Pod Insights Summation with Auren Hoffman Fmr Verisk CEO Scott Stephenson - building a $40 billion data company About this episode Scott Stephenson is the former Chairman, President, and CEO of Verisk Analytics (NASDAQ: VRSK), where he led the company's transformation from a nonprofit insurance data collective into a global data analytics powerhouse with a market cap of over $40 billion. In this episode of World of DaaS, Scott and Auren discuss: Data monetization and defensibility strategiesAI's impact on data businessesLeadership succession in data companiesEffective board membership Looking for more tech, data and ven...
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, fellow data nerds. My guest today is Scott Stephenson. Scott served as chairman, president and CEO of Verisk Analytics from 2013 to 2022. Verisk trades under V-R-S-K on NASDAQ, has a market cap of over $40 billion. Since leaving Verisk, Scott served as chairman of Avetta and most recently became chairman of One Incorporated, a digital payment solution
0:45 Auren Hoffman? for the insurance industry. Scott, welcome to World of DAS.
0:49 Speaker 01 Thank you, Orin. Good to be with you again.
0:52 Auren Hoffman? So you are the second second timer on World of DAS. So this is a huge honor.
0:58 Speaker 01 I guess, I guess so.
0:59 Auren Hoffman? Yeah. Now, we first talked when you were CEO of Verisk. Now, how do you think Verisk has done since you've departed?
1:07 Speaker 01 I had a few hopes for the company when I was stepping out in 2022. Really, there were four things that I was hoping would happen and they were the things I was talking about, actually, around the time of my departure. Those were, number one, I was hoping that we would sell the energy business. The second one was that in light of having changed the contours of the business, because there were other non PNC businesses that had been sold even before that.
1:33 Auren Hoffman? PNC is property and casualty.
1:34 Speaker 01 Property and casualty, right. So the majority of what the company was, in light of that, the second thing I was hoping for was that the costs would be refactored to reflect a more focused kind of a business.
1:46 Auren Hoffman? And cost being refactored, meaning like reduction in workforce and stuff?
1:50 Speaker 01 Yeah, in line with, and mostly center costs and mostly in line with the fact that the shape of the company was changing a little bit.
1:58 Auren Hoffman? So center, meaning like at the headquarters, operations, those types of things, G and A?
2:04 Speaker 01 Yeah, shared technology and things like that. The third thing I was hoping for was there was a project. Called the reimagine project that was actually initiated in 2018, 2019, but it was basically about trying to completely rethink the most legacy part of the whole company. And just to come up with fundamentally different ways of serving customers and adding value. And the fourth, which is kind of a little bit more tactical, but by the time I left in 2022, the company had been out in the public markets for 10 years. And so I felt that there were some matured processes that ought to be in place. One of which was to provide guidance to the markets. We were now a well seasoned kind of stock.
2:43 Auren Hoffman? Sir, what do you mean by providing guidance?
2:45 Speaker 01 Annual guidance of.
2:46 Auren Hoffman? Oh, that wasn't happening?
2:47 Speaker 01 It didn't.
2:48 Auren Hoffman? Oh, interesting.
2:49 Speaker 01 Does now.
2:50 Auren Hoffman? Okay, got it. Okay. Why wasn't it happening before? I thought like most companies do that.
2:54 Speaker 01 Well, usually when companies begin to go public, they take the opportunity, at least for a while, to talk to the street in the ways they would like to. And it frequently doesn't involve guidance right out of the box. But by the time you're in 2022, the company had been public for 10 years. To me, it seemed like it was time to move in that direction. So those were four things I was hoping would happen and they did happen. So they have happened. So I'm happy with that. Yeah.
3:20 Auren Hoffman? That's amazing. When you have a new CEO, did you work with the new CEO? How's it work? Or does the new CEO, classically, the Mark Andreessen advice for when everyone gets a new CEO, throw the old CEO under the bus, right?
3:34 Auren Hoffman? That's one way of doing it, right? Another way is we're going to be collaborative. What happened in your guys' transition?
3:39 Speaker 01 In the case of Verisk, the person who succeeded me was my CFO. And so he had been with the company for several years. I felt really strongly a desire to see an internal successor, if at all possible. So I had been active for years, actually, in creating a couple of internal candidates. And the board chose one of them.
3:59 Auren Hoffman? Okay, cool. So the board like ran a process and then they had maybe some internal and some external or something?
4:05 Speaker 01 It was more focused on the internal candidates.
4:08 Auren Hoffman? Okay, got it. And usually the CFO is not always the best leader of the company. Maybe in this case, because it's more financially oriented. Why do you think in this case, the CFO was a better pick?
4:21 Speaker 01 I think there were a couple of good candidates, but the board felt comfortable with him. Lee Shavel is his name and he was well known to the board. I think there was a comfort factor there.
4:31 Auren Hoffman? Now, you were there for a while. When you look at it in hindsight, what are things you would have done differently? How do you think about your journey there?
4:38 Speaker 01 Maybe I can tell you what I consider to be a couple of the lessons that I take away from the journey. And maybe I can just take a second and create a little context around all this, because it's actually a story which played out over a long period of time. But I do think that some of the things which developed, I just think they're good things to think about for our DAS entrepreneurs who are on our call here today. So it actually, the precursor company to what became Verisk came into the world in 1971. And it was basically the PNC insurance industry deciding that these agencies that they had at the state level that were basically stood between the insurance companies and the regulators in order to facilitate data flow from the insurance companies to the regulators. But the insurance
5:22 Speaker 01 industry collectively in 1971 said, these state level things, these state level agencies that exist, let's pull them together into a national organization.
5:30 Auren Hoffman? It was a nonprofit at the time, right?
