Yet Another Value Podcast - Clifford Sosin on Cardlytics $CDLX

Episode Date: September 10, 2020

Clifford Sosin from CAS Investment Partners discusses his investment track record, his investments in Herbalife (HLF) and Carvana (CVNA), and then dives deep into his current investment into Cardlytics (CDLX).

All right. Hello and welcome to the Yet Another Value Podcast. I'm your host, Andrew Walker. And with me today, I'm excited to have the founder of CAS investment partners, Cliff Sosen. So Cliff, how are you doing? Pleasure. Thank you for having me. Yeah, no, absolutely. Look, let me start this pod the way I do every podcast. And that's by kind of pitching you. You know, I did my podcast on Carvana with David Kim from Scuddle Blurb a couple weeks ago. And on the heels of that, I got so many requests there saying, look, if you want to hear more about Carvona, you need to have Cliff. He would be a great guest on the podcast. There's no one more knowledgeable about him. I'd actually never heard of you before, but I went and watched some of your videos on Carvana. And, you know, I think one of the things about great investing, it's kind of being able to see the future, maybe two, three, four years ahead of time. And I watch all these videos of Carvana from 2017, 2018. And it was, it was crazy. It was like you were talking in 2020 because you were just saying exactly where they were going to be, exactly how the thing was going to play out. So anyway, bottom line, Cliff is a super sharp investor.

I'm excited to have him on. His returns are world beating, and it's just great to have him on. So Cliff, maybe you could give us some background how you came, got into investing and everything. Well, I think like most people, I fell backwards into my career. If you went way back, I thought I was going to be an inventor. I was in, and I went to school to study engineering and economics under that theory. And as I studied engineering more, I,

I realized that the details of how to make, you know, a Gizmo work weren't as nearly as exciting to me as the business model around the Gizmo. And that led me ultimately to, you know, to an internship in private equity, which led me to think that this idea of buying businesses might be interesting, which made me think I was going to be in private equity, which, you know, led me to an analyst program in financial restructuring because that was the past that I thought would be sort of best to get into private equity and also avoid being basically an M&A analyst because that's not very interesting. and that in turn led me to a job as a distressed debt analyst, which led me to focusing more on bonds at a place that mostly did equities, which was UBS, and that's where I spent a bunch of time, and that eventually migrated to all equities, and then it was sort of somewhere in there that my way of investing developed, and I developed the confidence to think that I could probably do this on my own, and ultimately that's where where I was where I left when I started

CAS. So, you know, it was, I think, you know, the world is far too complex for anyone to have a master plan. And I think you're pretty naive if you're, you know, 15 and you've decided you know what you're going to do. The world is way too interesting for that. Yeah. No, I think back all my friends, you know, it's the classic freshman year of college, everyone comes in and they say they want to be a doctor when they're done. And by junior or senior year, you know, everybody's studying the art, so we're going into completely different stuff.

So, yeah. So on a total aside, the world would be, I think, I think we'd shuttle people into careers too quickly. I think it would be better, you know, I think there's very questionable value to a lot of higher education. I think Brian Kaplan is a great book about sort of the limited value of higher education. But I do think there'd be a lot of value to letting people sample careers. I think if people had a period in their life, which was sort of broadly accepted, you know,

you're going to spend three years, you're going to try six. jobs, they're going to be six months apiece, that's just what's going to happen. I think that could be incredibly valuable for people, and so to help, you know, a great way to make society more productive, happier, and so forth is to better match people to jobs. And we'll make progress on that over time, you know, polygenetic scores will help, but I think trial periods. I think people don't even know their preferences prior to trying them. I think it's very, very hard to know what you're going to think about something before you're actually there.

And so, anyway, I think there's a lot of room for experimentation. No, you know, I 100% agree with you. But I guess that's a conversation for another time. Let's go back. So after you left UBS, you founded CS, and what year was that? CAS, I founded CAS October. Our first day was October 9th of 2012. Okay, great.

Yeah. And one of the things... People ask, why not the first, by the way? Say again? People ask, why not the first? And the answer is we missed the deadline. The fund administrator couldn't take the wire in. Yeah, I didn't have my documents in a row.

And then people say, why not the 8th? And the answer was it was Columbus, so it would be the 9th. So one of the things I love about CS, and we're going to dive into it a lot, but you run super, very concentrated. You know, I was just looking at your 13F, and I think you could round up to 100% of your portfolio is kind of Carvana and Cardlittics now. So you launched 2012 running a super concentrated fund.

What are kind of your first big investments at CAF? Oh, geez. You know, if I look back over time, you know, the big investments have been in United Rentals, which eventually parolid into Ascheed Group, which we still own. We went for a bit of it, actually. And what else have been really big? We owned a lot of herbal life. For a while, herbal life, which we still own, really dominated, you know, the portfolio. Um, you know, then, uh, I guess, you know, or as herbal life kind of came to be less dominant in the portfolio, um, I think, you know, Carvana kind of came to be kind of the new thing. Um, and, and, and, and world was also a big thing for a time. And then about, more of the acceptance corporation. And then eventually, you know, we came to be more, um, kind of what we see now. Um, I would note that, uh, you know, the Carvana and Cardalitic positions are big, but, but, uh, you know, I think Carvana is, um,

mid-40s percent of capital or something, which is a lot. But, you know, if you add them up, they don't actually round to 100%. We have other things. I was just being loose with the, very loose with the rounding. Herbalife, so you get involved in there for a big time. Is this on the heels of the Ackman short, or are you invested in the, in Herbalife before Ackman's big short comes out? No, I invested in Herbalife, it was like January of 2013, and it was after Ackman

he did his pitch, I think, in December of 2012, if I have all my dates, right? Sort of right around New Year's, sort of right before New Year's, I think he did his pitch, right around Christmas. It was Christmas Eve, I remember, because the stock was way down. And I remember people, like, I had never looked at it, but people were like, he did this into the most e-liquid market possible. Like, I guess good for him. What a way to make your yearly numbers.

Fair point. But yeah, we, you know, this interview had been done. uh four years ago almost certainly the topic would have been herbal life because for a while herbal life was you know as much as a 30 percent position and um you know there's there's really quite a lot to like about that business and we still own a fair a bit it's it's gotten i it's gotten more expensive um and so it's less attractive as an investment um i also think that um you know my some you know there's a range of outcomes in terms of what you think might happen and i I'd say that gradually speaking, my sort of optimism has abated a little, but that, but that doesn't

mean it's not a really great business doing, doing really well, and we'll do really well over a long time. Are you following herbal life before Ackman comes out with the short and then the price kind of gets the stress, then you make the investment, or does Ackma make the short, you see the price of the stress, you kind of, I mean, anyone could see at the time, you know, it traded very cheaply, it had great returns on capital. There was the whole MLM thing and everything, but which kind of comes first to spark your curiosity. Yeah, I was aware of Herbalife the name in part because I'd lived in L.A. and I'd seen the sign, in part because, you know, it always showed up as sort of a quantitatively

cheap, you know, because of Garby name. I'd never spent a lot of time trying to understand it. And then Ackman did his case. And the thing that, you know, ultimately drew me to herbal life. I thought, I watched Actman's presentation, and I watched it, I've said to joke to people, I watched it for very much the same reason, people look at car accidents, you know, and as he was going through it, I thought it was very compelling. I sort of, you know, I sort of generally bought into the whole notion. He had my, you know, my sense of injustice, you know, riled up.

And then he made this point about the retail price that Herbalife's product sells out. Yep. And he said, look, you know, it's, you know, it actually sells much less than retail. And I remember sort of noticing that. It was he had some good data that he'd gathered to support that. And I said, you know, that's interesting. He'd scrape the internet or whatever. And I said, that's interesting because the price that he's saying it sells at is more than the distributors are buying it for.

