Justin Yoshimura - Resurrecting & Rolling Up Retail - [Joys of Compounding, EP. 35]

Alright friends, today's episode is brought to you by Portrait, the AI research platform built by fundamental investors, four fundamental investors. Now we've become huge fans of both the people and the product behind this exciting company, and so we are fired up to be collaborating with them in season 3 of the Joc. For founding Portrait, David Plahn spent a decade as a Bicide analyst at some of the world's leading investment firms, including our friends at SlatePath Capital and the BALPOST Group. Now even as some of the best resource firms in the world, David kept bumping into a reality that may resonate with many of you, never having enough time to research all the compelling ideas across his desk. This tension inspired him to build Portrait. The research platform that delivers the equivalent of an army of capable analysts, ready to help you focus on the most actionable ideas with unprecedented speed and depth. Portrait isn't about replacing judgment, it's about giving you back time for the work you actually love, developing novel ideas and building conviction through deep, creative, value-added research. Portrait is what makes Portrait unique. They provide nuanced idea generation. Portrait assesses the same qualitative attributes we value as investors to identify businesses that fit your specific mental models. It's not just screening, it's pattern matching with soft qualitative attributes in the way great investors think. Portrait also gives you custom research reports. It generates comprehensive primers and lays out bull bear cases in minutes rather than days or weeks. The system is trained to help you process pressing research questions, delivering final outputs at the caliber you would expect from a member of your team. And finally, Portrait offers intelligent thesis monitoring, tracking thousands of data points across entire value chains daily, extracting the key insights that actually drive the business you follow. The result? You will become much more prolific in your sourcing, analysis, and monitoring capabilities taking steps toward experiencing the joys of compounding even more abundantly. So visit PortraitResearch.com to start your free trial and see how David and his team can help add productivity and creativity into your team's process. Before we get to the episode, I want to tell you about a product I use every night, the pod by 8-Sleep. For years, my co-host Rick has been bragging about his sleep quality and the transformative nature of his 8-Sleep pod. Despite respecting Rick's genius on many fronts, my wife was skeptical about a bed-dowmaker cold, so I reluctantly agreed to continue without 8-Sleep. Fortunately, as a gift on our recent wedding anniversary, she agreed to my simple ask, "Do we give 8-Sleep a shot?" Needless to say, it's been instantly effective. From the very first night, our sleep quality and duration have been off the charts, and with a predictable cadence that's transformed this incredibly important and restorative part of our lives. The 8-Sleep pod mattress cover is super easy to set up and operate, and surprisingly, even though you can cool or heat each side of the bed, my wife has become obsessed with cooling her side of the bed, and we're getting our best sleep in years. If you're trying to get serious about the quality of your sleep, and fully embrace the joys of compounding, you have to give 8-Sleep a shot. Find out why so many of our good friends like David Centra, Patrick O'Shaughnessy, Brent Beeshor, and many others love their 8-Sleep. And get up to $350 off when you go to 8-Sleep.com/joys. That's 8-Sleep.com/joys. We're still on the same mission. To study greatness in order to help you find and compound your life's work, is fast as possible for as long as possible. I'm Paul Buzer, and I'm Rick Berman. For your hosts, in each session, our teachers will be some of the world's most compelling people from across the vast range of human achievement. This show is brought to you by Pine Grove Studios in collaboration with Colossus. The hosts of the show, Rick Berman, and Paul Buzer are the co-founders and co-CEOs of SATA Grove Holdings and co-CEOs of SATA Grove Management Company. All opinions expressed by any of Rick, Paul, or their podcast guests are solely their own and do not reflect the opinion of either SATA Grove Holdings or SATA Grove Management Company. This podcast is intended for informational purposes only and should not be relied upon as the basis for investment decisions. SATA Grove Holdings or clients of SATA Grove Management Company may maintain positions and securities discussed in this podcast. Our teacher today is one of the youngest, most impressive, and under the radar of leaders in all of retail. His name is Justin Yoshimura, and he's founder and CEO of CSE generation. Now you probably never heard of CSE, but you likely know some of its beloved brands, including Sir LaTob, Backcountry, One King's Lane, and many more. As a holding company, CSE does over a billion dollars in annual revenue. Is profitable and that AI enabled roll-up thing you've been hearing more about recently? Yeah, they've been doing that long before, was even a thing. Now Justin's an iconoclast in just about every sense of the word. In terms of formal education, he pulled that plug around the age of 16, not even bothering to graduate from high school, let alone college. Instead, he started to follow his curiosity and passions in a way I can only describe as radically authentic. He's motivating his own curriculum through entrepreneurship, learning by doing and refining his own point of view on what actually drives enduring businesses. No analyst training program, no associate of this company, then VP of that, he just got to work building, compounding by stacking up early reps as a founder. And so I guess it shouldn't be any surprise that he took that first principles orientation and applied it to how he's designed CSE. By building a serial acquisition holding company designed to buy, own, and operate retail businesses over decades, supported by a deeply aligned shareholder base who share that same ambitious orientation toward compounding without a clock. What makes his story especially relevant now is how CSE has approached harnessing the power of AI, pioneering a truly AI need of operating system for commerce. Not AI as a feature or a set of experiments, but AI that converts core workflows into scalable, autonomous operations, reshaping how decisions get made, how inventory moves, how pricing evolves, all in service of helping incumbent businesses not just survive, but win again. As Justin shares, CSE acquires retail brands and transition, often out of bankruptcy, and then re-architects them on their agent orchestrated operating system they call Genesis. And appropriate name by the way because the results have been absolutely outrageous, one might even say biblical. Figurally turning lost making companies into profitable ones in just the first year of ownership. Now I think Justin's story is particularly valuable to us because it challenges some of our default assumptions about how talent is best cultivated, how companies ought to be structured for greater alignment and impact, and how struggling incumbent businesses can successfully adapt as the greatest technology the world has ever witnessed begins to unfold. With that I hope you enjoy class with our friend Justin Yoshimura of CSE Generation. Hey Justin, great to have you on the Joc. Paul and I were thinking about where to begin this conversation and we got some counsel from our mutual friends over at Althos, Anthony and Ho. And one of the things that they see is foundational in you both in you as a person and you as a leader is just this alchemy between what I guess on some level is a very traditional upbringing mixed with some very non-traditional decisions you've made and the path that you've carved for yourself. So we thought the best place to start is just your beginning and maybe just a thought experiment. You're writing a book about your life. Tell us what chapter one looks like. So basically I grew up in LA. My dad is Japanese American and my mom is from Osaka, Japan. They actually had me when they were pretty young. You guys also I think had kids young, but my dad ended up working in finance and going to Kellogg, but I was born before that when you actually used to sell light bulbs for GE. So I grew up in between. We actually lived in Compton, which is not a nice part of LA. And so my dad was like, you know what? I don't think I want to raise a family in Compton. I'm going to go into finance and go to business school and have a serious finance job not to say that selling light bulbs is not serious because there's actually a lot of hard work as well. As I was growing up though, I did pretty well in subjects at school that I liked, but in subjects that I didn't like, I just did very poorly and the teachers