5:32 Speaker 01 It was a nonprofit. So 1971, all that stuff gets pulled together, nonprofit, and basically remained in that same condition for 25 years until 1996, when it was reorganized. But two really important-
5:45 Auren Hoffman? And you were, at the time, you were like a partner at BCG. They were one of your clients, right?
5:49 Speaker 01 No, they were not one of my clients, but you're right. I was at BCG at that time. And actually, I didn't know the organization at that point. But the importance of that first phase of the company, I think, was fundamentally two things. One was the solutions were being put out, the then set of solutions, much smaller than what's there today. But they were being put out at, they represented a good deal of value, because being not-for-profit, essentially, the customers were getting these inputs at cost, and they were reliable products. So customers felt comfortable increasingly standardizing their own processes around what it was that this company, which then was called ISO, was putting out. So
6:30 Speaker 01 that was one thing. And the second thing was, the company organized itself in a way that it really had a lot of interaction with the customers. Customers were involved at the governance level. There were many, many standing ways that customers were able to provide input into the products and solutions that the company was putting together.
6:50 Auren Hoffman? How did it go from the nonprofit to for-profit? How did the equity get distributed at that point?
6:55 Speaker 01 And this is an important part of the story. So because enough trust had built up, literally over 25 years, the industry was eventually approached by the company and said, "Hey, we reorganize and become a stock-based entity, a for-profit entity. Hopefully, we will be able to improve the performance of the company and reinvest back in and bring you yet more solutions that would be helpful to you." And that was an argument which took a while, but the industry eventually agreed. Essentially, at the 25-year point, yes, okay, we agree with that. As the company was getting reorganized, there were two classes of stock. There was the class of stock that was given to about 250 insurance companies
7:37 Speaker 01 in proportion to the amount of use they made of the solutions in there.
7:42 Auren Hoffman? So basically, how much revenues?
7:43 Speaker 01 How much they were spending.
7:45 Auren Hoffman? Right, so one of them was a million and the other one was 500,000. The million would get twice as much stock as the 500,000.
7:51 Speaker 01 That's right, directionally, that's exactly right. And that class of stock had 90% of the economics of the business, meaning it owned 90% of whatever the value of the company was. But there was another class of stock that was put together, and that was 10% of the economics, and that was an ESOP. And the ESOP, as it turned out, had a fair amount of the voting rights at the board level. So you have a company which becomes fairly agile.
8:18 Auren Hoffman? Basically, employee-controlled pretty quickly, which is rare for a company of that size.
8:23 Speaker 01 Correct, although it was much smaller back then. But that's exactly right. That's why that was the first 25 years. The second 25 years runs to the end of 2021, and that was when I was there for almost all of that. But you needed that foundation of depth with the customers in order to be able to make that kind of transition. And so I just really called that out because I used to frequently say that I felt like I was standing on the shoulders of giants because they had spent all this time sort of creating this depth of relationship with the customers and a degree of connection and trust, which enabled it to be reorganized in the way that we just said. So basically, you have a new company in 1997. It doesn't make any money.
9:05 Speaker 01 It doesn't grow very fast, doesn't do business overseas, didn't write a line of customer-facing code. It had to change. And that analysis, in many ways, was basically actuarial science. There might have been five people around the company that used generalized linear models to kind of model something, but-
9:23 Auren Hoffman? Virg has grown a lot. As we're recording this now, the share price is roughly like $275 a share. It's 40 billion market cap company, which is amazing. But is there something you think you could have done where it could be $1,000 a share right now? Was there a 10x thing that you're like, "Oh, I wish I could have done this," or something?
9:43 Speaker 01 Yeah, we went through some evolutions where in 2009, we went public. 2013, we became a comp owner of the S and P 500. And at the beginning of the journey for me, if you had told me that we were going to be an S and P 500 company, I would have just thought you were crazy. How in the world could that happen? But what I would say about the journey is we had a playbook that fundamentally sort of featured five steps in it. The first one was to try to get as much proprietary data as we could. The second was to turn the data, whether proprietary or public or mixing them all together, into what I call analytic objects that became kind of industry standard, scaler value kinds of things that
10:24 Speaker 01 customers would use, pull into their decisioning. The third was to have enough service and support, but not too much. Enough so that we could help customers get value out of their use of an industry standard solution that their competitors were also using. And to be close enough to see how they were using it so that we could make it better. The fourth was to wrap all that with software, whether it was analytic software or more kind of workflow software to get our signals into the customers' systems. And then the fifth was to expand the solution sets. Because once you've created that software link with the customers, it's like you put a tool belt around their waist, and now it's easier to hang new products on that. So those are the five steps. And I'd like to think that
11:07 Speaker 01 we did a good job at each of those five points. But I really think what got us as far as we got, it became a 35 billion- Million dollar total enterprise value company by the end of 2021. By the end of the second 25 years, I think what was really the most important thing was the rate at which we were running around that circle. I think we were running good plays at each point around that circle. But I think it was actually the rate at which we were running around the circle that was kind of different. So could we have run faster and would that have created, you know, a 70 billion?
11:38 Auren Hoffman? Or is there a different circle?
11:40 Speaker 01 Yeah, or maybe a different circle. I don't know, but.
11:42 Auren Hoffman? Verisk is one of the most successful data companies, but really almost no one has heard of it. Of all the S and P 500 companies, it's one of the least well-known company of the investor class, right?
11:54 Speaker 01 Right.
11:54 Auren Hoffman? Now, is that being relatively unknown? Is that a feature or is that a bug? How do you think about that? Because you could think of it both ways.
12:01 Speaker 01 Yeah, I think that for the public markets, I think that the investors who ought to be interested in a story like Verisk's and Verisk's find each other. I don't think there's any real gap there. And then otherwise, the profile that you get from being public, I mean, I, a lot of it is.
