Pod Insights Summation with Auren Hoffman Sayari CEO Farley Mesko - mapping global financial crime About this episode Farley Mesko is the CEO and co-founder of Sayari, a financial intelligence and supply chain risk platform that serves major financial institutions, government agencies, and corporations. In this episode of World of DaaS, Farley and Auren discuss: How geopolitics are reshaping global supply chainsThe evolution of financial crime networksBuilding data products for government agenciesWhy entity resolution is the future of intelligence Looking for more tech, data and venture capital intel? Head ...
Transcript Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.
0:02 Auren Hoffman? Hello, fellow data nerds. Welcome to World of DAS. I'm your host, Orin Hoffman, CEO of Incubate and GP of Flex Capital. Discover more episodes, get weekly data as a service news, original content, articles on data and more at worldofdas.com. That's world of DAS, D-A-A-S, dot com. Hello, data nerds. My guest today is Farley Mesko. Farley is the CEO and co-founder of Cyri, a financial intelligence and supply chain risk platform. Prior to founding Cyri, he was the COO of the Center for Advanced Studies, a national security research organization. Farley, welcome to World of DAS.
0:40 Speaker 01 Thanks, Orin. Great to be here.
0:42 Auren Hoffman? I'm very excited. Now, the last few years have brought massive changes, especially in like the global supply chain. What are you seeing from kind of your unique perspective?
0:52 Speaker 01 Some really big thematic trends driving demand for transparency. Transparency both in the financial network side of things, which is part of our business, and in supply chains as well. And obviously, I see those two things as fundamentally related. I think if I had to map back to one big thing that's driving most of this, from our perspective, I would definitely point to the strained geopolitical environment.
1:22 Auren Hoffman? Is it US-China and we have to understand the seventh layer down supplier to something?
1:29 Speaker 01 Or is everything from trade tensions with China, as you point out, to the Western efforts to restrict the flow of technology and finance to Russia for their war in Europe, conflict in the Middle East, everything we're seeing in the market kind of maps back to this strained geopolitical environment. So that's the one big thing. And then, as you think about how it trickles down from there, you look at sort of the knock-on effects, China example that you mentioned, that seventh layer supplier, that manufacturing capability, the financial flows, that has to go somewhere. As we think about strategic decoupling from these geographies
2:09 Speaker 01 or whatever, that has to go somewhere. So what we're seeing, from our perspective, our clients, is a very pronounced fragmentation. Our clients are very much accelerating the pace at which they diversify their supplier base away from these traditional centers of gravity like China.
2:29 Auren Hoffman? I could see a lot of scenarios where, okay, something was being manufactured in China and now it's still being 98% manufactured in China, but it's 2% finished in Vietnam and now it's got the Vietnam label on it. I mentioned these things are just incredibly hard to track.
2:45 Speaker 01 They're very, very difficult to track. There's probably all kinds of enforcement and policy related things we could talk about in terms of what the customs authorities have to do, in terms of looking at value add across the supply chain and tracing individual components, different techniques, technologies for looking at literally down to the DNA profile of where materials came from.
3:04 Auren Hoffman? There's a lot of incentive to lie all across the chain.
3:07 Speaker 01 For sure, yeah, from a tariff perspective, absolutely. From our perspective, and what makes things interesting to us as a data company, talking about DAS, is you have this fragmentation, you have this shifting of capacity into Southeast Asia or Latin America, Middle East, Central Asia even. And in some ways, that's great because it's not China. And in other ways, though, it begs the question, just because it's not China, does it necessarily mean that that's safer? And that could be what you're talking about. It could be this minimal value add and kind of using terror evasion strategy.
3:47 Auren Hoffman? More diversity probably does mean safer. So if you're getting it from 13 countries rather than two, it's probably more likely you have some redundancy there, as long as it's not all coming from like really the same source.
4:00 Speaker 01 It definitely makes things more resilient. It also definitely makes things more complex. And I think that complexity, those open questions about transparency in some of the places where you do see this capacity kind of tailing into, pair that with a more engaged consumer and a more active regulatory community putting pressure on sanctions and trade restrictions. It does create a pretty gnarly risk and compliance environment.
