AI Marketplaces with Jack Greco (Co-Founder, ACV Auctions) | Verticals Ep 4

Hey everybody, welcome to this week's episode of Verticals on Lucicina. So I'm in the founder seat. Welcome back everyone. I'm Nick and I'm a VC. Awesome. We're gonna dive right into this week's Vertical Pulse. We got a bunch of great topics and we got a fantastic Vertical Titan coming on here shortly. I'll actually gonna hold back on telling you who it is until he shows up. So keep the suspense alive. Yeah, I love it. So let's start on this week's Vertical Pulse. I think Nick's got our first topic here. So let's do it. Yeah, absolutely. So, you know, this is a piece some of you may have seen. It was published by our friends over at Melo Ventures. The title here is AI's UI problem is actually a new era of software. I think as, you know, everyone listening is familiar with L-Lems and what they allow where they're strengths lie. And that's really, you know, if you think about it, that's text, that's voice, that's content generation, new modalities, right? If you compare what that enables to historically, you know, what SaaS encompassed. So, you know, checkboxes, drop downs, a whole new UI to learn. What we're seeing is kind of a different face of software in this post L-LM world. So, can you interact with your software, your platform, your solution via voice? Can you interact just via natural language text? And so, what this piece goes through is just really what that enables. They go through really three errors of enterprise software, right? The traditional system of record, a system of engagement that's really focused more on, you know, intuitive UI and natural language. And then finally, systems of work that are just fully integrated and you don't even have to worry about, you know, hey, do I have to go in and enter something in a, you know, kind of data object or whatever. So, I think what's an overall takeaway here is we look at some of these, you know, kind of solutions that are coming out, they're AI native. There is less, it more naturally folds into someone's workflow. And as a result, there is less friction of adoption, right? So, especially if you think about the perspective of a vertical user. So, let's say you run, you know, a dentist practice or a veterinary clinic or a retail store, you don't have time day to day to sit around and say, hey, what's, you know, what's figured out how to use a new solution every week. My dad was a doctor. I know he, with his practice, he just, he said, I can't change practice management systems more than every couple of years because it's a huge training undertaking, right? So, I think the overall takeaway and what it looks, I'd be interested to hear your thoughts on is, are we seeing UI change in a way that makes software solutions, AI native software solutions easier to adopt, and therefore are growing the market opportunity? Yeah, this article reminds me a lot of Dave from TideMark's post on the system of action. I mean, look, I think, I always try to look at these things from the customer seat, right? And so, you run up your dad in the dentist category, you know, looking from the customer vantage point. What has happened with software sprawl is a horrible customer experience, right? So, you know, it's, it's the more buttons, the more things to click, the more things to learn, especially when you, you look at the service company that, that all of us are providing software to, you know, the software is, is not supposed to be a pain in the ass. It's not supposed to be there to click a bunch of buttons and fill out a bunch of texts and make sure the, the year system of record or your CRM is up to date, like nobody likes doing that. And so, you know, I kind of think about this as like how it's, it's taken actually quite a long time for this to evolve into something that's very simple. You can just talk to or type to and it actually takes action on your behalf. And so, that's very complicated, but I think the holy grail here is almost kind of like what Uber did, right? It made it so simple, you tap a couple buttons and you get a car and, you know, right and out of your doorstep. And I think if vertical software founders can think in that, that regard is obviously that you're doing a ton of highly complicated things to make that experience so incredible and magical and, you know, there's no friction involved very little. That's the holy grail. And so, you know, I like where you're going with this. I think, I think we all need to do a better job of, um, of thinking about, you know, how can we have an agent that somebody can just talk to and it, and it fills out the CRM. And it does the manual, you know, click the checkbox work that, that no service provider actually enjoys doing. Yeah, 100%. I mean, I think the primitives are already there, right? I mean, look, I don't know if you use granola like a note, a note take or a scribe, but between that and my email and decks we receive, right? That's about 90% of the information that goes into our CRM. Even still today, I, I largely have to enter it myself, right? So, um, I think the, the lift of maintaining a system of record or whatever we call that in the future system of action system of engagement is coming down. That's my takeaway. I like it. All right. I got, I got one for you here. So a company called extend just raised around that I wanted to talk about. So extend is a provider of spin management software for banks. And I don't want to zoom too far in on extend. I, I want to zoom out a second and talk a little bit about Brex and ramp, who obviously sit in the horizontal category. But I think it's really interesting. I believe ramp just raised that was at $39 billion valuation recently. And, you know, very focused on spin management on kind of financial AI agents, Ray, a lot of startup founders are using Brex or ramp. I use Brex at my business, but I've seen more and more over the last few years vertical specific Brex and ramp businesses. There's another company that's doing this in the healthcare space called nitra. I've seen a few in construction. And so I'm curious to get your take from the investor seat on the verticalization of Brex and ramp and, you know, and specifically with AI, you know, becoming a key element in some of these business, some of them are AI native businesses that are doing this. Like what's your view there? Are we going to see that happen across a bunch of different industries? I think so. I mean, my personal view is it's, look, the reality is not all software has to be vertical. And in a lot of cases vertical and a vertical approach can be better. But, you know, if we're just talking about, you know, general expenses and, you know, kind of employee charge cars and things like that. I don't think necessarily, but when you have, you know, what what vertical markets provide is great surface area for building specialization and defensibility. So, you know, and let's let's take construction, for example, right? You know, you have a couple great companies out there that, you know, offer this for for lenders or to give it to their developers. And there are really unique considerations there because, you know, you've got to figure out what are you going to buy? Where are you in the process? Is that going to be validated for your lender things like that, which are very vertical specific and ramp is never going to build that. That's where I think we can see these platforms really gain purchase is where there is vertical differentiation. So, I guess my answer is kind of mixed. I think you will see it in some categories where that exists. And others where they're just taking a ramp and trying to sell it into a particular vertical. You know, maybe the need isn't there. I don't know what you think. Yeah, no, I'm with you. I think there are definitely verticals where the workflow and the process is very different than your classic spins spin management. And I think all those verticals are really being service today by, you know, they don't even know what Brexit ramp is. So just the like virtual card to them is an innovation. So, yeah, it's going to depend on the specific industry, but if there are key workflows or a bunch of suppliers that you can leverage group purchasing capabilities, drive down prices. And you competing against like the Bank of America credit card that the, you know, Main Street Service business is using that sucks. Right. I think it's interesting. Plus, I mean, here's another maybe parting thought is let's see you offer a larger platform, whether that's for financial management or other aspects of the business. Maybe that allows you to offer better terms on the Fintech, right? Whether that's, you know, interest rates on a bank account or perks on the card, but, you know, that can be a loss leader for a larger business, right? You don't want to compete against these ultimately. You want to, you know, you want to compete against software. So that's, that's a super interesting one though. Have you guys done anything at your fund? Have you invested in anything in that or adjacent to that category? I don't know. No, I don't think we've done anything quite yet where the card was kind of the leading wedge. You know, we have seen that as kind of a later on product, but we have looked at things and are actively looking at things in that space. And again, I think it's, if that is hard for that to be your standalone products because that's not probably where you're going to build differentiation. But, you know, it is part of a larger offering. If that's like a core part of the day to day workflow in that world, absolutely. All right. I'll