Pod Insights Summation with Auren Hoffman YipitData CEO Vin Vacanti - why hedge funds dominate data usage (and corporations don't) About this episode Vinicius Vacanti is the co-founder and CEO of YipitData, an alternative data intelligence platform that analyzes billions of consumer data points daily for institutional investors and corporations. The company reached unicorn status with a $475 million Series E from Carlyle Group in 2021. In this episode of World of DaaS, Vin and Auren discuss: Why hedge funds dominate data usage versus corporationsHow AI is transforming data processing costsThe durability of data businesses over 120 yearsWhy...
Transcript 10,556 words · 2 speakers
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 worldofdas, D-A-A-S, dot com.
0:25 Auren Hoffman? Hello, fellow data nerds. My guest today is Vin Vikant. Vin is the co-founder and CEO of YP Data, an alternative intelligence data platform that analyzes billions of consumer data points daily for institutional investors and corporations. They reached unicorn status with a $475 million Series E from Carlyle in 2021. Vin, welcome to World of DAS.
0:52 Speaker 01 Thank you. I've been listening to many, many of the episodes, so I'm excited to finally participate.
0:58 Auren Hoffman? Oh, this is awesome. Well, you're a data nerd, so I'm really excited to have you here. Now, your clients are some of the most sophisticated investors in the world. What do like some of these hedge funds understand about data that maybe a corporation hasn't yet figured out?
1:16 Speaker 01 Yeah, I think that's a really interesting question. Oftentimes, the question behind that question is, well, why aren't so many more corporations adopting data the same way kind of hedge funds do?
1:27 Auren Hoffman? Yeah.
1:28 Speaker 01 I think it comes down to not doesn't it's not necessarily that hedge funds are smarter in using data. It's more that data in and of itself has no value. The real value is if it solves an actual business problem that you have. And what's unique about the hedge fund market is they actually have an opportunity every single day to make a business decision on a security that they buy more, do they sell more, do they hold, et cetera. So every day, there's a decision to be made. And so for them, data can inform whether or not they should make that decision. Oftentimes, people believe, well, if I have this interesting data corporations, of course they will use it. But my question is, well, hold on. How will that corporation generate more revenue as a result of using that data?
2:12 Speaker 01 Or how will it make them much more efficient as a result of that data? And that's usually where people, well, I think it's valuable, et cetera. So I think it's less that hedge funds are more sophisticated in their usage of data. It's more that hedge funds have like a real business value from accessing that data in the way their corporations may not, depending on which end market they are in.
2:34 Auren Hoffman? Yeah, I would assume like most companies, like a B to C company can make decisions about pricing pretty often. Maybe back in the day, they'd make pricing decisions once a year and now they can make pricing decisions daily or sometimes discriminate versus, you know, Vin might get a different price than Orin or something like that. It seems like there's a lot of opportunities to do it there. There's a lot of opportunities to do it in marketing. There's a lot of opportunities to do it in labor. There's lots of things where like there's dollars there. But yet for whatever reason, it doesn't seem that people are using it at the same degree.
3:07 Speaker 01 I think that's a very fair point. I think it comes down to what are the three most important things that, take that example of a DTC company? What are the three most important things they're trying to do? Ultimately, it's usually in that example, it's marketed.
3:21 Auren Hoffman? Yep.
3:22 Speaker 01 They're trying to grow their user base versus like, you know, change their price by a dollar, down by a dollar. Now that can impact the lifetime value and all those things. But far more important to them is like, how do I just get a much broader consumer base for my product? And so it often does come down to where is their biggest business need right now?
3:41 Auren Hoffman? Yeah, we're working with one of the biggest quick serve restaurant chains in the world. And what they did is every week they would determine the store hours for every store in the world. And, you know, obviously the default would be to leave it the same. So if it was open at 7:00 AM, it'd leave it the same. But they would look at all the competitors, what they were doing. And maybe if a competitor was opening, they might move it to 6:30 to gain market share. They might move it to 7:30 to save. Money and get more profitable. They're constantly tweaking things like that. And then, of course, labor came into play. Maybe they had a labor shortage, maybe they had a labor gut, and they had ability to move things as well from the labor standpoint. So it's like a very sophisticated operation. And I probably met with at least thirty other large firms
4:26 Auren Hoffman? that no one else was doing anything even near to that level of sophistication. This is five years ago.
4:31 Speaker 01 Yeah, that's a really interesting point. Maybe I learned this actually on your podcast, but there's this notion that if you take any domain, any end market, there's like for every a hundred companies, there's going to be a few that are like extremely data savvy that see that as how they win.
4:46 Auren Hoffman? Yeah.
4:47 Speaker 01 But then the majority of them just don't. It's not how they see their competitive edge. They may be going crazy on the menu items themselves and like what kind of ingredients. And so I think there's also an element of like, not every company will win just through massive usage of data. Different companies will see themselves as having different strengths, where maybe their marketing strength and they're all over that. Or maybe, like I said, it's like the food quality. Some of them are on the data side. So different people will have where they want to focus on it.
5:16 Auren Hoffman? And true in hedge funds too, right? So if you're Renaissance or you're Point 72 or something like that, you're going to be investing a lot in data. If you're Citadel, you know, if you're Berkshire Hathaway or something and you're just like have a very extremely long duration, maybe data has no value at all to you.
5:33 Speaker 01 Yeah, that's, you know, it's a great point. Why isn't every hedge fund aggressively buying every kind of possible data set that there are? Because there are some that are really aggressive there. And see, that is kind of their right to win, et cetera. And I think it comes down to two things, which is one, the point we sort of just talked about, which is not every company sees data as the way they win. They may see it in other ways, other things that they do that's more unique to them. But I think the second was that when all this data became available, it became a real problem for the industry in some ways, opportunity and problem. Because you felt like, oh, do I need to use this data? And if I'm not using this data, am I at a competitive disadvantage relative to everyone else? And that's really where like our company came in and other companies. Companies like us came in, they said, you know what, it's okay.