didn't like me. And then I was diagnosed with ADHD. But my parents were basically in denial that I had ADHD. So then they just told me that I was out of control and that I would become homeless and ostracized from society, like someone in my extended family. They weren't trying to be mean. They were just honestly concerned because the school would always be calling and all of this stuff right saying I was horrible. And then I didn't really get along with my mom because of cultural. She just didn't understand what it was like in my opinion to be like an Asian American kid growing up in the US. She didn't really speak English going to school. So it kind of a struggle and it became pretty clear that I would not graduate high school due to various conflicts with school faculty mainly. And I recall actually one of my teachers told me that I could never be hired, have hold a job or anything like that. Thankfully though, my dad eventually recognized as well that I should just do my own thing. When I was a teenager, he just supported me and saying, okay, fine, school's not for you. I thought you would go to Northwestern or good school, right? But he loves Northwestern because Kellogg changed his life. But fine, you should just go to your own thing. So I would say that's our chapter one between. You've got a rebellious spirit. You clearly have some unique giftings, maybe also some elements of shortcomings that don't fit the rubric of traditional school system. I think going back 15, 20 years, that was treated differently. I think it's unfortunate. You talked a lot of people around that era and ADHD was much less understood than today. Curious when you're connecting the dots looking backwards, what do you feel like are elements of that upbringing? Again, in part, that upbringing that maybe you were pushing back against to some extent. But when you look back on your childhood, your parents, the kind of culture that was in the water and your family, anything that you feel like is important to understand in terms of what you have carried forward that's led to some of your successes and just led to, in general, the kind of person you are. Yeah, so I would say that it was very interesting because I was introduced to compounding and a lot of the concepts that you guys cover, obviously, on the pod by my dad, right? Because he started working in finance. That's all we would talk about, to be honest. We would talk about other things. But we would mostly talk about current affairs and macro and like company finances. Even when I was 10 years old or younger, I thought I wanted to actually work in finance. My dad had actually also said, "Hey, you probably won't do well in finance and you should look at all these great entrepreneurs." And then I met some of his clients that were entrepreneurs as an early teen and then I was like, "Yeah, I guess that's what I'll do." But initially before I discovered technology, which I've been passionate about, obviously, for the past 20 years, I was doing everything from selling fireworks, writing erotic novels and publishing them on Amazon, all kinds of schemes. It was more of like entrepreneurial hustling. And then I discovered that actually if you apply that in a compounding technology business, that would actually be much more durable than just being an entrepreneurial hustler. And I would say that that's definitely what's carried forward. Same more about that entrepreneurial spirit, because I feel like that's an interesting other side of the coin to a rebellious child, right? Or somebody like, "If you're not going to do well with authority in school, you probably wouldn't have done well at a GE or a large bureaucratic institution of any kind, which in some ways limits your choices down to if you're going to do something in business and finance. The entrepreneurial path becomes a little bit more obvious. But if I'm not mistaken, you were really young when you were actually doing some of these things and just say more about how you navigated and cut your teeth and building little businesses and maybe some of the learnings from those early shots on goal that have been formative for what's ultimately been this sort of life's work in CSC, which we'll get to." I gave myself by saying this, but I was in my early teens. I was, I think, 14 when cell phones became a big thing. And before it was like car phones and these huge, gigantic bricks, right? But then cell phones actually became cool. The first cool phone I still remember was called the Motorola Razor V3. Right, before cell phones came out, I was actually selling fireworks. The long story short, there was a very high demand for fireworks and the gross margins and fireworks are actually extremely high. However, it was a problematic business for various reasons. And then so I said, okay, you know what? I should start selling cell phones. So then I started sourcing cell phones worldwide and I started selling them on eBay. When I was in middle school, we were doing over $10,000 a month. But the problem was that eBay was taking all of the profit and so I was actually making very little money. eBay does no incremental work to get that sale. I'm doing all the incremental work. My mom would even help me oftentimes, right? If sales were too strong, I couldn't actually ship all these products myself. So my mom would ship products for me. But I realized that if I had actually owned the customer relationship and built the marketplace, that I could make more money. So I stopped selling on eBay. I started to learn how to code and started to build my own marketplace. And that was when I was 15. I got a handful of questions about your stint as an erotic novelist, but I'm going to table them. How about a fireworks salesman? Erotic novels, dude. That was before generative AI. A teaser was that I actually had these people in India helping me with erotic novels. It was really an interesting experience in my life. All right. We have all the content we need for round two whenever we decide to do this again. One more question just, and sorry to stay on these early days. But I've met a lot of young people who go through those kinds of perceived setbacks, being told and reinforced that you're not doing well in school, you're not achieving the things that you should be achieving. There's like an irregularity to you. Just curious, looking back on that clearly that was something you've been able to get through and to maybe even marshal towards some of your successes. But just say a little bit about how you internalized that period of a lot of what I would consider to be like negative reinforcement and got beyond that and into embracing those things that you were really good at and that you really were curious about and that you enjoyed. Some people are really smart. Some people are really good at coding. People have these talents. One of the few things that I really have that I feel makes me unique is that I just don't really care what people think. And also I could come back from setback very quickly. I can recover from any amount of turmoil typically within a day. The next day I'm honestly truly fine. And I don't actually have suppressed emotions. A lot of people that think that they can recover from setback. They actually are just holding it in right and it causes all these problems for them. But after decades, honestly, I actually get motivated by it. I really enjoy being the underdog and being contrarian and people thinking things about me. It gives me this rush. So I just channeled it to further my causes. Sounds like Rick on the tennis court and what people say about him. Thankfully, we're not going to get into tennis again today. I was checking out the CSE generation website. Recently, we're 10 years into you building your magnum opus. And it's so fascinating, even just the tagline around building the AI platform that's reinventing commerce. It's this huge bold vision. But bringing together with Rick, his unpacked by your early life, bridge how you got from there to CSE. What was it about business that you learned about structuring your company? Maybe through other entrepreneurial endeavors, then made you want to found CSE about a decade ago. So after the marketplace for cell phones was acquired by family office when I was 19, then I applied to all these different schools. And I didn't get into a single college. But I actually applied to Y Combinator and I got into Y Combinator, but because I didn't get into any colleges, if I'd gotten into Harvard or something, maybe I would have had to think about what I wanted to do. But I was like, okay, I literally have no other option than start another company. And so I did that. It was called Five Honor Friends. We raised venture capital when I was 19 or 20. And then we grew the business to become one of the leading cloud-based loyalty program SaaS platforms. And a few years later in my early 20s, it was acquired by this company called Merkel. And Merkel is this random company that nobody has ever