And, you know, it turned out that he was finding kind of corner cases where people were selling it on the internet or whatever. But even then, people were selling and buying the product. the internet for more than the distributors would buy it for herbal life. And I was like, that's weird. Uh, you know, that's just in his world, distributors were buying the product. They were struggling to sell it. They were failing to sell it. They were failing to sell it. They were failing to return it, which you can do. And then they were losing their money and ultimately being very unhappy about it. And yet, there you are, like on, you know, at the time, eBay, you could just sell it and get your money.

back and so you know it just struck me as you know you had to believe that these distributors were unaware of the internet or you know and it just didn't make any sense or you know just didn't make any sense and and that was the first thing that where i sort of wondered you know where acumen was over time i came to really understand the air black business better i mean people don't i don't think people really appreciate just how powerful a business it is it's um i'd like to say you know people tend to think about, there's a natural tendency to think about the world the way people think about buying a jet engine, right? Very rational, quantitative. But a lot of great businesses work on the aspects of our minds that aren't computers, you know, the sort of evolved mind

that humans have. And herbal life, there certainly is a rational component to it. There is a business model. You can build a business. There's rational incentives that work to reinforce the idea of building a business. However, I think what makes herbal life so powerful is, is, is its ability to motivate people very much in the way a religion does. In fact, people who've criticized herbal life have called it a cult and I've sort of laughed that, you know, everyone else's religion is always a cult, right? And so if you're not in herbal life, it looks like a cult because, you know, everyone else's religion always looks like a cult.

But in the case of verbal life, the business, the sort of motivation of a belief system, that you are, if people in certain religions view themselves as good people who contribute to the community and live good lives and help others as a part of their in life. And to say that to themselves and to say that to others and to reinforce that for themselves, they participate in this group that is sort of focused on that ethos. Well, that's a lot of religions. Herbalife, it's about living, healthy, active lifestyle, you know, avoiding bad foods, and then losing weight, and then sharing those gains that you've had,

improving your life with other people. And that's an ethos, too, that people reinforce for themselves and then share with others through their participation in this organization. And it's an incredibly powerful, you know, way to motivate people. And sort of once you see that, and once you understand that, and we've tried all kinds of different ways to get at the core of that, but then you sort of realize that this is a business that's going to last for a very long time, very much in the way that religions last.

And then questions like, oh, well, did you know that you could buy similar, you know, soy protein powder at a GNC for a dollar less? It's like, well, yes, but you can also buy, like Catholics can get wafers for less at ShopRite, but that isn't really at the point. So, you know, anyway, that's the idea. No, look, that's great. The Catholics can buy wafers.

That's one of the best analogies I've ever heard. I'm going to have to start using that going forward. Look, we've mentioned, I think we've mentioned five stocks so far. You know, Carbana, Carlytics, world acceptance, herbal life. I think we might have mentioned credit acceptance in there. I can't remember it. You know, go ahead. United Rentals.

United Rentals. And I own Ashtey Group today. my United Rentals position I sold a few years ago to buy Ashton. But of the five we've mentioned so far, I'd say four of them have historically been mammoth, kind of short battleground stocks, right? You know, huge short positions, herbal life, huge short positions, big long positions. And actually all of them, I think, have worked out very poorly for the shorts. But, you know, it just, are you particularly attracted to short battleground stocks?

Is that just where you see, see value? Is there anything there? Only by accident. At one point, I noticed this and ran a screen of highly shorted stocks to see if I could find something interesting. It seemed very hard. So, you know, at no point in my sort of investing process have I ever sort of tried to sort of look for, you know, look for things that are battleground stocks to be in with. But on the other hand, you know, when you look at the other way, it's not surprising that that's where you end up because, you know, people don't give away great businesses with bright. futures at cheap prices unless they have something to worry about. And I think in all of my

investments at some point, there is something that people worry about that I think, you know, if I'm successful, I'm able to show isn't really a good concern or is either miswaded or misunderstood or whatever. And that, you know, is kind of the, you know, that kind of gives rise to a good investment. Yeah. No, you know, I was just thinking, I've been thinking a lot about my historical investments and returns. And one of the things that struck me is I barely, rarely made a lot of money buying a company that everyone thought was good at kind of 12 times earnings and just waiting out. I'm not saying that can't be done. But most of the times I've done something done really well has been you buy something. Maybe it's cheap, maybe it's optically

cheap. Maybe it's not optically cheap. But you buy something, but there's a big, a lot of doubt about it. You know, like cable's been one of my investments. And a couple years ago, people were cord cutting was a huge worry. And wireless internet replacing it. And, I didn't think those were big issues, and it turns out they haven't been. But, you know, that type of doubt is where I think the real, if you can get an edge one way or another. So, I don't know stocks, the edge is there, right? The goal isn't to buy the good businesses. Exactly.

The goal isn't to buy the good businesses. The goal is to buy the good businesses that other people don't think are the good businesses. Because that's, it's a paramutual betting system. Yep. And it's not about, and so, you know, that's what you're observing. And so, you know, the idea that there are long sellers. who want to dispose of something at the wrong price, you know, is probably somewhat correlated with the idea that there are short sellers

who want to bet against, you know, the investment at the same price. So it doesn't shock me, you know, that I've wound up in things that are highly shorted. On average, I should note, I think there's been some empirical work. I couldn't cite it for you. But I do think that on average, shorts do underperform. I think that's right. And so, you know, it's interesting. you know, one thing I've always thought was interesting was, you know, we don't do any sort of empirical sort of top of the funnel screening. It's all very organic. But I do think that a lot of people put a lot of effort into designing screens that optimized for mean performance of the portfolio that you'd build with the screen. I wonder if there isn't value in building more screens that optimized for the probability that a given stock within the screen is in the top desial.

of performance, as opposed to for mean performance. So then you would hand a portfolio of stocks that have lower mean performance maybe than the average, although if it was lower than you could short it, but have maybe average mean performance, but within that have more volatile outcomes. And then you would hope that you could hand that to a qualitative analyst who could then sort the winners from the losers. So basically throw out the middle of the distribution in some sort of efficient way. I haven't actually done that. I just sort of go one at a time, but maybe that's an efficient

way for someone. No, I think that's a really interesting thought. And I think that's exactly correct. So another thing, we've mentioned a lot of different business types here, right? Like Herbalife, consumer focused, consumer focused, I don't want to, I guess, multi-level marketing, but Herbalife consumer level marketing. Carvana, use car, retailer online. Cardlytics is, we'll talk about it a little bit more, but, you know, it's connecting merchants with consumers through banks. That's a lot of different types of stocks. So how are you screening and finding these different types of stocks?

One at a time. You know, there really isn't a way I could describe some sort of systematic process of finding things, certainly at the sort of top of the funnel. You know, what I would say is that where I get a little more. systematic is after I sort of spend a little time getting to know a business, I start trying to be fairly rigorous in the way I think about that business, you know, using various intellectual tools that I've picked up along the way to try to basically construct a sort of a mental model of how the business and that industry competes and kind of what ultimately gives rise to that business is enduring, you know, competitive advantage. So in the case of verbal life, I walked you through briefly, kind of this idea that there's a set of human motivations that you can observe in other areas and that definitely exist, that I, the hypothesis was and still is, that those are kind of at the core, a big source of what drives herbal life and in some sense differentiates it in the minds of people who are involved.

If we had a longer conversation about Sunbelt Rentals, which is on Biashthe group, you know, we would talk at length about the economies of scale in the construction equipment rental business. We'd also talk about, you know, something called Cornell oligopoly, which is this idea that in the rental business, you provision a certain amount of equipment in advance of need, and then you maximize the yield on that. And so the decision-making process is quantity, then price, not price, then quantity. And the game theory of that in kind of the performance. competitive that equilibrium is very different than the game theory of that in a in in in the protron dogoply which is where you say price then quantity so there there are you know all of these tools and

you know in some sense i've sort of joked with people that like a good investment it's kind of a gadget you know it's it's like you find some quirk about how the world works and you know some companies through evolution they sort of find some some way to take advantage of that quirk you know human beings have this tendency to form groups and to, you know, and to sort of have belief systems and to sort of develop values about themselves through their belief systems. And so if we can create a belief system, like we could probably get them to give us money, right? That's a sort of cynical way to look at herbal life. It's not how it happens, but sort of what it turns into. Um, you know, or, um, you know, you could say, um, uh, you know, it turns out that, um, you know,

in construction, rental, that, you know, that, that if you're buying lots more of the equipment you can get at a cheaper price. And so if you're just a really big, you know, you can, you can basically deliver the same product. And all of those things, you know, it's sort of a quirk, right? It's like, oh, well, because of the phenomenon of declining average cost of production of construction equipment, you can build a business that's better if it's bigger. And so once you identify what kind of that quirk is, uh, the world that can make. Hey, Cliff, I think we lost you for a second. Uh, uh, gadget that the company, uh, can you hear me?