6:19 Speaker 01 We're gonna go and collect all of that data, bring it under one roof. And then we're gonna make sure you have the best, greatest conclusions from that data. So that you can go back and do like the things you are kind of good at. And you don't have to worry about building some crazy data team to take all of this data. And I think you've seen that all over the board and all the industries. There are companies that have developed these great data assets. But to get to the true long tail, it often requires somebody else kind of doing all of the work for people. And then serving up the full meal versus just providing the raw ingredients.
6:53 Auren Hoffman? And one thing in the pipeline, this might be like later on, is joining disparate data sets. So you can ask like deeper questions of the data. How do you think about that? And how do you think about these different join keys to do it?
7:08 Speaker 01 Super important. It's often the case that the customer, they don't care about the data. They care about solving the business problem that they have. And you're like, well, I can't, I can solve part of it because I have one data set, but I don't have, they're like, I don't care. Get it. And so our customers on the fun side, they want to know the most there is to know about Netflix or about Uber or Amazon or Tutof Surgical, or whatever the company may be. And it is our job to go get all relevant data sets, bring it all under one roof. And then bring it together and deliver our best insights on that company based on all the data that we are sort of observing.
7:45 Speaker 01 And I think it's one of the challenges that a lot of data businesses have had in terms of scaling their business is that if you're just providing one of those kind of raw inputs in. It's not that valuable to them because you're asking them to do a lot of work to take your raw input and combine it with a bunch of other raw inputs to ultimately get to some sort of insight. And so I think it's really important, our perspective is we're not here to sell data. We're here to solve their business problems and it's our job to make it work on the backend.
8:16 Auren Hoffman? And data itself is also like super messy. Like a lot of the data sets that you're buying or using, I assume have lots of errors in them. They have bugs. There may be, there may be fraud in some of the pieces of the data you have to root out. You know, there's lots of mistakes in all the data sets. So even some of the best data has lots of mistakes in it and other types of things. Like, how do you figure out, like, okay, we ought to get as close to the truth as possible, but nothing completely predicts the past. So how do you kind of figure all that out?
8:51 Speaker 01 Yes, great question. So the problem with the data that we work with, which is alternative data, is it's not intended for this.
8:59 Auren Hoffman? Yeah.
8:59 Speaker 01 It's, it's kind of exhaust data, and so it's extremely messy. And as a result of that, it takes a tremendous amount of effort to turn that data into something that is a reliable signal, insightful about a business. And so the problem we've always had as a company is we've always had data on millions of companies. And the crime of our company is we would just throw all of that away and focus on just a few companies. Because the cost of processing that data for each company was so high. Because the strings, you don't know how to match it to the right vendor, to the right SKU. The data can be off, there could be gaps in the data, et cetera. So it worked for us to pick stocks.
9:40 Speaker 01 Because if we could really do a great job for Uber, people would then pay us a lot of money for a subscription to just the Uber product. And so you can think of it as like, and I know you love charts, you can think of it as a chart where there's like a certain return on investment for each company. And if you do public tickers, the return is very high. The moment you go off of public tickers, the return gets really low.
10:05 Auren Hoffman? Yep.
10:06 Speaker 01 And so the ROI was never there for processing data for the millions of companies. And we only did like 500 companies, 1,000 companies. What's changed all of that is AI. And so now, using L L M's, you can start tagging this data at scale in a way that would have taken a human kind of in the loop mechanism to do that. That's what's changing about our company. You know, today, we are the source of truth for real time company performance data for a bunch of stocks. Like, if you want to know right now, how is Uber doing Netflix, Amazon, there's no better place than us to provide that estimate. And that's why our business exists. We are on a mission to become that for every company, not just the public tickers.
10:50 Speaker 01 But we could have never done it had it not been for AI, lowering the cost of cleaning that data and making it reliable across millions of companies. So that's why we're kind of really excited about this next journey for the company.
11:02 Auren Hoffman? One place for data, which is somewhat interesting, but maybe isn't a massive market, is like, understanding my competitors, understanding the competitive set. And everyone may define that differently. If you're Starbucks, it might be everyone who sells food and drink. So it might be a very, very broad set. For something like Uber, it still could be pretty broad, because obviously, there's Lyft, there might be different transportation services, there's DoorDash, or these other kind of things that are out there. You know, maybe ultimately, they even see like a Tesla as a competitor and stuff like that. But there hasn't been a lot of money in understanding my competitors, because it's somewhat nebulous. Have you been able to like, create products around that?
11:43 Speaker 01 Yes, great point. So you're exactly right. I think anyone who thinks about data is like, oh, my God, it's amazing. You're going to be able to see how you do versus all your competitors. But if you ever actually try to sell that to somebody, they're like, okay, you know, I guess, kind of interesting. I guess I lost a little share here, but it's because of this and this, and it doesn't really matter. Plus, they already know what their competitors are doing, because they talk to lots of people and all that stuff. So it's not something that people are willing to pay a tremendous amount for. So it's kind of hard to build a business off of just straight up kind of competitive intelligence. That's a like, thousands, maybe small tens of thousands kind of opportunity. And it's really hard to have the high churn business and all of that kind of stuff. That's why I kind of go back to this beginning where I kind of the example we give at the company is we always talk about.
12:29 Speaker 01 When Google Earth came out, I remember it coming out and I thought it was the coolest thing ever. You were like a satellite, you could go to the Eiffel Tower, your house and all this stuff. But then at the round, literally the same time, Google Maps came out.
12:40 Auren Hoffman? Yeah.
12:41 Speaker 01 So much more boring. It was like, it was just like a, you know, like a map or whatever. Guess how big Google Maps is versus Google Earth today?
12:47 Auren Hoffman? It's crazy.
12:48 Speaker 01 That's because Google Earth doesn't solve like a real problem. You don't actually need to be a satellite on a daily basis to go around and look at things. You need to get from place A to place B all the time. And so Google Maps is far bigger. So I always sort of talk to people when they're kind of building data businesses. Don't get caught building Google Earths.
