Matt Ober: What the Buy Side Really Wants

Matt Ober was looking for unconventional datasets before alternative data became an established industry. He has led data strategy at WorldQuant, served as Chief Data Scientist at Third Point and now works with data providers through Social Leverage and Initial Data Offering. In this episode, he explains how the market operates from the fund's side.

The buying process differs sharply between quant and fundamental investors. A quant team may begin by examining coverage, history, frequency, delivery and point-in-time integrity. A fundamental analyst usually begins with a company and a specific investment question.

We discuss what vendors should provide during a trial, why external research cannot replace a fund's own testing and how poor documentation can waste a limited evaluation window. Matt also gives his views on pricing, renewals, distribution and the importance of maintaining good relationships with buyers.

Transcript

SPEAKER_01 0:01

Welcome to Selling Signals, the podcast focused on how businesses actually monetize and sell data. Each episode, we interview an industry insider to hear their experiences and lessons learned.

SPEAKER_02 0:11

The series is powered by Valcus, the company that transforms your data into investment-ready intelligence products.

SPEAKER_01 0:17

If you enjoyed the episode, please subscribe wherever you get your podcast.

SPEAKER_02 0:23

Today's guest is Matt Over, a well-known name in the alternative data industry. From leading data strategy at WorldQuant to helping build data capabilities at Third Point, and now through initial data offering and social leverage. Matt has spent more than 15 years at the center of how investment data is sourced, evaluated, and commercialized. In today's episode, we're stepping inside the buy side to understand how hedge funds think about alternative data, what separates great providers from the rest, and where Matt sees industry heading next. Matt, welcome to the podcast.

SPEAKER_00 0:50

Thanks for having me, guys.

SPEAKER_02 0:51

Yeah, welcome. Great to have you on. Maybe I know I just gave you a bit of an intro there, but I think you have such a vast experience across the buy side. I don't think I could do it justice. Do you want to sort of give the audience a sort of blend of the buy-side experience you've had and some of the roles you've held there and some of the responsibilities, just so we can sort of set the scene there?

SPEAKER_00 1:11

Sure. So I uh I actually started my career at Bloomberg in San Francisco. Uh quickly moved to New York, mainly like analytics help desk. Back then, everybody uh answered questions on the Bloomberg Terminal, but you got paid to essentially learn Bloomberg and learn about the markets. And then I quickly uh joined World Quant, um, quantitative hedge fund under Igor Sholczynski, part of Millennium Partners. Back then we were, you know, a little less than 100 people out in Old Greenwich. You know, over the six years I was there, I grew to 600 plus. You know, Igor always had this vision of, you know, hire the best talent everywhere in the world and give them access to more data than anybody else in the world, and you can manage more money. So we opened up offices around the world and I ran the data strategy there and I helped launch uh what was World Point Ventures, where we invested in a lot of the companies that we also were uh using, traveling around the world from everywhere from Russia to China to everywhere in between, just thinking about how we can consume more data than everybody else. This is in the early days, probably before alternative data was a thing, before the you know, all the alternative data conferences and everything. Um, so I did that for about six years, and then I joined Third Point as the chief data scientist working for Dan Lobe. Um at the time it was about 20 billion in assets, doing everything from equity credit, structured credit, macro, and about 2 billion in privates. And I essentially ran risk data and analytics and thinking about how do we use data and information to make better investment decisions. You know, very different model where you know a small position is 250 million in size, and a big position is like you're taking over and becoming activists and Nestle, Campbell Soup, some of the biggest companies in the world. And so like the use of data is probably easier to think about just from a you know a non-investor perspective. Like how do you invest in Starbucks without understanding like credit card data and app data, right? Everybody buys a Starbucks coffee on their app. Whereas in the quant world, like you know, you're thinking about ones and zeros and patterns and uh speed and a lot of different uh use cases. Um third point experience was fun, just given uh you've got to really understand a business and then think about all the levers you can pull from a data perspective. How can you be you know the data science team within Starbucks, but be smarter than they are, and then also know their competitors so that way when you approach management and you make an investment, you essentially can tell them things they don't even know about themselves. Um I did that for over five years through COVID, uh, great experience. And then I joined Social Leverage, which uh is a C stage venture capital fund uh started by my partners Howard and Tom. Howard Lindzen is probably most well known for being the founder of uh StopTwits, which is like the Twitter for finance. I actually cold called him 20 years ago when I was working at WorldQuant, uh, got to know him and you know invested alongside him on the uh venture side. He invested in a lot of data companies, uh Robinhood, eToro, alpaca, why charts, chart IQ, uh fiscal now. And so um was a friend, became an LP, and then ultimately decided to join them as one of as their fourth partner to invest in early stage fintech, but that also covers like data and information services companies. So I get to you know do the venture stuff I was doing at WorldQuant and at Third Point, but full-time and you know kind of build for the long term, right? We're seat stage investors, so you gotta think 10 years out. Uh gives me a lot of excitement to uh you know not focus on uh the day-to-day returns.

SPEAKER_01 4:53

And uh we we'll obviously get on to um one of the initiatives you're working on now, initial data offering a bit later in the podcast. But just to take you kind of right back to to the early days, um, you were obviously working in the alternative data industry essentially before it was even a category, before that that name had kind of uh permeated the ecosystem. And people now who've maybe only been in the industry for three, five years will really have no idea of in the pre, shall we say, new data, battle fin, eagle alpha days, what data sourcing even looked like. So I'd be really fascinated to hear in the early days, what did the hunt for an alternative data set look like? How did you go about sorting those data sets? But but also maybe how did you convince data vendors who perhaps didn't even view themselves as data vendors at the time, like license their data to you and think about this kind of completely different market to probably the market they're operating in from their core business?