4:31 Auren Hoffman? If you think of all these relatively new sanctions on Russia, we've seen all these great stats where it's like the French exports to Russia have gone down by Y dollars and the French exports to Armenia, Kazakhstan, Georgia, et cetera, have gone up by that exact same Y. And so clearly they're just moving it and maybe marking it up. Probably hurts Russia a bit because they have to pay an extra 15% or something for those same goods, but they're clearly still moving to the end source.
5:00 Speaker 01 If you kind of think back over the last couple decades, we're all familiar with the rise of sanctions as a policy tool. There's been sort of a proliferation of lists and everything else. But I think that at some point, there's a realization that you can't put everything on a list. You cannot keep up with naming the individual companies in Armenia or Georgia, wherever. It's crazy. And so one of the things that we've noticed, and this is also very interesting for us as a DAS company, is everything that we just talked about has kind of manifested itself in policymakers utilizing more narrative-based
5:44 Speaker 01 restrictions, narrative-based sanctions and policies rather than list-based. So government basically tells you, there's all kinds of people and companies that you can't do business with. They fit this parameters of this narrative policy that we put together for all kinds of reasons. You can't do business with them. We're not going to tell you who they are. It's one thing to kind of go to industry and say, okay, well, this guy is a terrorist. He's ISIL. Put him on a list, screen it for him, et cetera. It's quite another thing to say, we're going to come up with restrictions around an entire section of the Russian economy or an entire region in China or an entire set of goods and components that could be
6:26 Speaker 01 used in dual-use technologies and put restrictions on that. There's no list to screen against. And industry kind of has to figure out how to do that. Again, this question of how do you screen in a non-list-based world is a very interesting question as a data company and something that you could sort of sum up our business in a nutshell.
6:47 Auren Hoffman? And even in a list-based world, it's super hard because you have these companies that are like Russian dolls going through. They have nested LLCs, even if LLCs are all based in Delaware, which they're often all over the world. But even that Delaware LLCs are just incredibly hard to penetrate. Sometimes they're the hardest ones to penetrate. You have all these different entity swappings and stuff like this. Even Microsoft doesn't have like a simple corporate structure nowadays. It's like common to have these nested LLCs. It doesn't necessarily say you're a bad actor if you have them, especially overseas when they have their own reasons to have nested because they're trying to evade their own government sanctions, their own government taxes, or something like that. How do you deal with all that stuff?
7:31 Speaker 01 We used to keep track of the craziest corporate structures we've ever seen. I remember it was like 12 or 13 or 14 layers of ownership that we were kind of able to map out on one of them. It was just really sprawling stuff. There's probably a broader point to make here, which is just that whether we're talking about this phenomenon of company structures or money laundering topologies or whatever, the most capable of the illicit actors are not going to stand out, right? I mean, they know enough to know to not stand out. And so also kind of what you're talking about where it's like we have maybe something is sanctioned, but maybe it owns something that's not on a list. It all falls back into
8:12 Speaker 01 that category of the narrative. The treasury says you can't do business with these people if they're owned this percent or control this in the EU. And how do you screen for that? And so for us, we can pretty easily take a narrative sanction, a description or a trade policy, a trade restriction. And we've collected this massive haystack of data, company data and trade data, individuals tied to those companies and shipments. And we can basically slice it and dice it based on those parameters and effectively create lists where they don't otherwise exist. And that's our business in a nutshell. And it's great because industry has spent the last
8:52 Speaker 01 Two decades focused on optimizing the process of comparing, I mean, to your point, even screening itself is not a solved problem, but you spent decades optimizing for, take this list from the government and match it against my list of customers or the declared end users of this thing I'm shipping or whatever, and make sure there's no intersection. And so then what do you do when that doesn't exist? And that's kind of where we come in.
9:23 Auren Hoffman? One thing for to do that in manufacturing, it seems like a next level more difficult to do that for commodities. And a lot of these, especially commodity traders, energy traders, et cetera, they're well-known for themselves also trying to skirt the rules, move things around. Very few of them are based in the US for that reason, because they don't want to be completely under the US. They're usually in Switzerland or Dubai or Singapore, some other type of place, Hong Kong. How does one deal with those types of entities?