pull up the next one. Yeah, let's move on to the next. So, you know, we have, we have a piece here, you know, by Janelle Tang. You know, she writes some great stuff. If you guys haven't come across her sub stack, it's, you know, next big tang. Love the title. This piece is much do about margins to grossly overhyped AI gross margin debate and really where this, this comes from many of you may have seen. Bessimer's state of AI report 2025. This in the, you know, in the tradition of, you know, I, I lean at Cowboys unicorn, no manclature. They've, they've introduced some new ones for the AI era. So supernova is shooting. stars, you know, that really just kind of try to capture the profile of some of these extremely fast growing AI-native businesses, vertical and otherwise. But Janelle was talking about in this piece is really the question of gross margins. So AI-native software businesses, or just AI-businesses, as we're all calling them, they have lower gross margins, right? And part of that certainly is inference, the cost of inference. So, you know, we have a new infrastructure layer in foundation models and every time you want to, you know, kind of whatever it is, generate your content or analyze your audio, you've got to pay for that in the same way that you pay for AWS for computer storage. Now, early on, that can be fairly costly. The question really, Janelle Rapples with here is, how does that margin profile scale? And how does that align with the sort of valuations that, you know, AI-native businesses are getting today? Now, on the latter piece, just hit that. Obviously, you know, in Luke, there's a kind of bar chart a little bit for their down and AI valuations. But, you know, these businesses are getting incredible valuations. Now, early stage venture, it's not always based on multiples in terms of how companies are valued. We all know that. But I think the ultimate question, the ultimate bet you would have to be making is one, either businesses can increase these margins over time, whether that's getting volume scale on your, you know, your inference costs, whether that is automating human in the loop aspects, that maybe the LMS didn't cover to start. And the or that the market opportunity is so much bigger that it doesn't matter that there's maybe, you know, 10 or 20 points lower margin at steady state. I guess maybe one final thing I'll layer on here and then Luke, I want to get your take on this margin question. I've seven conversations with Jake Saber over at Emergence, who made a really interesting point that I think is relevant here, which is these businesses are growing so quickly. And these sees are investing earlier and earlier because they want to get in front of the winners, right. But as a result, we haven't really had as many cycles. And let's say at a series A, you know, the company, let's say they went from zero to 10 and they're raising a $50 million series A or 30. You haven't really had cycles of, you know, churn and retention. You haven't really been able to see like what is the customer's success cost going to be. And so in a weird way, Jake was saying this hyper growth almost introduces more risk. And I think that's very true margins. You don't really know where margins are going to land. So I mean, look, how does this hit you? Yeah. So I think every VC or at least most of them is just assuming that the margin problem is going to get figured out and resolved in the price will go down over time as it did. And you know, with every kind of major new technology that's come out. And I think that's probably right. But, you know, 25% margin for if you are a SaaS company pre AI with the 25% margin, like you'd be untouchable. No one would even invest in in the business. You know, the payment skies have gotten hammered in the public markets on valuations because they're sitting at 40 to 60% margins. So, you know, look, do I think this is gets resolved over time? Yes. But, you know, I'm looking at this chart right now and Sierra raised at 225 times a revenue. So like there's gotta be some sort of, you know, yes, will it come down? I think that's right. But we can't pay crazy. You gotta be at least somewhat price conscious if you're in the investor seat. At least that's my take. So that's the first point. The second point, and this is really around stickiness is I think a lot of the AI companies, specifically the horizontal AI companies that are that are built on top of, you know, other businesses, LLMs. What they don't talk about is they have horrible retention, horrible, horrible. Like they're using an ARR number and they're just taking this month's revenue and time'sing it by 12. And, you know, what that's missing is, yeah, we went from zero to 100 million in like 30 days or whatever the craziness that we're seeing on X. But I want to see like a true cope. I want to see a true cohort analysis of how this looks over time. And, you know, again, that comes back to the why why I believe and I think you share this belief with me that the future is vertical because yeah, you may not see that growth rate. Although we have seen it with like open evidence and in healthcare, which is incredible. And that's been historically a brutal market to to scale in. But I think even if you're not growing as as fast as some of these quote-unquote supernovas as Bessamer Cosm, retention is the lifeblood of SAS Jason Lumpkin has been screaming that, you know, from the rooftop for years and I agree with him. And so if you're a founder and you're you're really focused on the growth rate, that's great. You should be. But you got to remember that retention is as important as growth. And that's my view. 100 percent. 100 percent. I think ultimately, you know, if you look at open evidence, I think Harvey is maybe not as good of an example or magic school. These are all businesses that have built up really impressive adoption numbers, right? So I think for the first two, Harvey and open evidence, they've got kind of a third of their overall user base, whether that's big law in the case of Harvey or, you know, the case of open evidence physicians, got a third of all of them quote-unquote using it. Magic school has 80 percent of US school districts using it. This is consumer-like adoption. But let's think back to consumer businesses, right? Their retention just doesn't look anything like B2B SaaS. And so, you know, that's I think that's a real push and pull. We should all pay attention to. 100 percent. 100 percent. I'm with you. All right. Let's jump to the last one here and then we'll get our guests on the horn. Who's who's a killer, by the way? I'm excited for everyone to everyone to meet them. So meta recently dropped these new AR glasses. You know, we've been hearing about AR and all the buzz around augmented reality forever. These are the first glasses that don't look bulky and like painful and if you warm around in public, you'd be, you know, looked at like an alien. And so they're really they're really good looking. They come with a wristband, which is how I think they've made them not as bulky as everything else. They partnered with Ray Band. I'm excited about these. Like, I almost pulled the trigger on the Apple ones and then I stopped myself, which I'm glad I did because I feel like they're in everybody's, you know, drawer somewhere now. But my question for you, Nick, is and I was my brain was spinning a little bit on this one is number one. Is this finally going to open the gates to, you know, widespread adoption? And number two, what are some kind of rapid fire, like vertical AI businesses, vertical software businesses where you could build a company on top of these? Yeah, it's interesting because I think over the period that you and I Luke have been in you know, around the software world, there've probably been three different hype cycles around AR, AR/VR, right, or metaverse, whatever it's being called in any given time. And I think the challenge is always that you need the hardware, you need consensus on the hardware in order for the software to flourish, right? So, you know, if folks have used what's Apple's product on this, you know, their AR hardware, I forget the name of it for some reason. Apple Glass is supposedly going to be the glasses, but I can't remember what the one before is. Right. So I think the challenge with it is like these things are very expensive to manufacture and therefore they can't be widely adopted just yet. So they may be used in certain enterprise use cases, but without that kind of larger access, there's not really that much support for development ecosystem and it's just hard. So I think this is this is one of those technology jumps in which at some point, you know, we're going to be able to you know, compact all of the electronics and actually come up with a great wearable form factor. And then the ecosystem can flourish, whether this will be it, I have no idea, but I think there's no shortage of of use cases. And frankly, there are there are already businesses that are offering services today that could just adopt this platform, right? You see a lot of them and certainly in, I mean, healthcare is one, although probably in more niche use cases like actual actual surgery, but I mean, I guess another, I think they even had a picture on there of someone like looking at a tomato, certainly supply chains. So you think about like pick and pack and I mean, they're, I think they're pretty, pretty widespread use cases. It's anywhere you have, call it field workers or workers who are doing their job on their feet. This could be relevant. I love the quote that the hardware has to be consensus for the