SPEAKER_00 5:54

Yeah. I mean, I think it was a it was exciting back then and obviously very different. Um we were very good at Google searching, but it comes also back down to like, you know, the three things that I think we worked on the most was like brand building as like a big data buyer and approachable. You know, there's I think a lot of people that aren't in the hedge fund space, like look at hedge fund guys as, you know, not the most approachable, not nicest to work through, work with. We took meetings, we gave feedback, you know, we thought relationship building was important. You know, I always tell people like the relationship you build today, 15 years from now, that guy might be CEO or president or whatever it is, you know, like so like we we did a lot of that in the early days. I think we were firm but fair. Like we would tell it how it is to the vendors, but like we weren't assholes about it, right? Like we give feedback, like listen, you have nothing here, and or your the price you're asking for makes absolutely no sense, and we can tell you that because we're spending you know more money than anybody else on data. Um I also think like we showed up in person, right? Like, and I think that's more and more important, especially in an AI world that we live in today, which is like in-person relationship building. Like I went to Russia, I went to China, I went to weather conferences, I was in like you know, the weirdest parts of the world for events, obviously New York and everywhere else, but like going to these different places and then like you know, kind of like in the venture space, like maybe you don't invest in a founder, but that founder had a good experience and introduces other founders. Vendors are obviously always doing the same thing, right? Like, hey, it didn't work out for us, but like I have three buddies that have other companies, and then like you know, you're convincing them to sell data, but a lot of them were look are looking for revenue, right? And if you're you know explaining how you're gonna use it and giving some sort of feedback, which is obviously very rare in the quant space, it goes pretty far. And we moved with speed, right? It's nice if you get the data to us and we're paying you quickly. Yeah, of course. Money talks.

SPEAKER_02 7:54

Absolutely. I I the I think later on the podcast, uh I'd be love to delve into that a little bit more about how that's changing and sort of the more corporates that are coming to market. First out, I want to sort of unpick that sort of buy-side quant versus fundamental piece. Um I don't know too many people that have spent such a vast amount of time at a sort of T1 quant shop like World Quant and then moved across to a very pure fundamental firm and sort of lived those two experiences. So I mean, maybe as sort of someone that worked very closely with sourcing data providers, working on the data, I mean, if you say investing in data providers in those seats, what were sort of the biggest differences between the way a quant thinks and uses data versus a fundamental firm?

SPEAKER_00 8:41

You approach from a quant perspective, you're approaching data as data, right? How much depth and breadth do you have? How much history do you cover? How many companies do you cover? How many tickers are there? How many data points? How often does it update? What's is it point in time, you know, all these things that is really about the shape and the type of data. And like, you know, where is it being sourced from? How is it delivered? API, flat files, you know, nowadays, MCP, all these different things. Where in a fundamental world, you don't really start with data. Or I don't believe that the good investors should. You start with a question you're trying to answer or a problem you have. What's the most important KPI for the company we're investing in? We're investing in Netflix, and the only thing that matters is, you know, the time when I used to look at Netflix was like subscribers. And it wasn't US subscribers, it was, you know, new subscribers in India and different countries. Or, you know, we're investing in Nestle, and the only thing that matters is the coffee business. And like we need to better understand the coffee business versus all the competitors out there that do coffee. Start from a fundamental question and KPI. I think that's why it's hard to go from quant to fundamental without fully like getting inside the mind of a buy-side analyst. Like, let the buy-side analyst who's making all this money and making the ultimate decisions, what keeps them up at night? The only KPI that matters. I remember the first time I talked to a consumer analyst one time, he said to me, the only thing that matters is french fries. He's like, if you can tell me how many french fries are gonna sell, how much it's gonna cost to make these french fries, like that's gonna make or break my ear. It's a very different approach for a data sourcing team than like, hey, we've got all this cool sentiment data on news, or like, hey, can I show you these app analytics? They don't care. Like they're not like maybe if they were using it to source changes and like they're looking forward to like come up with new idea generation. But like, you know, in a concentrated fundamental book, like new idea generation is important, but is data really going to find that? It might over time, and we built things to do that to find trend changes and get ahead of things, especially for shorts and everything. But like from a long only from a long perspective and you know, getting deep into a company, there are KPIs that matter the most. And so like the data approach to sourcing is very different.

SPEAKER_01 10:56

And and if if you were kind of advising a data vendor who's looking at obviously like the quant versus the the fundamental space, we it's obviously quite clear when a data set might be interesting to quant, you've got, as you've already highlighted, like long-tail history for backtesting, broad coverage of tickers, um, obviously the closer you can get to like directly impacting something like on the on the you know top line or or share price, et cetera, the better. Um, and then if you want to actually target those quant funds, particularly the large ones, have dedicated data sorting teams, you get in touch with them and the process is pretty kind of well-worn by this point. If there's a data vendor who think perhaps, you know, we we maybe don't have that kind of breadth, we're not quantum applicable, but we have a really, really specific uh niche that we think, you know, we we can tell you how many French fries are being available, for example. Um, how would you recommend that they go about kind of targeting those individual fundamental firms? Because 99% of the time it's not on anybody's LinkedIn what type of analyst they are, you know, that they're, you know, do consumer staples, discretionary. And that's kind of the minimum. But let alone the you know, the stocks that they're actually looking at, you know, that um all that kind of stuff. So how would you recommend that a data vendor go about targeting the right kind of fundamental people if it is really that kind of um if the other way around it would be don't call us, we'll call you type thing.