12:18 Auren Hoffman? Maybe you can recruit better people. Obviously, you can make up a lot of cases that it should be better known. Maybe the stock market, maybe the stock price jumps a bit. You have more retail people in it, et cetera. Of course, not being well-known also gives you all these advantages too. I don't know which way that's out.
12:35 Speaker 01 One of the things that comes to mind as you ask that question is not infrequently when I would go and visit the CEOs of our customers. In a nice way, they would sort of tease me about our performance. They can see that we were doing well and.
12:48 Auren Hoffman? Right, right. And they're like, hey, you're raising prices too fast or something.
12:54 Speaker 01 Right, exactly. I'm over here trying to make my company go. So I guess I always thought of the profile in the public markets as being a mixed thing. And the messages we wanted to get through to customers in order to help them to see the value of what we were offering. And hopefully to encourage them to adopt. I felt like that was in our hands, whether we were public or private. So I didn't worry that much about being a household name.
13:18 Auren Hoffman? Yeah, I know CEOs go through this as well. So I know many CEOs I know, they're like, okay, should I? I'm not on social media or I am. And there's pluses and minuses of being well-known, both as a company and being well-known as CEO. Obviously, you want to be well-known to your customers.
13:32 Speaker 01 Yeah, exactly.
13:32 Auren Hoffman? And your market.
13:33 Speaker 01 Exactly.
13:34 Auren Hoffman? Flexport, I'm not a customer of Flexport, but I know so much about them because they're out there. The company's out there. Ryan Peterson, the CEO, is out there a lot. I read them all. So I kind of know a lot about the company, even though I'm not a customer. There's no reason for them to market to me. Maybe when they eventually go public, maybe I'll be more likely to buy their stock or something. I'm not sure.
13:54 Speaker 01 Yeah, I hear you. There's another dimension of being known and what you're known for. And this also, I think, is an interesting point for folks that run companies, that share some of what goes on at Verisk or at your companies and et cetera. And that is in your desire to do as much as you can, to add as much value as you can. How far do you or do you not get away from what might be considered sort of the center of what you're doing? So it's been the case from the beginning that by far the largest revenue stream and the greatest part of what Verisk has been, has been serving the property and casualty insurance industry. But there were moments along the journey where we chose to try to move into other vertical data analytics.
14:39 Auren Hoffman? Where you guys bought Argus, you bought some of these other things.
14:42 Speaker 01 Right. And part of the story, by the way, which I think naturally sort of gets lost as you move forward in time is, the company went public in 2009 based on 2008 financials. And in 2008, revenue grew about $95 million. And the reason why that's interesting to me is about half of that growth was in the mortgage analytics business, not the core property and casualty set of solutions, about half of it. So it really is the case that when the company went public in 2009, it was based on performance, which was the growth part of which at least was pretty substantially mortgage analytics. And then in the time period from about 2010 to 2012 or 13.
15:28 Speaker 01 So shortly after we went public, the greatest excitement on the street where the company was concerned was the healthcare analytics business because it was growing much faster than the PNC business. It's just an interesting thing about you try as many things as you can to try to grow your business. But as I said before, by the time we got to the end of 2021, what I felt I was seeing, at least the perception of the company out in the marketplace, was your property and casualty business is so pretty and so optimized. Anything else you've got inside the company is just going to look like the redheaded stepchild.
16:06 Auren Hoffman? And by the way, it depends where you are in property and casualty versus the clear number one. It's not even a close number two in what it does, but there are other businesses where you don't want to get into being the number four player or something. So it also depends on like, yeah, it'd be great to have six businesses where you're all number one and two, right? Like, that would be amazing. It's just hard to do that.
16:27 Speaker 01 Yeah, and I think that in that second 25-year phase, we tried many. And I think what we proved was that it was hard to gain as much advantage. That five-point wheel, it was just a little harder to make that work in these other domains, maybe because of their structure, but also just the compare and contrast. Your point about are you known and what are you known for? I mean, when you've got something that is so strong and then you've got something which is not as strong, it just naturally draws attention to the thing which isn't as strong. And I concluded by the end of 2021, that around the time of my departure, that it just, the optimized business was a focused business. So the entrepreneurs that are listening to us today, you should think about, can you extend, can you change the surface area of your company?
17:12 Speaker 01 But be aware of the fact that in the end, what will get rendered is the strength of the thing that is maybe your core and then the strength of all the other stuff. And there just can naturally be a reckoning. I don't regret.
17:25 Auren Hoffman? And but also, like, you focus on, there's might be ways to focus more on the core.
17:30 Speaker 01 Exactly.
17:30 Auren Hoffman? And make that business way better. Maybe there's a way to expand it, maybe a way to grow it. As people like, if they try to diversify, then they're going to lose focus. It's like if you're an oil company and then you get the oil back in the seventies, bought all these financial services companies and stuff like that. They even own ESPN at one point, right? Yeah, maybe that was smart in some ways, but they were oil people. That's what they knew. Actually, their oil business started to decline because they were starting to focus on all this other stuff that was out there.
17:59 Speaker 01 Yeah, exactly. So it's kind of the both and quality of it in that in 2009, when we went public, if we did not have the growth associated with the mortgage analytics vertical, would we have gotten out or would we have gotten out equally good condition? And we enjoyed at least three years of real favor in the market because of the healthcare analytics business, not the PNC business, or at least not so much. So it's on the one hand, but then on the other hand, by the end of all things, you have to look at the relative performance of everything. And that's how everybody else is going to look at you, too. So it's, it's a both hand.