Yeah, I think we lost you for a second. Sorry. Uh, you can you hear me now? Yep, I can hear you perfectly. I apologize. Um, but anyway, I was just saying that once you figure out the gadget that the company's using to be competitive, then you can kind of extrapolate, try to extrapolate the company's longer term future and then just compare that to the price.

Perfect. Um, last thing here, and then, uh, we'll, we'll switch over to Cardletics. So we've mentioned a lot about businesses, about the business, right? But when you were in a really concentrated portfolio, I would think management is a huge consideration as well. So how do you kind of, when you're taking big swings at these companies, how do you kind of look at management teams and think about them? Yeah, my, my, I came, if you'd asked me that question five years ago, I probably would

have said that I think not at all to very little about management with the exception of trying to avoid basically outright, you know, frauds, you know, liars, cheaters, that's those stuff. Over time, I have, in part through experience and in part through some deliberate effort, I believe I've gotten a little bit better at thinking about management. And people tend to think about management as the CEO, CFO. I like to think about the culture of the entire organization, the guy or gal at the top being a part of that, but actually, you know, not necessarily the main determinant.

of what's going to happen. And if you don't believe me on that, just look at all the people who threw themselves against general motors and couldn't, right? And so, you know, what, you know, think of, so then you start thinking about how people behave every day in an organization. You know, people are responding, you know,

they're signaling sort of aptitude to one another. There's sort of a currency that people value when they signal to one another that is meaningful. and that's different across different organizations. So, you know, at Google, you know, they probably really think you're great if, like, you've got some wonky algorithm

that, you know, you're really proud of and you built some cool code, right? At, you know, at, you know, I don't know, Wells Fargo, the sort of what's... The more fake accounts you can open, the better you are? No, no, but, you know, it's probably more about, you know, running your business P&L

and, you know, and kind of making sure that you've achieved everything on their compliance checklist. That's, you know, important. There's, you know, differences in kind of, and, and so under, I think I've come more to think about these very large, you know, culture in some sense is the way you motivate people when you can't supervise them directly. And that's really what I'm talking about here is this idea that people are, you know, your people are trying to convince everyone else that they are a certain way. And the way they do that within an organization can vary a great deal.

And by the way, it can vary a great deal across, you know, different societies and across history. It's sort of amazing. When you start looking at it, it's pretty remarkable how malleable people can be. And so I think about each company as a little society. And within that little society, there are the norms and values and expectations. That society has the leader. The leader is in some sense, it can come from the outside. But, you know, the board is picking, you know, it's very organic in the sense that usually

the leader is almost a function of the society. The best person to run a given company varies a great deal by industry, right? The right person to run Boeing and the right person to run Starbucks, you know, are just very different people, Boeing, you know, and then also the other thing, which I should mention is the problems the company has to solve are very different, and then the processes by which the company solves those problems are different. And those lead to different, you know, what they call organizational habits, which are kind of instantiations of individuals habits. People tend to follow habits. You can then kind of align people's habits in a certain way and create organizational habits. Organizational habits are simply, hey, this is how we solve things. And so different companies face different types of problems. So Boeing has vast coordination problems. The person, you know, designing the Aileron needs to coordinate with the person designing the cockpit control systems. Who needs to coordinate with the person designing the tail wing and the person designing the fuel tank, right?

And these coordination problems are complex and very difficult to manage. It naturally lends itself to bureaucracy because there is no kind of obvious way to make these decisions without kind of having people sit around a room and make a judgment. Conversely, an organization like Starbucks, you know, this is an organization where, sure, there's some things that are centralized, elements of marketing, elements of of kind of purchasing but in the end you can push a great deal of autonomy to the node you can basically run have the store managers run their stores hold them to certain metrics you know measure them across you know compare them across one another try to figure out what the best ones do and and communicate

best practices but it's but it can be run in a very in a much more decentralized way these lead to very different um the people who prosper in these organizations are different the way to lead these organizations is different. The way that you sort of win within these organizations is different. In general, obviously, the Boeing type organization has a natural tendency to bloat and so fighting that. But you need bureaucracy. I mean, you can't build jet engines. That's not saying a Boeing, but you can't build jet engines without bureaucracy. And by the way, the organizations that you're liaising with also matter. If you're doing drug development and you're interacting with FDA,

FDA is a big organization and you're going to build a complementary organization. So the point I'm saying about all this is that I've come to think a lot more about how the people are assembled. I wouldn't necessarily say that I, you know, that I sit around and spend a huge amount of time thinking about, like, is this guy great, or gal, I should say, is this person great at this role? I spend some time at it. I'm generally looking for people who are making well-reasoned business decisions. I don't necessarily care about the outcome. I just want them to articulate to me a set of reasons that makes sense for the choices they made.

But usually I find that there's a fit component to it. It's like the person running the company needs to, the company has certain problems. The person running the company has a certain role. And the question is, does that person sort of fit in that role? And are they making kind of decisions that are suitable to that role, that is sort of sensible. And then along the same lines, like once you learn how an organization is built, like people up, that tells you what it can do. tells you what it can't do, right? Like, so once you learn that you, what you've got as a football

team, then the question, like, you obviously know they can't win a golf, right? And, and, and so, you know, Boeing is never going to be nimble and dynamic, right? It's just not going to be. It can't be, right? And if it were, it would be a disaster. I wouldn't want to get on the airplane, right? And at the same time, Starbucks will never be able to produce, you know, a carbon fiber 787. It just can't, right? And anyway, so, so those are sort of, I guess, um, that's how I think more and more about management culture and organizations and all the rest at this point. No, that was great. Actually, I thought up like 30 minutes of questions on just that while you were talking, but I think we might take the podcast in a different way. Just one follow-up

question is that, you know, the reason I was asking is Carvana and Cardlittics are both run by the founder, the CEO, right? So I was particularly interested in, is, are you really attracted to founder-led organizations. Oh. Um, I guess I have, that's an interesting question. Not intentional. Um, I guess I have been. Um, if you just look at what I've owned, there's been, I guess, those two,

trying to think of anything else I've owned. I've been founder. The CEO wasn't, he wasn't the founder per se, I don't think, but I think he had really grown the organization. You could almost look at him as a many. Where? Herbalife, am I remembering that correctly? Yeah, you know, Michael Johnson, you know, what I would say is maybe it's not so much about a founder.

And maybe this isn't necessarily an absolute requirement. But it's good that there's somebody in charge. And, you know, I don't necessarily think that it would be a negative, although it wouldn't necessarily prevent me from investing because it's sort of always, nothing's perfect, to have what you kind of think of as kind of a passing through rental executive who, you know, isn't really able to make decisions, you know, I don't want to say unilaterally because that doesn't happen. I mean, people, the executives think they're making all these decisions, but everyone around them is bringing them all the information and tipping them

in the way they want. But, but, you know, the idea that there's someone who's actually sort of relatively ensconced in control and can kind of make sensible choices and, you know, I think is important. And maybe you get that more with founders. But at the same time, you know, I've had, I've been invested in companies where that's not quite the case. And, you know, it's been fine. So, in general, I think the simplest way to think about it is when you ask the person about choices that they've made in running the business, do they articulate things reasoning that you think, whether it was right or wrong, is sound reasoning? And if they're making decisions based on factors you think are important based on sound

reasoning, you're probably going to do okay in terms of their ability to run the company. Now, sometimes they'll blow your mind, right? They say things that you just like, wow, I think of myself as fairly bright and, you know, you've thought a lot about this. And you know, you've just put a new lens on this that's totally different for me. And if they do that repeatedly, then, you know, you might have some real talent on your hands. And that probably is something that you should like. And conversely, if you find yourself asking these questions and you find that they haven't reasoned through things the way you would reason through them and that they've sort of missed a lot or they haven't thought about it right or they've waited the wrong factors and they sort of find that, you know, to be a persistent problem. then, you know, they probably are less talented.