software to follow. I think that's right on the money. I magic, do you remember magic leap, right? Magic leap had crazy hype, tons of money raised, and then they, you know, it didn't really work out. They brought in a new CEO, Peggy Johnson, who was, you know, formerly Qualcomm, then moved to Microsoft and moved to Magic Leap. And her promise, I thought made a ton of sense, which was let's move this to enterprise and let's focus on blue collar training, I believe. And if I totally agree with this notion that there's this huge class of the US workforce and the global workforce that hasn't had the power of the internet really prevalent in their day to day work, right? So people that are working on machines that are, you know, using their hands. And so once we get consensus on the hardware, I think there's going to be a huge amount of companies that are working on their own. companies that pop up that can really support this. So it's very exciting. All right, everyone. We are super excited to bring on Jack Greco today, as our guests, we are gonna be chopping it up on his background as a co-founder of ACV auctions and now the founder of Greco VC. So with that, Jack, why don't you share just with the audience a little bit on you, your background and how you got interconnected with our theme today, which is AI marketplaces, and we can take it from there. - Cool. - Well, nice to meet everybody. I haven't done one of these in a little while, so hopefully the risk gets kicked off quick. So yeah, I was one of the co-founders at ACV auctions. There was a B2B wholesale automotive marketplace that we started back in 2014, which feels like it was like generations ago at this point. Now that I see what's coming out today, I always joke like what took us forever to build with ACV, you could probably give a kid with a case amount and do in a weekend, he could probably build something even better than what we had for our first couple of years. So no knockout or developers, it's just amazing how fast and how quickly things change. The reason I'm here, Nick, is actually, I think you cold called me at some point a long time ago when you were getting you click-go, and you were like, yo, we should talk, and we built a friendship, and I now happily, and LP in your fund, and I get to, you know, jam with you on stuff like this, so it's cool. So happy to go wherever you want with it. Cool, man. Why don't we start, if cool by you, let's start with the early days of ACV auctions. So, you know, a lot of our listeners are founders, probably a good amount of them, thinking about marketplaces or marketplace like bottles. I'd just be curious to kind of go back to those early days, and like how you thought about the marketplace dynamics, then, and how it just evolved over the first couple of years. Yeah, I think it's important to realize, I'm not a car guy. You know, I'm an American guy that likes cars, but I didn't grow up in the industry. And so, you know, you'll get a mix between founders, specifically with marketplaces, they grew up in industry, and then grew up outside of it. I grew up outside of it. I didn't grow up outside the mechanisms in marketplaces. My day was a dealer. He was a antique dealer, right, which meant, there were wholesale with something he had used at times. He was a trader in terms of traded stuff between people. You know, we went to a lot of auctions when I was a kid. So, the mechanisms I understood the content I didn't, right? So, just take that into consideration when you consider my point of view. So, you know, early days of ACV, right? I actually remember when I was in high school, me and my buddies sneaking into automotive auctions, 'cause we just thought it was cool, right? There was one about an hour and a half south of us, in a place called Dan's on the New York. They grew up in a little farm town called Cannega. And, you know, you go there, it'd be cool, be, you know, it's just such an Americana culture. You know, chewing tobacco, cigarettes, you know, guys barking at each other, pants falling down, so the buck cracks out everywhere, right? It just smelled like gas and oil and like everything in 18 year old loved. And so, you know, fast forward to 30 when I met Joe and Dan and we decided to start this. I had that memory, that, you know, that recollection of my brain. And I remember marketplaces like, you know, I always look at the Americana misby trade, right? So, I look at the tax on everything, right? I look at the way what the market tells me. You know, and it didn't make a lot of sense. I was like, you got, you know, dealers take cars in on trade. And when they take a car in on trade, their money gets tied up in it. And I knew from tech, what does that mean for not even less car listeners? Yeah, or, you know, everyone that lives in New York and San Francisco and doesn't own a car. If you have a nice enough car, when you go to buy a new one, you could, you trade into the dealer, right? You're like, hey, I got this, you know, 2019 and I want to buy a brand new 2025. And instead of, you know, for the customer, they don't want that car. There's a reason they probably want to get rid of it. They also are buying the second most expensive asset that they'll probably ever buy outside of their house. And so they have to think about where the money is going to come from. So people trade their car in in states where there's sales staff that new purchases like New York. People love trading in it because it actually cuts down the amount of which you actually get taxed. You buy a brand new 40,000-dollar car. You traded a 10,000-dollar car. You only pay in tax on the 30,000-dollar difference. I don't have to worry about selling it. It's gone, you know, good dealerships give like a good experience. It's just part of buying a car. And at the time we started, I think there were like 30 million cars that got traded into dealers every year. So franchise dealers. This is a dealer that says Toyota or Ford or GM or whatever, right? This isn't Jack's auto barn down on the corner of, you know, what you'm called street, right? With a bar wire on the top of the chain link fence. Those are independent dealers. So his car's got trade in. And then dealers, you know, my dad just because of the business he was in. We happened to our store was on a road where, you know, we had a dealership right next door, Patrick Pontiac, P.O. GMC. I remember. And I got to know this guy over there, you know, again, remembering back to, you know, my teenage years and my late, you know, my early 20s. And I remember they're like any other business, right? Like you think these car dealers are rich, but they're not. They're all run a business in cash on matters. So when you got somebody that has a business that's trading in very expensive capital assets and they have money tied up, dealers need to free that cash up to go run their business, pay their payroll, pay for all those wonderful ads that we've seen on TV, throughout our childhood and adulthood, you know, and so they had to get rid of them. So, you know, I might trade in my 2019 car when I buy a new 2025. The 2019 car, if it doesn't get retailed, which it usually does by a third of the time I dealers a lot, that dealer picks up the phone and calls another dealer and tries to get them to buy it. Same way my dad was like trying to buy and sell in teeks or they would take it to a physical auction that was literally a place that looked like, you know, it would be like, they could be a two bay garage. It could be all the way up to like a 20 bay, you know, it looks like a transfer station for like goods. There's just a place where people bring cars usually once a week. They get auctioned off with an auctioneer, you know, like, hey, but right like the guys that say that it literally was there with the gavill, the car would roll into the lane 60 seconds. People would bid on it. It would either sell or it wouldn't sell and roll out. It just seemed busted. I was like, this is 2015. I mean, it's not today where like your phone is literally buying you toilet paper when you scream at your wife asking where the toilet paper is. But it was a point when you could cross buying stuff on the internet, right? And, you know, when Joe originally had the concept, you know, it went through a lot of versions. But the idea was like, we should be able to buy stuff, you know, instead of having our laptop like the one I'm talking to you on right now, you know, we should be able to buy stuff from our phone and use card dealers. This is the only way they can get, you know, we can get inventory is the buy at auction typically. So we just got to make it easier. I've used card new card dealers want to get rid of the cars use card dealers want to buy the cars. Let's just make it as easy as possible. We took a really buyer centric approach this because we knew that if the, you know, we knew if we could secure supply and we had this kind of like rabid starving, you know, demand base that the marketplace would just start working. So we did we built, which if you look at the early, the very first version of ACV literally looks like something you could probably make. How one of these no code systems and like a couple hours. You know, it was a picture of the car. We just literally had a text box and wrote like, this is what's right with it. This is what's wrong with it. And there were just buttons. You could bid $100 or you could like set your price and people would just bid on things. And it was almost like gave a fight at a low and put it inside on top the phone. You know, look early early days. There is a lot that goes on like we were we were not. I still do