heard of that's based in Baltimore, Maryland. And the CEO of Merkel would, in my opinion, be a great guest for the pod because the guy David Williams, who I learned a tremendous amount from, built the company with this long-term compounding mindset that he was going to build a technology enabled at agency, basically. So he bought Merkel for a few million dollars and then 20 something years later, it was essentially acquired for a few billion dollars by Densu Agis. And Merkel was structured similarly but as a C-Corp instead of a fund. And so he could have obviously done what he was doing as private equity, but he did it because of fun life and other reasons as a C-Corp. And so that influence was actually very influential in deciding to structure CSC the way that we did. Prior to that as well as a child, I think I mentioned most of my conversations with my dad was tied to business and compounding in like Warren Buffett and Brookshire Hathaway. My dad would speak of Warren Buffett like he was his savior. Because of that, I'd studied that, read all the Brookshire annual reports, et cetera, but then seeing it in person, actually at Merkel, the combination made a big impact. Why a retail? This industry that, at least from the outside, seems fairly low margin, extremely competitive. Maybe it's better days are behind it. And marrying that with these lessons of Brookshire Hathaway trying to build something for the long term with a potential return on capital. And then taking back to this eBay story where you do all this work, but there's just not a lot of margin or value for the role you had. Why was it that retailers rolling up retailers, building a permanent entity around that was this based on the way you're doing it. they want to go after? All admit, the part of it is hubris where it seemed easier than it actually has been. I would say things have gone well, but it's like way harder. I'm still literally grinding every single day. It's really crazy. But to answer your question, consumer spending is actually the largest segment of US GDP. When US retail sales in the fourth quarter of 25 was $2 trillion. And Walmart is actually a trillion dollar company. At the time when I started CSE, Walmart was not a trillion dollar company, but I was very bullish on Walmart, Amazon, Costco. There's these amazing, extremely durable retail companies. Retail commerce is definitely hard. It's definitely messy and it's definitely complex. But if you actually figure it out and you actually have a real flywheel in the category, you actually have a very good long-term mode that's actually very durable and even will survive all kinds of disruption. As evidence by Amazon, Costco, Walmart, being great companies that I think even in the AI era will actually continue to compound. I said, okay, Amazon has their own advantages. Walmart has their own advantages. Costco does. It's all extremely under-digitized at the time. And it's just a massive sector where if we actually created this operating system that continued to compound as you brought in more data into the platform that we could create a hundred billion dollar company. And at the time, I really just wanted to create a hundred billion dollar company. And it seemed harder to do doing that in software, pure software. Of course, if I were smart enough to realize that, oh, I should have started it. If I were really good at training foundation models, which would have started a generative AI foundation LLM company, right? And that might have been an easier way to get to a hundred billion than doing what I'm doing. Unfortunately, I didn't have that skill set, I would say, or foresight. One of the interesting things here in retail, despite potentially lower margins, there's lack of tech and AI prowess in that system and potentially inflated cost structures. Created a situation that was right for a holding company that could use organic and inorganic methods to grow. And I think Rick and I, we talked about this often on the podcast with Will Thornberg last year on here. He made the point that the total addressable market for buying up other companies with your free cash flow is way bigger than trying to run or build a company with perpetually, extremely high organic growth rates. And somehow, I don't know what those inflow would be interesting to unpack. Those influences of why and how you saw that as a viable strategy here. And then what the core components were, you just alluded to it at the center, an operating system that's just better when you bring together these systems because maybe one retailer can't do it on its own. They can't have a full tech stack like you've built your own internal AI functionality. Are there other reasons that you thought rolling up retail versus just say buying one and improving it was the way to go? Yes. So Will, who I've gone to know has been a great source of inspiration for me. I've read his book multiple times. I was introduced to him through our mutual friends at Alto's. We at CSE agree with that. So commerce companies have actually taken a step back. Have actually been a trap for a lot of investors, especially venture investors because you could grow organically very quickly. But then you basically hit this brick wall where you can't profitably grow anymore. Most brands effectively have a ceiling of how big they could get as a standalone brand. There have been a few companies, of course, the Nike's, Lulu Lemons of the world that have actually broken through that ceiling, but it's rare. So you could have a company go from $1,5 million, $20 million, $40 million, great organic growth rate, attract all this venture, and then it just stalls out as soon as it hits $50 million. Maybe it's not $50. Maybe it stalls out as soon as it hits $100. Maybe it's $200. And so that was where after realizing that doing it as one brand is really great if you could, as you just mentioned about the market of companies that could do that as one brand is actually very limited and is akin to catching lightning in a bottle. So I said, okay, if you could grow organically instead of focusing on maximizing your organic growth, you actually use a cash flow to then acquire another company that's synergistic that brings data and supplier relationships to the platform. That would be a better long-term return on capital than spending on performance marketing. I keep telling Paul, let's not talk so much about the virtues of inorganic growth because for whatever reason, very few people seem to appreciate it in the investing world. It's crazy. We're actually recording on the same day that just announced all birds was sold for $35, $40 million. You're just whatever a couple of years ago, this multi-billion dollar IPO that reinforces your exact point. And that's again, that's a brand. I've loved that brand for years. Maybe you could tell the founding story there of CSE. I think 10 years ago, I think it was a direct buy. Your first acquisition that got this all going, but maybe just set the table a little bit, give people a sense for CSE today, the scope, the brands, some aspects of size, and then rewind the clock a little bit and tell us some of the journey maybe starting with direct buy. The simple version is that CSE is a profitable commerce platform doing over a billion in revenue. We're scaling rapidly. We have 13 brands. We call our mission, we cheekily say that it's AGI for commerce, which essentially by that, we mean a self-improving operating system that helps commerce companies grow sales, while reducing costs at the same time, that gets better over time on its own. Today, we own a portfolio of commerce brands, but the idea is that by building the basically foremost AI company in commerce, that we could build a very valuable long-term compounding business. Where it started, about a decade ago, we acquired this company called Direct Buy in Northwest Indiana. So as part of that, I used to live in California. I was born in LA as I mentioned, then with Living in San Francisco. Then I actually moved to the Midwest, to Chicago. I didn't know many people in Chicago at all. I'd actually consider moving to Northwest Indiana, but there's no Japanese food in Northwest Indiana that's very good. I eat primarily Japanese food and I was like, I can't live in Northwest Indiana. So I chose Chicago and I would drive to Northwest Indiana and then use Direct Buy as a platform to basically then once we got into the game through that, we were able to start our compounding journey. And maybe just a little bit about the economics of Direct Buy. If I recall correctly, you acquired over $100 million in revenue, but what was the deal? And another really fascinating question will probably get to a later, how you win these deals, often out of bankruptcy, presumably going against large institutional distress investors, maybe just say a little bit about the nature of that deal and what tactics that you've developed in order to hone your skills on this inorganic side of the equation. It was