come and rebound for me for two or three hours. And then we go home and have dinner and do homework.

And one point, midway through my junior year, we were kind of talking while I'm doing shooting drills and he's rebounding. We're talking about college basketball, not about me to be candid with you, but about, I think, Syracuse. And he said, I shot the ball. He said, he He caught it and he held it. And he wouldn't pass it back to me. And he said, you know, you could play for Syracuse. And I just, Dad, I'm a mediocre player. I'm barely starting for a poor team in a not a basketball hotbed.

To give you your sense, Vermont hadn't had, you can count on one hand the number of people that played college basketball outside of the state of Vermont in the last 20 years. And I said, dad, that's, that's, I'm not going to be able to play for Syracuse. That's never going to happen. He said, you don't believe in yourself enough. If you saw what I saw, you could do it. You could do anything if you believed in yourself more.

And it sounds, I don't know if it sounds corny or whatever, but just he just said it with so much conviction and so much belief. And what I think about, I've thought about that moment at so many different points in my life. When I ended up going to Duke for college and I ended up trying to make the team. I ended up playing at Duke, trying to get into consulting, trying to get into business school at Stanford, a number of different things, even starting circle up. And I look back at that memory and at each of those points of my life, and I'm just so grateful that he in that moment and in so many other moments was able to inject in me confidence and a self-worth that led me to just want to try.

I think that that is so critical to being an entrepreneur. But more importantly, I'm just, as a parent myself now, I think what a special thing. Is there anything better that you could give your kids than the self-worth and confidence to want to try? So that's the kindest thing anyone's ever done. Good stuff. Sounds like your dad's a guy I'd like to meet. Absolutely. Thanks for all your time.

This has been super interesting. Obviously, this is quant applied anywhere is right in my wheel. house, and this seems like a really emergent use of that general idea. So thanks for your time and all the insight. Thank you so much for having me. Hey, everyone. Patrick here again. To find more episodes of InvestorFieldguide.com forward slash podcast. If you're a book lover, you can also sign up for my book club at investorfieldguide.com forward slash book club. After you sign up, you'll receive a full investor curriculum right away, and then three to four suggestions of new books every month.

You can also follow me on Twitter at Patrick underscore Oshag, OSHAG. If you enjoy the show, please leave a quick review for us on iTunes, which will help more people discover Invest Like the Best. Thanks so much for listening.