not believe ACV even currently is can should be considered technology company. It is a tech enabled service business that has a field service aspect to it more than it does like a deep tech play. Right. You know, we've taken the exact same mechanisms that happened in auctions for centuries. And we just were the one of the first we were one of five companies has started within like a one year period. We were one of the first that really went mobile first with it. And we were one of the first that took the whole light the whole process of figuring out like everybody had to do their own diagnosis on every single car. We were like, this is silly. We'll send our own guy out. He'll look at the car and then we'll stand behind whatever he says. So now everybody for the first time it was a flattening of the information everybody had the same info. And they now knew that most of the most of the you know most used car owners were the ones that went and bought it because they couldn't quote and quote trust anybody. But now they could trust us because if we were wrong, we were going to put our money where our mouth was. If we said this car runs and drives great and you got it and didn't shift into second gear. That was on ACV. There was no longer on that employee who what you pay him is the amount that that car costs, you know, you would have to pay him for 12 months to pay him the amount that you just bought that car for. And it's not on yourself. So it's alleviation of your time. So we made easy to buy stuff and we kind of fixed all the uncertainties and surfaced all that information, which again made even easier to buy and also lowered the risk of purchasing these I would call them snowflake assets. Every single one is different. You know, so it was the long answer sorry about that. But you know, no, can I can I dumb it down for a second just just for me trying to get it in a good experience or understanding of it. So you basically have a I go in I trade in my car. Yep. The you said something interesting, which is a a third of the trade-ins actually get resold at the dealership at retail wise. So the other two thirds are though, that's before you guys existed. That's I'm picking up a phone and I'm just calling my network of people and saying, "Hey, I got this car, come get it off my hands." So a third gets hold of the deal and these are very round numbers and these are from the back of the deal. - Yeah, understand. - Okay. A third of the cars, I was like, "Hey, this is I'm a Toyota dealer. This is a three year old camera. I'll clean it up, I'll sell it on my lot." Okay. A third of the car is, "I'm a four dealer and I just took a Silverado and no F and Way that I'm putting out my lot, but my buddy in the South Towns will buy it. So I'll sell it to him and he'll be able to retail it." Right? The last third was, "I don't know anybody. Maybe it's an older car, it's got issues. I'll send it to auction and those are the ones destined for primarily used car dealers, independent used car dealers." So think of it as a third of 10 million, 10 million, 10 million, 10 million, right? And those physical auctions, the last third, the last 10 million, they were operating, maybe that number was 12 million or something, that doesn't matter. The point is they were operating across the United States. There were two big players, manhime and edessa, when you took a, we literally took a map and we did manhime and blue, a dozen red and we just kind of put dots where they all were and it literally looked like a political map. There were red states, there were blue states and there were purple states and like, but that was it, there was a two party system, right? And because they both had made so much money so long, as long as there's enough healthy competition between them to keep number three from showing up, they were gonna split the pie. Who knew who was gonna, I mean, manhime was always bigger than edessa, but they were the biggest players, right? And that's the world we entered into. Being in, until we made the argument, we said, "This is silly, physical auctions run, your car gets to run across the auction block one time a week." And if it doesn't sell, you gotta physically move it and then hope to run somewhere else or wait the week. We're like, "We'll run it. If it doesn't sell, we'll run it again in four hours. Who cares? Doesn't have to move." And if it sits on your lot and somebody happens to come and want to buy it retail, it's not gone, you don't have the opportunity cost, right? There was physical auctions, we dubbed the term sole, sole consignment. You have to give that car, it needs to go to somebody else and then they have the consignment on that vehicle to sell it. It's physically gone, right? It's really hard to call an auction being like, "Hey, I actually have somebody here that wants a nine-year-old Jetta, which I know I sent you and it's going on the block tomorrow, send it back." Like that doesn't happen, right? So it just, it also allowed us to, you know, we found out the days they were coming to pick up the cars. Let's say they picked up cars for auctions on Tuesdays and Fridays. We'd make sure we showed up Monday and Thursday, right? And we're like, "Okay, we want to see him first." And it was a really easy sale. There was no enterprise contract. There was no negotiating. It was, "Can I go give you a couple numbers on your cars?" And they're like, "Sure, you know, we went." Again, we did have a dealer's license, so it wasn't, we were, you know, the proxy for these people bidding on our platform. We would load the cars ourselves, we'd get them up there. We'd ask them ahead of time, "What do you want for this car?" "I want 12." And then, you know, there's the 12, that's ridiculous, you know, MMR, which is like a trade book, says it's 11.5. Well, and you knew that they were always bullshitting with you. You know, so there was a human version of it. We go outside, it was a 20-minute auction, so you could go launch like three cars in an hour, go get something to eat, wait for them to sell, you know, wait for them to get bid up, come back and be like, "Here's the numbers I got you. You want to sell them?" They're like, "Wow, you just sold those?" We're like, "Yeah, I remember we used to like, I would carry a checkbook on the rare occasion, I would do it." And I would just write a check, and be like, "Here you go, when it catches, I'll come pick them up." They're like, "Wow, this is real. It's so easy to make somebody a believer." I mean, these weren't the reasons why, we weren't like, "Let's build a business where there's no conge, like we didn't take look at it that way." These were just natural dumb luck things that came from the way we built the product that once we realized we were doing it, then also, and we just did it a lot more, and that's how the company grew so quick. Got it. And the shifting of liability piece too is probably just another example of that, where it sounds like it became, "Hey, I give this to these guys. Now it's on them versus the auctioneer calling back and be like, "You give me a piece of shit car." It's lower the threshold of access, right? So it's easier to access this stuff, increase the amount of information so I can make a better buying decision, and then reinforce a guarantee on the back end of how, like, you could now transfer your trust to somebody else because there wasn't that big of a, like if I trusted you and you failed, it was actually better if you told, if ACV said that this car runs and drives right, and it doesn't, then if I went by myself and I thought it did and it doesn't run right. So now you've got people being like, "I'd rather buy it on ACV." If the same diagnosis I give, the ACV rep gives, and they're wrong and that diagnosis is wrong, ACV's gonna pay me in one case, but I'm just gonna be left home the bag and the other. I would rather buy all my stuff on ACV. So you started to see these buyers, they're like, "I just, I only buy an ACV." They would retool their business because they started growing these businesses outside of a solution for the risk that would normally exist with being wrong on a vehicle assessment. So fundamentally change the way, like, use car dealers. If they became committed and devout to, and this is for online platforms in general, I thought we had the best, the most comprehensive, the high-scared T, but that was what made such a swift migration over to digital was, it wasn't that you could bid on your phone like that's cool, but you could basically go to any other option, bid on your computer, the UX wasn't as good. It was that we created this unified, or excuse me, is uniform, clear set of information and then a guarantee it backed it up. I really thought that was the most special thing we did. But I imagine that was, I mean, that must have taken some tech on the back end in order to be able to have confidence in that guarantee, right, or was that still pretty? No, I mean, we would train guys, threaten them to high-health that they couldn't screw stuff up, and have them just run through, like, did you walk around the car of the paint meter? You gotta remember, if the average car, let's say 40 people looked at a car to physical auction, and there were, those 40 people each had, how much time would they spend per car, a minute? Who knows, like, there's, at a physical auction, you can't drive the car, they start it, they'll drive it forward, drive it back, turn it off. That's it. If you're running for like two minutes, that's it. Maybe it's a cold start, maybe it's not. So you got 40 people spending a minute per car, that's 40 minutes of the man time that's