And you need to think about how that might impact the performance of the company. The other thing is that, like, different companies have different requirements for management. I mean, management always matters. But, you know, a monkey can run, you know, a stabilized apartment building. Yeah. Right. Whereas building a new company a la, you know, Carvana, is very difficult and requires a certain level of business, you know, acumen, you know, that

if it's not there, you won't succeed. Retailers, you know, obviously notoriously sensitive to the person running it. You know, on the other hand, you know, I think Google's had a lot go right over time and the management has probably been net additive, but certainly, you know, Google's search engine, you know, I mean, they could all go home, right? Yeah. So, anyway. If it was 2008, anyone could have run Google search in it, right? Like the moat was that big, all the other things. But like with a, as you said, with the Carvana, it took someone with vision to say,

hey, this isn't going to work if we just do one market. Like the economics don't make any sense. I need to vision what happens if we do 200 markets and build this whole thing. No, it makes tons of sense. Makes sense of sense. Anyway, let's turn into one of your portfolio names. Let's talk about CardLittics. This is obviously a high conviction name for you.

You know, I think it's approaching 40% of your portfolio. You own about 16% of the company. You know, it's just looking at the filings. I think you bought $25 million more of the stock in early August that prices kind of around today's level. So this is obviously still a very high conviction name for you. So what did you just dive in? You know, what is Cardlittics?

How do you find it? Why are you so bullish on it? Sure. So I'll step back. Just correct a little thing. It's not quite that bigger percentage of the portfolio. Other things have gone up some, which has. a denominator effect on that but uh but um you know what cardlytics is you know people aren't

necessarily familiar with it and and so um what it is is if you open a banking app and you're fortunate enough to bank with your bank of america wells fargo chase truist um and most of the top banks um if you open your banking app you'll find um in your primary spending account so if you use your debit card a lot, it'll be in your debit card account. If you use your credit card a lot, I'll be in a credit card account. You'll find offers to save money at various retailers and restaurants and, you know, travel companies and so forth that are serving the categories in which you spend money. And Cardlytics is the, basically the company that is powering that on behalf.

half of the banks. And what makes cardlytics interesting is that it, so it brings a great deal of value to all the members of the ecosystem. So advertisers love it because they can advertise in a brand safe space in a very targeted and measurable way. So Starbucks, can identify people who drink coffee or maybe who go to cafes and other sort of Starbucks competitors that are near Starbucks but don't frequent Starbucks enough and they can present those individuals with offers which you know the consumer experience is that you tap the offer and then it's instantly connected to your payment card and then when you go and you buy the product, you instantly and sort of seamlessly get the discount. So there's no further work required

by the consumer after tapping the offer. And so, can I just pause right there for, so I go into my app. I open up and it's, it's that thing where it says, hey, click here for 10% back at Starbucks on any purchase over the next 45 days. Cardlytics has powered that. And then Starbucks, is Starbucks trying to target specifically people like, do they offer that to everyone? or are they only offering it to, hey, this person went to a Dunkin' Donuts last week and that Dunkin' Donuts was around the corner from Starbucks. Let's push to them to try to yank their spending from Dunkin' Donuts to Starbucks. Yeah, correctly. It's highly targeted. So the targeting can be done based on anything that I could know about you from, that your bank knows about you.

Yep. So, you know, think about everything I could learn about you from your credit card statement, right? Your debit card statement or both, you know, obviously there's name rank and serial number and whatnot, but then there's also, you know, where you spend money, when you spend money, et cetera. And so Starbucks could be much more precise than that. It could be, look, we want to make, this person based on historical behavior tends to be very open to offers, tends to switch easily. And so we want people who tend to switch easily, who we have between 20 and 30% of their wallet share where they spend at least $100 a month in the category. And where they spend, they spend at least, you know, a third of their spending is within two miles of a Starbucks.

You know, you can get that targeted. And then we're going to present those people with 5% off. And everyone who meets those criteria, but they're less susceptible to switching, we're going to present them with 15% off. And, you know, everybody who's in the, you know, 70% to 80% range, we're only going to target those who are easy to switch and they only get 5% off. And it's going to be different if you're in the Starbucks loyalty program versus not in the I mean, you can just keep going. The sort of ability to go in depth and targeting is vast. And then what happens is you can imagine there's a hundred and, there's going to be nearly

160 million MAUs shortly with the onboarding of U.S. Bank Corp. And so they might say, okay, well, based on how you'd find it, there's, you know, 10 million people who meet your definition. And what we're going to do, if you have unlimited, if you're willing to spend as much as makes sense, what we're going to do is we're going to target, you know, nine million of those people. And we're going to have a million that we hold out as a control group. And I'm just making up these precise numbers, which get the idea.

And then what we're going to do is after we've presented the offer, we're going to look at the difference in spending at that establishment between the group that was given the offer and the control group. And we're going to measure to the penny the incremental sales that are generated by the campaign. And so the second really interesting thing about cardlytics is that these offers are marvelous for advertisers. The returns that they're able to generate for advertisers are incredibly measurable and typically between four and six times what you spend. So if you spend a dollar, you can typically expect to get between four and six dollars of incremental sales.

And this isn't like some sort of bogus. I've got some attribution marketing mix model, you know, where like some PhD tells me it's right, but it's really not. Right. This is not that. This is actually, you know, properly randomized controlled trial. There's no question about incremental, incremental sales lift. And you can see it, right? You can see Andrew's credit card. Last month it was Dunkin' Donuts, this month it was Starbucks, right? Like there's no better proof than that. Correct. That's right. That's how it's measured. You can actually see it going, you know, in the card statement. So we've mentioned why it's great for advertisers. Why? Why Why is a bank choosing to do card analytics?

Because we mentioned they work with seven of the top ten banks in the U.S., right? J.P. Morgan, Wells Fargo, and Bank of America have 20% of the credit cards in the United States. Why can't Bank of America just say, hey, I don't need to include cardlytics. I've got enough data. I can go do this all by myself. Yeah. This is an area that I spent a lot of time on to try to get comfortable with. And what I learned is that reach matters a great deal in advertising.

And you can think about the importance of reach ultimately as arising from the fact that the advertiser incurs some amount of fixed cost to manage a channel. And so as a consequence, advertisers are not particularly interested in small channels. And of course, on the flip side, the network manager, the Cardlittics, has a certain amount of fixed cost to build and run the system as well as to manage the ad reprimiser. And one of the things that, you know, the more sort of tension you put into Cardlytics, the more you can get out of it. You can think about all those tests of different groups that you might do over time.