SPEAKER_00 12:29

I mean, listen, I think network and Rolodex are important in the uh investor space because like you gotta figure out who the consumer analyst is at, you know, a multi-strap firm like Millennium to like who's sitting in the consumer seat at third point, right? Obviously, you can go through 13Fs and filings and start to figure out who's investors in these companies and then try and figure out how to get to that person. Most of the big shops are gonna have gatekeepers. Again, it's like you know, it's like a networking thing. It's no different than like the recruiters. The best recruiters in the hedge fund space have mapped out every person on the desk, right? They know what they do, who they are, they're trying to poach them and everything. The best salespeople have to be the same way. You walk into a millennium or any of these big shops like a Citadel, like all these different teams gotta know who does what and when. I think like the filings and what they're invested in is important. And listen, this is kind of why I uh put that initial data offering platform together, which is to be able to just announce it to everybody, and like a lot of them are just subscribers to see things. And then you gotta make relationships with these data uh gatekeepers, right? All these firms, at least the largest ones in the world, all have some sort of data team, data sourcing. You know, in the past it was like the market data rep, right? And like maybe they're not like just buying, they're just ones going through the paperwork, but like building a relationship with them, whether it's at an event or a happy hour, catering lunch at their office, they can tell you who the players are, right? Yeah. So I think it's a hunting type of uh approach versus like you know, the quants. It's like give them a white paper and some insight on like how you can use the data. If it's good, they'll buy it.

SPEAKER_01 14:09

Interesting. Sorry, you just said something at the end there about um a white paper. I I know I we hadn't uh discussed it previously, but I do you think that is what uh vendors should be doing? Should be they they be doing their own kind of back test on their data, or should they be saying, we think it's these kind of insights that are in our data set? Because it seems to be quite a controversial point in the industry at the moment. Some people are on one side of the debate, some people on the other. What what what are your thoughts?

SPEAKER_00 14:33

I think the more information and the better the packages you deliver, the quicker you're gonna get people to make some of those, right? Like in the early days of selling to quant, like, give me your API docs, give me your marketing material, and give me any white papers that you've written, but like arguably that maybe you've had academics and students' PhDs put together, package that together with me with like all the historical data, come back in 30 to 90 days, and like it's a yes or a no. If you don't do all of that, it's fine, but like then I've got to have my team figure out what the best use cases are, which like listen, like there's enough technology now, like that's possible. But like if you've done some of the work, if you've got a half dozen white papers that have been written by academics that have used the data for different use cases, like you're just speeding things up. Now, like it's I I just think like you put together a better package, whether we believe it or not, whether the results they show are real or not, like at least you've kind of like put frameworks together versus like if you hand somebody a garbage bag of data and you tell them like you want a million dollars for it and they've got 30 days to test it, like good luck getting a quick response versus like here's a polished document with everything, and like here's some use cases, here's what we've seen work, here's how it's really correlated with you know, top-line KPIs for the consumer industry. You know, like you put more effort into it, it's a more polished product, it's gonna be a little bit easier to get off the shelf and sell.

SPEAKER_02 16:00

A counter-argument to what James just said, I I had a conversation with uh a vendor the other week uh was saying that they have an internal quant team, quant team comes from uh from the market, so it's credible. Uh, and they they feel that to your point about being more polished, it also gives them more control in the sales cycle in a point of time where a lot of salespeople lose control. Where if they know that a uh a uh researcher is sort of going A-Wall and not thinking about it in a similar way, the likelihood of that deal closing diminishes drastically from their research. Um, so it was a nice sort of viewpoint from a control element where, as you well know, there's a lot of dark time for a salesperson when a quant is trialing a dataset.

SPEAKER_00 16:45

Yeah, I think the flip side argument to that is like if any of the quants are researchers that the vendors are even decent, they would have been poached already by the buy side. So the like the uh respect that the end quant is gonna give the vendor for like their internal team is very low. Even though it's helpful, you know, like there's gonna be like, yeah, that's fine that he did this, but like it's not realistic because he didn't think of these 50 other things. Like you can't execute those type of trades, like you can't get that type of volume, you know, like your simulation and backtesting environments okay, but it's not like you can't actually you know produce the returns that they theoretically put together. So I think it's all like helpful from just like a design phase and from like a sales perspective, yes, like you can hold them to like a 60 to 90 day window or something shorter because like we've done some of the work, even just like the documentation, right? Here's the fields we have, here's the frequency of updates, you know, like making it like easy for them to just get off the ground and run.

SPEAKER_02 17:52

I understand the argument of um uh the best talent being on the buy sides uh given the money that's being paid. But is there an argument to be made that a a strong quant sitting on a provider side that spends 90% of their life over the last five years in that single data set thinking about how that can be applied to different parts of the market that does provide something value to someone on the other side of uh the table?

SPEAKER_00 18:18

I think there's value. I just think it's uh there's only so much value. Like I think I learned this when I worked, I worked at Bloomberg and then I went to the buy side. You don't fully know what's going on on the other side. You don't know the intricacies of what it means to short a stock. Like, how does it actually work? Like, what are the costs of like borrowing costs? Like just like all of the little nuances versus like, you know, at a vendor, like they can't teach you everything because they a lot of the times nobody's been on the other side. So I think like it provides value and direction, and I think there's a value to having it, but to a point.