there. You know, and then, you know, they gotta walk the whole auction, maybe wait for you looking at cars for four or five hours before the auction starts. Well, you know, in four hours, if I'm looking at, you know, let's say it's really two minutes of car, I'm looking at 30 cars an hour, you know, that's, you know, 60, 120 cars that I look at in four hours. I've exhausted by the end, you know, and I don't know what I'm doing either. We would put a inspector out there that would take 20 minutes per car, but one person that'd share all the info. So now all of the sudden, that car got 10 times the amount of minutes that eyeballs were looking at it. It walked through an actual list of things. We were able to do things on the dealership lot. Like, we weren't really supposed to drive them 'cause we weren't sure too, but like they would drive it around the dealership a couple of times saying we're just trying to find a good place to take a photo to see if we can get into like, third gear to make sure it wouldn't slip. Like there were, so it was way more, hey, assume, and we were getting people a lot of the guys that were our vehicle condition inspectors were car guys. They were guys that lived at AutoZone and advanced auto parts in Napa, did their own work and probably had detailed businesses in the background. Like they knew the fundamental nuances. I give Joe, one of my co-founders, Tonicredit, he was a car guy for forever. Like he knew all the things to look for. We would build databases and be like, okay, Subaru's have timing chain issues between 800 and 180 and 100,000 miles. You know, so how do we look for that on this model? Right, that was the kind of stuff it wasn't. It was like guys walking around with stuff you could have bought from literally advanced auto parts for $100 worth of tools, spending the time on the car. That was it. Yeah. And was the business model on day one? Was it commission base? So when you sold, you did not charge anybody, anything unless the car sold. We came and did all that for free. That was the point. We could come shoot, we called shooting, photographing and doing the condition, the CR on a condition report. We might shoot 100 cars and sell five of them the first time. And we're like, okay, you know, selfie is 150 per car at $750. So like you did all the effort, $750. And I'd say, no, we did all of that so you realize that we're gonna be your partner forever. And we only sold five cars today and that's because you overpriced them and if you can get the price, we'll come again, but you just, you know, you got to be a little better on the pricing. And over time we learned how to negotiate because the truth is nobody knows how to price this asset. This isn't pork bellies or butchers a week, a weed on the murk, right? This isn't like a stocka near a stocky stage. Everyone's different. So there was a very gentle education process where you had to take, use car dealers were probably wildly insecure people because their ego's, the defense around their ego was massive. You had to say, "Yes, you know everything. You do." But if you want to sell this car, you know, looking at how many people looked at it, we made a very clear here's how many people looked at it. Here's how many bid on it. Here's how many went through photos. We didn't lie to people. We said, "This is how many people are on the other side." So, this is just what the market is. You were probably right five days ago. So, use our app and you'll be right every day. You know, maintain that position of a. that I'm the present kind of knowledge that people believe, or at least they believe that they had to have. So, in some subtle way, it was like, "How do you make them continue to stay on the stool of which they put themselves on?" A lot of psychology. So, Jack, I remember a moment, it's supposed to have been 2016 or 2017. I was talking to Matt Brennan, who at the time was it, Bane. He's like, "Hey, man, you know anything about this company, ACV auctions, they just went from, I don't remember what it was, like one or two to 12 in the last year." And I was like, "Fuck me, man." I like, 'cause I've met you guys a year before, and I don't know, whatever, my partner. I don't even remember what happened. But, what was, like, could you describe, what was that moment that must have been some sort of unlock? No, look, that's like AI growth. Yeah, yeah. No, seriously. Look, I mean, I sit on a couple AI boards right now, but I think. That's a whole 'nother conversation. We'll talk about that right after this, right? This is what I told people. I said, "My grandfather used to plant tomatoes. He had tomato plants, and people go, "Where in the house is the story going?" And he said, "Shut up, just listen to me for a second." Right? You put it tomato plant in the ground. If he's sit there, you wait, nothing comes up, right? Eventually, it sprouts, eventually it starts making tomatoes. And it makes a one, then it makes a couple, it makes a couple. And depending on how you take care of that plant, it'll make more and more and more, and it takes a while. Really, assuming environmental conditions are the same, tomato plants grow the same way. The thing is, it takes a lot of care in the beginning to get them going. And the question is, you gotta believe it's gonna happen before it starts happening enough, right? So all we said was, "ACV did not benefit "from any type of exponential growth "or any type of economy of scale. "Once, we hit a critical mass, "and a critical mass was not very big." Again, there is a. What happened? The cars sold in Washington State do not matter to people that live in North Carolina by and large. It costs too much money to move cars. Think of it like a dollar a mile. The margin's not that big for them being that far away. So within a region, a watershed, as I would call it, you had to build up a critical mass. But the truth is, all that growth was just from fundamentally putting out these linear growth models, these tomato plants inside each of these cities. You have a territory manager. The deal was, you have to close two dealerships a month. That's it. One every two weeks. You gotta get somebody one every two weeks to start using ACV. We did not mandate a lot of stuff around it. Ideally, that meant that we would sell 15, 7 and a half cars per dealership per month. Okay? Now, they might not start at 7 and a half, but and that was a pretty low number. Right? Dealerships are moving 30, 40, 50 wholesale units a given month. Some were removing 100, some in Florida are moving like 500, some in the town outside of the town I lived in, which was even smaller farm town moved like two. Right? There was like a massive, but like that. You was just keep putting the seeds in the ground. They will all start producing a yield. And if you look at it, it was phenomenally consistent. It was just, we hired more people and we retained our customers. That was it. There was no like, oh, amazing. Yeah. Like did it help that we people started to know who we were before we showed up? Yeah. But at the same time, we also started going to the territories where we weren't the first digital auction going in. I would have rather been the only digital auction. No one knew where we were had to educate everybody along the way because the places that we were first are the places that they do the best than the places we enter second are the places we do not do the best. And it is that easy. The logic is that simple. You know, your first loves you to love is what I used to tell people. So, um, but that was it. The unlock was it looks giant, but that's because we just started hiring four or five territory managers at a time instead of one every three months. And then all of a sudden, look, I mean, was it a perfect straight line? No, it really looked more, you know, it was like, you know, low. And then it would, it would cut up because there was momentum with the inner market. I can go into all sorts of economic fun little drawings of the way things grew. And I could say it had to do, you know, behavioral economics. They had to do with, you know, like bell curve effects. They had to do with all those different stuff. But the truth was, we just did something and we did it loud. You know, we would show up at dealers and, you know, show up and buy Mepizza. And if they throw you out, buy Mepizza the next day, too, like just keep doing it because everyone's eventually going to try and use this, right? And guess what? I bet you, I mean, I've been in ACV in a while, but I guarantee the first territories we were in, which was Buffalo, Albany, Rochester, Syracuse, Bigampton, Scran, like Middle-Nower. They still, I mean, we're still absolutely in the market leader in those. And I bet you they are still the halo accounts that everybody points to. They were going to go raise money. They would point to that cohort again and say, "Look, this is what happens when you're there first." You know, so, but that was it, man. You know, and that's, I bet you when we start saying apart, you know, these AI native companies, it ain't a lot different. It's like, are they really making more money per, is there dollar marketing really buying them that much more? Or is this, you know, nobody pulls the cheese apart and really figures out what's going on between it. Like, you know, there's, I remember reading your gazes, one, when it came out. And thinking, this is a super interesting business model, but it feels like it's going to take a while to people really understand how this works. Like, I think Uber got so far in advance of talking about their geographic strategy that almost like the market understood