Cardlytic, you know, also on kind of their side in terms of fixed cost, is building a self-service platform, which is really important to allow ad agencies to resell this to their customers as well as to allow the customers to do all of their own, you know, targeting and therefore to get more, you know, granular. And so all of these things, what they do is they cause it to be the case that advert, it's much more efficient to distribute this product if you have a lot of reach. And conversely, and so as a consequence, and of course, also fixed cost to running the system. And so those two things mean that by pooling their efforts, the banks are effectively able to attract many more advertisers who spend much more in advertising in the channel

much more efficiently and to share the fixed costs. And what that means is that all the advertisers, they want to be where the consumers are. That's where the bank. are. And so they want to be with the guy who has all the banks. All the banks, they want to be where all the advertisers are, because I should have mentioned that this is very good for the banks. The banks get, they get a revenue share, which is about half of the money that the advertiser spends after getting off the amount the consumer gets. They also get, it turns out consumers really like it. And so they get a benefit in terms of spend and they get a benefit in terms of retention. And those two effects are actually quite large. And so it's, it's a benefit.

important to the bank to have, you know, the most offers. And the way they get the most offers is by bundling their offers together. So the banks want to be where the advertisers are. The advertisers want to be where the banks are. And then there's a non-trivial amount of technology that has to get built over time. They're sort of card analytics being partway through that. And that takes a fairly long time to build and makes it even harder for someone to think about moving or switching. And then other things have to happen too in the fullness of time you know ad agencies are going to learn to use the cardlytics you know direct channel they're going to get involved in selling that they're not going to be particularly excited to learn to use another platform uh which is much smaller um you know

that that sort of stuff so when you start looking at it you come to realize that um you know there's definitely um this has a lot of the makings of a natural monopoly and i guess uh as you're kind of internet there's a lot of data learning there too, right? Like CardLix three years ago, they've got so much learning. So if JP Morgan ever wanted to pull off, you know, A, they'd have to recreate all the technology, but B, you know, their offers would be much more primitive compared to CardLytics with three years of data learning built out and everything. Yeah, there'd be that. But then also, you know, say you're sort of a marginal advertiser at Cardlittics, right? And, you know, you can spend, you know, not a lot,

a hundred grand a year in the channel. You're basically, you know, let's say 200 grand year. you're basically just big enough to be relevant right well chase big as they are you know they're not that big so suddenly you know you you wouldn't be able uh chase really wouldn't be able to serve that advertiser um and as a consequence uh you wouldn't use them chase wouldn't you know use uh wouldn't wouldn't have your offers which would mean that you know while that advertiser in itself wouldn't matter when you added all those advertisers up there'd be less offers flowing through chase's system and as a consequence chase would have fewer offers for its customers.

And so they would see, you know, higher, so that would give Chase a competitive disadvantage to, you know, to Bank of America. Chase would see sort of higher attrition. They'd see lower card spent. They'd also get a smaller red share. And for all of that that they lose, they would also have to incur the fixed cost of the sales relationship to,

you know, to maintain, to actually sell the advertisement to that person. And then, of course, that's before you consider the time and cost to build to maintain the technology. And, you know, that's sort of where we are now. In the fullness of time, I think that, you know, what you can imagine doing with Cardlittics versus what's currently done with Cardlittics is fairly substantial.

So, you know, one challenge, for example, that Cardlittics has is that big advertisers or potential advertisers like Target, their discounting budgets are actually at the category level. so and this makes sense because target has a variety of departments within it you know and categories and those categories have different margins so target does not want to offer 5% off at you know or 10% off at target because if you use that 4% margins right so you'd be selling groceries out of loss if you did that but toys that's right or home decor right and so what target wants to do

is offer 15% off home decor and a dollar off Target private label tomato soup. Or better yet, a dollar off Campbell's soup funded by Campbell's. Yep. At Target. Now, to deliver on that, ultimately, and they're not there yet, right? But to deliver in that, ultimately, if things work, because I hope they will, Cardlidics is going to have to end up doing integrations into Target on their back end. And there's going to have to be a lot of work kind of done in the target end to learn how to use this and integrate it and all the rest.

And those integrations, you know, in addition to the self-service platform, which is a big technology build out and the ad agencies learning to use it, will be sort of yet another piece that, like, makes it such that JP Morgan would never be able to switch. And along those lines, you know, the, along those lines, the, the cost of, I shouldn't say never be able to switch. I mean, if they needed to, right? But along those lines, the value that Cardletics can create by having the scale to be able to do that, like to get to be big enough that Target would care to do that creates a huge amount of value for the banks, for consumers, for Target. And so that kind of is in a nutshell, kind of in one very simple example, why everyone wants to pool and use one provider. No, that's fascinating. And like, you know, I hadn't even considered this, but it's something like Starbucks, right? Starbucks can say, hey, 5% off if you order from

Starbucks and 25% off if you throw an extra shot of espresso into your coffee, right? Because the espresso's huge margin. Or better yet, time of day. Yeah. So target people at 2 p.m. Correct. Or you could imagine Starbucks might say, look, in some areas, our stores are at max capacity. So we don't want to do lattes, but we'll offer a dollar off ice coffee. right um so there are you know as you start thinking about getting deeper with the advertisers um and there's a lot that has to happen for this all happen right you have to the the offers right now are not sufficiently um uh big or they don't have pictures i mean right now it's just the Starbucks logos is like 5% yep yep but um with the improving UI they're rolling out and improved as you

know they're rolling out a new improved uh UIUX uh with right now with um US Bank Corps and I suspect over time the other banks will adopt that and that will continue to improve and that will allow them to have, you know, pictures and text and the ads units are going to be a lot richer. And then once the ad units are a lot richer, now you can communicate within the ad unit. So now you can have, instead of just 5% off Starbucks, it can be a dollar off iced coffee at Starbucks between 8 and 11. Yep. And that would allow Starbucks then in turn to get way deeper in terms of what they spend, right? So now they can find people who maybe we have 100% of their wallet chair in the afternoon, but they don't buy any coffee from us in the morning,

and they're near stores that have surplus capacity in the morning. So we're going to offer them, you know, some offer in the morning. Or maybe they're near stores that don't have surplus capacity in the morning, in which case we're going to offer them only a discount on iced coffee, you know, if it's ordered through the app, because that's something that we can do really quickly and, like, it doesn't impair our ability to make the line move, right? So you can really go as deep as you want. One thing that strikes me about this, And I know CardLix talks a little bit about they want to expand more to being able to offer this to small and medium businesses, right? But when we talk about getting granular enough where you're doing enough data, where Starbucks can push you to underpopulated stores at certain times of the day to sell higher margin products, like that takes a data analytics team on the Starbucks end.

So is it going to be, is this really something that only the big retailers can use? Or do you think CardLytics eventually builds it on their end where they say, hey, outsources to us, we'll run all the data on our end and we'll direct the high. high-margin products to you. How do you think that plays out? Yeah. So that's where you're talking about the self-service platform. Yeah. Right. So the self-service platform ultimately gives the tools to advertisers to do this on their own within the rules set by card analytics and approved by the banks. And so you can imagine the self-service platform making it possible for someone like a big advertiser, like Starbucks or something, you can go as deep as you want, right? You can have at, you know, Linux teams you can do all that work. Um, you know, another layer would be

something like, you know, uh, ad agencies and there, you know, maybe they're going to get, uh, less handholding directly from Cardlix that might be limited by the functionality of the self-service tool, but there'll still be lots they can do, you know, you can imagine discounts in time of day. There's a lot you can still do even just within, you know, you can imagine a fully built self-service tool. The self-service tool, by the way, is like in beta right now. So all of this stuff is future state, right? Um, the, the, and then when you talk about sort of smaller businesses, the good news is big advertisers instead of medium-sized advertisers

are the vast majority of advertising. To go eventually after the smaller advertisers, there's two ways you can conceptually do that. A few ways you can conceptually do that. You're going to need a distributor. You're not going to be marketing them directly. Ad agencies can do it. You can also imagine building it into

the SaaS software that people sell for merchant acquisition or just merchant acquirers. The banks also can be your partners here and sell into small businesses. All of these are ways to go after it. And I suspect to be really successful with small business, ultimately where this is all, like this is way down the road.

And all these things, by the way, as you think about why it would be difficult for a, why the banks want to pool their resources and why it would be difficult for one to do it on, all these things can't happen if a bank is doing it by itself, right? But as you push your way down, you know, further down the road, you know, you could imagine someone productizing this in a nice way. You know, we sell software that is used by car washes, you know, and our car wash software as a variety of functionality for car wash, include car washes,

including, you know, we've built some functionality to help you target people in your community who go to car washes that, like, aren't yours. And as a part of that, you know, you're basically buying ads on Cardlittics through us. And you're getting a piece of the revenue, and Cardlix is, you know, getting the rest. And the bank is getting, you know, everyone's getting a piece.