9:54 Speaker 01 Certainly the sophistication of sanctions evasion, financial crime and trade-related illicit activity, the maturity and complexity of those schemes has increased. There is both a professionalization in the sense that the illicit actors are becoming more professional, and also a professionalization in that they are getting more support from professionals. Those are two different things. I think politically, again, to come back to my previous example, it's all well and good to say, okay, this guy's a member of ISIL, let's put him on a list.
10:37 Speaker 01 The world can pretty much agree on that with some exceptions, but can pretty much agree on that. It's quite another thing to say, well, here's a sector of someone's economy or a commodity coming from this geography that should be restricted. And then going and trying to enforce that, I mean, it puts you in a very difficult position as a multinational business. The framing maybe that you made was a little bit more insidious, but it could very well just be that a multinational broker, trader, whatever, is just trying to figure out if I have a sanction from the US over here that says I can't do this thing and then a policy over here from China that says if you do this thing, you will get sanctioned. How do you navigate that?
11:19 Auren Hoffman? Yeah, exactly. They have competing interests. Historically, a lot of the money flows have flowed through US banks. That is potentially starting to change. I'd love to get your thoughts of like how the money flows are changing.
11:32 Speaker 01 I've heard this about the demise of the US dollar before. I can't say it feels super different this time around, so I'm skeptical that the primacy of the US dollar is really under threat right now.
11:45 Auren Hoffman? You could still trade in dollars and still do bilateral, though, and not go through a US bank. You could have ledgers on both sides, or you could have Tether or Crypto or some other type of thing too, which is dollar-denominated. There's probably other ways, because the dollar does seem like a safe way. Everyone kind of understands it around the world to trade. Doesn't mean you have to route it through a US bank.
12:07 Speaker 01 It's true. I think it would be very difficult, at least if you're talking about large-scale, industrial-grade money laundering, to avoid, I don't want to say the US, but at least the Western financial system, which sort of falls under the loose jurisdiction of the US because of the cryptocurrency networks that run through here. If you look at some of these DOJ indictments recently around the Chinese, they call it flying money or some flying money service, and it's tied with crypto, it sounds very scary at the headline level. And you dig into it, and it's like the vast majority of the money is moved through more
12:48 Speaker 01 traditional trade-based money laundering topologies. If you're trying to access tumor market here in the US for your drugs or whatever it is, like it's going to be very, very hard to avoid that. Again, that's sort of US-influenced.
13:03 Auren Hoffman? It's really interesting, the Chinese money laundering, you have one side where you have all these Chinese people around the world who are laundering money to avoid China because they're trying to get money out of China, they're trying to move money around, they're trying to move money in a way for the Chinese government not to see it. And they've gotten so good at it that they're really good at money laundering for everybody. These Chinese money laundering folks have become this Goldman Sachs of money laundering because they're helping all these Chinese expats. You can see all these like weird things happening from once and we maybe we benefit from Chinese laundering money out of China, but then we also get hurt by the opposite.
13:39 Speaker 01 That's right. It's one thing to sort of cheer from the sidelines when financial flows are coming out of China, but you have to think about what kind of infrastructure is being created to do that and how else it could be repurposed.
13:49 Auren Hoffman? From your standpoint, you don't see like crypto yet because it's still a very, very tiny percentage of all the money flows.
13:56 Speaker 01 You alluded to this. First and foremost, nobody knows the scale of the problem of money laundering globally. People argue about it all the time, go on LinkedIn, a new chain analysis, state of the market or whatever report comes out, or a new IMF study comes out and everybody's talking about what percentage of crypto transactions are illicit, how does that compare to the traditional financial system? If it's not clear, I am convinced that nobody knows even on the crypto side or on the traditional side, the actual volume of money that is laundered globally every year. Regardless of that, the ratio or the
14:39 Speaker 01 percentages of absolute terms, there can't be any comparison between the volume laundered to crypto and the volume laundered to the traditional financial system.