actually crazy. So basically we acquired Direct Buy. Direct Buy was a unique acquisition because it's hard to build this operating system. You need to own the company's first. You can't just build in a room. Otherwise you wouldn't really know what the pain points are and what you would need to build. So we just wanted to get in the game. The problem was that I didn't know anything about restructuring or anything of that nature. This is how we cut our teeth. The sponsor in Direct Buy, it was HIG, owned the company. They had actually foreclosed on the company because they had owned the debt. They didn't want to deal with it because it was rapidly declining. Direct Buy's business was unique in that there wasn't really a way to really grow the Direct Buy business. It was just a way for us to get started. We were like, "Okay, this isn't the ideal company that we want to buy, but we were actually able to acquire it for a few hundred thousand dollars and then assuming all of these liabilities that had been racked up, there were all these unshipped customer orders." They just didn't want to deal with that. They're like, "Hey, if you'll somehow deal with making these customers whole, you could just have the business." We acquired the business out of bankruptcy and that's where it all started. It's interesting that the starting point isn't necessarily indicative of where you are even three or five years down the road that own 10. We talk about this with serial acquisition strategies. One of the incredible benefits, if you're truly investing in building with duration, is the implied optionality to change your mind and to learn things and pursue new strategies with the additional cash flow that you generate so that you can essentially renew the company through an organic measure. It does require a unique set of investors that are willing to go along with you not only for that duration, but also to give you that kind of latitude. Maybe say more about your fundraising journey, how it started, and what you were looking for in your capital partners. One thing that I learned mostly personally as an entrepreneur, the vast majority of investors I've worked with have been good, but as an investor in technology companies, I realize that most investors are actually negative value add. The reason is that a lot of investors are funds. The incentives are not necessarily aligned with the entrepreneur or with even the company's own schedule of let's say how long the company needs in order to have a good outcome. They might have their own issue where they're struggling to raise a fund because they haven't returned enough capital to their LPs. They'll say, "Hey, start pushing everyone to try to sell the company." For me, I really wanted to, because this was a crazy idea, right? Commerce, retail, very contrarian. I was like, "This could totally take longer than I expect." I really need to have patient investors who have a long duration and are also contrarian themselves. Where their identity is like, "Okay, we want to invest in a little bit, not just straight up down the fairway SaaS company that's growing at a certain rate, rule of 40, or everyone, eight term sheets, everyone's chasing the same things." One of the great partners that I had known actually since I was 19 years old was this guy Carrie Lai from Conductive Ventures. He used to be at Intel Capital and IVP and he decided to start this new fund, Conductive with his friend Paul, who used to be a Kleiner Perkins. Carrie was an early institutional investor in the company. He introduced me to Anthony Lee at Alta's Ventures. I met Anthony, I guess, about six years ago, through Anthony and Alta's and Ho. I got to meet another group of great investors with similar long-term duration and thinking such as yourselves. We've been involved in more than a dozen holding companies. We're there day one. Those first few years, there's always a lot of learnings. If you get the structure right in a mindset, then typically that evolution and that's a mix of mistakes, successes, luck and some failures, that leads to an even better phase. That sort of modeling perpetuates from decade to decade. I didn't heard this version of it from you thinking about Alta's and Ho and Anthony and Han. We actually hadn't known them from our previous lives. It's very fortunate how this all came together for us and partnering alongside you four or five years ago. We were introduced by Doug Hamilton, who helps with Howard Troult's family office. We had just heard Ho NAM on with David and Ben, our friends had acquired. And we put these things together. We were doing sound and fascinating and we were very interested in just. But we were as interested to get this shot to meet the Alta's guys. So we ended up doing a reference on you. But it ended up just becoming a situation where we became very good friends with them. And we agree that if you have fund structures behind things like this that might take two decades, it doesn't really work. We were just blown away at the way that they think about. They were the first venture firm to become an RAA so they could hold things a lot longer. And so then seeing how they were viewing you and backing you around after round and weathering things through bankruptcy and ups and downs. And then being your biggest fan is now we'll get to this next phase or building when we think is the most compelling AI native solution within commerce. It was a very fortunate set of circumstances. What do you think is the secret sauce of this group together that really gives you that feeling that you can evolve and change because it's one thing to say that you have this structure that seems like it's built for patience. And yet you have changed your mind a lot and the world has changed a lot. And some things have not worked and other things have what has that been like now in your 10? Well, we're the strengths of this model or what have you had to evolve? I would say that there's been a lot of crazy things that have happened to the business, including having dealing with the massive fraud, the COVID, boom and bust. There's just been a lot of unforeseen challenges. And another massive one is really tariffs. Had I known all of those things when I started CSE, I would definitely not have gone down this path because those are such three massive body and head blows that were just so painful that it's shortened my life probably by several years. On the flip side, there's also been the big tailwind, I would say. So there was a temporary tailwind during the COVID boom, but that was basically just a sugar high. That wasn't real. But the real tailwind of this business is actually the advancements in AI. And so what's occurred was that with COVID in particular, it causes massive inflation. Everything is more expensive than it was pre-COVID. And so for many companies, COVID was actually negative, a net negative when everything was said and done. For example, in the commerce space, because of inflation, store wages are higher than pre-COVID, outbound shipping is higher than pre-COVID, warehouse operations are higher than pre-COVID, right? They're all these different marketing costs are higher than pre-COVID than the pre-COVID trend. And so there are all these things that make commerce a worse business today than in 2019, 2018. Thankfully, though, there's this counterbalancing factor, which is AI. And so if it weren't for AI, but all this other bad shit happened tariffs and COVID inflation, commerce would actually be really in a dark place. We would have completely shifted our strategy. But due to AI, it's what's made it still be not only as interesting, we believe more interesting. Let's dig in here just a little more on the transitions throughout these one-off events, but things always happen. But then the more steady-state challenges for mid-market retailers, like operational inefficiencies, high-scu complexity, thin margins, there's inventory risk you've talked to us a lot about, limited engineering and resources. How has AI helped to solve those problems for the typical problems that not only haven't gone away, but they seem to have gotten worse? Yeah. OK. So if you think about, let's say, why people like software companies, a lot of people like software companies because there are actually many ways not that complex, but a lot of software companies, let's say, have 75% gross margins. The thing is a lot of retailers might have 50, 60% gross margins. Some have lower margins than that. But on a gross margin basis, retail is actually not necessarily a bad business. Everything you listed is basically a death by a thousand cuts. So you might start with a 60% margin, but then you buy too much inventory. You have to mark down. You have bad employees. You have fraud. You have all this waste. So then your contribution margin, you have high-marketing costs. So after all that shit and your payment processing costs, your contribution margin is actually very low. The dollars that flow through after