And so that's how you can imagine going really down market. There's going to be a certain amount of that. But all of those pieces, I mean, that really kind of gets to the core. And all of those pieces require, my joke about Cardlittics is you can do almost anything. I mean, movie theaters could basically give away highly discounted movie tickets to people who don't usually go to their theater for the movies that aren't going to sell out, right? Especially if they notice that that person, when they go to the theater, makes two purchases, the first being the ticket, the second being popcorn, right? Now you've got someone who tends to buy popcorn who doesn't come to your theater and you're going to give them a free ticket to come to your. theater. Can that be done conceptually? Yes. Is it either there? No. I mean, they're not there.

But I think eventually this is a channel where they're going to build all of that. And that's how it becomes a really big business. Yeah. No, when I came into this, obviously, I had done some work on and stuff. But one of the things I was most skeptical on was why does it a bank build this themselves? Obviously, some of the proof is in the pudding. They added Wells and J.P. Morgan over the past year or two. So some of it's in the pudding. But as you say that, that's stuff like no bank's ever going to be able to match that. Let me just turn back. I want to make sure me and everyone are crystal clear on the economics. So Starbucks gives me a 10% offer, right? Hey, spend a dollar, we'll give you 10% cash back on it. Spend $100, 10% cash back, whatever.

What are they paying specifically to the bank and what are, I guess what do they pay to Cardlittics in the bank? And I think Cardlittics in the bank split it pretty evenly. But where does Cardlix in the bank get their take? Yeah. So the way to think about it is on average, if an advertiser spends a dollar, about 30 cents goes to the consumer. That's the offer. You know, that might be 10% off up to whatever or, you know, 5% off or 15% off or whatever. The remaining 70 cents is roughly split between Cardlittics and the banks. And so that gets you kind of, you know, sort of 34 cents for Cardlittics, 36 cents for the banks, and 30 cents for the consumer.

Now, it's worth pointing out that that dollar is what Cardlittics calls billings. Yep. Um, revenue is the 70 cents after you net off the, the consumer, uh, portion. And then price off it would be the 34 cents after you give the bank. Yeah, minus a teeny bit of cost related to computing and stuff like that. So if I get a 5% offer or 10, let's make them at the fee, a 10% offer from Starbucks. So 10 cents in this case would go to the consumer, the 10% off. And then actually 20 cents on the back end is going to the bank and Cardlittics.

So I apologize. My phone is connecting randomly to my Bluetooth and I lost you half of that, half that question. No, no, no. So I get a 10% offer, right? So Starbucks is technically, I guess, paying me 10 cents of a spend. There's another 20 cents of spend that's going 10 cents maybe to the bank and 10 cents to Cardlix. But that's kind of the way to think about it. That's roughly right.

And now it's fair to note that this very is a great deal by campaign. Yep. And so you might imagine there. is to have richer offers, so maybe the advertiser is spending a dollar in 50 or 60 cents is going to the consumer, and the rest is being split between credit analytics and the banks, but that's maximizing value for everybody. And conversely, you could imagine situations where the offers are thinner, but for whatever reason, very effective.

And so maybe a smaller offer that, you know, maybe the consumer is only getting 10 cents, but there's an averaging effect that's going to happen across the portfolio, and so you can see what that's been over time. I guess, and there's been no cases of banks or anything leaving them, right? City Group didn't leave. They got, they fired City. The company was, they had city. Cities implementation was very bad.

It was basically hidden deep in the apps. And the company was, Cardlittics was really struggling at the time because it's been very hard. as you can tell, it's been very hard to get this business to the point where it's been sort of hitting takeoff speed. It's been a real slog. And so Cardlytics actually got rid of Citigroup. That makes it because there's a line. No FI partner has ever unilaterally terminated its use of our solutions. And I kept looking at it and I was like, that is some interesting language. I bet you that there is a partner who has left at some point. That's a line from their 10K for, yeah.

A frequent question. I got tons of questions here, but I'm aware of time. I'm going to have to let you go in a second. But I think the most frequent question I got, because again, this is a little bit of a battleground stock, right? And I think this has been really elusive to the bookcase. But a lot of the bears have said, if this solution is so good, if the return on sales for the customers are so good, you know, in COVID, especially in April, I think tons of people shut down all of their spend here. And partly I think that's attributable to, hey, you know, a lot of this. is a retail and travel-focused thing. A lot of those were absolutely closed during April, but I think it was even bigger than that. So why were so many campaigns shut down or paused during April? Well, a lot of reasons. I mean, first, you're right. Retail, dining, and travel are some of the biggest categories they have, which is bad in terms of how they were impacted by COVID. You know, the other thing, which is subtle is, you know, They're, you know, they're giving, these are offers, you're giving people money. And so you can imagine if you're running advertising campaigns, these are things that you might

in a very different world want to retest because your assumptions about incrementality can evolve and should evolve when the world gets very different. And so there's a fairly, you know, you can, you know, if you're going to run a campaign and then measure it, given that the campaigns are 45 days and, you know, they don't have a self-service platform, and they don't run short campaigns yet. All these things are going to evolve. But because the product is still sort of very, it's insertion order driven, it's sort of 45-day fixed campaigns, you know, you can see how that would tend to slow people's ability to iterate as the world is moving around them. That will get better, but that's where they are. Also, in terms of onboarding new customers, again, there's a fairly lengthy process. You're going to kind of, you know, there's going to be some sort of selling process where you get to know the product and make a decision. There's going to be a pilot that's going to run for 45 days.

There's going to be some process by which afterwards it's analyzed, and then Nielsen's going to produce you a report that says how much, it confirms Cardlittics's measures of incrementality because people don't necessarily trust Cardlittics to do their own homework. And then now it's been, you know, a few months. and now you're going to like decide to maybe lean into this more in order to drive more business. And so, you know, there's just a fair bit of, it just takes time for things to ramp in this channel in a way that it doesn't necessarily in channels that people are much more familiar with and channels where people aren't giving away, you know, discounts to consumers in channels where people, where they can test and learn faster. because, I mean, they'll be able to do that with Cardlytics, but they can't yet. So those are all things.

Another thing, phenomenon, which is complicated, is, you know, the pandemic created this dichotomy where some places were sold out and other places were, you know, empty. And so if you're sold out, the last thing you're going to do is discount. Yep. And conversely, if you're empty, you can discount all you want. It doesn't really matter. And so I think that there probably was some effects of, like, you know, people who were using them who didn't use them because they, you know, they fell into one of those two camps. So in some sense, the good news for the subset of customers who had good news wasn't really, you know, good news for them.

So all those, I think effects, you know, probably explained. I, for one, wasn't, you know, particularly surprised. at what revenue was in the second quarter. I didn't have particularly high expectations. You know, we'll see what happens going forward. But I think that, you know, I didn't expect, you know, all the new customers, they were bringing all the new logos. It's just the pilots, you know, are great and they're going to lead to a lot of revenue,

but it's just not, doesn't happen yet. Last one on Cardlake. So, look, today I think, you know, it's obviously done pretty well over the past couple years they've added two of the biggest banks in the in america they've uh you know there's seven of the top ten banks now when i look out like where is the growth coming from is this just average revenue per i think it's monthly active user or monthly active login or whatever is this just going five times higher as they get more and more targeted offers or is this an international story how do you how do you kind of think about that well let's just talk domestically for a second

um so in terms of growing the the obvious paths to growth at first they can they can add more MAUs. They have about $150 million. There's $160 million. There's probably another $100 to $150 million more that they could conceptually get. And I think in the fullness of time, they'll get some more. You think basically every bank is on this platform at the end of the day or the decade or whatever you want to call it?