13:06 Auren Hoffman? Yeah.
13:07 Speaker 01 The thing I always love is people always put data on a map. Look at all this data on a map. And I'm like, so what? What am I supposed to do with this map? And so instead, go find an actual business problem that they have that this data can go solve. That's going to cause them to put real dollars in their pocket as a result of using that data. And so that's why I like competitive intelligence is not like a great market because it doesn't actually put dollars in their pockets.
13:35 Auren Hoffman? It's also not sure, like, what do you do with it? It's not as actionable.
13:39 Speaker 01 Exactly. The actionable, do they take a business decision the very next day because of what you're putting in front of them?
13:46 Auren Hoffman? And do you agree the most valuable data is the temporal data, it's data that's changing?
13:52 Speaker 01 No. I mean, there is data out there that is extremely valuable, that is largely historical data. So for example, take Zoom Info. Having contact information for all of these business professionals is extremely sort of valuable. The fact that, you know, one week later, how much has it really kind of changed? Just a little bit, but it's the corpus of that is very valuable. A lot of this stuff, think about AI training and Reddit data and how valuable all of that corpus of Reddit data is. The fact that the very next month, some new information comes out is not as kind of valuable to it.
14:33 Auren Hoffman? World of DAS isn't just a podcast. World of DAS membership is a private invite-only community for founders and executives of important data and AI businesses. It's a place for off-the-record, high-trust conversations focused on the biggest questions data leaders are facing. The community is about peer learning, curated intros, and access to private events. Everyone in the group is building or leading something meaningful in data. If that sounds like you, apply to join at worldofdas.com. That's world of D A A S dot com.
15:14 Auren Hoffman? Well, but do you think it's really that valuable? Like, Google paid a lot of money for the Reddit data for two years, right? I think it was, let's say, $60 million, and they got it for two years. When that ends, do you think they're still going to pay $60 million for it? Or is it going to be like $6 million for it?
15:32 Speaker 01 Comes down to the value. There was a tremendous value in the historical data and training these models.
15:37 Auren Hoffman? For sure.
15:38 Speaker 01 There is some value because the world changes. New products, new opinions, new things. But you're right, that it doesn't feel like that month of data is worth as much as like the historical data was worth. And so from that perspective, I agree with you. Now, for us, you know, a lot of people look at our data. You know, our data's value isn't the history of the data. Our data's value in the fact that we're going to see it next month. Because if company performance data changes dramatically, everyone wants to know what's going to happen next month, not what happened in the past. And so for us, the temporal data is very valuable.
16:13 Auren Hoffman? Yeah, some of the data that you get and lots of other people get are get by like crawling sites. And there's this kind of like cat and mouse game of crawling, and there's cloudflare, and there's all these kind of like anti-scraping, anti-crawling. How do you think that converges over time?
16:29 Speaker 01 You know, I used to stress out about that a lot because we originally started, we don't do, it's a smaller part of our business today, but we started by collecting data from the public web.
16:40 Auren Hoffman? Yeah, and crawling like eBay and Amazon.
16:43 Speaker 01 And I used to really worry that those companies were going to make it harder and harder for people to sort of access them. The reality is that there's so much business value in collecting that public web data that as hard as they try to make it on that side, there are companies that have been created that are just as entrepreneurial and just as innovative to help other organizations make sure they can go get that data. So there is this thing where like, the harder they make it, the more of a business opportunity there is for people to go help other organizations and we're not, you know, it's like millions of companies are collecting kind of public web data.
17:20 Auren Hoffman? Have you heard of this company String?
17:22 Speaker 01 No, I have not.
17:23 Auren Hoffman? Oh yeah, they're like an amazing crawl company and they get so in the weeds of just that one problem, yeah.
17:29 Speaker 01 Exactly. So that, look, if there's business need and it's legally not an issue to collect that data, then ultimately the companies on the other side will also rise up.
17:40 Auren Hoffman? No, most of the really big, well-known data companies were started over 40 years ago. You have the Morning Stars, the Bloombergs, the Standard and Poor's, you know, all these companies that are out there. And in some ways, like you're competing with many of them, maybe all of those ones I mentioned. How do you think about those companies?
18:02 Speaker 01 One of my favorite things to say about data businesses is if you look at a bunch of businesses 120 years ago, and are they around today, you're going to find that data businesses far outperform every other category of business that is still around.
18:18 Auren Hoffman? Dun & Bradstreet, right? Abraham Lincoln worked for Dun & Bradstreet, right?
18:22 Speaker 01 Yes. Oh, you took my quote. I was about to say that. That's my favorite quote.
18:26 Auren Hoffman? Okay, I'm sorry. Like, that's, that's insane, right?
18:29 Speaker 01 Also Moody's was a part of Dun & Bradstreet. You know, that got kind of spun off in 2000, et cetera.
18:35 Auren Hoffman? Oh, I didn't realize that. Okay.
18:36 Speaker 01 These, you know, Standard and Poor's has been around for forever, obviously Reuters, et cetera. So these businesses are around forever.
18:43 Auren Hoffman? They're very Lindy essentially, right?
18:45 Speaker 01 Yes. What was very fortunate for us was that when we started YB data, this didn't seem like a big market, alternative data. We were doing web scraping and like selling specific tickers to investment funds, et cetera. So it kind of became, it was kind of like a little snowball and it kind of started rolling downhill and getting bigger and bigger. And so now we just, you know, we have a bunch of data. We've built our own panels. We have these exclusive relationships with a lot of other companies, et cetera. So now it's not so easy for these bigger companies to just launch kind of, and none of them have. There's like a very big barrier to entry to this space.
19:23 Speaker 01 While it is huge, it's not so huge because these companies are so big that for them, it makes like a big, like they need huge revenue to kind of move the needle for them. So we've kind of nicely been in that slot of under the radar, but keep getting bigger, keep getting bigger. And so now it's kind of like too late for them to kind of enter the space because there's just too much that we have done over the last 15 years.