14:46 Auren Hoffman? We're talking probably trillions in the normal money movement and maybe billions in the crypto movement.
14:54 Speaker 01 This is very, very early days, Sarah's stuff, but we would get this question from investors all the time. And we had this graphic of the estimated amount of money that was laundered through crypto. There's a bar graph and on the one side is crypto and it's like hundreds of millions of dollars, something like that. And then going like the other side, all the way up and then onto the next slide, a bar and then onto the next slide, the next bar, onto the next slide. And then the top of the bar, like six slides later is the volume of money laundered through the traditional financial system. First, it doesn't mean that we shouldn't tackle both. Both are important. I just think we still have a ton of work to do in our ability to stop dirty money through more traditional channels.
15:37 Speaker 01 But crypto never struck me as a super logical or efficient way to move tons of money. If you kind of put yourself in the shoes of a professional cartel launderer, I would have to imagine that just thinking about shipping containers of bulk recycled cell phones or something like that, that has no market value, kind of make up whatever value you want and ship that and somebody can buy it for something else, some other amount of money and the difference can kind of be the transfer of value. That is, to me, one of the only reliable ways of actually transferring that much value.
16:16 Speaker 01 I think for crypto, there's probably use cases of small volume, high value, fentanyl precursors comes to mind, really something that you don't need, it's not talking billions and billions and billions of dollars at a time or tens of millions at a time. I would hesitate to even call them edge cases because they're well-defined use cases that we have to worry about. But I do think even in situations like that, you look deep enough, you'll probably still find some alliance on the traditional financial system too.
16:42 Auren Hoffman? One of the problems is for many jurisdictions around the world, if you just think of emerging markets world, almost 100% of companies are doing some sort of money laundering because they have to from their own government. They often have the two sets of books. My guess is like every company in Russia has always had two sets of books, but that's probably true of like many, many, many countries that are out there. They're trying to avoid their own government. Their own government may have certain onerous things, or maybe that's more of an accepted practice there. Money laundering becomes the norm for many, many, many companies. And maybe even the majority of companies around the world have to engage in it. Then it's like, okay, well, what level of money laundering are we willing to accept to go through?
17:26 Speaker 01 For that reason and other reasons, it can be very difficult to collect data and get transparency in places like that, Russia, China, et cetera. One of the ways I often describe what we do is arbitraging what people think is available publicly versus what is actually available publicly. In many of these environments, there is just far more information available publicly. If you know where to look and whether that's the legacy of communism, a very highly bureaucratic, record-keeping culture, whatever it is, there is just a lot of that. It's not necessarily easy to find. You have to be prepared for foreign language, non-machine-readable, semi-structured, unstructured data.
18:09 Speaker 01 But it does typically exist more so than people think, both in terms of the regulators, but also or compliance teams helping them open up that data so they can look at it. Then also the people who are actually creating these shell companies or whatever in the press place. Like, I don't know that they fully realize that that record with that receipt scanned and attached to the back of it that says, I'm Chapo's kid and I created this company. And yet here's the receipt for which bank the money was paid out of, you know, scanned right onto the document. That stuff does exist and it does exist at scale.
18:46 Auren Hoffman? Partly to connect this jigsaw, I imagine entity resolution is going to be incredibly important. Like, how do you think about entity resolution? And I assume in your case, the entity is the corporate entity.