SPEAKER_02 18:54

Understood. The we spoke a lot about the trial process. You finish that, you go, I want to buy a data set. How does a decision-making process at a fundamental firm differ from something like a quantum uh a coin firm?

SPEAKER_00 19:08

I mean, to get a quantitative fund, like you can see what the results could be, what how much money you could put behind strategies, you can understand how correlated it is to other strategies. You just have a ton of more metrics. You also can say, like, hey, we've bought a lot of things similar to this. This is the price point we're gonna pay. We're paying 60 grand a year, maybe 90. Anything above that, it's not worth it because there's all these other things that we have that are correlated that can provide similar results. Fundamental shop is much different, right? Like, are we using it just for one analyst and one position? If it's a multi-billion dollar position in an activist investment, I don't care what we're paying. Because if it's really moving the needle and we need it tomorrow, we need it.

SPEAKER_02 19:51

So you feel like sometimes contract values on the fundamental side could actually be a quant ACV.

SPEAKER_00 19:58

Yeah. I mean, uh there are things we bought on the fundamental side that were expensive that I never would have paid close to at the quantum side.

SPEAKER_02 20:06

Interesting. And if you go a level deeper to that, uh and think about you're a salesperson looking at a fundamental opportunity and a uh a quantum opportunity, and you're thinking about mapping out decision makers. Like, who are the key personas that a salesperson should be thinking about that I really need to be influencing those groups of people?

SPEAKER_00 20:28

Listen, I think at the quant side, it's there the the big shops have data sourcing teams, right? They have teams that are testing and backtesting, they have a process. At least they have. I mean, through the amount of years that this has been going on, the big shops have a good process, right? I think at the fundamental side, like you just gotta get the buy-side analysts interested. PMs probably don't care. Listen, if you can get to the PM, great, but he's looking at a lot of things. But it comes down to the actual, you know, the buy-side analyst that's making the like decisions and using things. And then, like, obviously, a lot of these firms have data science and data teams that are helping with a lot of those details, but I think if you can get the actual analyst inside excited, but you know, you've got to play the game of uh making sure the gatekeepers feel special, right? You don't want to go around them. That's why I think the nuances of sales into the institutional world is each firm is different. Do they have a gatekeeper? What's the process? Does it need to go to the gatekeeper to go to compliance before the analyst looks at it? Is it better to go to the buy-side analyst and have him tell the date the gatekeeper? Like who like each firm has a different process. And I think like the nuances of if you're a new vendor and you don't know that, like you've got to tread lightly to not like piss off the wrong person because then you're in like the doghouse of getting your data to actually be looked at.

SPEAKER_01 21:54

And and you mentioned um uh pricing. Um, this it it's always seems like one of the most difficult things to do in the alternative data industry. And obviously, as you say, on the quant side, there's, you know, you can run a back test, figure out what sort of ROI you might get on the data set and kind of back out of there on when you're the buyer. But as we've already discussed, the sellers maybe don't have the capacity to figure out what sort of alpha there is inherent in their data set. And so their ability to do the same thing is obviously somewhat diminished. You then have, I suppose there's, you know, comp, like you're you're a transaction data provider in Geography X. What do similar transaction data sets go for? But again, that information is often not very available to you. What would what would you recommend to like a a a new buyer? Are they just going to like start with a figure and see like like throw enough on it and then see what sticks? Is it that kind of thing? You mean asking a new buy from the buyer perspective or from the vendor perspective? Sorry, vendor's perspective.

SPEAKER_00 22:52

You're you're sorry, you're new vendor and you're uh you're looking at the Yeah I mean listen I think they need to like build a network and to kind of ask around like how do people like think about like the you know the value of the data. Like I think like as much as you read the headlines of multi-million dollar data set sales like how many of those actually get done is very far a few in between right like it's not a common thing that people are paying seven figures on it. So it depends on the firm size right is it being used by Citadel and Bridgewater across the entire firm is it for a small pod is it being used at 0.72 for just a consumer team we know like there's so many nuances to the pricing based off of the use cases and who's using it and why they're using it. It's going to go to other groups right is it going to be sticky because it's going with the quant team and like reality is if you get in there like you can up the price every single year versus like the tough part about selling into a fundamental shop like if it's a niche data set that's helping me on specific companies I might just not be invested in those companies anymore. So I don't care. Or I do care but like now the value for me is like a tenth of what it would be versus if I had a huge position.

SPEAKER_02 24:00

I was listening to your uh episode on odds are open and I think you touched on this a little bit about how when you were sat on the quant side that um you were more tracking the return on investments on each data product that you licensed versus looking at the stock market. My question my the the thought that came to my head around that is if it's so easy not easy but if it if quants are quite accurately producing RI figures on data sets why is there not more transparency around pricing on the vendor side where it's you know the the the quant isn't going to tell you well it doesn't similar like they're going to tell you what the value of the data set is internally and therefore it's a bit of a chicken or an egg in terms of what you do price it. So like how come that was what why is there no transparency on on that ROI number?

SPEAKER_00 24:53

What's the value of the quant giving that information up? Like why what incentive do they have to help the vendor? Yeah I I don't have a good a good uh response to that I guess but you you you're correct it goes back to the whole thing of like you know I think one of the things I used to hate the most is when a vendor would be like hey like you know the price is depending on like you know your AUM as you grow we want to grow and I think like most most buy side firms would be like you got to be kidding me like our success is not going to be dictated by like you as one vendor like like unless you think like our whole firm is being built off of you the fact that you like kind of position yourself and say that like hey as you guys grow we want to grow with you in terms of like what we get paid listen you want to keep that internally as part of your calculator on how you calculate what you're charging us like more power to you I'm on the other side as an investor in in companies and you can do that. Because like you think our success and returns are based off just your data set. Like how many strategies really are just based off of one data set? I would say very few if any goes to the whole thing of like when vendors say like you know like we're thinking about starting a hedge fund because our data's so good.