19:47 Auren Hoffman? And if you think these companies like S and P, Dow Jones, Reuters, they've really been built by M and A. They're very good about buying companies for, you know, X times EBIT. And because of the efficiencies, they can make the company more efficient. They can also stuff the channel. They've got great salespeople. You know, even though they're buying at X and the X might be high, like they can grow that EBIT really fast on both sides of the operation. Have you thought about using M and A as a lever now that you're starting to be at scale?
20:18 Speaker 01 Yeah, for sure. Well, you know, on that point, I think the part that people sometimes don't give them credit for, it's not just about efficiency. It's not just about the distribution. It's actually a lot about the data that they acquire. They get to integrate that data in a bunch of products that they have that make those products better. In some ways, that is the real reason that acquisition makes so much sense.
20:44 Auren Hoffman? So every time I buy a company, it's making their other other products better as well.
20:49 Speaker 01 Yeah. Data businesses are like that, which is it's really about getting more data into the organization and then improving as many products as you can, which gives you more customers, more breadth, and then you go around and you buy other companies. You know, we acquired one of our largest partners back in 2021, which is this business Edison. And that was great for us because it kind of gave us the ability to work more closely with them, you know, and build out that panel and collect, be able to parse the data better and all of those types of things. And so we're always interested in potential kind of acquisitions, but it has to make sense in the sense of, is it going to make our existing products better versus just another company where we're going to try to drive it through our
21:32 Speaker 01 sales platform and those types of strategies?
21:36 Auren Hoffman? It still seems that data is not the most understood of business models. It's not like everyone understands SaaS and all the different metrics around SaaS or something like that. When you talk to investors who are buying data businesses or thinking about investing in data businesses, what do they most misunderstand?
21:55 Speaker 01 I have a huge gripe about this, and I think they misunderstand three things. First, this snowball effect, which is as the data asset gets bigger, your ability to improve all the products, which brings more customers, which funnels it right back into the data. And so data businesses historically don't grow that fast. They're not the company that's going to grow 200%, 300%, but they're consistent. They just keep growing and then the barrier to entry on those data businesses are very strong.
22:23 Auren Hoffman? Yep.
22:24 Speaker 01 The second thing is when you think about AI, software is actually now getting really hit by this notion that if you think about a business, there's data, there's processing, and then there's distribution. And AI is really starting to chip away at like the barrier to entry on the processing of data.
22:42 Auren Hoffman? For sure.
22:43 Speaker 01 Which is for us, an opportunity because it allows us to make our data more actionable, like what we were talking about with the using AI. And so that's the second thing I feel like investors don't understand about data businesses is they're actually very AI resilient because the AI models are only as good as the data that they have access to. And if you're a data business, then you're sitting on data that is, as long as it's not available in the public domain, it's hard for an OpenAI or an Anthropic to kind of replicate your business, hard for a three-person company to just use LLMs to create workflow solutions, et cetera. The third is around costs. When you think about the accounting system that was created, which is GAAP and all that stuff, the way it kind of works is there's this concept of gross profit.
23:28 Speaker 01 And it's like you sell a good for $10 and then what did it cost you to kind of produce that good? Let's say it cost you $5 to make that shirt and so your gross profit is $5. And a lot of people wonder, well, what's the gross margin and how's the gross margin scaling and all of those things because it's almost like a proxy for EBIT, for like long-term EBITDA margins. But it's also a sense of like, is this a good unit economic business? The thing about data businesses, which is unique, is the COGS don't grow as the revenue grows. And so you actually spend a ton of money creating the data product, but then the incremental customer, all of that falls to the bottom line. And so when you look at, you can have data businesses when you first start that have
24:10 Speaker 01 they have negative gross profits because they have all this fixed costs in producing their data, but that gross margin will scale amazingly well. So early on in our, in our kind of existence, our gross profit wasn't that great. Today, it's gotten so much better because of this scaling factor, but then if we go and get a new data set, it bumps it back down. And then we have to kind of scale that to kind of get it back to being a great place. And I feel like investors are just like, they don't get that notion that there is a fixed cost.
24:41 Auren Hoffman? Well, also, it just depends on like where should you put it, right? I remember having this big argument with one of our auditors. They wanted to put all this stuff in COGS. It's not COGS because if I had a SaaS company and developing software, you wouldn't put it in COGS. It's an asset that I'm building and I'm selling many, many times, right? You just put that in my general operations of my business. It's either everything is right about data businesses, but everything is wrong about SaaS businesses because they should be putting more things in COGS or the opposite because like all their assets that they're building, they're building this like amazing workflow, but none of it like their gross margins are 90%, quote unquote, whereas the data business is the same thing. They're building this like amazing data set, but they have to put that in COGS for some reason.
25:22 Speaker 01 Yeah, so my advice always to people in the data business is really focus on this. Get it right. Have your story right around what is a fixed cost. Show the very, the unit economics of an incremental customer and how much falls to the bottom line. For the data businesses, COGS are sometimes fixed. Most time variable for data businesses are almost entirely fixed. And so like you should show how the gross profit like can increase dramatically as the number of customers grow for that business.
25:50 Auren Hoffman? Yeah, that makes a lot of sense. A lot of data companies do struggle to scale at a certain point. Their growth basically gets to flat at some point. And this, by the way, it's true in software companies too. A lot of software companies, there's a certain point where they just struggle. They, they hit their TAM. What should companies do at that point?
26:07 Speaker 01 Three things. First, they have to be able to find new end markets for their data. They can't get caught just selling to one end market and that's hard. It's hard to go and source a completely new end market and figure out how to retrofit your data to kind of work for them.
26:23 Auren Hoffman? Especially because probably the first end market is the one that would pay the most for that data. Right? If you're selling into hedge funds and then like the next market you do, they might not want to pay as much for the data.
