Brian Yormak – Venture Investing in Data – Ep.186

Ravix Group — Ravix Group is the leading outsourced accounting, fractional CFO, advisory & orderly wind down, and HR consulting firm in Silicon Valley. Whether you are a startup, a mid-sized business, are ready to go public, or are a nonprofit, when it comes to finance, accounting and HR, Ravix will prepare you for the journey ahead. To learn more, please visit their website at https://ravixgroup.com/

Hood & Strong, LLP — Hood & Strong is a CPA firm with a long history of working with search funds and private equity firms on diligence, assurance, tax services, and more. Hood & Strong is highly skilled in working with search funds, providing quality of earnings and due diligence services during the search, along with assurance and tax services post-acquisition. They offer a unique way to approach acquisition diligence and manage costs effectively. To learn more about how Hood & Strong can help your search, acquisition, and beyond, please email one of their partners Jerry Zhou at [email protected]. 

Oberle Risk Strategies– Oberle is the leading specialty insurance brokerage catering to search funds and the broader ETA community, providing complimentary due diligence assessments of the target company’s commercial insurance and employee benefits programs. Over the past decade, August Felker and his team have engaged with hundreds of searchers to provide due diligence and ultimately place the most competitive insurance program at closing. Given August’s experience as a searcher himself, he and his team understand all that goes into buying a business and pride themselves on making the insurance portion of closing seamless and hassle-free.

If you are under LOI, please reach out to August to learn more about how Oberle can help with insurance due diligence at oberle-risk.com. Or reach out to August directly at [email protected].

Interested in sponsoring?

(00:04:05) Brian’s passion for data

(00:05:55) Data is the new oil

(00:08:39) Categorizing data companies

(00:13:42) The importance of proprietary data to investment decisions

(00:16:49) Attractive API businesses

(00:22:19) Data Legislation

(00:30:05) Finding the right founders & companies to back

(00:38:10) What non-venture CEOs can learn

(00:41:06) People data

(00:46:45) What strongly held belief have you changed your mind on?

(00:48:17) What’s the best business you’ve ever seen?

Alex Bridgeman:  Brian, thanks for joining. One thing I want to learn more about is like why was data so interesting to you? You’ve talked about data and AI being your focus for last couple of years as a venture investor. Why did that start being so interesting to you? And how did that begin?

Brian Yormak:  Yeah, I’ve actually never even thought about it. So now I’m thinking about it. I would say there’s probably two things. One is so I started out in investment banking, did not particularly enjoy it, ended up applying for, getting a job at a firm called Fontinalis, which is a Detroit based venture capital firm started by Bill Ford, focused on smart transportation, autonomous vehicles, things in that world. I don’t even know if I had an opinion on data before I started there. I would say when I think about it, I view data to be logic, so it’s just like codified logic, digital logic. And I’ve always been interested in logic and logic puzzles and logic games, so my brain just naturally gravitates towards what can be done with the underlying information. And then I think what happened was, when I was focused on autonomous vehicles, I recognized that the way an autonomous vehicle works is by leveraging data. So it’s leveraging camera data, LIDAR data, radar data, IMU data. And I think the thing I thought was so interesting is that the machine was leveraging real world data, and that felt incredibly novel, and it felt like something that hadn’t been done before. And so, when I spoke with my boss at the time, the way I articulated it was like there’s a merging between the real world and the digital world. And he never really understood, I think he kind of got it, but he never really understood what I was saying, probably because I didn’t articulate it particularly well. But there was some trend there that was, wow, this is a very interesting moment in time where we can codify a lot more to derive insights in the same way that a person derives insights. And I’ve just latched on to that trend for the past decade or so now.

Alex Bridgeman:  Yeah, the phrase like data is the new oil has been kind of tossed around here and there. Where do you feel like that’s true and not true?

Brian Yormak:  Yeah, I think I wrote something on this a while ago. So, I’d say I built like a whole narrative in my head around like you mine oil, so like data exist, data is kind of like unrefined oil, like what’s in the ground. Then you have these refining pumps, and you have whatever mechanisms to transport oil. And so, I think, to me, it’s actually a fairly apt description because data and oil in their initial raw form are useless. It is kind of like everything from that initial form to the end use that is super interesting to me, which is like how do we pull the data from the system? How do we refine it so that it’s super useful? And so, we spent a lot of time in like that middleware layer, which is like, okay, what is the data? What is it relevant for? How should it be used? How is it normalized? What is the frequency of usage? And I think like oil, there are a lot of different use cases for it. So, you can refine it in a lot of different ways based off of the end use case. I think it’s pretty apt.

Alex Bridgeman:  Yeah. Is there anywhere where that’s maybe overblown or data is less helpful or less relevant?