SPEAKER_02 26:17

We can all laugh about that but like that the reality is is like how many data vendors have actually started their own fundament I think you we had a conversation uh in the sort of pre-recording about consumption models um and you see quite a strong advocate with it obviously CarbonArk is a a business that's quite pro-consumption arcs and I believe IDO and CarbonArc are are partnered. What's the conviction there?

SPEAKER_00 26:43

If you think about the quant use case do you think that's got could go to consumption models or are you more viewing it from the lens of bringing in the longer tail of buyers that can't afford the the high license fees I don't think it's as much the quant we can get back to that I think from a fundamental perspective it's really simple right like think about like uh an activist investor I become activist in Starbucks I want to buy the data when I need it for when I want it for what I need it for at that time. If I go activist on Starbucks and I really need to understand the coffee markets in the Philippines I'll pay a hundred thousand dollars for that data. But I don't need that tomorrow and I your data set might only cost like sixty thousand dollars a year for an annual license but I'm gonna spend a ton today because that doesn't matter to me because I need the information when I need it and I want it right now. And I think that also like if you think about a world of AI agents like and your agent has a wallet and you're a buy side analyst and you need information your agent is working on your behalf finding that information like hey we're doing a deep dive into Starbucks and there's three private companies Phil's coffee and a couple others that might go public soon. We don't have any private market data the agent goes out at finds a place that can buy it for it it's going to cost it $10,000 to get it right now. It's a no-brainer for buy side analysts to spend $10,000 on that day to get the information that they need. They don't need that information on a regular daily basis. So I think like that makes a ton of sense to me. I think on the quant side like listen if you have alphas that are like on the shelf that you don't use every day and like you know it's part of a a rotation strategy where it goes into something and goes out of something and from a consumption basis it makes more sense and you're not spending as much money today as you need as you need to or maybe for back testing or if you're you know building things you're only paying one to use it. I think the quants can get there and you know they're more mathematically inclined and can run a lot of forecasting models and it might make sense for them to go consumption based. But everybody else that's not like even think about my seat right now on you know the venture side like I'm looking at so many different opportunities on a daily basis. I'm going really deep on an opportunity like I'm there's there's almost zero data sets that I want to buy right now on a regular basis. I'm happy to spend the money when I need it for what I need it for. Right? I have an access to CarbonArc and it's fully consumption based and I spend five grand a day obviously it was worth it. But I don't need I don't need to buy maybe the data I'm buying from him on a regular basis because like once I make the investment I'm a seed stage company like I don't need the data or maybe I passed right we didn't make the investment. So I think that that's where the world is going I also think that in a world of AI we all want to consume a hundred X more data. We want to pay a lot less for that data but overall we're gonna spend more because we're gonna have access to more we're gonna do more our intelligence as a as a world is going up right the amount of information we can all consume now is unbelievable right all the ideas you have can be automated now. Like the only thing holding back people from being more successful is their creativity.

SPEAKER_01 30:01

Sorry you said two things that I just want to um uh think about because they could completely change certainly the business model if you're selling to the like the fundamental investors one is the the pricing because I um thought that if we do go into a consumption based model just kind of in in all circumstances with from a fundamental perspective the data vendors would be making less money so kind of why do they do it? But you gave some pretty compelling uh suggestions that actually no it they might make less ARR but there are several situations in which they might make $100,000 literally for giving data across three tickets and they might make that in in a day. So I I'd like to I hadn't really thought about how that would then impact the industry because you're no longer doing it on an ARR basis. So like how do how do those sorts of data vendors then sort of plan their or forecast their revenue? That's one point. The other is then how does that link into the whole traditional way uh vendors in the industry at the moment actually think about whether or not they want to buy a data set which is through a trial you're not you're not if you're a vendor and you're getting paid potentially like one off a large lump sum, you're not going to agree to let that uh vendor trial the data, get what they need and then not pay you anything. So how do you think those two things are massively going to change in the industry, at least on the kind of fundamental side of things?

SPEAKER_00 31:27

On the trial side there's always ways around that right like you don't give them the most up-to-date data quads like you know new vendors are always worried about that like give the quads the data up until April of this year. You can't trade on stale data in the quad space same in the fundamental space right like I'm not gonna make an investment in McDonald's today if I only have the data through March right in backtest understand like on a two year stack through different periods it's 85% correlated to top line sales in North America you know like we can go through all of that. So I think you can get around those pieces. I think like listen the ARR piece that's that's tough right like for public companies that's the biggest thing they struggle with how does the CFO forecast that private markets like you're a little bit more flexible like you have to deal with your investors if you're venture backed but like otherwise like you you can experiment more quickly. I think that like you have to start to think about it like transaction volume, right? Or you have to start to think about like listen we have a floor it's you know some of these companies are like hey we're gonna charge X for access $1,000 a month so we have some ARR. When wake up one day and we made 100 grand and then one month it's still just that flat size. And so like you have a blend of your your actual revenues um but I will tell you that talking to a lot of the public companies they're all like kicking the tires on consumption models how does it work but like you know look at it from an AI perspective like outcome based pricing is happening everywhere.