26:34 Speaker 01 And look, it depends. If the data is new, if it hasn't existed before, it's possible that there is a far bigger market that you're not aware of.
26:43 Auren Hoffman? That's true.
26:44 Speaker 01 Because the issue with the hedge fund market is there aren't 20,000 hedge funds to kind of go sell to.
26:49 Auren Hoffman? Yeah.
26:49 Speaker 01 But in these other markets, there are tens and twenties and 30,000 companies to kind of go sell to. So the first is an end market. The second is you should really talk to your customers and say, okay, you're using our data for what? Because like I said, no one wants data. They want a solution to their problem. And so for instance, I'm going to give a stupid example, but maybe they're using the data for location selection. So then you should try to understand, well, what are they doing exactly? Once they get our data and can I do that part too?
27:16 Auren Hoffman? Yeah, and they might be marrying it with three other data sets. They might be doing this other thing. They might be providing some sort of data science on top of it, et cetera.
27:24 Speaker 01 And nope, I find you go up the stack, you get 10 times more money. And then the third thing is a lot of people are like, wow, I can only have this one data set. You know, I'm not going to use any other data set. This is what I sort of specialize in, and I think that's a mistake. Because ultimately, that's the other way to not get caught is start adding incremental data sets that when combined with your data set, make the product better. And then some of those data sets can launch completely new end markets. And I think people sometimes get caught thinking like, well, I'm only ever going to do this one data set. Start looking for tangential data sets to marry to your existing data set.
27:57 Auren Hoffman? Now, most of our companies in the data world, AI has significantly reduced their costs of creating the data. And you were mentioning before that has allowed you to go much longer tail in the number of companies. But I would say at least so far, these companies have not yet passed that cost on to their customer. So for most of our data companies, their EBIT has grown, their margins have grown, et cetera. These are competitive markets with probably many competitors who are also investing in AI. How do you think that changes over time? Like, are we going to start to see some price compression over time with some of these data companies? If the cost of the creating that data does get cheaper over time?
28:40 Auren Hoffman? Or are they going to try to layer more services to keep the price where it's at? Or how do you think that kind of changes?
28:46 Speaker 01 Well, no one's going to lower their own price unless they're doing it to attract like a new logo from like a competitor. And so some of that was definitely going to happen. But I think on the AI side, you know, a lot of people are very focused on the efficiency side, but the far bigger opportunity is the ability to develop new products using AI that historically you could not. While an EBITDA margin percentage, you know, the rule of 40 is you add revenue growth to EBITDA margin, the reality is the market doesn't value those equally.
29:16 Auren Hoffman? The growth is at least 3X on the other side.
29:19 Speaker 01 Exactly. And so the advice I always give to people with these data businesses is yes, there's efficiency there, but think of like, how does AI, by lowering the cost to process this data, what else could you be processing with this data that would allow you to develop completely new products?
29:35 Auren Hoffman? Yeah. But price is a weapon. I mean, you can grow your business. If you're selling the same thing as somebody else, I mean, obviously, this is hard to do an apples to apples comparison on data, but if you were selling the same thing at half the cost, it would be a lot easier to sell it.
29:49 Speaker 01 It would be. I think the bigger opportunity there is less about selling it to the existing customer base, obviously at a cheaper price to win share, more by lowering the cost to produce it, does it open up a completely new market that historically would have never paid you 30,000 a year, but they'll pay you 5,000. And then all of a sudden now you've got 100X the potential customers to go after at a much lower price.
30:15 Auren Hoffman? One kind of a data business, which I always think is interesting, I'd love to get your thoughts on, is the expert network business. I'm sure you've thought about that a lot over the years. Like, how do you think that's evolving?
30:25 Speaker 01 So the expert network business, amazing business that they created, extremely profitable business that has a network effect in terms of you build out this network of experts and then funds, but corporations, they need expertise and it's really painful and they reach out to you, you find them an expert, it's already in the network, you kind of spill them, you get a cut of that, et cetera, that makes a ton of sense. And then it got very competitive.
30:51 Auren Hoffman? Yeah.
30:51 Speaker 01 Because anyone could build an expert network in some ways because I'll find you experts because a lot of it is just hunting people down, et cetera, LinkedIn, et cetera. But then Tegus showed up and Tegus, it was genius what they did, which is they said, you know what, we're going to do the calls at cost because the real value is can we convince you to allow us to publish the transcripts? And that was super interesting because all of a sudden people were like, sure, I'll do that. And then they built up the library. It was almost like they used like a different business model. It's like a, it's a disruptive, you know, technique, which is you just change the business model.
31:24 Auren Hoffman? Right. Again, they use price as a weapon.
31:26 Speaker 01 And it was price, but it was really, there was another, they started to build a network of data as a result of that. I think what's interesting is in this world of AI, AI is only as good as the data has access to. You can kind of assume AI will get access to anything that's in the public domain. These expert networks are actually sitting on an interesting opportunity because they have data that's not accessible to AI, which is all these calls and all this expertise that is getting poured through. And I think that is an interesting angle in terms of creating unique proprietary data. And so when you look at some of these large financial services companies, they're starting to be like, oh, what's our unique proprietary data? And I think expert networks are sitting on an interesting, potential, unique proprietary
32:11 Speaker 01 data set if they can convince their customers to allow them to kind of, you know, start publishing these transcripts.
32:18 Auren Hoffman? One interesting data set that, you know, people have talked a lot about, you know a lot about is credit card and debit card data. Outside the US, what other markets have robust credit card and debit card data?
32:31 Speaker 01 I think you have some stuff in Europe. I think there's some stuff in APAC. It's hard. The, the real challenge is there isn't as big a market need as you would think. Because when you look at like the public tickers, which is hedge funds really drive the majority of the demand for credit card panels. You know, most of the investable companies are in the United States. There's not that many in these other regions.
32:56 Auren Hoffman? But wouldn't you want to know like how they're doing? And some of them are global. I mean, if you think of like Netflix, Uber or Starbucks, or, you know, you just go down the list, like, wouldn't you want to know how they're doing in these other markets?