you have to deal with all that are thesis. And we've actually proven this thesis to a large degree, even though we're in the early innings, is that these companies have this problem because it's a complicated problem, but they just don't have the right tools. They don't have the right process and they don't have the right people to attack this because the world has changed. And they're still operating as if it's a decade ago. And the problem that you listed can largely be solved with AI. What does the competition look like for new deals now, given that? So you have this confidence and you have these tools internally at CSE, which we got to dig into. Rick mentioned early deals. Maybe we're out of bankruptcy. Oftentimes you're paying very little. I know maybe you can get into the backcountry deal, which was fascinating in the last 18 months or so, the way that huge top line revenue, but just essentially no net margin available. How much do you have to fight for these deals versus there's really no player that has the tools you have? So there's much more value that you can see and you become the preferred buyer. So we're in chapter three, chapter four of that, or maybe third or fourth inning in terms of being the preferred buyer. And the reason is this, we believe that our ability to win and grow even faster than we're growing is tied to being able to pay more money as well as actually go up and down, meaning we could do more acquisitions because of AI and that we could actually afford to pay more for the acquisitions that we do. And so first, generally speaking, there's not that many people that are trying to do what we do because this is a contrary in play. And a lot of the distress debt funds and people that used to play in this space, they're very wary because they've been burned over and over again because the private equity and distress debt playbook is very challenging in the commerce and retail sector. There's usually maybe only one or two competitors in every process. So almost every process we enter, there's maybe two or three serious parties of which CSE is one. In terms of the companies that we want to buy, we're successful in acquiring 25% of them. We believe our judgment has gotten a lot better. And so if we could double our close rate, which we can, by just paying more money, that would allow us to compound much faster. But in order to pay more money without lowering your margin of safety, that is what we want to do. We want technology and AI, specifically, to be our margin of safety. And so if our platform were better, we could pay more, and actually have lower risk than even if we pay less and have a slightly worse platform. That is the inflection point that you will see CSC really take off. We have a very clear path to getting there. You mentioned just your judgment improving around prospective acquisitions. Maybe that's an opportunity just to double click on what you really look for. I mean, what kind of a company makes for a great CSC generation investment today. What are the hallmarks? What are the pitfalls that you've learned maybe the hard way or that you've figured out that you avoid just say a little bit more about your learnings around the inorganic growth portion. I think I mentioned earlier, right? This has been quite a humbling experience. There's been countless amount of things that we have learned. But one of the main things that we've learned is that some businesses are just impaired by structural realities that even strong execution is not going to overcome. In 2019, we acquired a brand as an example that had sourced two thirds of their products from China. We acquired it and their technology, their data was horrible. So we said, "Okay, if we implement our platform, which is not nearly the capabilities of what we have today, but we'll be able to add significant alpha where not a great business would be a decent business." What we didn't realize was the whole decoupling with China that Trump had started that Biden then continued. The effective tariff rate when we bought the company was 5%. And then within six months, it went to 30%. On a hundred million dollar company was, let's say, 40 million dollars of COGS. If you suddenly have a 30% tariff on about 24 million dollars of COGS, you can't on a 100 million dollar business pass along $7 million of incremental costs to the customer without dramatically impacting sales. That example popped me that, "Okay, you can't just be focused on the tech and data and execution piece. You actually really have to be macro aware. The right acquisition is not a business that is brittle or a business with poor unit economics. It's an ideal acquisition. It's a business that has value, has customer relationships, and the unit economics are actually still decent without this kind of crazy macro exposure." Yeah, how does the learnings from that apply to the backcountry? If you're willing to share details, it was about a year and a half ago, you had learned a lot through some of the other acquisitions. So this has been like 24 AI models are getting quite good at that point. What did you see at that point where you could apply almost like a phase 2.0 of the CSC playbook? When we acquired backcountry, it was doing about 500 and 70 million in issue revenue, but losing 10% EBITDA margins. The LTM EBITDA loss prior to us buying it was negative 55 million. Obviously, very few companies can sustain that kind of EBITDA loss for very long, and so they needed to find a new home. In the private equity firm that owned backcountry, they were very thoughtful actually. Oftentimes, these sponsors don't really care what happens to the company and the stakeholders and they just pretend like it's not happening or whatever, and then it just spirals out of control and it liquidates and everyone loses their jobs and the suppliers get hosed and it's horrible for the customers, the suppliers and the employees. But the private equity firm really wanted to ensure that that didn't occur and they were proactively trying to find a good solution. So we were able to work with them. It was a large acquisition. It was a big bet that our platform could actually take something that was so bad, so unprofitable and actually quickly get it to a basically positive EBITDA, which we've done. It was a really great case study that we could actually increase a company's EBITDA margins by about 1000 basis points within 18 months of buying the company. It's just a lot of hard work to buy something that's negative 10%. So what we would prefer to do is buy something that's break even and maybe get it between 8% to 10% EBITDA margins. Backcountry was perfect place because they didn't have the China exposure. It's a really great brand. All the CSE criteria from an AI platform fixing the problem perspective. But now to go from negative to break even and then positive, obviously, if it were slightly negative, for example, it would be an even easier world for us today. And so that's where we're focused. We had a shot a few months ago to meet up in London for dinner and I think we actually had some Japanese food, so you were happy. And you brought out your MacBook and showed me the secret sauce. We can share as much as little as you want today on CSE, but essentially it was a homegrown AI chat interface that has all the CSE data and where you've created the name image likeness of CSE. And I think I'm really struck, Rick and me that this is AI native software built from scratch that doesn't really touch the medium versus non AI oriented company at all. This is the melding in between where when you think about backcountry or surlotab or Zee Galerie or one Kingslaner, the longstanding brands, lots of people, lots of old processes, you have spent a lot of resources and time building this internally. Two questions come to mind and you can share anything else you want about this. But one is how did you build this thing? What was the purpose? Maybe just share a bit of what it is internally and how you guys all use it. And then two, I think the biggest use case outside of maybe pricing and some business metrics you can input. It's really about talent. Inputting the people that come with your companies that are in your companies. And then if you think about a business system across CSE, the way that people can move across companies in this harmonious leveled up way where the advantages really start to accrue to a much bigger set of companies. How that actually works and why you built this thing. One of the architectural decisions that CSE made early on that has really paid off very well was we really tried to avoid training our own models. So whether it's generative AI or machine learning, we basically want to just be the platform where you could plug in any model, right? And it's easy to just swap in and out. And also not just that, but there's these what are called eVals and everything from taking a step back like orchestration eVals, which is evaluating right that it's actually not just