18:36 Auren Hoffman? Now, Verisk and Visa, they both started as nonprofit industry consortiums. What are some other examples of companies that made that transition from nonprofit to for-profit? What can we learn from those?
18:46 Speaker 01 Well, I don't know that there have been a great number of them, but.
18:49 Auren Hoffman? Yeah, I mean, obviously, there's like OpenAI right now, which has just done that or is in the process of doing that. They're probably the most famous one at the moment, which is making that transition.
19:00 Speaker 01 I always thought that a reasonably good analog were Visa and Mastercard. And ultimately, you're going to get judged on, are you actually creating value for customers such that they want to buy your solutions? And then if they do, then you're going to get rewarded for that. And then the performance that the market likes will follow. So what I would stress is if you started out in that not-for-profit mode, were you creating value for your constituents then with really good solutions that are being very competitively priced? Because you're not even trying to make money. So you're selling them essentially a cost. And are you really, really engaging with the industries you're serving in a way that there's trust and relationship building up?
19:45 Speaker 01 There was 25 years of that before I even showed up. And that's really valuable. So I really encourage our entrepreneurs, whether they're starting not for profit or they're starting for profit, just to lay that foundation. It's going to serve you well.
19:58 Auren Hoffman? Now, when you think of companies like Visa or Mastercard, they're sitting on this treasure trove of data. How would you advise companies like that to monetize it or think about monetization, that type of data?
20:09 Speaker 01 Well, they think about it. First of all, they definitely want to do it, but they are appropriately very sensitive to how their customers feel about them reusing the data, repurposing the data. And I would imagine in both cases that they have gone about as far as they can because they know that it's a good idea if they can. But the pathway to monetization is going to be to anonymize the data, tokenize the data so that it's highly secure, so that there's just no shadow across the idea of doing it. And then you probably need to think hard also about what's their incentive for you to monetize the data that they've given you access to? Do you need to provide them some incentive so that they'll lean into you doing it?
20:55 Auren Hoffman? Which could be giving them access to the insights or giving them price breaks or something like that?
21:00 Speaker 01 Or both, right. In the case of Verisk, the same contract in which a customer agreed to buy our solutions, they also gave us access, they gave us rights in the data in order to be able to create the solutions. And then when you're negotiating the contract, we were very open to discounting the contract, at least to a degree, in order to have or even expand data rights.
21:22 Auren Hoffman? Okay, interesting. Now, so many companies struggle with monetizing their data effectively. Do you have a framework for determining when data has some sort of genuine commercial value versus just when it's just quote unquote interesting?
21:37 Speaker 01 I would sort of walk to the other end of the question that you're asking, Orin, and kind of put it this way. Well, when I began my career, which was quite a while ago, if I had wanted to get a petabyte of storage capacity, probably would have cost like $8 million. I mean, the kind of thing that would rise to the level of the board of directors deciding to approve it. Now, today, if you and I wanted to get a petabyte of storage capacity, we could rent it in the cloud for what, $10,000? Wherever it's going at.
22:04 Auren Hoffman? I actually don't know the cost, but it's very low.
22:07 Speaker 01 Exactly. And so my starting point is, in this world that we're in, where it doesn't cost very much to first order capture the data, well, then grab the data. If you're in front of it, grab the data. Always be asking.
22:19 Auren Hoffman? Most people already do. They're already like getting the data. They're already, maybe not in a smart way or something like that, but they are capturing it.
22:26 Speaker 01 Where I would go to next, actually, is be capable of using it in a very high-velocity, low-cost sort of a way, and be capable also of bringing in the next data sets and putting them up alongside the ones you've already got.
22:43 Auren Hoffman? That I would say most, that's where most people fail down. Joining the data sets is non-trivial, and there's some entity resolution. Also, it means you have to understand not all your data is good, some of your data is bad, so then you have to be fixing the data.
22:57 Speaker 01 Precisely.
22:58 Auren Hoffman? Data-munching, et cetera.
23:00 Speaker 01 Yeah, so basically, create your data as flat and denormalized as you can, and then build really good data fabric in which you can actually reach in, pull it up, and begin to make use of it. If you're not spending very much in order to grab the data, and you're not having to spend a lot in order to be able to access the data, then to me, the premium goes to be in close touch with your customers, show them something that is novel, that you just came up with, see if it's starting to resonate with them, and then pay a lot of attention to, don't just back your truck up to their loading dock and give them this new synthesized set of signals, but understand what they do when they pull it off the loading dock into their own environment.
23:43 Speaker 01 And now, maybe you're really getting some insight in terms of what do they need?
23:48 Auren Hoffman? What are your thoughts on this? All these companies now have chief data officers. What's your thoughts on that?
23:53 Speaker 01 I think that a chief data officer can be very difference making at particular moments in the journey for a company. I would say especially if a company has not really given a lot of attention to monetizing data or is not yet really well organized to make use of it. Once it has it, then your chief data officer can be your point person.
24:17 Auren Hoffman? I'm in the data world my entire life. I have no idea what a chief data officer is, what it does. Is it just a glorified salesperson? Is it a glorified BD? Sometimes they have this huge org under them. Sometimes they have like one person under them. Sometimes it's some random free consultants in a company. Or I know what the head of engineering does, a person who like, or all the engineers, the chief product officer does. But what do they do? Why is it even valuable?
24:41 Speaker 01 Well, let's start with that. What's the purpose of the role? To me fundamentally.
24:44 Auren Hoffman? To me, the purpose is just like, oh, give someone who's smart a title. We got to give them a cool title with a C in it. They're kind of a glorified spokesperson or something and they sit on some panels and stuff.