33:08 Speaker 01 For sure, but the cost of building out that panel,
33:12 Auren Hoffman? Yeah.
33:13 Speaker 01 It may be a little cheaper, but not that much cheaper to do it in these other regions. It's still pretty painful. And so then you have to say, what's the ROI? If the ROI, if there were a bunch of publicly traded companies in those regions, then it would be interesting. But if it's just one of the 74 markets that Netflix is in, you know, it doesn't really move the needle to know that one.
33:35 Auren Hoffman? But then why does the US move? If the US is 30% of the Netflix revenues or, you know, somewhere in that range or something like that, why does the credit card data even matter that much? If it's fluctuating, if it's just US where it's like for Lyft, where, you know, the US market might be 98% of their data, you know, their, their revenues or something like, maybe that makes more sense if credit card data is, is actually indicative of their performance.
34:02 Speaker 01 I don't think it does matter that much for Netflix. I think you gotta have like a global perspective. I think the real thing with Netflix is look, a lot of people use Netflix. And so really what it comes down to is how are they upselling the Netflix customer basis? Like how are price increases being handled? You know, how's the gaming stuff going?
34:20 Auren Hoffman? Yeah.
34:21 Speaker 01 It's much more like a competitive kind of thing. It's less about predicting like in the US, the exact top line. It helps a little bit and there's still some interest. You know, if they fall all of a sudden, if they stop growing, if they decline, those are noticeable things. That's news, but you're right. It's not as, it's not as valuable.
34:38 Auren Hoffman? Yeah, interesting. Now, I remember you taught me a story how like you famously taught yourself to code after, you know, you had some outsourced developers back in the day that didn't work out. And we'll link to it in the show notes. You wrote extensively about this. Nowadays, you know, we have all these like AI coding tools, right? That are out there. You've got great UI coding tools like Bolt New or Replit or Vercel and stuff. I imagine it's even like easier to get up from zero to one on coding today. Like what would be your advice to non-technical founders?
35:13 Speaker 01 You're exactly right. Basically it's been getting easier every year. Like AWS, when I first started, I didn't even have access to AWS. Like I was doing a server on one-on-one and it was just like a disaster trying to manage that. I had no idea what I was doing.
35:26 Auren Hoffman? Even just putting an environment on your laptop, even just a few years ago was super hard to do.
35:32 Speaker 01 Yeah, I remember we had these servers running our website at the time. And I had one, if the server went down, I had one command that I would run that would boot it back up. If that command didn't work, I was like, well, it's over. I literally don't know anything else to do if this one thing doesn't work. So it was way harder back then, but now like, you know, like putting something in the App Store. If that was painful, but now that's gotten easier through all of this type of stuff. Some people say, well, I don't need to learn to code at all, and I'm not sure about that. Because the problem is it's great for building like a simple prototype. But what you'll find is as the prototype gets a little bit more sophisticated, and then you actually start having real people using it, you go to change a thing and like the whole thing breaks.
36:13 Auren Hoffman? Yep.
36:14 Speaker 01 And you're like, whoops, what happened? And then you undo it and then you go to change it again, it breaks again. So it is important that I think you have a decent understanding of what's going on. But it makes it so much better in terms of not having to know all the crazy syntax and a lot of like weird stuff, et cetera. But there's also like some security issues. If you put something out this way, like it's probably you've exposed your whole database and you have no idea, et cetera. And so I think it is important that you still kind of learn the technical basics to this. But I think it's critical for companies where the product is really the thing. If you were Groupon, you didn't need to learn to code. Like you can build a little website that can collect people's payments and that's very straightforward.
36:57 Speaker 01 It was really a sales and marketing challenge. But if you're building a product for consumers or a workflow product for businesses, I think you gotta build that yourself as a founder. Because if you're not, and if you're just outsourcing that, it's never gonna work. Because you need to be iterating like on an hourly basis. It's like, oh, I want to change this. It can't be, okay, let me reach out to my developer. They're gonna come back in a month and change it. You need to be able to type it in and see it reflected in the next 30 minutes.
37:24 Auren Hoffman? Yeah. And even what I found is it used to be, we're talking like a little over a year ago. If you're a product manager at a company, you would like mock something up and you'd take it to some engineers. And then you'd work with them on like a V1 or something. And then you'd get somewhat workable demo. And then you talk about it internally or something. Nowadays, the product manager does all that themselves. They can literally get to at least the demo. No, it's not hardened. It doesn't have the security. It's got tons of other issues, et cetera. But they can literally get to a super visual, very beautiful solution right off the bat. And then they can go to the engineers. Like I say, hey, make this better. That product manager has so much more powerful than they've ever were before.
38:08 Speaker 01 You know, it's funny you say that. I think that's true everywhere. The engineer is a lot more powerful than they were before.
38:14 Auren Hoffman? Right. That's true. Because also I found like a lot of engineers I know more like backend engineers. And they never could do frontend. And frontend was like magic.
38:23 Speaker 01 They know.
38:23 Auren Hoffman? They're like, I don't do that stuff. And now they're like, oh, I can do it. They just like get it done immediately. You're right. Each person now can do more themselves than they could do before.
38:33 Speaker 01 Yeah, that's exactly right. And I think, look, it's gonna keep going. I also think it's why it's such an exciting time for entrepreneurs. Because the hardest thing as an entrepreneur is like the why now. Why now? Because you're also just doing the same thing everyone else has done. And there's such a great why now. And companies, blank AI, we'll try it. It doesn't matter what it is.
38:52 Auren Hoffman? Totally.
38:53 Speaker 01 People are afraid not to try it because if they don't, they get left behind. Now, of course, there will be retention issues with all these companies that go from zero to 100 and then back down to 10 after three years.
39:03 Auren Hoffman? Yeah, we're seeing that in our companies as well.
39:05 Speaker 01 Yeah. So people will try anything, but then are they going to really retain the product afterwards is the real question.