18:58 Speaker 01 I would go so far as to say that one of the more important pivots, quote unquote, that we went through in our journey as a DAS and SaaS company. Was probably three or four years ago, we've been around for 10 years. We spent the first few years of the business really focused on the data acquisition side of things and just professionalizing that, automating the processes for collecting and refreshing our data. But a few years ago, we sort of realized that at the end of the day, we had to be an entity resolution company because, and I think this is true in almost any data intensive
19:41 Speaker 01 business, but at least part of and probably a significant part of your raw material inputs are going to come from somebody else's data. And that's often commercially available, maybe it's publicly available, but the only data that you're going to 100% own is data that's generated on your platform. And then you have all the questions about monetizing customer data and privacy implications. But if you rely on, to any significant degree, data that could be acquired publicly or commercially by someone else, by a competitor of yours, or somebody who with enough time and resources,
20:23 Speaker 01 given a five-year time horizon, how do you think about building that durable moat if you're necessarily dependent on other people's data? This is a question we've spent a ton of time thinking about. And so one of the best ways of creating that durable barrier to entry, that moat, is to invest the time and the resources in both the business logic side of entity resolution and the infrastructure and scale side of it. So being able to do this against really, really large sets of data, both of those things done well, I think can really underpin an entity resolution component to your stack that serves as a differentiator. And the goal there, of course, is
21:08 Speaker 01 Even if you didn't originate the individual sort of components there, if you can be the one who does the best job of stitching those things together, then you can create this derivative product, a unique derivative product on top of it. And then at least you have sort of two legs of the stool. The data itself is necessary, but not sufficient as a piece of the moat. But then for the secret sauce of how you join things together, the business logic, any AI layer you put on top of that, the scale and the processing side of things that I mentioned, that gives you a much better foundation to build on.
21:43 Auren Hoffman? I assume in any resolution, in your case, the false positive is way worse than the false negative.
21:49 Speaker 01 You're absolutely right. And this is sort of what I was getting at when I was talking about, invest the time and the resources to build business logic and train models and build infrastructure that works for you because you're going to know your data better than anybody else. And so in our case, yes, a false positive is worse than a false negative. And so we bias RNA resolution models a little bit more on the conservative side.
22:14 Auren Hoffman? And AI is getting better and better at providing tools to help on that. So I imagine a lot of things you may have done eight years ago is no longer advanced today. How do you deal with that world changing?
22:26 Speaker 01 It is very much changing. This is true of the user rules based approach or an AI driven approach, or some combination of both, which spoiler alert, that's where I would say the world is going, or certainly where we're going. But the annual resolution, in addition to all, everything I just said about creating that mode, good annual resolution allows you create a network effect in your data. So obviously your data holdings get more valuable as you add more data, like that's definitional. But done well, adding new data then also because it increases the connectivity of previous data sets that you had, it increases the value of each of your old data sets, every new data set that you add. That network effect is very powerful.
23:10 Speaker 01 And if you then professionalize the process of really getting new data sets onboarded, ETL fast, you create that really nice flywheel effect of just adding new data. Value increases and being able to throw around words like flywheel and network effect is also a good thing when you're raising money for best VCs.
23:31 Auren Hoffman? Asking questions across data sets makes each data set way more valuable because you can ask more deeper and deeper and deeper questions. Any one data set is going to be really not that useful at all by itself. That's the problem also even evaluating how valuable the data set is. If you evaluate in a vacuum, you may think it's not valuable, but then when combined with these other four, it may massively increase the other four's value.
23:55 Speaker 01 Yeah, think about the example that you talked about a few minutes ago. Seven layers deep of suppliers across presumably seven different countries trailing back into China. That is an import data set from the US and maybe a corporate records data set to show who owns the company here and transshipment data from Ecuador and value add data that was added in whatever Vietnam, tracing back to a manufacturer in China, which is a Chinese import-export data set. But then who owns that manufacturer is a six data set and then going on and on and on.
24:31 Auren Hoffman? We're back to the Russian doll there.
24:33 Speaker 01 The Russian doll of data. Yeah. If you break any link in that chain, you really, really diminish the value. And so the flip side is if you continue to add data, the ER gives you the ability to unlock more value.
24:45 Auren Hoffman? Time-based data can add a ton of value as well.
24:49 Speaker 01 In general, I think that temporal data is more valuable than non.
24:55 Auren Hoffman? Certainly, the price of temporal data is way more expensive than non-temporal. Yeah.
25:00 Speaker 01 If you are sourcing any part of your raw material, as we talked about a second ago, from places where other people could get it, as we just discussed, you really need to think about your mode. And temporal data, to the extent it is ephemeral, meaning that it disappears once something changes and there's no natural change log wherever you natively sourced that, that can be very valuable.
25:26 Auren Hoffman? Because you need the history. So if you're the only one with the history, then that's huge.