24:55 Speaker 01 Well, hopefully not that, but I understand. I understand. I understand why you say that. Your data should be thought of as an asset. And there are a lot of people that are advocating for different things you're trying to get done inside your company. There are people that are advocating for the customers. The engineering crew is advocating for changing the processes or rewriting the software to accomplish a certain set of things. So you've got people that are advocating for different things. The chief data officer should advocate for the maximum value of your data asset. Should maximize the value of your data asset.
25:31 Auren Hoffman? So in some ways, they're the capital allocator for your data. Their goal is to, to understand that. Yeah.
25:37 Speaker 01 You get data. Many companies get data in a lot of different ways, which can be a little bit haphazard. In other words, some of it comes from you run your processes. Some of it could be that you buy it. But is there kind of a comprehensive view of, look, this is what we've got. And this is what it's worth today. If we added to it, if we did something different with it, here's how and why the value of it might go up. Somebody should be thinking about that. To me, that's fundamentally what the CDO ought to be doing.
26:06 Auren Hoffman? Now you left Verisk, I think, three or four months before ChatGPT came out. And so like the entire world in some ways has changed in the last two and a half years. How should companies be thinking about the world of AI now?
26:21 Speaker 01 So I don't think there are too many companies out there that would see themselves as being direct competitors of an OpenAI or of an Anthropic or any of the people that are kind of at the level of the large language.
26:35 Auren Hoffman? There's probably 10 or something.
26:36 Speaker 01 Right, exactly. I'll just side comment, by the way. Those foundational pieces are extraordinarily valuable. I do think it'll be interesting to see how much of the profit, the surplus that gets generated by AI methods flows to the providers of those foundational layers versus those who sit on top of them and essentially try to make use of them almost as tools as they try to sort of up.
27:01 Auren Hoffman? And how do you think, right now, there's roughly, let's say, 10 of these companies that are out there spending gazillion dollars each on building these models. Now, in some cases, one is the clear leader. Anthropic, right now, at least, is the clear leader for code. So if you want to write code, pretty much everybody builds stuff on top of Anthropic. The clear, obviously, leader for the consumer sentiment and using the consumer's OpenAI or maybe Grok's starting to come in and stuff like that. But if they're roughly the same, then I assume then they're just going to keep lowering their price, right, if they're roughly the same and then it gets out of the way.
27:35 Speaker 01 And the end around of somebody like DeepSeek, who basically went open source and then we work hard to tune it up. I personally think, now I'm biased because my work has been in vertical spaces, trying to be intimate with the data and trying to be intimate with the customers. But I really do think that a lot of the value here is going to be experienced by those who can kind of use those things almost as tools, as inputs, and then find a way to generate insight that's really valuable and not easy to get to. So it's kind of like small AI rather than big AI. And that's a lot of what I'm involved with today, the companies I invest in, et cetera.
28:16 Speaker 01 That's where they sit. So I, I would expect that the kind of disruption that came when DeepSeek showed up, I don't think that's the last time we're going to see something like that.
28:25 Auren Hoffman? In the data world, you have any resolution, that's something where AI can play huge, you have risk modeling where AI can play. Any type of models that people built before, where you built these very, very, very specific type of models. I assume that AI could probably make much better, allow you to do quicker. If someone spent like five years building this very, very, very proprietary model, they're going to have some competitor come at them.
28:50 Speaker 01 Yeah, no, I agree. And it's, I think that my own thesis would be that if you are a company that is already a leader in its space, and there's kind of a content richness to what you do, you ought to be able to amplify your advantages in this AI moment. In this AI motion, you ought to be able to. I'm not sure every company is. In fact, I think that many of the incumbents are trying to, I mean, they know they need to be doing something. You'd have to be brain dead if you're not doing something on AI. But I think a lot of them are still.
29:22 Auren Hoffman? Just like chief data officers, they all have chief AI officers too now.
29:26 Speaker 01 Exactly.
29:26 Auren Hoffman? I'm not sure what those people do, but yeah.
29:28 Speaker 01 Well, I'm not sure what those people do either. Right. But let me put it to you this way. When AI is doing everything that it's supposed to be doing, it just presents as software. It's like, oh, okay, that's automated and it does what it's supposed to do. And when you think of it that way, and you think about software chunking along, even if it's highly automated, we always worry about, is the human in the loop and in what way? So there has to be adaptation here where let the AI take the first draft and the second draft, and a human is in the loop somewhere, but take the efficiency of it for sure. And you mentioned code writing before. That's one of the strongest motions right now is code writing.
30:04 Auren Hoffman? I assume or any type of writing, in some ways, AI itself is just by default deflationary. So you should see it lowering costs in your organization, but because of that, your competitors are also doing that. So that means you're going to probably have to pass some of those savings on to your customers by lowering prices as well.
30:22 Speaker 01 Do you think AI will end up being deflationary at the national economy level? Do you think it'll kind of have that effect?
30:28 Auren Hoffman? I don't know if it'll be for companies that are in the data world where your production is data, your production is software or something that will allow you to do a lot more with less resources, with less dollars. Even the average engineer today is probably at least 30 to 50% more productive than they were a few years ago. Everyone is becoming more productive. So that's in itself deflationary. That engineer you spent a lot of money on is way better than they were. And you're not really spending that much more on that engineer today than you were two years ago. So I think we'll see scenarios where companies are going to, they're going to become more profitable in the short term, but then at some point you got to pass those savings on to your customer.
31:09 Speaker 01 Yeah, customers will expect that. I think you're right about that. And at the intersection of another thing that has been at work inside of our knowledge intense companies, and that is offshoring to human beings. So what happens with that motion? And then you overlay AI on top of that. I mean, it really is an interesting question about talent.