39:11 Auren Hoffman? One of the things that's happening, I think, at least in the market, whenever you see these high growth things that are happening, at least for our companies, they're growing so fast. Is they almost always that means there's like four amazing competitors are also growing so fast. And then their products are also amazing and also getting better on a weekly basis. And so the customer might be moving, if they made a decision to buy you six months ago, that probably was the right decision six months ago. That may not be the right decision today, given like how much these products have diverged over time.
39:42 Speaker 01 And I think the other challenge is what's the moat that they're building when they, when a customer starts using them? How easy it is for that customer to just switch to the next hot company that comes along? It's hard to build a moat when you're going from zero to 100 because you're just kind of like, you know, you're not integrating deeply into their environment. And there's no sales process that kind of someone else has to now go through, et cetera. And so like, that's the thing I would definitely encourage. Get very aggressive with building out a moat for the business.
40:11 Auren Hoffman? One of your biggest shareholders is now Carlisle. What have you learned from them?
40:16 Speaker 01 Yeah, for sure. Honestly, we were traditionally, we were like a VC backed company until Norwest invested in 2019 and then Carlisle invested in 2021. And we've learned a tremendous amount from both of them, which is about running a great business. Because, you know, previously with VCs, it's very much just, you know, you want a 10x, you want to get as big as possible, as fast as possible. And they really kind of came in and told us the power of running like an efficient business. And the reason why that's so powerful is when you run a business efficiently, that means you have extra cash that comes from a customer sort of signing up. And what can you do with that cash? You can self invest that cash into the business, which for us meant let's go sign more data partnerships.
41:01 Speaker 01 Let's grow our panels, et cetera. And it kind of created like a much more accountability around how efficient we were being. I remember when Norwest came in, we were spending all this money on AWS. They, you know, they were like, hey, you guys should be more profitable than this. And then we were like, okay. And then we looked at our AWS thing and then three months later we had cut it in half.
41:20 Auren Hoffman? Wow.
41:21 Speaker 01 Just through efficiency. And like, there's a lot of businesses right now that are listening to this podcast and they've never once really tried to make their cloud cost more efficient.
41:30 Auren Hoffman? Yeah.
41:30 Speaker 01 And if they did, they would instantly find like just free money coming back to them. It's actually one of the reasons, you know, one of the products that we built for our own internal purposes was this notion that we had all this software subscriptions. And it costs a lot of money. It was very hard to kind of manage all of it. We didn't even know everything that we were buying because the ERPs don't organize it nicely by vendor, et cetera, what you're. And so we built this product internally that went through our entire ERP and then figured out all of the different software that we were buying and created a nice web application where it listed them all with renewal dates and all this stuff. And we were like, wow, this is so nice. And then we thought, huh, maybe other companies would like something like this, too. And then we were like, well, but we're not going to charge for it. You know, the business model is
42:15 Speaker 01 if we can get enough companies to do it, then we have like a data set of what all these different vendors are sort of doing that.
42:21 Auren Hoffman? Yeah, it's co-op.
42:22 Speaker 01 Yeah. And so we spun it off, still in the company, obviously, but it's a new product called SpenHound. We have almost a thousand companies now that use this product.
42:30 Auren Hoffman? I love it.
42:31 Speaker 01 To help them organize their expenses better. And now we even have like a procurement expert team.
42:37 Auren Hoffman? By the way, I actually created a business that did exactly this with like QuickBooks.
42:42 Speaker 01 Oh, wow.
42:42 Auren Hoffman? Back in the day, we sold it to G2, but it was the exact same idea.
42:47 Speaker 01 Yeah. So now, you know, that's been awesome because we now have visibility into the B2B side. So we can tell you which accounting firms are doing well, which software companies are doing well.
42:57 Auren Hoffman? Amazing.
42:58 Speaker 01 We found the ERP data a disaster. I don't know if you remember this, because people just type in things and you don't know what is it referencing?
43:05 Auren Hoffman? Oh, yeah. So hard to normalize.
43:08 Speaker 01 That's where the LLMs come in.
43:10 Auren Hoffman? Yes. You can go much deeper. You don't have to just do it for a specific vendor.
43:16 Speaker 01 Exactly. And so that's what we're working on right now, which is why I sort of made this point of we are the source of truth for how Uber and Netflix are doing right now, but our long-term mission is to be the source of truth for every company.
43:30 Auren Hoffman? What other things internally are you tracking that you don't think most other companies track? If I go into Vins dashboard, are there particular things that you care deeply about that maybe, you know, the average company doesn't?
43:44 Speaker 01 Well, I think every company is unique in terms of what is the key bottleneck to their current growth. And I believe very much you manage what you measure. And so we have this concept at our company of scoreboards, which is every initiative, there's a scoreboard. And the scoreboard is something that tells you, are you winning or are you losing within one second of looking at it? So it's always where we are, what the target is, are we behind it, red, are we above it, green, what is the North Star for that initiative? And then we have the sub stats doing that. We have scoreboards all over the place at the company. So everyone knows, are we winning or are we losing by how much? And we also have forecasts for like the next three months. And it's always our best guess as to what it will be. And then is it, are we winning or losing? And then we send it out every
44:29 Speaker 01 week. So every week it says, how did the forecast change? Did it go down or did it go up since the last time? That has been awesome. That's like the dashboard that we have at our company. And so that's how we know whether or not we're winning or losing across all of our initiatives.
44:44 Auren Hoffman? How do you think about people? Because coordinating lots of people is really hard. The best case scenario, you've got maybe an end communication problem, maybe worst case scenario, an unsquared communication problem. It makes things, you know, the more people have, just things just by default move slower in a company. So how do you think about actually the employees? Are you trying to keep the number down so you can move quickly? Are you trying to silo them? Are you trying to two pizza rule it? How are you thinking about like organizationally?