SPEAKER_02 32:55

Usage based is happening everywhere right like all these innovative companies are doing it CFOs are going to figure it out or like you know they're gonna miss on their estimates I mean James raises a good point you think about the advertisement world um that is a essentially a large flow of cash running through that it's if if you know in a week's time everyone stops advertising there's a lot of companies that just die because a lot of that is based on quote unquote consumption although it's more like cost per media. I I guess my my question to you Matt is there is a big trend right now of big corporates coming to market because of the sticky ARR, the the retention rates of of this business if that gets turned off when we go to consumption border which I think all the points you raise around us going to an agentic world and the those workflows makes a lot of sense does that disincentivize those businesses coming to market or yeah what are your thoughts on that?

SPEAKER_00 33:56

I don't think so I mean listen I think there's going to be still a lot of people that want a flat rate it's this much per year. This is how we're doing it and you'll have a lot there I think like I always say to the vendors like why not offer both? Because like you have a bunch of companies that now have access to your data when they want it, when they need it through like a CarbonArc or maybe there'll be other platforms eventually then you might wake up to all this newfound revenue and you're in front of people you're never in front of like if tomorrow XYZ hedge fund goes activist on SpaceX and you need this specific data how long is it going to take to get through compliance and legal and data delivery and all these things you're not getting that data by the end of the day. Right? Even the fastest moving firms it's not going to happen. Right? DDQs all that if that's all done and it's always on the shelf think about it we all should have a shelf with access to thousands of data vendors and as the guy that has to get the data in front of like a PM like there's no excuses anymore. Right? Like that's the world we should live in we should all be able to get access to the information as quick as we can actually move like why how great is it would it be to sit in an investment committee meeting in a boardroom you have your phone on you and somebody says well what happens if we were to go all in on this company like how do we think the numbers are looking and be like well we don't pay for it but like I have Claude on my phone access to Carbon Arc or whoever it is I can ask the question my AI wallet an agent can go spend and has access to spending for that drop 50 grand on that conversation but now we're making a 500 million dollar investment because like the data looks really good. Versus like hey let me go source that data I gotta call this guy hopefully he's not on vacation we've got to go through DDQs, compliance, legal back testing, ingestion the engineering team's gotta get it in we've got to put it in Snowflake or data birth you know like wipe all that out yeah you you you create quite a compelling compelling picture on the it's a game changer these days I mean I walk around New York City going to meetings with my own investors and I've got Claude on my phone I can talk to my fund admin about like how much money are those investors invested in me across all six of our funds what what's the name of the entities they're invested out of and also tell me the last three conversations they've had with me or the other partners from our from our CRL. I can do that with voice with my AirPod in walking New York City and within like three minutes get a full answer.

SPEAKER_02 36:37

I don't even have to prep anymore yeah right moving on site because I appreciate we've we've gone deep there we've talked a lot about how vendors need to sort of try think about consumption models but you've probably worked not necessarily directly but spoken to thousands of data providers in your time what have you found separates the good from the truly great relationship building at the end of the day you've got to build a relationship with these people the best vendors if I reached out to them I could get a hold of them whatever I needed to we really needed access to this data or we had a question and we're asking for a favor.

SPEAKER_00 37:22

It's relationship building market did this great right back in the day like market owned so many things before like merging with SP right they owned like the VI number on your car through like the mobility spinoff to like you know they had the data explorers like short interest data to CDS stuff. You have one contact there he makes sure he gets you access to whatever you need into the the you know the person within the firm that like knew the most about that. Small vendors same thing right and so like hey like I really need to look at this ticker send me a sample gets it to you within 20 minutes. It's relationship building like I mean the whole world is built on relationships right it's it's what you it's who you know and then what you know right like if you can get access and to information and get the contacts and build those relationships that's really what matters I mean obviously you have to have a product sell. If I have seven new sentiment providers and you could argue that 90% of the investment world couldn't tell you which one is actually that much better. Right like I always laughed when like you know a longtime Ravenpack client friends with Armando and those guys like the new companies would come out and be like we're 10 times better than Ravenpack our sentiment blah blah blah blah blah I wasn't at a pond fund anymore. I'm like I don't know maybe it's better I don't spend my day thinking about like natural language processing techniques and like what's the best of the best but I'm not gonna go and try and like displace what I have but I'll tell you this I have such a good relationship with those guys like I'm not getting paid to go try and displace that it works it's great you should be happy that I'm even taking this conversation right so like if you think about it from that perspective like it's all about building those relationships I get see I I agree so it's small worlds I I think getting a foot and I completely agree.

SPEAKER_02 39:19

How does that compare when you think in a sort of a quant lens where there is a black or white you produce enough alpha for us to invest versus right I mean in the quant world like it comes down to like the numbers speak for themselves is it truly point in time do you cover a thousand tickers?

SPEAKER_00 39:38

Do you have 500 data points? Are you delivered daily versus intraday? How many times in the last year since I've been applying to like the API or the feed go down or there are errors right like how quick do you respond when things aren't working at 4 a.moting and offering new things like it's a different game because like you know like you could arguably have like a whole long checklist of like if you meet these 50 criteria it makes sense for us. Hey we don't trade the European markets I could care less about your data set being global or we do we do trade globally like what else do you have outside the US moving on to another topic that I was intrigued to to pick your brains on.