31:29 Auren Hoffman? I think the first place AI attacks is everywhere you've offshored. So I would be at least short term, short on India and Philippines and stuff like that. Cause I think those guys will get hit the fastest. Anything you're BPOing today is probably something where AI can do a decent job. And where if you're paying someone in the United States, like a gazillion dollars to go do something, maybe AI is just going to augment them, make them a little better. Whereas in a place like any, it could actually replace that person.
31:59 Speaker 01 Remains to be seen.
32:00 Auren Hoffman? You spent a lot of time in the insurance industry. It's gone through a pretty crazy past few years. I think 2024 was the first underwriting gain in a while. What's the overall state of the industry?
32:11 Speaker 01 I think it's always innovating underneath. If you're looking at it from the top down, I think that there is change which is occurring. But it's interesting, the insuretechs, the companies that started out as they basically felt that their major innovation was the ability to differentiate risk and to attract customers, the so-called insuretechs. They haven't swallowed the industry and they've been at it for a while now. In fact, some of them. Kind of have plateaued a little bit. So you had those kinds of changes. There's been a lot of speculation that maybe the capital markets would just step in and disintermediate. And what effect would that have on global reinsurers? And there's been some of that. There is some securitization, but there's still a role for global reinsurers.
32:56 Speaker 01 So I think that structurally it looks more like it used to and probably will continue to than not. The profitability that you were talking about, we had this real major effect coming out of COVID, which was inflation in a lot of factor costs. And so now you're trying to, so now you're trying to settle a claim. And wait a minute, how much does a two by four cost in Philadelphia? Wait, what?
33:19 Auren Hoffman? Totally right. And the labor, of course, yeah, has gone up so much.
33:23 Speaker 01 So I think when you're looking at 22, 23, you had a fair amount of that. But I think the industry has caught up with some of that.
33:32 Auren Hoffman? The industry is also just exiting markets. So in many markets, when I lived in San Francisco, I had one option. I had only one. Some of my friends, depending on where they live, in California or in Florida, they have zero options. They have no way of insuring their homes today.
33:47 Speaker 01 Well, there are the risk pools that the states put together as the insurer of last resort. The Southern California fires, the terrible fires that happened. I think when the story is all told, one of the really major parts of the story will be how many people were underinsured or not insured at all.
34:05 Auren Hoffman? Cause I think a lot of their insurance companies have pulled out over the years. I have a friend in Hillsborough, California, they built their dream home. And then they realized right after they built it that no one would insure it. So they were close enough to a fire zone that nobody wanted to take that risk. And when you have certain properties of certain sizes in terms of dollars, the state's not going to really come in either. So they're essentially self-insuring their own property.
34:30 Speaker 01 Probably in some of the higher risk places, people still like to run to the coasts. They like to be on the beach and build expensive houses there. So you're putting yourself in harm's way. You're consciously putting yourself in harm's way.
34:45 Auren Hoffman? Yeah, you consciously are doing it. It's just tough. It's really hard. Now, you recently became chairman of One Incorporated, which processes almost $100 billion annual payments across almost a million vendors. What's broken about how insurance payment flows work today?
35:03 Speaker 01 I just think that as an insurer, you naturally are giving a lot of attention to things like segmenting risk, selecting risk, pricing risk. That's the guts of what you do. So you naturally need to send money out in the form of claims. You also receive money in the form of premium payments. But if most of your attention was going somewhere else, then I think it follows that maybe your methods with respect to payments in and out just haven't modernized as much as maybe some of the other things.
35:35 Auren Hoffman? There's so many people who the insurance companies could only pay them by check. There's all these stories about how the check got sent to the wrong address or got sent to their old address, which was literally burnt up and stuff like that. They couldn't access that check for many, many weeks. It's just crazy.
35:51 Speaker 01 That's sort of the legacy of where we have been and the way we have been.
35:56 Auren Hoffman? I thought they were sending the checks because they were just trying to keep the float for as long as possible. So they're just being jerks and sending the checks. But you're saying they're sending the checks because they just don't have another alternative to send the payments.
36:08 Speaker 01 Also, you just want a feeling of security with your insured, just that they feel good about the form of the payment they're getting and et cetera. So if you're going to get the insureds to be, I'm thinking of consumers, to be comfortable with being paid in other ways, you may have to do a little bit of work. And for all of those reasons, focus on the modern methods, work the network, make sure that the endpoints are more than happy to affect desire to receive payments in these more modern ways. That's a lot of work. It creates the opportunity for a third party to come in and focus on just this one thing and really get it right and really get it optimized. That's why there's the One Inc., which is growing very, very rapidly because now it just seems to be a moment when a lot of the insurers are willing to open up
36:53 Speaker 01 to a new and different way to manage payments.
36:56 Auren Hoffman? Now you're now spending more and more of your time on boards. What are your thoughts of how to be a good board member?
37:03 Speaker 01 I would start with the thought that the most important work that a board member does is not in a board meeting, it's between board meetings.
37:11 Auren Hoffman? What is that? Is that getting to know the company? Is that understanding what's happening?
37:17 Speaker 01 Yeah, I think it's doing the work to really understand what the issues are. And then it is being available in a way that makes sense to the management team to actually talk things through. A board meeting is going to be more for sort of like ratifying decisions, basically, but actual deep work and thought. So we're trying to figure out how to reprice what it is that we do. Okay, sit in on the pricing meeting if the management team wants you and contribute your thoughts to that. Advocate for the company out in the marketplace. If you know people that could be customers or partners, do the work. Advocate for the company. So the most important thing is it doesn't happen in the board meeting.
37:56 Auren Hoffman? Mark Zuckerberg at Meta, he's got two boards. He's got his traditional board that we've all heard about. He also has a product board. Now, some of the board members on the traditional board are also on the product board. But these are very, very deep into product stuff and they do product reviews with him. And they're all outsiders who, and they spend quite a bit of time with the product. And his company, like product is the most important. You can imagine something similar in another company who instead of product, it's whatever is the most important thing for that company.