like unleashing AI without any evaluation criteria where agents are then evaluating the work of other agents, workflow design. And then finally with the machine learning decision layer that actually makes decisions. It's not just like AI said to do this. And then it gets pushed into production right there needs to be the decision layer. Even in commerce, people trying to build their own models, their own chat bots, stuff like that. There's so much investment going into AI that the previous supervised learning way that commerce companies that were using AI were doing things is actually not really the right way. And we should actually just build everything so that as AI models improve, we'll just continue to improve. The problem is that AI doesn't work if you don't give it the right context. So the context is so key because the more context you give the agent, the better the agent, you need context and eVals. And without a platform to easily like what is good in commerce? What are the alpha? Who are good employees? What does good pricing look like? What is good purchase? What's good planning look like? Without those examples, how can you just have an agent do the work? You don't literally type in a prompt that says update all my pricing on backcountry.com the way you think we should do pricing. That's not what anyone would do, but people think that AI should do that. And then they have AI. Then they're like, look at this horrible result that AI came up with. You didn't give it the right context. This is so crazy. There's no decisioning layer. There's no rollback of anything that goes wrong. We have a very strong engineering culture. Obviously, most of CSE employees are actually our biggest function is engineering. It's not finance. And so because of that, we were very well suited to build this platform versus most acquires are very finance happy with very little engineering capability. It's helpful framing. I want to read a little quote. You had sent out all your shareholders that I think encapsulates this, but then left us with a cliffhanger. What is this actually going to look like? How long is it going to take? And so you said to conclude, the organizations that will dominate the next decade won't measure success by how many dashboards they maintain or how many AI pilots they've undertaken. Instead, they'll be defined by how deeply they embed intelligence into the operating system of the company itself. and build towards AGI. In the future, every employee, whether they're a merchandiser, engineer, or CX agent, will work hand in hand with a network of agents that surface insights, validate decisions, and execute actions at this machine speed. Performance will be recognized continuously, knowledge compounds across teams, and the cost of change approaches zero. The result isn't incremental productivity. It's an entirely new organizational physics companies that scale software. Say, "What does that happen?" We need the labs to go faster, but I'm certainly not smart enough to speculate when that happens, but we really do believe that will happen. There's already a shift where two years ago, our best employees at CSE were people that used AI really effectively to help them do a better job. So they were using AI as a tool, but now our best people, they are the ones helping AI do a better job, where AI is actually doing 80% of the work product, and they're ensuring the labeling is correct, they're ensuring the e-values are correct, right? That piece, and they're doing the architecture. So essentially, they're helper of AI. Right now, the next step in our best managers, by the way, a few years ago, were people that were really good at managing people. Today, our best managers manage agents and people. Just being good at managing people is only like half of your job now, versus before it was 80% of your job. In the future, it might actually be 20% of your job, and 80% of your job is actually managing AI agents. And so we're actually far along that paradigm shift already. We could actually see the next step. We can't really say necessarily when it's coming, but it's pretty clear, and we have more confidence than ever, that it will come over the next several years. One of the things that comes to mind, just hearing you riff about this epoch of AI, is just how quickly things are changing and how fast things are moving. And it makes me think that something we haven't talked about yet, culture is maybe even more important in circumstances like this, where on some level, you are reacting to what the technology can do to what's possible. You're building for the current environment, but you're also trying to anticipate where everything is moving. Maybe just say a little bit about how you think about culture in the context of just generally, just how you've tried to set and reinforce the culture at CSC. And then in this AI world, as you're seeking to be a leader, and by the way, like we should say, we've heard it from several folks. What you've done over the last couple of years in terms of your ability to navigate the AI era so far. We're obviously just in its infancy is pretty astounding. It's certainly impressed your investor base, but I know the autos folks have paged you for teachings to their team, and some of their portfolio companies, etc. Just say more about the culture piece and what is been most important in getting this first phase of the AI era directionally correct? Yeah, so the culture piece is really key. You can't have right strategy without culture, but like in culture now, I actually think it's even more important than it was pre-AI era. And I appreciate the kind comments and I would say that it's really our culture that has allowed us to have the success that we've had with AI. But in terms of changing the culture, you could say whatever you want to say. I like talking about how Enron apparently had in every conference room, or it was like a disproportionate amount of posters that talked about the importance of integrity at the Enron headquarters. It was interesting. I've heard it firsthand actually from people that visited about just integrity everywhere, but obviously there was a lack of integrity with that company. And so you could say whatever you want. You could say your AI first. You love AI. The end of the day, the culture is actually what people do. It's what the organization does. Not what you say that it is. So the only way to actually change the culture is to hire people that actually raise a bar of where you want to get to and then to part ways with people that are not. Because if you don't part ways with the people that don't want to be part of that culture, the new people you bring in get deluded. You actually can't change the culture. And then the second thing something I learned really early on in my life is a value of incentives. And this is something that Charlie Munger, I would say, had been an inspiration to me in teaching me about this topic through all of his great writing. Incentives are so important because if you don't have aligned incentives, then the chances that somebody, most people, not everybody, they care about their personal incentives when it comes to work. And maybe they care about the mission, other things, right? The incentives are certainly important. So if the incentives are not properly structured to achieve the desired outcome, that's also not going to change. If you say that you're AI first, but you promote managers who don't actually know anything about AI, who are not AI first themselves, who don't use AI that are not managing any agents and production, then you're actually telling the company that hey, you don't actually have to be AI first to be in a leadership position or to be promoted. So the people underneath that person and other people see that and they'll say, great, yeah, AI seems cool. I'm not really going to experiment with it. I'm not going to use it to transform my vertical because you don't need to do that in order to get promoted. On the people side, are you mentioned being technical, heavy, which is advantageous you greatly? There's the debate that's growing around what are the most consequential skills? There's folks out there telling folks to tell their kids, forget traditional education, forget pursuing white collar jobs, become welders. Has anything changed yet for you in terms of the kinds of people that you look to bring on to the CSC team and the attributes skill sets in this AI world? First, really try to bring people. What we mean by technical people is a lot of times people think that oh, that means people that know it a code and people that we think people that know how to code are very useful. But we really want to hire people who think from a first principles perspective and systems. So people that basically have a systems-based thinking and Ray Dalio in his book Principles does just a phenomenal, I read the book Principles actually every year. And the reason why is it's in my opinion one of the best books