it by that time. But it's very, it's kind of similar in a way, right? Yeah, I mean, look, like we're not, Uber was going to evolve. You know, like, you know, originally it was, hey, I want a black card to show up and I got money, you know, I did, I had a little bit too much fun in the club and I really don't want to get a doey, right? Or my dad doesn't want me to get a doey. So he gives me money and I just, you know, for this to happen. And then it went from that to, you know, like, like anything else, and went from that to bigger cities. I mean, look, I lived in New York, dude. I mean, who was it? Was it Cuomo or Patecchi? One of them, like, didn't let, outside of New York say, didn't let ride sharing come here until like, not that long ago, like 10 years ago. It took until Bloomberg. But like, the thing is ACV fundamentally, now they're doing all sorts of different things, which I don't want to get into that because that's not like a grumpy old architect who somebody like fucking paid it to house they loved purple, right? But I think they got away from their fundamentals a little bit of fundamentally is the same business. This is, this is, there's a lot of money tied up in something inefficient. We make it more efficient. We try and optimize user experience. You know, the term trust and transparency is something that we kind of, I don't want to say coin because everybody has set up, but since the very beginning. And the idea is just to make it a very easy process. But it's not rocket science. That's why when we went public, I was like, I'm not going to say anybody because I own a whole bunch of stock of this thing. But like, this is a service business that's got technology on the outside of it and makes it run efficiently, right? We're never going to hit 70% margin in this business. Like, not even a 70% margin off that 4% take rate or the rate that we took off it. And that's okay. It's also not going to get AI to out of existence. It's not all the sudden, like the guys that go and view cars. I mean, I'm an investor and a couple technologies that use other tech, other pieces of optic base and other types of technologies to do car, a condition scanning and stuff like that. But still, like, it's a human business. Marketplaces are still until humans are making buying and selling decisions. Like, they're human. You have to know the human psychology and the psychographic nature over there. So let's let that's a perfect segue. You know, I mean, for those who might not be familiar, Jack is a also not just a marketplace founder, but a prolific investor. So you've got, you know, a huge perspective. What do you, yeah, what are you seeing in AI marketplace? Is it like, is it changing things? As you just mentioned, in some cases, it's not like, what's your view on the future marketplaces posts L.A.? So I don't really use chat to PT or large language miles. I don't. I exclusively use them ask stupid questions to, well, I use them with, you know, if my son wants like something funny, then I ask it for some stupid and funny. But besides that, like, I just use it to like, remind me of things. It's like, it's like having like a secretary. I'm like, where did I put it? Like, where on the internet is the NVCA doc for this thing? Right? Like, so, so please realize it basically a lot of it when it comes to that. I don't know enough about AI. The way everybody talks about it to know what the end state of it's going to be. I feel like that it's a science. And if I was to talk about it, it'd be me talking about thermodynamics, which I also understand. I like to get a warm car in the winter, but that doesn't mean that I understand like key transfusion and all that kind of stuff. So I don't pretend to do that. I think I'm at an advantage acknowledging this is a science that I don't know. I will tell you this. Now the way people talk about AI, I realize how many companies I have worked in in the past that had the chromagnum version of AI in it and just never knew how to productize or market the name correctly. To me, this is what I think. As I just said, as long as humans are making decisions, everything is human-based sale. I will also note it terrifies me when I see these things of like, especially the visual AI of like, you know, it's like a picture of Tom Cruise, but with like Eagles wings, like fighting like a Velociraptor, it looks real like I'm like, okay, it's crazy. But what I tend to focus on, like again, I'm a bricks and mortar guy, you know, I don't think AI, I think AI does everything that's not at the end. Right? I think the ends, you know, the relationships with a customer, you know, the human aspect of things still exists. And AI just kind of shrinks all the work that happens in the middle and makes it really easy. You know, it outsources it to some degree. And so I'll give you example, like I said on the board of a foundational AI company called Akito. Super cool company. I have a passion for, you know, for healthcare. I think it's in an a human right. We don't have enough doctors. We can't just call China and ask them to ship us another, you know, 10, you know, we'll lift the tariffs to send 10 more, you know, boat loads over. And I think it's a very tough thing to diagnose because I do truly believe like while you could be wearing your Fitbit and your Woop and your Whip and whatever else, like there's still a very human centered component of being able to self diagnose what's going on. If you're in a trusted environment, right, we're going to go back to trust and transparency again. But Akito is really interesting because it's taken, it's the first real time that AI has enhanced, greatly enhanced the ability for like a cardiologist to see more patients. And for those patients, they have as good or if not better clinical outcomes, period, like early cohort sort of like three times better, right? What is it doing specifically Jack that's making that possible? If you think of this is for cognitive medicine, so think of, think of your mom or your grandparent going to the doctor. I used to be my grandmother's caregiver and my mom's side of the family, my grandmother's side of the family is history of cardiac issues, right? She was in congestive heart failure for like 40 years and still lived to like 96. But it didn't help that I mean, it didn't hurt. She was like four foot tall too. Like literally her heart just had to like move blood like one cell's length, right? You know, so, but you know what they did is they said, and again, I love these ridiculously simple things. You know, I was just talking to the guys this morning about it. All right. You know, the average visit in a cardiologist is 26 minutes. That's how much time it takes a cardiologist to went up to get one more patient in. That includes coming in and having the initial conversation with the patient to figure out what's going on. They all have different ways of sussing out the information, taking that information, then running it across their own brain of what potentially are there any issues? Are there any red flags deciding if those red flags are validated again, going into the corpus and medical knowledge that's there, then making conclusion, diagnosing on what's going on. And then taking that and saying, well, I think it's X, I think you have congestive heart failure. And so because of that, here's the care plan that I think given all your other factors and the way the winds blowing in the fact that it's a Tuesday that this is what we're going to do for a healthcare plan and then sending you on your way. And then immediately having to do that again in a batch process, right? You know, they don't want to figure out how to just like do like a thousand visits at the exact same time. So was it cute or do it? It takes that, which is just a sequence of steps. It's a process. It breaks it up into individual steps. It identifies with those steps each are. And then it says, do I do this? Does the doctor do this? Does a medical concierge do this? Does an LLM rip through this? Do we build another piece of technology to evaluate it? And just like businesses that are thinking about using like an AI BDR agent, but like a human that's a salesperson, you're able to take the idea of care. And you're able to say, well, maybe since I'm serving a large Burmese population, instead of me being, you know, an Italian doctor from South Jersey with a nasty, you know, like kind of demeanor, bad bad sign manner, I can take that medical knowledge enough of it, put into a tablet, put that tablet in the hand of a young Burmese woman that's got maybe a barista at Starbucks and has great interpersonal skills, take that 26 minute meeting of which the first six were giving out what's going on. Stretch it out to 40 minutes. Let them ask you all the questions, surface all the things that could possibly be there. And then when the time of the doctor needs to come in, all that investigative work is done. That's the first part of it. It run against the potential things that could be there. That's at least outlined and then it's there's lines crossed out why they might not be there. And then the doctor chooses, do I, is there enough knowledge here? Is there enough information for me to be able to either maybe the diagnosis is all but made. And by the way, if that is correct, here's care plan based on what you prescribed in the past. It's there. And they decide what they want to take off the shelf is a pretty made meal and what they want to say, no, I need to go in and talk to the patient about this one. The access to triage process. So in a day when they the average cardiologist, he's 14 patients, you know, maybe