25:30 Speaker 01 Correct. Back to my example of like the ISIL designated terrorist, if that person got designated by the US Treasury and then the next day resigned from their five companies, but that native registry had no record. Once it changed, there was no change log, right? And so that just disappeared into the ether. If you were the only one that had that, that would be tremendously valuable. There is a half-life to that value of that data. I mean, it is definitionally historical, backward looking, not sufficient to get into things strategically, but that aspect of temporal data is definitely very valuable. As a DAS company, and particularly in the red tech risk tech space, we see a few different variants of this temporal data trap.
26:14 Speaker 01 On the one end of the spectrum is kind of the one-time value. Every risk and compliance workflow, I just say every, but most risk and compliance workflows to some degree have some kind of screening step where you have to evaluate a large population for risk. And then you find some things and then you triage that and then you look into some stuff and then you whatever. And so a lot of it starts from that screening bit. And so many companies fall in this trap of building a product that delivers a lot of one-time value. The customer gets diminishing returns once that initial screen is complete. So if you're not thinking about that reframing of the temporal aspect to data, but if you're not thinking about that ongoing value, we've seen countless times SaaS companies fall into that trap.
26:59 Speaker 01 But then the other side, this is also something that we have the scars to attest to here. But there's a different trap on the other end of the spectrum. It's sort of one time is on one end. I think real time, quote unquote, is on the other side of the spectrum. This is not true of all use cases. My point is really, you need to know your use case and you need to know how real time you need to be. Because most companies aren't changing daily. Most use cases do not require real time data. And so that temporal high-velocity data is definitely inherently more valuable.
27:42 Speaker 01 It's inherently better at delivering the ongoing value that we talked about because it's changing all the time. But it's only true up to a point for the vast majority of use cases. I mean, there's a point of diminishing returns where if you strive for real time. You really need to make sure that the use case actually demands that because real time is very hard. The faster you go, the harder it is. And that was, I think, something that we learned. Our sanctions data flows in every 24 hours. That's probably the highest velocity data that we have. The sanctions data is comparatively tiny compared to our overall holdings. There's 20,000 things, maybe 50,000 things against billions. But trade data, that's hundreds of millions of transactions. That's like in every two
28:25 Speaker 01 to four weeks, we update that data. And again, extended resolution work, you're able to create more value in the historical data, et cetera, et cetera. And so we're able to monetize that, you know, offer a monitoring capability that's tied to kind of like every few cadence. We have great retention. Typical data is definitely more valuable, but there is a point of diminishing returns there.
28:45 Auren Hoffman? So many companies are trying to sell on a national security and they run into this issue where they get all these like additional requirements that they have to do. And from the product manager or the CEO, who might be like the product manager's perspective, these requirements add massive costs and then they often degrade the product. They make the product worse, not better. They feel bad about charging the government more. They're a patriot. They don't want to charge the government more. And they also feel bad about offering them a worse product than they give to other people. I don't know if that's a legitimate criticism, but how would you allay the fears of some of these folks that are working with government and national security world?
29:24 Speaker 01 If you're going to pursue work for the government, just know what you're getting into. You're not going to be able to run a typical SAT playbook. It takes months of paperwork even to set up a free trial. You can't give things to the government for free. Creates an unfair competitive advantage. And we have a lot of government business. It took us a while to figure out how to, as close as we could approximate to a SAS or DAS playbook in the government. So you just need to know that the traditional playbook won't necessarily work. It takes time. There's no hack. There's no shortcut. I mean, look, you can do the accelerators. You can do the SBIR programs.
30:06 Speaker 01 The government is objectively getting better at working with startups than it used to be. But at the end of the day, you're going to move at government's speed. And you're going to have to do what the government asks you to do. Over time, you build up enough of a customer base. And if you're intentional about how you work your way into analytical exchanges and briefings and become a hill and whatever, you build up enough brand awareness that people will trust you when you say, you're telling me you want to do this, but it sounds like you want to do is this, and here's a better way to accomplish what it is you're trying to do.