that talks about systems-based thinking and in the age of AI that's even more important. Because now it's not just about what you can do, but it's about what you can get agents to do for you. So before, if you're an individual contributor, for example, it was all about how much code you could write or whatever it is. But now you're not limited, but what you can do during the workday. We look for people that truly understands that okay, A comes in, you want it to be C when it comes out. So what needs to happen in the transformation phase? What are the bottlenecks? What are the potential failures? And how do you evaluate that A actually became C? In a modern company, before it was a company, was about the people. But now it's a company is actually about the people. Certainly very important, but it's a people technology workflow and all orchestrated by agents. So if somebody isn't technical enough to reason with machines, these agents are going to be running wild and could be acting on its own. So you really need someone to understand and sure that the system is working properly because there's actually more failure modes with AI. Rick has been known to break out in song. I don't think that's going to happen now, but if you want to, Rick, feel free. I'm bringing the street out of Compton by Jay Z and Eminem. You've some of the most entertaining letters to shareholders we've seen. You mentioned a grown-up song about realizing the world is changing, accepting the cost of adaptation and choosing to move forward anyway. It's a beautifully said man. We got to listen to Jay Z. When you find these people and you've validated this, how do you keep them? If they fit that mold, it reminds me a couple of us had a chance to spend a half a day with Luca Ferrari of bending spoons over in Milan recently and their year 13 to 15ish. So just a little bit ahead of you and they were in the wilderness for a while, trying to figure out their operating system. They buy a lot of companies and then they have a mothership of talent that they deploy in the companies and then bring them back. What does that look like for CSE and how do you keep this excellent talent from here? As you keep this playbook of buying small, medium, some bigger enterprises, but CSE becomes very large. How does someone who's really talented not get lost in that? CSE, we've been fortunate to have very few regrettable departures where on the management team, for example, actually the tenure is very long. People really stay here. There are several key things about the culture that really plays a critical role in retaining people, but I would say that it goes back again to the culture and the incentives. So here, for example, we really focus on being a meritocracy. You could be early in your career and you could actually end up with a lot of responsibility where most people that are in charge of big PNLs at CSE. I previously never had any relevant experience that even remotely similar. So if you look at all the brands and who runs each brand, you will see that not a single person actually had former PNL ownership experience prior to coming here. In large part by the way, the reason we could do that is that we believe that most people that have this we would rather have back these outliers and then have the technology platform to de-risk it. There's higher beta when you do that. If we didn't have the tech capability, we would just put these people in these positions and some might go really well and other times it would just flame out. We've now gone to a point where it's actually not that risky because of the feedback loop is so quick that if someone's not doing a good job, we could actually move them very quickly. And to answer your question, like the job is very interesting because you're not really held back by titles or tenure or age or whatever it is or race or whatever biases that many organizations have, here you could just be put into a big position relatively quickly. So it's a true meritocracy and you don't need to wait because we're constantly buying companies, you don't have to wait until your boss leaves. So most companies because of this hierarchal structure, there's companies where by the way the tenure is really good, the management's very competent. And so the vice president and the sea level people are never leaving because the colleagues are really good and the company's doing well. How do you actually overtake your manager or at least be at the same level as your manager, unless he or she quits, it's actually very difficult. But due to the nature of how we're structured and we're adding a company a quarter, there's plenty of opportunity for capable people here. And so that's what's really helped us retain the talent. Yeah, if you think out further, let's say the next decade, we've been bringing our friend, we'll throw him back into the conversation here. He often talks about the idea of flywheel and the self perpetuating cycle of cash flow coming from businesses that feed the next acquisitions. And you've outlined talent feeding the system too. And you get these beautiful things going not only from a balance sheet perspective from many different angles, where is CSC in that journey? Where does that look like 10 years now? You're over a billion and combined revenue now, EBITDA margins are going up. There's cash there to make new acquisitions. But how do you think about balance sheet potential? Maybe IPO in the future, other ways to think about taking this to that grand vision of scale because it's such an enormous tan in the US alone. A completely agree with Will. The one thing that Will actually didn't talk that much about is the data flywheel as well. When you actually have this proprietary data, not just data, but the way to actually measure, right? So data plus eVALs, as you add more data and more integrations and more nodes into the platform actually becomes even more powerful. Because all of that being said, we've actually now reached the inflection point of the first inflection point. So there's kind of two inflection points that I see. The first inflection point is when we don't need to raise capital to just continue to do acquisitions. And so we generate enough capital, enough cash flow and liquidity now, where our last four acquisitions, we've not had to look for any equity capital. And unless we were doing a major acquisition, we don't see the need for any equity capital to do any acquisitions. Despite, by the way, we are investing heavily in AI and talent. And so even with the heavy investments we're making, yeah, it's impacting our EBITDA, but we still generated enough liquidity and cash where it's okay. Now that we're at that point, that was a big milestone. And then the next thing is our organic growth. We're hitting record organic growth numbers than we ever have. And so what that means is that we have not only figured out initially, we were using AI. Our focus was on using AI to fix the inefficiencies and commerce. Because if your unit economics improve, you could then have a better foundation to grow your revenue. Because obviously if you're unique economics, you have a leaky bucket, you don't want to grow your revenue and then the water goes in the bucket and it leaks. And now because we fix so much of the leaking that we've actually shifted our AI focus on growing growth profit dollars on an organic comp basis. That is showing tremendous progress. But again, we're only in the second inning. Our organic growth is a few hundred basis points higher than it was before. That's huge. What if we could add another 400 basis points to organic growth profit dollars? If we can, then that goes back to the previous inflection point that I mentioned where we could pay more for companies because we have this magic mouse trap to grow right now. We have the cost side dialed eight out of 10. But on the growth side, it's only like a three out of 10. And so if we get the three out of 10 to an eight out of 10, we could win 80% of the deals that we want to win. This might be the most incredible part of the equation for me, at least just that organic growth acceleration. I've heard you in the past talk about one of the pitfalls that oftentimes investors private equity VC make in trying to force feed brands that are not intended to organically grow at the rates that those investors would like them to grow based on their economic models and their time horizons. But you're getting this while running extremely lean efficient. You're driving obviously efficiencies through AI, but you're also tapping into the promise of the revenue acceleration opportunity, which is extremely rare right now. I think something that a lot of people are very interested in curious just what you see on the horizon for retail. How do you think it's changing? What do you expect to see over the next couple of years? And in that context, what are going to be the best kinds of companies that CSC is the right home for it? Are they great brands that leadership