Yeah, I think so. Yeah, in the fullness of time, I don't see how a bank will be competitive without having, you know, the ability to basically get consumers free money and also to generate extra. revenue, which sort of subsidizes their overall economics and allows them to be more competitive elsewhere without it. So there's sort of two sides to it. One is that the channel can get bigger, and the other is that the monetization within the channel can get a lot better. So right now we're talking about the channel getting bigger. So the first thing that can happen is more people. The second thing that can happen is there's a meaningful,

difference. There's meaningful growth in like online engagement and mobile engagement in banking. And you can look at the various stats that like a JP Morgan or Bank of America put out or whatever. But use of the mobile app is growing, you know, kind of double digits or

whatever. And there's a, you know, as people go from, you know, banking once a month on their desktop to, you know, downloading the app and opening the app on average quite frequently, like 11 times a month or whatever. You get this meaningful growth. in the portion of banking customers who are engaging with the app, with the offers.

So another way to think about it is that, you know, my father-in-law, you know, would never see these offers, but my sister, you know, would see them all the time. And as, you know, my father-in-law, people like my father-in-law become a smaller portion of the population and people like my sister become a bigger portion of the population and people gradually adopt the behaviors of my sister, that gets you a second leg of growth. And I think in the fullness of time, that probably can double the size of the channel. The next thing that can happen is the implementation can be improved. This is, again, on the size of the channel front.

We've talked a bit about, you know, making the ad units more engaging. There's just a very long list of things you can do between alerts and notifications and more engaging offers and where they're presented in the app. And all of that has to be done in concert with the banks. And all the banks are kind of at a different place, kind of in this journey. But there clearly is kind of an up into the right trend in terms of getting a bigger, a better implementation at each bank, which in turn gets more of the bank's customers to notice and use these offers. Bank of America is I think their oldest bank, right?

Or one of their oldest banks, obviously their oldest major bank. Is the engagement and revenue per user noticeably higher at Bank of America versus kind of the, I don't want to say J.P. Morgan or Wells because they're still kind of on the test, but there are other cohorts of banks. Is it much higher at Bank of America? So they don't share exactly what the engagement metrics are at different banks. That said, they did talk a lot. If you go back and read the 2018 transcripts about the heterogeneity in terms of revenue per monthly active user, different banks. And so there's fairly meaningful heterogeneity across banks. I don't know exactly. My sense of the Bank of America is better than many, but I don't actually know precisely. Okay. That's right. But certainly they're all kind of on this journey of kind of getting better over time. Along with that, I'll note that there is a, you know, even a Bank of America, they are seeing growth in of the Bank of America consumers, the portion that engage with the product.

there's a certain just sort of discovery phenomenon that occurs over time. So all of that is on sort of the channel getting bigger. And if you start multiplying all those things together, you get sort of, you know, several times bigger. You know, be it sort of three or five or whatever the numbers, but sort of a fair bit bigger on that side. That is all nice. The really big driver is more revenue per monthly active user. And the way I might phrase this to you is that, you know, back prior to bringing on Chase in 2018 or at the beginning of 2019, so in 2018,

the company had, it was about $2.30 or 50 cents, if I recall, of revenue per monthly active user. And the way to think about that is that, you know, Bank of America customers, who for the most part made up the ecosystem at that point, they opened their app and they were shown a certain portfolio, of offers, you know, 30 offers typically, and they had a propensity to activate them and a propensity to redeem them, and that led to a certain amount of revenue per each of those. And the question I would ask is, you know, if they opened that app, and instead of seeing 20 offers, they had, you know, 500. And if instead of the offers being kind of randomly

assembled, you know, down a sheet, they were pretty and exciting and interesting, engaging, and they were organized by, you know, by category. And if instead of them being, showing the same offers basically every time because they're 45-day campaigns, you see the same offers for 45 days, but they were the most interesting, you know, 200 offers to that person and that the person didn't activate it. They were then shown a different offer the next time they opened the bank, which typically is like the next day. The question would be how, what portion of those offers would the person engage with? and, you know, what would happen to revenue per monthly active user. So that's, and my, the bet is that, you know, when presented with vastly more offers,

people would activate sort of proportionately more and therefore redeem sort of proportionately more. A very simplified way to do it is someone might have used it at Starbucks, but they didn't use it at Lowe's because it wasn't a Lowe's offer available. Yeah. Now there will be a Lowe's available to do it Lose. The question you might follow on with is, well, are there, Can advertisers, are there kind of enough advertisers to fill out all those offers? And here, you know, I think the evidence is pretty clear that there is.

The, you know, the company has talked a fair bit, and you can read it about, you know, just how underpenetrated they are with advertisers. You know, they shared the staff that, you know, a large department store whose name started with an M in reported earnings, you know, the morning they said this. You know, how this advertiser, I think the number was they could spend 120, but it might have been 150 million in the channel if they wanted, if they max it out. And that worked out to be something on the order of 40 to 50 basis points of, you know, that advertiser's revenue driving something on the order of 2% in parental sales for that advertiser. And so the point is that if every, you know, the mental exercise I do is if every advertiser, you know, spent 50ips of revenue. you know then obviously there's five trillion of you know retail sales like

all over it's a very big opportunity uh and obviously that 50 dips was at a time you know with only 100 million i may use and the implementation wasn't what it could be and like there's just a lot and you know there's just a lot that that can be bigger so you start multiplying those things you get to very big um numbers in terms of the uh long-term opportunity um i think it's great point like i'm just looking you know i i opened up my i have changed I opened it up to see it. And one thing I've always been struck by, and as you said it, it made tons of sense. It's like, the top offer I get is Jimmy Johns.

I don't live anywhere by a Jimmy Johns. But if you offered me, you know, if you offered me the sandwich placed on the corner, I would probably activate that and be pushed to it. You know, the second offer on the border? Don't live anywhere by that. You know what the 10th offer is? Rite Aid. And there's a Rite Aid, a block further than the Walgreens that I always go from around the corner.

If they pushed that to the top, like I had to actively click through to see that. If that was the top offer, Rite Aid could probably steal some incremental business from me, Or, you know, so, no, I think it makes tons of sense in giving smarter and more I'd be even more sophisticated about it. So if you think about the ecosystem of the offers, there's the top four offers, which kind of draw you into the rest. And then within the rest, there's kind of the ordering of the offers and the presentation of them.

And so they've done no optimization here. Right now, they're essentially presented randomly. But you could imagine, like, there might be offers that you're willing to almost do at a loss that it would be on the first page that would drive people to see the rest of the offers. Yep. And the whole ecosystem could be better, right? And so the idea of sorting offers you know, in a much better way, the idea of charging advertisers more for better placement, for more reach. These are all, you know, the idea of doing a better, like a way to grow engagement, right, is to find the people who aren't responding to your offers and like give them a really

good offer to get them to discover it, right? All this sort of stuff to sort of optimize this hasn't been done. And yet, I would note, you know, the company on a trailing 12-month basis in the Bank of America, basically Bank of America, in Calendar 2018, generated, it was, let's call it 250 per MAU. And at the time, they had 40, or 50 million MAUs. Now they have 160 million MAUs. And so if you just, you know, if you think they can do a little over three bucks, which is trivially more, given digital growth and everything else we're talking about here, um, you know, then, then you're talking about a business that would do, you know, $500, $600, $600 million of revenue, a couple hundred millions,

so $250 million of gross profit and would make, you know, $100 million a year of net income. Now, that's not, you know, that would be sort of just barely enough to cover the $2 billion market cap. But that all that assumes, basically, is that they're able to get to where they were a few years ago. And a few years ago, they were radically underpenter. I mean, they didn't even have a luxury, you know, they didn't even have luxury or travel as categories. Like, like, you know, I mean, there's that the self-service platform, which access is something like 70% of all advertisers, is just barely in beta and now.