30:47 Speaker 01 But that simple process of requirement solicitation and iteration is very, very difficult in government. And it's even difficult if you go, as I said, through the accelerators, through any of the SBIR work or any of those early stage issues for government because you're going to be building, you're going to be intermediated by some kind of a program staff, whatever accelerator or group that you're working with. They're going to be intermediated by the R and D folks who own the budget where the agency support to that R and D, to that accelerator goes. And then they're intermediating the actual end user who's sitting somewhere in a SCIF or is forward deployed or whatever, who can't talk to you.
31:30 Speaker 01 And so that game of telephone of the requirements is incredibly difficult and time consuming way to build software. There's no hacks. It just takes a lot of.
31:40 Auren Hoffman? You guys brought on TBG as a growth partner about a year ago, and they took a big piece of the company. In some ways, it was like a partial exit. How did that manifest itself? How do you think through that?
31:51 Speaker 01 It's been a really great partnership. I mean, we'll be effectively a year since signing next month, which is hard to believe. There were a few reasons that we wanted to transact at all, and then why we wanted to do a partial exit. I think you could probably bucket those into two or three different categories. On the factors internal to the company side of things, I think there was a pretty strong sense that we'd been at it for, at that point, eight and a half years, something like that.
32:23 Auren Hoffman? I assume you had some shareholders that were like, hey, I would love to get a return on my investment.
32:27 Speaker 01 Totally. I mean, as a CEO, being able to return capital to shareholders and get liquidity for yourself and your early employees and all your employees. So thread that needle between returning value to the employees and the shareholders who got you to where you are, but then also preserving that upside for the next transaction. That is a really nice needle to thread as a CEO. The reason we did it, as I said, we've been at it for eight and a half years, and I think there was a realization among our leadership team that if we really wanted to set the company up to continue to grow at the very fast clip that we'd been growing and do that in the future and really get to the next level, there
33:11 Speaker 01 needed to be a general sort of up-leveling of the organizational maturity. We weren't going to do that on our own. We weren't going to do that with some kind of external accountability. I think different companies attack this in different ways, but for us, you kind of know what needs to happen. You know there's more strategic work you need to do. You know you need to rethink how you package and price your products in the market. There's all these big strategic things that you kind of know you need to do. But if no one's there making sure that you do them, it is very, very hard to self-motivate your way. And things are going well, too. It's just very hard to self-motivate your way through that. And you're so operational.
33:53 Speaker 01 You're thinking a year out, a year and a half, two max, maybe. There's nobody saying, what is the five-year strategy? And what does the org chart need to look like in four years? And each quarterly milestone to get through there? Like, how do we really think long-term about this business? It feels totally hand wavy when you're in startup operational mode because you're just so focused on the next year or so. So we really needed somebody to kind of push us to do that. And it's been awesome. I mean, our team has really stepped up. It's made us all better executives and just as a company, more organizationally mature. The other big reason we did this was just if you look at the dynamics of the market
34:34 Speaker 01 that we're in, highly fragmented, you could count on probably a couple fingers, the number of companies over a hundred million in ARR, whatever arbitrary level gets you to maturity. We have the effects of all the 2021 money that propped up all these businesses, unreasonable valuations, and everybody is still in this all in grab mode, highly competitive, propped up by VC money. And I think that that's fine, except that with heavy competition comes a lot of price pressure and you're gonna see a separation of companies that have good economics and sustainable business model versus those that don't. And we really wanted
35:18 Speaker 01 to position ourselves as an acquirer, as a consolidator, as the industry inevitably starts to consolidate. For other CEOs who might be in our position for any transaction, but obviously something like this is very consequential, just know what you're trying to get out of it. We sort of had a pretty good idea of what we needed as a business in order to get to the next level. And we designed a process to get us there.
35:43 Auren Hoffman? Now, a few personal questions. You are famous for having like a big Movember style mustache, 12 months out of the year, not just in November. That was definitely big in like the 1850s. Do you think it's gonna make a comeback? There's not that many mustached people anymore.
36:04 Speaker 01 It was totally on a whim 18 months ago. I just. Was like, I'm gonna do this. And here we are. And now I feel like I can't stop.