SPEAKER_02 40:26

I mentioned earlier and I'm sure you you're very familiar but there is an influx of bigger organizations looking to monetize data not just to the investment world but to to other verticals too as they look to to find differentiated revenue. The problem I've always kind of thought about when a big corporate comes to market is to our market explicitly is you've got all of these businesses currently in the market that have built a data product for the investment use case or are a data business collecting data for the purpose of being a data business. Not necessarily the exhaust data that has been generated from you know offering payment solutions to a business, offering procurement solutions, being a reviews platform. And what often happens with those businesses is they're driven by their sales organization. So I think about my time at Trustpilot they did really well in banking and e-commerce and therefore because of the products they offered they ended up collecting a lot more reviews on those businesses than any other and there was this big skew in the data et cetera et cetera as you maybe in sort of like a quant lens how do you sort of manage that bias does some data set become unusable and how does that work in in in a quant lens I think the bias stuff like you know it's good when the vendor knows their bias sort of in like the consumer transaction data sets that like hey how does it map to consensus?

SPEAKER_00 41:52

Like is there a bias towards the cokes? Is it you know ultra high net warmth people that are using this you know where you're getting your transactions from or is it a lower demographic? Is it in the south versus you know wherever it is I think understanding your biases is obviously helpful from a vendor perspective. I think it goes back to my comment of like if you're a corporate trying to sell your data to the investment space or for other industries like are you building a data pro data products or are you just trying to say like we have all this data but we want to sell it like I think if you're not building data products like get what you you know you get what you paid for type of thing like if you're not gonna put in the work like don't expect people to like be jumping hand over fist like just give you money. I would argue with a lot of those corporates unless like you need this to be like a huge revenue generator for the firm like find distributors and partners take 5050 go put it inside of a carbon arc and find a few other partners that are going to build products around it.

SPEAKER_02 42:49

Let the people that know the industries that they you want to sell into do what they do best and collect your check versus like putting all these resources into like trying to create like a quant product like there's a ton of people that love to sell quants all day long and build products for quants and like have the distribution already like how much is it worth all the time and effort to do all that in-house sometimes I would argue it's not also it's a distraction right so I think it kind of depends and also you could test the waters like it's not there's no there's no problem like getting out there and like oh wow we're seeing a ton of revenue from this like let's take this in-house over time where we've gotten like the for the top 50 now as we expand and build more products because there's more demand for other things we do the rest in-house right yeah I've I've always thought about some of these roles being more partnership like I think about some of the conversations I've had with businesses that are monetizing and it's you think about some of the big telcos and they're saying oh we're seeing a lot of uh regional clickstream data I want to build a similar web like product to try and compete with them in that market. And it's like just sell the data to Similar Web.

SPEAKER_00 44:02

Or go sell it to the six competitors that Similar Web has, right? Or go put it in a Carbon Arc-like platform where they can map it and kind of position it for you, you know, or a whole host of other things. Right? Like sell to all of them.

SPEAKER_02 44:17

Talking of partnerships, what's becoming a bit of a theme at the moment, and I guess going back to the start of your career, is the old the alt D terminal. And there are more and more providers, especially on the consumer uh alternative data side, that are partnering with Bloomberg. What's your opinion on that? Do you think that's good for their business? A lot of them are arguing the fact that you know if more fundamental uh hands are touching the data that that makes more opportunity for a quantitative investor. What's your views on on those types of partnerships when people are already in market?

SPEAKER_00 44:50

Like partnering with Bloomberg and putting their data on the terminal, you mean?

SPEAKER_02 44:54

Yeah, like Apptopia recently announced that they're they're they're partnering, you've got Placer, I think it might be similar web that's on on there, but like yeah, like is that net good, net bad?

SPEAKER_00 45:06

I mean, listen, I think distribution is good, even if you're putting a piece of it there, like you don't have to give everything. Assuming Bloomberg's not paying for that, and it's just like another way to get more eyeballs on it. I think it's a great way to get more eyeballs than like to drive like inbound. I also think like every vendor wants to sell you alpha, but you really want to be beta. Put on a VC hat. I don't want to invest in an alpha data set because we live and die by making money for those investors. I want to be beta. I want every hedge fund and every buy side and every sell side bank to have to have my data to do their day job. Right? Like, because if they don't have the data, they don't do their day job. You can't invest in consumer companies now without app data and consumer transactions, right? You could argue that like without having access to those things, you're investing blind. I would feel that way as an allocator, especially if we can prove that it's you know correlated and it gives insight and so forth and so on. So now that's beta. That's great as the vendor. That's sticky, recurring revenue that's never going anywhere. You're selling alpha. It's a tough game.

SPEAKER_02 46:18

You said that on the um on the Audit Open Podcast also. And one of the things I thought about the you want to be uh beta is how do you get from alpha to beta? It is that partnering with people like Bloomberg where you're just so now ingrained in the market that people just have to use that data. Like you say, arguably if you're investing in consumer, you need a transactional panel. Like what was what was what's that tipping point? What are the key criteria to tick for someone to go, yeah, that's no longer a an alpha product?

SPEAKER_00 46:52

I think it's distribution. I also think it's positioning. You know, in the early days of consumer transaction data, it was like so much of everybody highlighting on the tickers like how accurate they were on the beats and misses and like predicting the earnings numbers. Like let the buy side guys and like the guys that are good at that do that shit. Like if you're doing that for them and you're wrong, you have more to lose than letting them do it themselves and not charge and charging a little bit less, right? Because, like, yeah, I mean, listen, there's also like the it's a conversation also of like, what do you want to be as the vendor? Like, are you building a big business? There's a lot of great three to five million in revenue data businesses that are lifestyles, they're not investable from like an investor perspective into a private company, and like they're never gonna have a huge exit, and like they might not even have really any buyers that are out there. I mean, we've seen this a ton, like there's tons of companies that have been around a long time that sell data that like there's no real exit because they're never gonna get the like 10x revenue number as an out as an exit, right? They're gonna get like something small because like it doesn't grow much, but it's sticky. So like you gotta kind of decide what you want to be. Um, but I think it's it's distribution and positioning. Like, don't be afraid to be in a lot of places and be known. If you want to sell alpha, it's great, it's just a different game.