And it's going to be in the hands of a couple of out agencies right now. Like the ability to go after the, you know, they were able to achieve that $2 plus event per MAU, you know, in a time when use of apps was way lower in a time when, you know, the offer actual presentation. was a lot worse and it's going to be without accessing the vast majority of the advertising market through ad agencies or even act or even going after you know some of the biggest categories in advertising travel and luxury and and whatnot and so you know the idea that they're not going to be able to surpass that by fair bit strikes me as fairly unlikely and then the question really becomes and what I don't know the answer to but the question becomes how big does it get in the fullness of time how you know can it can it be in me Facebook

does, what is it, $140 for MAU, you know, one way to think about it is, is it wildly unreasonable to think that Cardlytics could basically, I mean, if we think about an MAU as kind of analogous to a person, which isn't quite right, it's much more complicated than that. But, you know, if Cardlittics basically had $100 of billings per MAU, they'd be basically bribing the average person 30 bucks to move something on the order of $450 of spending from one place to another. Home Depot versus Lowe's, Outback versus Chili's, Hilton versus Hyatt, et cetera, just $450 out of their wallet of $50,000, $40,000, right? I guess we're talking about average people and look at GDP, it's a little lower, it's more like, it's a little lower, it's more like,

you know, 40,000. So can cardlytics shift, you know, even a little bit of that spend? kind of population wide. And if they could do that, then it would be $100 of billions per MAU, $70 of revenue per MAU, and the business, you know, would do $20 billion of revenue. And so that'd be pretty cool. And then there's international, which would also be excited. Two more questions. One on this, we're running long, but I'm just super fascinated by this.

Doesn't Cardlittics take go up over time? So right now, the rough model we said was a third to consumer, a third to Cardlytics, a third to the bank. I mean, as they scale, couldn't you? see over time Cardlittics takes the lion's share of it? I think

it's very important that Cardlittics maintain great relationships with its banks. Okay. Now, over time, I think there may be situations where

there are costs that might allow the ecosystem to be made better, that Cardlittics might share with the banks if we're thinking

way out there. so you know you think about um let's say they're going to go after a bunch of small businesses and they know that's going to make the ecosystem better but they know there's going to be higher distribution costs associated with that that might be a situation where you have a conversation with your bank partners about what's best for everybody um but i wouldn't expect them to necessarily come out and just start trying to just take economics from the banks Because in the end, I think it's the banks are their partners, the advertisers are their customers, and, you know, obviously consumers are the banks, you know, customers. I'd also add another point, which is similar, though, to pricing, which is that right now, the returns they offer are extremely good.

Five to one is extremely good. There is certainly no doubt that over time pricing optimization, as that when this channel were fully grown, they could sell the same offer at a, you know, three and a half to one return. And that would just mean that they would get a bigger take, right? So, you know, if they go from charging a dollar for that advertising to $1.50, the consumer still gets $0.30. And so a dollar, you know, $20 is then split between them and the bank. So their share goes from, you know, $0.35 to $0.60, which is a fairly big increase.

So I think that's another big area. that there can be, you know, games. It makes total sense. It just strikes me that, you know, over time, if the argument is, hey, J.P. Morgan, like, we bring, you could try to do this on your own. It's going to bear a lot of expense on your end. And by the way, your offer is going to be way worse,

and it's going to bring in way worse revenue. Like, that is the type of model where we bring in something so much better. Over time, you have significant pricing power. And I get the banks of their partners, but it does seem like over time the banks take rate should go down because cardlidics brings so much value there. Cool. on cardlytics?

Well, one thing I might note, and one of the things that drew me to it as an investment, is, you know, I do, despite being invested in car around on cartilics, and I do think of myself as a value investor. And one thing I think people struggle with is this idea of how do you underwrite your downside risk in these situations? And in the case of cartilics, I actually think that it's helpful to try to invest. the situation and you say okay well what is the likelihood so right now we have the situation there advertisers are getting a great deal consumers are getting a great deal um banks are

getting a great deal everybody is very much benefited what is the likelihood that this company in some enduring way goes materially backwards such that they aren't able to kind of even come close you know to where they were before and and and then the banks are uninstalling it customers aren't using it, et cetera. And, you know, you get really creative and talk to this of radical changes in payments. You know, we all use Bitcoin. I mean, there's things you consider it would conjure up. But for the most part, it's very difficult to, I mean, the banks, you know, tend to leave

even products that their consumers don't use. If a very small, someone of consumers use a product, that doesn't cost anything to keep it, they tend to sort of leave it. And this is a product that a fairly wide swath of customers use, and it's fairly good for the banks. And so the likelihood they're going to start ripping this out, just seems very low. At the same time, advertisers who are getting these great returns, these incremental sales, especially as it becomes bigger and more important,

the idea that they're going to start, never mind, like, we don't want the money. I mean, that's wild, it's not going to happen. And, you know, consumers, I mean, they're going to keep taking the discount. So I just don't see how this goes really backwards. And in that sense, I think about, you know, it's a very, in that instance, sort of safe. It's very safe, actually, I think. You know, it might not grow up to be the $100 billion market cap company that I hope it will be, you know,

but I think the likelihood that it grows up to be, you know, not a $2 billion market company strikes me. It's just very low. I'll push back. I think the way it would break is if you're wrong and if there's no network effect, right? If the banks can easily do this themselves, then obviously this trades at, as it should, it trades at a big multiple of kind of invested capital, right? nobody's ever going to value a SaaS for it. But if the, if the banks look at this and say, we could easily replicate this, then they're going to have huge pricing power and that sort of

and I think you've explained all the reasons why the banks would not. But I would push back, it's not that this is probably. And the banks all looked at, the banks all looked at doing it on their own. I mean, Jason Morgan. But, but, but, you know, and it is conceptually possible to build their own sort of duplication of where card critics was a year or two ago. It just costs a lot, takes long time, and wouldn't be efficient. rest. But the bigger thing that the banks realized was just this, the, you know, you've a distributor who's selling, you know, six times as much reach. And that distributor, not only do you leverage that person's cost, but then also on the buying side, it's just really efficient

to buy six times as much at a time. And, and then, so that's kind of the simple math of it. But, yeah, look, you're right. If that's wrong, but, you know, on the other hand, though, my joke has been said, you know, banks tend to be fairly sticky and their choices, both directions. And look, one of the challenges of Cardlittics is that the banks are very, very conservative. And they can tend to sort of, like, the things that Cardlittics could do if they were Facebook is very different than what they can do. And, you know, and so that can be frustrating, but it's also the big advantage.

I mean, you have banks that are, they sort of send a pick a direction and they sort of tend to follow it. And so I'm not particularly, you know, you're right, though. That's probably a good risk to think about. And the other thing is, again, the proofs in the pudding. J.P. Morgan signed up over the past two years. Wolfs Fargo signed up. And I think Bank of America signed an extension recently.

So the proof is kind of in the pudding. And U.S. Bankrupt. Yep. Yep. Cool. Last question I ask everyone. This has been absolutely fantastic. So thank you again for coming on. Last question is anyone else you'd be interested in hearing a podcast with?

Oh. Have you spoken with Rob Van Alley? I have not. Oh, he's at RV Capital. Okay. And he's just the, I feel bad answering that kind of, you know, in public because invariably the many people that are probably great that I'm not giving their name will be bad. But Rob, he came to mind just because he, you know, has a wonderful way about him. I have to joke.

he's the kindest investor, I think, out there. He's really quite funny and happy all the time and brings a really great perspective to investing. He runs a large and successful one-man shop that, you know, I think you'd enjoy, even if he wasn't great on the podcast, I think you'd enjoy speaking with him. I also think he has some great perspectives. I remember, for example, I asked him some question about disconfirming evidence, and he said, look, you know, people's average mistake. Everyone's worried about disconfirming evidence. Everyone's worried about, you know, you know, thesis creep.

He said, but, you know, if you look at what people do, their average mistake is they sell too soon. He said, so, so I just fall in love with my companies and I enjoy it. I thought that was so brilliant. I mean, so he's dropped pearls like that, which I've found to be very fun. So, and I've changed the way of thinking. I said, I heard that. I said, you know, he's right.

So now I'd probably fall in love with my companies more and I enjoy it. So. No, this is the hashtag never sell, which I'd love to talk to you about more at some point i'll certainly have to reach out to rob but this has been so much fun this has been a blast and i've learned so much about cardlitics just just on this podcast so this was great thank you for coming on and we'll hopefully have to have you on again sometime thank you cheers