38:28 Speaker 01 And board members, there's literal governance kinds of things that board members do. But I think really helping the company to go. So Meta public, I would say especially if you're in a private company framework. You're not having to worry about much of the public governance things.
38:46 Auren Hoffman? Yeah, then you might as well just dive in and help.
38:47 Speaker 01 You're liberated. You're liberated. Work on the business. Help it to be a better, more rapidly growing, more profitable, better for customers. Like work on that.
38:55 Auren Hoffman? I'm sure at Verisk, you had a wide variety of board members from insanely helpful to maybe not that helpful. Was there something about the personality of the people that were insanely helpful that you could pattern match somehow? Like if you're recruiting board members, if you're advising CEOs to recruit them?
39:13 Speaker 01 Well, I think it's not particularly helpful in a board meeting when an individual board member will make a speech or speak ex cathedra and here's the thing that you should do. So first of all, those things I talked about before, like interacting between board meetings and the things really helping to get the thinking tuned up and advocating for the company, those are really important things. But I would also say the ability, when you're sitting with others and the ability to put out an idea, not as a settled conclusion, but more, have we thought about this or that? Now the CEO, she or he can really kind of take that on and take that away. And yeah, maybe we haven't thought
39:55 Speaker 01 about that enough. Thank you. You know, there's a way that you can push people without being pushy.
40:00 Auren Hoffman? Okay, that makes a lot of sense. Now, if you think these more PE-backed companies that are out there, I have this belief that like in the 80s, the best-run companies were the PE-backed companies. But now today, the worst-run companies are the PE-backed companies. They're just super bloated and stuff like that and bureaucratic. But you've had a lot more experience with PE than I have. Would you agree? Would you disagree? How do you think about PE-backed companies?
40:26 Speaker 01 Well, I think that a broad brush really doesn't fit. When you look at the league tables for PE firms, you can see substantial differentiation in performance. So let's start with that, that there's not just one monolithic way that it works. I think maybe the dividing line a little bit more is there are moments when a company could maybe really take off if there was a focused, thoughtful program of investment to try to build on or build things that aren't in place right now. And if that gets missed because the orientation is, well, let's make sure that above all else, that EBITDA is mounting rapidly. And so there's a strong orientation towards only cost takeout. Or another play
41:11 Speaker 01 that I think it's very reasonable to think about is price. Should price be raised? So those kinds of levers are naturally going to get thought of on almost all occasions. You're just naturally going to ask Those questions and there's no reason to resist asking those questions. But what I'm trying to say is if that's the playbook, but the company actually could be something much more than it is, which probably would lead to even more value increase. Let's make sure that we don't miss those ideas. So what's in the company's playbook? What's in the PE company's playbook? And I think there's different personalities to the different firms.
41:46 Auren Hoffman? Yeah, they're all different. A few personal questions. You're one of the small group of executives and CEOs and who talk openly about their faith. We talked about this a little bit on the last time you're on world would ask three or four years ago. How is that kind of faith informed your leadership?
42:02 Speaker 01 Well, I guess I would say a couple of things. So I'm a Jesus follower and I have been for my entire adult life. And I would say two things about that journey. One, I have spent countless hours sitting in groups of people. And we all have the same book open in front of us and we're all looking at the same words. And we're going over them and what does it mean? And then we come back to them and you come back to them again and again and again.
42:29 Auren Hoffman? Like a Bible study or these other types of things.
42:32 Speaker 01 Right. And also maybe some creedal things that you would say in the average worship service. But my point is I know the effect that it's had in my life to come back to the same big ideas, the same verses, the same words over and over and over again. It actually changes you. It changes how you think. I've always had it in mind when I had the opportunity to provide leadership, state the big ideas and then come back to them over and over and over again. That's why I felt so strongly about authoring something called the Verisk Way. I would open many, many meetings with people around the company by going over some part of the statements in the Verisk Way, because I know the power of it in my own life to go back.
43:12 Auren Hoffman? And even then, I imagine with a biblical text or verse or something like that, you have many people who might interpret it different ways. And then even yourself, you may interpret it different ways over your life.
43:22 Speaker 01 Right. So keep talking about it. Exactly. I'd say the other part of it for me is I went through a major change in my life when I was 23. I did not grow up as a Jesus follower. My family didn't go to church, but that change occurred for me.
43:36 Auren Hoffman? What was the story? How did that happen?
43:38 Speaker 01 So I was about a year out of college and everything circumstantially in my life was fantastic. I had a great job, great circle of friends, just everything was, I'd had a wonderful time in college, et cetera. And I just realized that I wasn't happy. And it was scary to me because it's like, wait a minute, if I'm going to get everything I want in life and I'm not going to be happy, then maybe I'm never going to be happy. So I started describing my condition to anybody who would listen to me, basically, but especially friends. And one friend said, well, if I got you a Bible, would you read it? And I said, sure. So he did and gave it to me and he put a bookmark in and a place in the Bible, the Christian Bible, that the four books which summarizes Jesus when he was on the planet,
44:22 Speaker 01 the fourth of them is the book of John or the Gospel of John. He put the bookmark in there. So I think I took it home that night and started reading it. And I thought, wow, this Jesus guy is just like the coolest ever. And that was it.
44:33 Auren Hoffman? You had never read the Bible at that point?
44:35 Speaker 01 I had never read the Bible.
44:35 Auren Hoffman? Oh, my gosh. Wow. Okay. So you really had grown up quite secular to not have read. Okay. Cause interesting that. Wow. Okay. And then after that, you just got kind of hooked you in or?