45:12 Speaker 01 So we definitely try to two pizza rule. You can feed everyone with two pizzas. We really try to break stuff up into components that work like in their own way. And then we give them as much ownership and autonomy as they need to do what they do. What we don't want is 150 engineers and then 20 product managers and all this stuff, all working on some big, big project, because that stuff like really falls apart. It's like really hard to that way.
45:39 Auren Hoffman? Yeah.
45:39 Speaker 01 So even that, you try to then create API. So this team is owning this and they have an API that they sort of integrate with another team and then they have an API. And so an API can be an actual API or like kind of like a metaphorical API. What it works for us is to just try to break people up into as small a group as is reasonable so that they don't need to coordinate with all these other people.
46:02 Auren Hoffman? Yeah, that makes sense. And now that you have like all these sophisticated people on your board, how do you run effective board meetings?
46:09 Speaker 01 Yes, it's very hard. I sort of see the board meeting as an opportunity for three things. First, it's really for the team itself, which is every quarter we have to stop and think about our business from an analytical perspective and kind of pull yourself back a little bit. And you're like, okay, well, what's working, what's not working? Is the strategy right? All of that kind of stuff. So 90% of the value of the board meeting is actually forcing us to analyze our own business. And then in the actual board meeting, there's two things we're hoping to accomplish. One is alignment with the board. So get everyone on the same page, like they get it. They understand what is happening with the company and they also are aligned with where the company is heading.
46:51 Speaker 01 And then the second is feedback, because the benefit of the board is they've seen so many other companies. For instance, it's just talking to you. You're like, you have this interesting position because you've seen all these data businesses. And you're like, uh oh, you need to do this because my other company X in this other domain had the same problem. And you should sort of focus on that. Now, every company is unique, so it doesn't always translate one to one. But I think that's like the value of having so many observation points across so many other companies that you should be able to bring in stories from other people, go to market, product expansion, AI usage to all the other companies as well.
47:30 Auren Hoffman? Yeah, two more question we ask all of our guests. The first one is what is the conspiracy theory that you believe?
47:36 Speaker 01 Very hard question. I know enough to say I don't know what I don't know. And when I look at all the conspiracy theories, I don't know enough about, like, I'm not in the inside. I don't have those details to know whether it's true or not true. But I would say generally speaking, I don't believe in almost any of them. Because I really believe in like the Benjamin Franklin quote, which is if you want three people to keep a secret, two of them have to be dead. And in my mind, these conspiracy theories often involve so many people to make them kind of plausible. And so in my mind is like, how in the world were they able to keep this thing secret for so long? Someone would have spilled their guts in one way or another. And so that's why I generally am not as big a believer, but I also, I don't know these things.
48:19 Speaker 01 I wasn't involved in the JFK assassination, so I don't know what happened and who was involved and who wasn't. What intel do I have to have an opinion?
48:27 Auren Hoffman? By the way, that's what someone involved would say. They would say they weren't involved in it.
48:31 Speaker 01 Yes.
48:33 Auren Hoffman? All right, last question we ask all of our guests. What conventional wisdom or advice do you think is generally bad advice?
48:39 Speaker 01 The one that comes to mind is a lot of people say you should be really focused when you're working on a business. Like, don't go do a ton of things. Stay very focused and do that one thing very well.
48:51 Auren Hoffman? Yep.
48:51 Speaker 01 And I think that there's a nuance to it, which is I believe you have to be at the right place at the right time. And the only way I know how to be at the right place at the right time is to be at lots of places at lots of times. We're big believers, the things that are really scaling, we're very focused and we continue to deliver that, deliver that really well. But we're big believers in having a bunch of bets, small little bets that we are making here and there. Many of which just die right there and then. We get a little customer calls, whoops, that's not good, that's not good, that's not good. But those little bets are how yipit.com, the daily deal aggregator, became a 700-person yipit data business.
49:31 Speaker 01 Was because it was one little bet that we were making that we weren't even sure if it was gonna work. And so, you know, if you look at our history of a business, the business has grown really well. But if you actually look at all the things we've done, it's basically one little bet every year that did 90% of the work. And then the other little bets all died.
49:52 Auren Hoffman? It's very tail oriented.
49:54 Speaker 01 And the only way I know how to do that is to have lots of little bets. And I know people say, oh, you should be very focused, you should be very focused. I'll tell you the second one is people say, hire great people and get out of the way. I do not believe that at all from a conventional wisdom perspective. I remember I read this post by this guy, Dave Kellogg, and he talked about how there's a thousand things any company can be doing right now. The reality is 997 don't matter. There's about three things that actually matter for that company. And I believe when you're kind of running a business, you should know what those three things are. And you should guarantee those three things go well. And if that means you have to go seven layers down and talk to this person who's working on that project and guarantee that thing goes.
50:36 Speaker 01 Well, you just have to do it because no one in the business has the context or the authority that the CEO founder has over that business. And so they are the only ones that can go in there and completely change the game on that initiative versus hearing about it six months later and knowing it didn't work. And you're like, why didn't we change the game back then? Of course, they didn't change the game. Who are they to change the game?
50:59 Auren Hoffman? Right.
51:00 Speaker 01 And so they didn't know what you know. And so that's the other conventional wisdom that I think, I think people feel like I'm not supposed to get in the weeds, like find the three most important things and get deep in the weeds on those three things.
51:12 Auren Hoffman? Awesome. This is great. Thank you, Vin, for joining us at World of DAS. Again, Vin, you're the CEO of Yipit Data, super cool company, one of the coolest data companies that are out there. I follow you on LinkedIn. I definitely encourage your listeners to engage with you there. This has been a ton of fun and super interesting.
51:28 Speaker 01 All right. Thank you, Auren. I love what you do with the world of DAS. There's not that many data companies. It's really helpful. There's endless SaaS companies, not that many data companies. So it's really helpful to have this podcast and all that you do for the data community. And I really appreciate it.
51:45 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 The LM Brief: The Syntax Illusion Next The LM Brief: India as the Next Epicenter of AI? Page updated October 31, 2026 · transcript and summary are machine-generated