two of them have complex cases, but you don't know which two. And so it allows you to get through those. The ones that don't need as much, you're still there. You're still giving them same level care. You're allowed to the doctor that can then spend more of their actual time on the complex cases. Plus you free up a bunch of bandwidth. You can go from 14 to 40 patients, you know, you go from 14 to 40 patients. The complex ones get more doctor time. Maybe that does not need everybody feels good about that because now I'm you get more revenue. Yeah. They're doctors making more money. Yeah. Yeah. Like, you know, I want to sit down with somebody that I can communicate with because if I'm not open, honest, genuine and direct, then there's no way for me to get the information, you know, maybe it takes 10 minutes for me to talk to you to say, I haven't had a erection in three months. I can't feel the bottom of my feet. I think I either got a zit on my back or it's a cancer's lump. Like, who knows? But the point is you've got to get, you've got to get people to be open on and transparent. And it all comes down to the UX and the UX is human, right? So like, that's what I think those are the home run. I mean, I love a keto, invest in it on the board, right? I'm also biased because of all those reasons. But yeah, I think it's going to transform healthcare. I mean, we're like, we need four times the doctors to see everybody in the US. We can't, we're not going to invent those. Like, they're not going to lower the standards of what it takes to be a doctor. So let's just take the knowledge of being a doctor and just know where to put it. That's when I think of, but, but when I think of that just to be clear, no one, no one agrees with it. I'm like, that's blue collar, that's field service. That's like, it's not AI. AI is there that does a lot of the work, but it's a modern day version of regression analysis, right? Like, I know anybody that knows about AI is going to say I'm an idiot right now, which I have and I admitted that at the beginning, but, you know, I think it's kind of, but I mean, I think what, for me, what LMS enable that is most interesting, I mean, whether it's this example or marketplace is more generally kind of comes down to that. I mean, if you think about that transparency, trust, transaction framework that a lot of people use for marketplaces, it is in that first part because exactly what you described, like, okay, I maybe only talked to my GP, like, I don't know, once a quarter. And when I do, it's a 30 minute conversation. What if that changed where I could literally call, you know, my AI version anytime, talk to them for two hours if I wanted to? And I still have the doctor for the, like, true trust part, but like that information surfacing, we can now do in a way we just couldn't before. So I, like, I kind of abstract this to, I don't know, maybe your retailer that's looking to figure out what you want to get for your fall collection or your, I don't know, like, someone looking to buy a car. How does this change information sharing? To me, you still, the user in the API that my data as a human plugs into is not a computer screen because I don't know how to communicate cross species. Like my, I'm 41, okay? Like my son is 10. He's going to know how to do it. It's going to be way more native for him. But I'll tell you what, if you told me, here's a screen. Tell us how you felt today, Jack. Like, I'm not going to be, I'm going to be like, get this out of me. Like get this out of my face. But if there's, I mean, I'll go to the full extreme and I'm not saying this is what we're doing. But like, if there is a gorgeous woman that even though I'm married, like, is fun to look at and talk to, like, take like a bartender. Why do people go to the bars? They can drink at home. They go because they want the interaction. And what will they say to those people? They will say, it's not just the alcohol. They will pour it out. That's what you need. And I still think so. I think AI enables people to stop being doing the mechanical functions and doing the, and focus on human functions of things. That's where I, that's where I get really excited. because it's not, you know, SkyNet and like terminators and like whatever else, it's, wow, like when people talk about a retisons potentially happening, I don't think it said everybody's gonna turn into DaVinci and Michelangelo. I think it's that people are gonna be able to, for better or for worse, get to function more in the, what are normally the little moments of life where we get to be human, the jobs will become human jobs. We will become like the service industry, right? Weightresses are still gonna be waitresses and they're still gonna get great tips that they know how to take care of somebody. But like the AI version of what goes on in the back room is that everything's pre-ordered. We don't throw anything else. There's robots making the food like it doesn't matter. The front of the house is gonna stay the same. The back of the house is gonna change. I think I, and Bluuck, I want to get a review too, but I, I think, I agree with that point that I think over time jobs are gonna get more human because that's what we're good at. But I just, here's an interesting anecdote. And I may have mentioned on the pod before. So sorry, I list this as a repetitive. But if you guys tried Bordy, you know, that thing that you can call and then it talks to you and makes intros. You know, check it out, Bordy.ai, right? It's, it's kind of a, I don't know what their belly ends to, but it's kind of a fun thing. You talk to it, it asks you about your background and then it tries to just make intros to other people. It's talked to. You can just call it. And I think what I found was, I probably ended up talking more to that thing than I would have if someone I didn't know called me and said, hey, tell me what kind of intros do you want? Why? Because I wasn't worried about offending it. I wasn't worried about its time. It's interest in me is unlimited. And now when it tries to make an intro and I say, no. And then it says, oh, yeah, I remember you said 18 months ago, you said this one thing like the memory is infinite, right? So it's it's, which is also terrifying, but it's definitely terrifying. But like when I think about, who do I trust to make it better entrance for me, Bordy or someone I just met 10 minutes ago, it's going to be Bordy because it I've told it a lot and it remembers absolutely everything I've ever told it. So I wouldn't trust it to deliver care to me for healthcare, but to intake information, I would. So, you know, when I think of the bell curve, where do you think you are on the bell curve? I mean, you're a VC on a podcast about AI. Okay. So just to be clear, you're probably a little ahead at the time, right? I mean, you know, this this is where, you know, I'm like, you know, literally I have a mirror next to me and I, I don't look at a mirror to see how I look. I look at the mirror to be like, that's who you are. Like, don't forget it. Right. So and the truth is, I think there's look, but again, there's there's a place and a time for everything. Yeah, that makes a lot of that makes a lot of sense, you know, that it makes a lot of sense why in that given interface, given who you are, I would look at it and I'd be like, yeah, Nick told me about this thing. The chances of me actually looking at board AI and this is no disrespect. You are about 0% right. So, right. So I don't blame right now, but all I'm saying is is like, I think that is very interesting. The question is, you know, and I don't do as much on the consumer side. There's a reason I should have said this, I primarily live in a B to B world, because I think humans are fickle and I have no idea how to market to actual humans. You know, or you know, capture them. We're a business I can understand. Businesses tend to be more like logical. Right. Let's like, okay, like we're not just jumping from platform to platform every single day. Right. So, and so I would wonder if the business version of you your business mind, right. Your operator mind would how long it would take for you to accept something like that. Like for example, like if, you know, if somebody was trying to reproduce ACV and they were going to use a piece of AI that remembered everything, would they, you know, personally they may like that, but would they trust it business wide because again, I think trust comes down to a different thing. You know, for every story I heard about what you told, I've heard stories of like people building like malicious AI that like try to people in the killing themselves. Right. You know, if you were my kid and you were talking to this piece of AI all the time, you know, if it's a little bobby from down the street to missing a tooth and chew and bubble gum. And I know his dad takes him out. Right. I'd be like, it's okay. But I'm like, oh my god, am I talking to this thing that's like manipulating my child's mind, right. This precious malleable thing with infinite wisdom and knowledge. Right. Owned by who owned by somebody behind a dog and in another place. Right. Yeah. A very good question. Well, Jack, we're getting towards time. Let me, let me ask one more and then Luke, Luke, round us out with a final question. But maybe just, you know, I'll let everyone come out of my like, you know, science fiction moment. Here and zoom back out to the real world. Like you've invested in a bunch of marketplaces. What is exciting you right now and has that change that they are. Is it really just the same like underappreciated markets. It's I think the