fails this AI transition? Are there other things that you think are likely to play into that? But maybe just talk a little bit about what you are focused on more kind of externally over the next three years. So I think everyone knows that commerce is obviously a huge market and we don't need to talk about that. But the number that I do like to point out is that commerce, excluding auto, which we don't have any plans to get into, is $1.5 trillion of gross margin just in the US on an annual basis, not trillion in sales, but gross margin dollars. At the end of the day, we do that to be the most important number because that's the real opportunity where you could have high revenue, but then no gross profit. That wouldn't be an interesting business. If you think about the 1.5 trillion, but then how much of that actually goes in terms of like durable free cash flow and durable EBITDA net income, whatever it is, it's actually pretty low because of the death by 1000 cuts that we mentioned earlier. In the future, because of AI, we think that a lot of the 1000 cuts that will make the 1.5 trillion turn into not that much profit today will actually completely change. It was kind of hard to say, but we think that for an AI first company on the CSE platform that we will be able to consistently increase EBITDA margins on a durable basis so significantly that it will result in a re-rating of what it means to be doing what we're doing. At that point, that's really the ideal time where we would like to go public, and we think that could actually happen in as early as a few years. That would attract a lot more capital and so on because if we could reduce the complexity and the risk associated with inventory, the reason why retail's difficult is really the inventory risk where if you over buy inventory, a good business could get liquidity constrained and totally spiral, things could spiral out of control pretty quickly, something that doesn't happen in software. So we could actually show that using AI, that's not going to happen. The biggest risk of retail is buying too much shit, buying too much bad inventory that if you could avoid that, then it's actually a much lower risk, lower downside, and more profitable endeavor and something that we're really excited for. One last business question related to what you just said. We can close things out. I know you're busy running this machine. This is awesome getting all these insights. You mentioned inventory and having too much bad stuff on hand is one issue. What about the demand side? Just think about customer behavior in the retail world. The switch from latent marketing to search engine optimization. Now to demand-driven AI-led customer acquisition, how do you use to envision consumer behavior changing in the next few years as it relates to retail or maybe beyond and they're going on offense and trying to think about top line for all of your companies. How do you restructure things marketing wise, brand wise, customer reach wise in the AI era? A lot of companies that are reliant on humans doing the searching on Google. Today, where Google works really well for them, where if a large part of your marketing budget is going to Google Pay Per Click ads, where someone searches for a Patagonia Puffer jacket. You are reliant on being the lowest price on the paid search ecosystem and people click. There's no customer loyalty because they just came for that and it was just a Google one and done. We believe there's actually significant macro exposure to that kind of business. The reason is that as customers, let's say, use chatbots in particular more to search. And if Google starts to lose that volume every year. So let's say the Google search is declined by 3% or 5% a year. That is really catastrophic for a lot of people. Because in a similar to a lot of media companies that were relying on Google for this free traffic and is that declined every year, they suddenly wake up a few years later and it compounds just like the joys of compounding up. It compounds down as well. And you're like, oh my god, after rives only 5% a year, but actually on a compounded basis after 5 years, I'm like totally host. And the medium of how people are going to discover products, there are people might not even buy products themselves. It might actually have agents buying products. He's going to shift. And I think that is very important to be one step ahead of that. Things are moving so quickly, right? I don't have a forecast of, oh, what specific commerce protocol or right, how exactly that's going to go. But it's definitely a rapidly shifting ecosystem. How do you think in closing young people, if they're like you and just skip high school and college, maybe it's advice for them, maybe it's no more normal in college or early career, or if you're generally in that point or founders, what advice do you just have for them given all this discussion around the structure of CSE? The skills you look for, the adaptive mindset, this wild opportunity and changing world, anything in closing you'd offer to folks out there other than the top folks making sure they call CSE? So you know what's been really cool? People talk about this in terms of, oh, software engineering jobs and so on. What happens to software? What happens to all these jobs because of AI? But what AI is really done, one of the main things that it's done so far by the way, I don't know what happens in 10 years. I'm certainly not smart enough. But I believe that even in the near future, over the next several years, what happens is that someone that's really good at their job that has as deep curiosity and true passion about what they're doing is way more valuable to me and CSE, than it was three, four years ago. Because if you can really be the best at what you can do, even as an individual contributor, we'll pay you more money. So for example, from an engineering perspective, CSE is less engineers today than we did a few years ago. But our engineering budget has not gone down because we pay people more money. We're hiring engineers from top tier tech companies like Google, which we weren't before because we couldn't afford to comp. We are paying our existing engineers that have been with us more money. They're getting bigger raises every year. We used to need to have kind of mediocre engineers to do things. Let's say something that works well on commerce websites is to add Gmail sign in. So if you allow people to sign in with Gmail, not have to enter just one click to get their email address, a lot of people Gmail will convert at a higher rate. That's not a crazy technical task that you need a genius engineer to do. So you would have someone spend several weeks to do a feature like that. Now that could all be done end-to-end with AI because you don't need the lower level work anymore. The people that are top of their field is actually just so much more valuable. So to conclude, I believe that people should really work on something that they're passionate about, that they have a real shot at being one of the best in that field. And specifically, by the way, we're hiring junior engineers. It's not just about being senior. We're hiring junior engineers, junior people, and we're paying them well because they're actually really good and they're unleashing with AI. The cost of learning has gone down so dramatically due to AI that if you don't really like what you're doing, you will probably be disrupted by AI because you're not going to be on top of how to deploy AI in that vertical versus if you really love it, I think that it's probably the most exciting time ever. There's a current prevailing narrative. This more scary about all the jobs that are going to go away with AI. This is really helpful just to understand this technology, like prior technologies, is a force multiplier. It's likely going to lead to levels of abundance in ways we can't really exactly predict. But what I'm hearing is sticking with the same fundamental advice that you would give somebody 10 years or 100 years ago, follow your passion, be excellent, pour yourself into something, get really good at it, choose a thing that you like because it's going to be a lot easier to become really excellent. It's something that you like and let the future take care of itself. But it's still fun just to have this chance to unpack the CSC story with you to unpack your own personal story. I know this is not something that you do regularly. What you've been building is pretty extraordinary. We're grateful to be a part of it and to call you a friend. But thanks for coming on and for just being so open, sharing so many insights with us. I know it's going to help a lot of people out there. So, appreciate your time, buddy. Yeah, no, thank you for having me and being great partners through this journey. Through the ups and downs, you guys have been there for us and it's made the difficult days better and the good days better as well. So, looking forward to continuing the partnership.