36:14 Auren Hoffman? You can't do it. It's like part of your brand now.
36:17 Speaker 01 I shaved the other day, probably a few weeks ago, and I forgot to trim your lip the same way when you're trying to grow a beard. And it looked so bad. And I was like, it scared me. And now I'm just locked into this.
36:30 Auren Hoffman? I like it. Two more questions we ask all of our guests. What is a conspiracy theory that you believe?
36:34 Speaker 01 There are so many that I want to believe. I am 100% all in on these UFO conspiracies that have bubbled up over the last few years. I'm so energized that it's gone mainstream. It definitely feels like the dam is breaking.
36:52 Auren Hoffman? Was a year or two ago when the New York Times started writing about it, which is crazy.
36:56 Speaker 01 Exactly. I know enough about how the government works now to know that a massive, sophisticated, coordinated cover up is probably not what's going on. But I do want to believe that there's more of a story and somebody knows something somewhere.
37:15 Auren Hoffman? Okay, I love that one. All right. Last question we ask all of our guests. What conventional wisdom or advice do you think is generally bad advice?
37:22 Speaker 01 I think it's a stretch probably to say this is bad advice. I think the thing that I tried to live for a long time was hire good people and get out of their way. This is sort of conventional wisdom in the DAS and SAS world. I think that this can be good advice sometimes in some situations. Maybe if you're a super mature business with incredibly well-defined culture, like Netflix or something. In general, even with your most senior team, in my experience, the best people that I've worked with want to be both led in the sense of, you know, hiring people to get out of their way, but also managed.
38:04 Speaker 01 They want to know how they're doing. They want to feel like they have a constant finger on the pulse of what you're thinking and overall company strategy and how that's evolving. They want your input on decisions. It's not sink or swim. They want to know. How what they're doing is contributing to that overall company strategy and how they can course correct if need be. I think, of course, like as the company gets bigger, you have to push more decision making down on the team. And this is probably where a lot of this is user error, rather than this is like generally bad advice. But I think that so many people hear people talking about this concept, hire good people the other way, and they sort of internalize that, abdicate the responsibility of management when it comes to the senior team. And I think they probably do that to the company's detriment.
38:49 Auren Hoffman? If you have many layers in a company, well, then like everyone's hiring quote unquote, good people and then getting out of the way, then it's hard to inspect, okay, how good are the people are they hiring? It could be many layers where people don't really know what's going on.
39:02 Speaker 01 You better have really good leading indicator KPIs because it can't just be you are responsible for this thing. I hired you and I got out of your way and it didn't work. That doesn't help you at all.
39:14 Auren Hoffman? Yeah. And the more people you have, the harder it is to kind of do that in a way. If you could design a company to have fewer people, or maybe you designed the company to have fewer layers. Jensen Huang has 40 directs or something. So he has very few layers in a company of that size.
39:29 Speaker 01 For sure. Again, if you are incredibly thoughtful, intentional, and probably a pretty mature business with really good leading indicator KPIs, then yes, maybe I could be convinced that this is really good advice. But in my experience, it hasn't been.
39:44 Auren Hoffman? Yeah, no, I'm with you. If you're changing anything, you kind of have to be in it to do it. All right. This has been awesome. Thank you, Farley Mesko, for joining us. We're all at DAS. I follow you on LinkedIn and definitely encourage our listeners to engage with you. This has been a ton of fun, super interesting.
40:00 Speaker 01 Thank you, Auren. This has been great. Thank you again.
40:05 Auren Hoffman? Thanks for listening. If you haven't already done so, please subscribe to World of DAS wherever you consume your podcasts, YouTube, Apple, Spotify, and more. And please help us get discovered by leaving a review and check out worldofdas.com. That's worldofdas.com. And of course, connect with me on Twitter, at Auren. That's A U R E N. Would absolutely love to hear from you.
Previous Magnite $MGNI CEO Michael Barrett - from linear TV to AI Next DoubleVerify $DV CEO Mark Zagorski - the $100B war against digital ad fraud Page updated November 5, 2026 · transcript and summary are machine-generated