companies. Okay. I like I like traditional business. I find a lot of joy in things. People would find boring. Right. I mean, if you think of the companies that I'm spending the most time with. That's the most exciting about like, you know, we just did a series a super excited about it over a pasture. Right. I think I showed you guys that awhile ago. Yeah. Pass for mediation. You know, like thinking of like extermination as a subscription for single family rentals. Is there very narrow. I'm big on narrow deep wedges. It's their narrow deep wedge. Right. Pass remediation. So like you pay monthly fee. And when if the bugs on the list, you can buy a short list, medium list, long list of bugs covered bugs on the list. All you got to do. Make a phone call or take a picture. A bunch of ways to submit it. Somebody shows up. Take care of it. No problem. No extra charge. But you pay Netflix. You know, directly through your landlord, you know, through your benefits package. You pay for pressure. Right now we're going through this thought process. I was just with them in Vegas, you know, we're sitting down with our new investors, IGP, which is cool group out of LA growth investors. Great guys. And we're talking through it. And we're like, okay, how are we going to use AI? Right. Like how are we going to use this? Like we have 70 employees. We just raise the bunch of money. We can go hire a bunch of people, but like let's be thoughtful. I'm where it goes and where it doesn't. And it's so great. I mean, I'm working so closely with them because I miss building stuff. Obviously. And it's just so fun to like look at this now. It's a former founder and somebody else's company and be like, oh, like look at all these new toys. Like how do we play with them? Where do you put them in? I would have been excited about past year or five years ago. I would have been excited about 10 years ago. Could have happened five or 10 years ago. Happened to happen now. You know, so the same things excite me, but now I get to ask, you know, the cocktail conversation is, well, you know, how will this affect, you know, you know, the way we do our service requests, you know, and we'll allow us to have different views into, you know, better unicognomics long term, you know, and stuff like that. Things are changing faster now. I feel like I think this is the thing. So, you know, the key to that I talked about without AI would not be as exciting of a play. It's the same kind of stuff. It really is like the same kind of stuff, a really good friend of mine. Colin. Colin heard CEO of mock the IO colleges that are raised back by the venture collective. Cato. There's great great VC to work with. He's got, you know, his last company was like a, it was a hardware product that bolted onto a combine and, you know, like a agricultural combine and gave it some base level of autonomy. And he said, you know what, we should do it at the OEM level and mocks doing it primarily an ag, but they're doing it in mining and lawn care in maritime anything. It's not on a road like a human road. They're having this full stack autonomy product, right? That for a marketplace guy, I find that very interesting. I get I come from an ag background. So it allows me. It allows anybody to create. So I do think it's cool, but I also think it is. A first thing I ask myself is if this is an AI play. How quickly is it going to get out AI by somebody else's player? Right? You know, yeah, it's a defense. It's like my son like love that movie cars when he was young, right? And there's like the hot rod red car. And then the next version. There's one that looks like a sweet badass like James Bond transformer car. Like how quick until I just my model gets outdated, you know, my, you know, I get outdated. And I think when you're in I love, I love things are blue, power and field service. I don't think those are changing. So for me, it's just giving me a different, a different vantage point. I now need to look at stuff through before I make a decision. Well, I'll round us out here, Jack, with a, with a final question if that works for you. Please any, you know, one of the hardest parts about marketplaces, you're basically, you know, have a go to market process on both sides. Right? So you're, you're trying to get users on both sides of that. I'm curious. If you have any tips, tricks, hacks for the audience that are maybe in their early days of building marketplace, like unique things that you can't read on a blog post or that an AI will tell you that you guys may have deployed or a company that you're, you know, invested in may have deployed that they really moved the needle. So I don't believe in this world that they're perfectly equal and they need to get done at the same time. You know, look, everything I look at as a marketplace. I walked down the street that 20 years ago, I would be, you know, banging on the door to get into a bar because of the girls that were in there. That's a marketplace. right? You have an establishment. How are you going to get the guys and girls in there? It's not that you try to get them at the same time. You have cheap drinks. The ones that girls like, they come in, then guys want to go into, right? Like it's a marketplace, right? I'm doing an event next week in Buffalo, right? Like I got seven VCs coming in from around the U.S. How did I get them to all fly to Buffalo? Well, you know, yeah, we have companies here, but the companies want the VCs. The VCs maybe want to see the companies, but they really want to raise money so they can go into companies from anywhere. The investors want to see that they're founders there. So it's like a triangle and you just have to kind of like, it's setting up a house of cards, man. Like for a marketplace founder, I'd say there's always one that has to be the dominant driver. And there's one that you have to find a way to coax along maybe even outside your business model to create demand, right? One of my best friends, my biggest angel investment in somebody else's company to date was TCG player. It was a marketplace for collectible trading cards, right? So super cool. Yeah. It's a cool company. Couldn't get a damn person to do our series B. So we ended up selling to eBay. Chatties should still be running it a whole nother podcast. But you know, like in that example, it's like, well, okay, like how do you get people to put cards on to a platform and then have people buy them? Well, the whole company started by them writing blogs about cool strategies. And then they actually secured the cards of which they started to sell. And when people are buying them, they're like, well, why should we be in an inventory game? Let's let other people sell them on there. And so you hot wire the marketplace. There's always a hot wire. There's always a cheap drinks for ladies. There's always a, I'll get a bunch of people with money here so that the investors will want to come. So the founders will want to come. There's always the will sell our own cards. And then, you know, there's always the Zappo story of running around and buying like there's always a hot wire story. And you got to figure out what your hot wire trick is to make the whole thing go, right? Your starter only has to fire when the car starts. You need a starter in order to get the cycle of the engine running. The starter may look completely different than by the starter and engine. They don't like anything like, but you need one to get the other one going. So like I never think that it's both at the same time. There is some hot wire. There's some starter. There's some kicker that needs to go on to make the revolution start happening. And then network effect is this centrifugal rotation of of everything. It's the volley, you know, but a serve looks different than every other part of volleyball. It does. Yeah. Pitch looks different than every other part of baseball. Like in order to get the thing going, it looks different. And I tell people, I love that hot hot wire. Find your hot wire. Yeah, it's good. You've just given us about 16 quotable bits that we can, we'll be, we'll be reusing. Awesome. Well, Jack, this was so much fun chopping up with you here as always. So thank you for taking the time. I love it, man. Yeah. This was, this is a blast. So thanks again. You guys are definitely worth spending time with. I appreciate what you're doing. Thanks, Jack. Good to meet you, man. It's a pleasure. All right. Take it easy, guys. Yeah, Jack. Let's catch up soon. Take care, man. All right. Jack is awesome. Nick, thanks for bringing him on the pod ACV oxen killer business. I think a lot of people have heard about it, but don't know the intricacies. So super fascinating for me to just get a better handle on it. And yeah, a lot of, a lot of fascinating playbooks, kind of tips, tricks. I think the, my favorite playbook is this hot wiring. Like in his view of everything is a marketplace. Like I had a complete different view because I haven't done marketplace. Personally, where you got to go to markets and he doesn't look like, look at it like that at all. So super fascinating, you know, tons of value there. So, you know, we'll be back next week as always. And I want to give a quick shout out to Paraphyn. Paraphyn's made our podcast possible. It provide ready to launch financial products to help your merchants grow. If you're a vertical software business, if you haven't checked out Paraphyn, go do that. Really drives retention, revenue and tam for, for founders. So big shout out to Paraphyn as always. And Nick, good seeing you, my friend. Good sparring with you and look forward to it again next week. See you next week.