SPEAKER_01 48:18

One thing I've realized that we haven't done yet is uh giving you an opportunity to talk briefly, um, because I think we're running low on time, uh, but about um IDO initial data offering. I think a lot of the data vendors would would love to hear more about that initiative and how that can help them reach more buyers.

SPEAKER_00 48:33

Yeah, I mean, this was an idea, it goes back like 15, 20 years, which was like my frustration in the data space was always like, how do you find out about new data sets coming to market? And I always loved like a website like Product Hunt where like you learned about new just products coming out of Silicon Valley and everything. And so I think I bought the URL 10, 12 years ago, and a few years ago, essentially launched it as a way as any data vendor in the world can announce new data sets coming to market. And now we have thousands of the biggest buyers from hedge funds, financial institutions, a lot of corporates and consultants that are all data buyers as subscribers, free for everybody. Vendors can announce new data sets or data sets that they never really marketed coming to market. It launches on the platform, newsletter goes out a couple times a week. Anyone in the community is interested, they can click a button and request an intro. Automatically, those that contact information goes to the vendor, they can go do a deal. IDO doesn't touch the data, it's not a consultant, doesn't want to look at it, it's not taking a commission. That's a nice partnership with CarbonArc. So if any of those vendors want to also launch on CarbonArc's platform and also like try and create monetization and you know really be within their framework, like it's a simple click-through when they're launching on IDO to fill out the DDQ and go that direction. But it's as simple as that. I think uh, you know, it's probably like two and a half years old. And I mean, honestly, it's been great. Like the best vendors for getting a couple dozen demo requests within like the first five minutes of launching. Well, I had vendors saying, like, you know, I made a few million dollars off of that announcement. And so, like, I always say to the vendors, like, if you're not doing it, like, why not? Like, there's so many long tail in uh funds that I've met through that platform that I had never heard of before. Managing 800 million dollars based in Wisconsin, they buy data, who would have known? The French fry investor that I talked about that like is interested in something really weird and random. Also, a lot of the vendors are using it for their own partnerships and distribution. And so, like, yeah, there's like there's some paint services on it, like you can do like premium advertising and little things like that, which helps just pay for the platform. But you know, my day job is running social leverage on the venture capital side. This is just a great way to build a great network around the data space, and a lot of our vendors, a lot of our companies that sell data have used it and found success from it as well. Um, it was a way for me to curate the network that I've built also from for a long time.

SPEAKER_02 51:03

And just to reiterate what you're saying, I've been a user of your platform from my seat of new data when I was a trust pilot. I plan to launch some of the data sets that I'm working on now on that platform, and I have made money from your platform, so I can validate what you're saying.

SPEAKER_01 51:16

Yeah, it sounds like a no-brainer.

SPEAKER_00 51:17

Yeah. I mean, it it surprises me sometimes that I'll meet some people and they'll be like, we did seven figures in revenue after launching an ideal. I didn't even know what their company was, right? I think like now also it's really interesting to see like which ones launch and get a ton of quick traction. Also, the flip side of it is like a lot of them will get like no requests for intro, and then I'll hear from the you know the CEO or salesperson be like, we got like 40 emails on the side, because like listen, a lot of hedge funds are secretive. They don't want to click a button because they think like maybe I'm gonna sell that data, which we've made it pretty clear. Like, we have no plans to ever sell the data on like who's clicking on what and doing everything. It's also all automated at this point. Like, you click request and intro, that goes to the vendor, and like good to go. Um it, I will say this it's you know you found the right product market fit for something like this when like the biggest names from the biggest funds are clicking request intro within three seconds of a vendor launching, right? Like, so that's where you kind of know like even the guys that have been doing this for 15, 20 years that you can probably never get a meeting with are still opening their IDO, and like the open rates are you know 60, 70 percent, which is like unheard of for like a newsletter. Yeah, I'm a beehive investor, I would know that. And so like it's pretty cool to see, you know, the value of it. And like also we keep it pretty simple, right? We're not trying to like stuff things down people's throat. I don't want you know, Reuters to come out and launch every data set that they've ever had in their whole life on the platform, right? It's not supposed to be a catalog, it's supposed to be uh a place to launch either new data sets or you know, there's a lot of vendors out there that we both know that like maybe have never marketed or positioned things that they own. Um, and they can now like kind of like bring that out to the world. I think bar chart just like a launch some data yesterday in the commodity space, pretty unique. Listen, there's probably not a thousand people that want to buy that, but for the 40 people that do want to buy that, that's like eye-opening for them. And then you can search on the platform, right? Do what happened to what launched in the past.

SPEAKER_02 53:18

That open rate, you probably could charge a premium for the advertisement.

SPEAKER_00 53:22

Those that want to advertise, just lots available.

SPEAKER_02 53:27

Awesome. Well, this has been a great conversation. I really appreciate you coming on. Um, I've listened to a lot of your content over the years, and uh yeah, I've always wanted to delve a bit deeper on in some of the stuff you've covered, so I appreciate you coming on.

SPEAKER_01 53:38

Yeah, thanks for having me guys as well. Yeah, that's been great, Matt. Thank you very much.