44:44 Speaker 01 What is it about all that that informed the way I led? So one was commit to the dominant ideas and go back to them over and over again. The other is to believe in the ability of people to be transformed. So invest in them, invest in their growth, in their development, believe in them when you give them a bigger job, stay close to them, help them. But having experienced a form of personal transformation myself, I just know that it happens.
45:08 Auren Hoffman? So many people also, they rise up when people believe in them. That's the missing ingredient that they need to believe in them.
45:16 Speaker 01 So leadership for me was and is very personal. It's very personal.
45:20 Auren Hoffman? Yeah, that's cool.
45:22 Speaker 01 Yeah.
45:23 Auren Hoffman? Almost the opposite question, but a question we ask all of our guests now. What's a conspiracy theory that you believe?
45:28 Speaker 01 A conspiracy theory that I believe. Oh, wow. A conspiracy theory I believe.
45:35 Auren Hoffman? You don't seem much of a conspiracy theory guy.
45:37 Speaker 01 I don't start out that way. I'm not looking around corners in that way. A conspiracy theory that I believe. Maybe if you have another question, something will pop in my head.
45:47 Auren Hoffman? It's interesting that you don't have one that comes. I probably have 50 in my head.
45:51 Speaker 01 You have 50?
45:51 Auren Hoffman? Oh, yeah, yeah. But I love it. I have all these crazy conspiracy theories about art and wine and lots of random things that are out there.
46:01 Auren Hoffman? Private equity, venture capital. I got them all. Last question, last question for all of our guests, which is also new since you were on, which is what is the conventional wisdom or advice do you think is generally bad advice?
46:11 Speaker 01 I want to say this carefully, but I think that there's a lot of advice which is given, especially to younger people, which is follow your passion. If you're not passionate about it, don't give it any thought. I understand where that comes from. And if your giftedness naturally leads you in the direction of fill in the blank, whatever it might be, that's probably a pretty good indicator that you might want to be interested in that. I've long thought that the three things that correlate with somebody being really happy in their work is they're good at it, they find it interesting, and they can attach some higher level of importance to it. It serves the world. I feel good about the fact that what I do shows up in the world.
46:52 Speaker 01 If passion means I naturally go in this way anyway, then you're probably going to hit at least the first of those three, which is I'm good at it. But I think that the issue that I have with that is, okay, now you've taken a job with whoever and you're in your 18th month or something, and somehow it's just not feeling completely right. Of course, you should keep your eyes open. Of course, you should look around. But the problem is, I think the follow your passion thing can also make a person a little bit inert. I haven't found it yet, so maybe I shouldn't do anything. Or I'm in a place that it doesn't feel like the greatest fit, therefore I'm going to kind of mail it in. That doesn't serve you at all. No, do excellent work wherever you are.
47:34 Auren Hoffman? Though there is something about to do excellent work, it's hard to do it unless you have at least some passion for the job and the work and the company and the people around you, or at least some sort of combination of those. If all those things don't really line up and you're only doing it for the paycheck, it's hard to do excellent work. Or do you disagree?
47:54 Speaker 01 I don't disagree in that you need to have all your energy available in order to be able to do excellent work. I just think that I'm looking for balance. I'm just thinking, because that advice tends to, I've heard it at so many college commencements, for example, and I just think unqualified, it doesn't actually provide enough guidance.
48:13 Auren Hoffman? Well, also, it's almost scenario where you probably have one passion and you should follow that particular passion. In some ways, it's almost should be opposite. Work at something and then hopefully you'll develop a passion for that thing. You may not know, let's put it on no. I can't imagine most people are like, I'm passionate about accounting or something when they're 18. But maybe later you actually do develop a passion for that and you do develop a passion for that. That's great. And probably within us, there's many, many, many things we could develop. There's not like just one thing that we could develop a passion for.
48:44 Speaker 01 If you think about it, if the topic is steward your own career, then one of the things that is true is we have a terrible information problem. Because of all the things that a person could probably do at any given moment, we know a small fraction of them. You just don't know everything that you could potentially go do. If you're trying to make a rational decision right now, you haven't seen everything. So something that it seems to me is constructive is be excellent where you are, keep your head up, keep looking around, be open to moving towards something that maybe is a better expression of what you're good at, what you find interesting, et cetera. It's not like you have to remain rooted wherever
49:25 Speaker 01 you are. We have a huge information problem. How do you know what you could do? You don't know everything that's out there.
49:31 Auren Hoffman? What do you do when no one's looking? What do you do in your free time or something like that? When I even think of my kids, my son, whenever he has downtime, he's in a big book. He's just reading the book. He's not playing basketball, not doing his instruments, or he'll do anything possible to not do the practice for his instruments or whatever.
49:51 Auren Hoffman? There's so many things he could be doing, but that's probably an area where he's probably more likely to be passionate. Whereas other kids, when they have free time, they're like shooting basketball hoops, or they've got something else, or they're building Legos, or whatever it is that they're doing, you start to see what they start to do. And then you have a sense, maybe more likely, their passion is somewhere in that route, somewhere in that world. All right, this has been amazing. Thank you, Scott Stephenson, for joining us at World of DAS. I follow you on LinkedIn. I know you're not really on X and stuff yet, but we'll get you there eventually. And I encourage our listeners to engage you there. This has been a ton of fun, great second time guest on World of DAS.
50:28 Speaker 01 Thanks for having me.
50:29 Auren Hoffman? Absolutely.
50:32 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, D-A-A-S, dot 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 a16z’s Martin Casado - building with AI Next Anduril Co-Founder Trae Stephens - Hard Tech & Modern Defense Strategy Page updated November 3, 2026 · transcript and summary are machine-generated