Brian Yormak:  Well, I think it’s oil exists in kind of like a fairly defined way. Like there’s a certain amount of oil that exists in the world. And now we figured out ways to extract more oil through things like fracking, so maybe it is near infinite, but there is theoretically like a finite nature to it. Data is infinite, more or less infinite. And so, there’s a lot more data that is like completely useless for a variety of reasons. And so, I think it’s like data is an endless resource, which to me means that it’s far less valuable in almost all instances but can be extremely valuable in a narrow set of instances. And so, I think it’s really just like a definition of the product set. And people talk about, the one example that always comes to mind is people talking about like connected vehicle data. And there were a bunch of companies that popped up in that space. And the reality is that I think McKinsey came out with a report that connected vehicle data was going to be worth $20 billion. And then it’s like, okay, bbut what does that mean? Like anonymized connected vehicle data, like across a fleet, like who was paying for that and why. And so we looked at a lot of investments in this space, we ended up investing in a company called Motorq, connected vehicle API. But the insight for them was it was not about vehicle data broadly speaking, it was about individual vehicle data, which is okay, I’m a fleet operator, I want to know where my vehicle is, the performance of the driver, or the health of the vehicle, etc. The aggregate wasn’t particularly relevant, it was actually on a per vehicle basis. And so, I think a lot of people get the concept wrong because they don’t think about the actual utility of the data; they just kind of stop and hope data is valuable, which leads to a lot of business failure.

Alex Bridgeman:  Yeah, that’s kind of an example of the one category data business we talked about, which is taking this previously kind of unavailable, untapped pool of data and making it online and making it available. We talked about a couple other categories, that being one of them and another being data that’s maybe unorganized that a company can come in and organize for someone to use, maybe it’s their own internal data. But broadly speaking, how do you group data companies by category or type of data or however you categorize them?

Brian Yormak:  Yeah, so I think this gets to our thesis as a firm, so I’ll kind of walk through how we got to our thesis because it breaks down the categories. So again, if you think about an autonomous vehicle, to me, there were kind of three core technological thematics or three different components to that vehicle. One was the data capture system. So, the fact that the car now has a camera or radar, LIDAR system means that you could capture data from the real world and structure it in a form that is usable but that has a hardware component that is actually pulling that data in. You then had improvement in the middleware layer. So, you had movement from on prem compute to cloud compute, and then in this instance, movement of microprocessors to the endpoints of the vehicle, which means that you can generate a ton of vehicle- a ton of data from that asset, and you can push it to the cloud, so you have nearly infinite storage to be able to process that data. And then you have improvements at the AI or the algorithmic layer, so ability to apply algorithms to now these massive real world data assets to infer insights from that data. And so that’s how we break it out, which is there’s creation of data, organization and orchestration of data, and then algorithmic based applications that sit on top of that data. Now, a lot of companies have multiple crossovers. But for us, we consider a hardware-based solution to be in that data creation world, which is how do we think about the sensors needed in this environment to start digitizing that asset. We think about API companies as kind of that middleware layer, which is okay, this data asset has come online or multiple data assets have come online, so how do we think about portability and normalization of that data? And then we think about AI broadly as kind of the application layer that sits on top.

Alex Bridgeman:  Do you think those companies are the most interesting to you, the ones that are taking new datasets and making them online either via API or creating that data set themselves, versus companies that are organizing maybe more conventional data that’s been around for a while, but there’s new ways to look at it and analyze it or build analytics on top of it? Do you view both those- one business, one set of companies or opportunities as more attractive than the other, generally speaking?

Brian Yormak:  The way we think about it is kind of sector dependent, depending on the maturity of the sector. So if you take more mature sectors like finance, finance is digital by nature, or largely digital by nature; you’re dealing with kind of a made up construct which is dollars moving. And so, you’ve had digital banking for a while, which means you can go to your Chase Bank, and you can log in, and you can see what’s in your account. And then Plaid and companies like Plaid emerged in the 2000s, which meant that you had interoperability of data. And so, I think there were a ton of new interesting applications that can emerge because you had newly formed access to data infrastructure and more dynamic data. So one of our early investments in this space was in a company called Pedal, which is a cashflow underwriting credit card. Now, the interesting concept is we’ve been relying on FICO throughout, whatever, the past few decades in terms of underwriting risk, but the reality is like there’s a very interesting component to hey, if I can look into your bank account, see cash inflows, outflows, how much are you getting paid, what are you paying for rent, what does that look like, that should be a very relevant barometer as to the creditworthiness of a borrower. And so, they spun up credit card targeting, light file creditors, immigrants, college grads that didn’t have access to credit, but where you had this new underwriting mechanism. And so, I think the application layer is very interesting when you have the new emergence of that data stack within the category. The reality is that most categories are less mature. So like the data assets haven’t been digitized yet. So, if you take, we invest in a company called Nanit, it’s a computer vision enabled baby monitor. And the idea is that there’s a ton of really interesting data going on with your baby, like how are they sleeping, in what position, how are they growing. But that’s never been codified before. So, Nanit is able to continue to iterate on product because they actually can pull the data from the camera. It’s like a really cool product they rolled out is a nightlight, and the nightlight is triggered based off the baby’s sleeping behavior. So okay, we see that the baby actually sleeps from 12pm to 3am, so we want to trigger a light to come on at a certain point. But those concepts where the data has yet to be digitized, I think the hardware component is really interesting. And I do think some of the largest businesses will be built on the heels of hardware centric businesses because they can own the whole thing. So, like you own the data asset because you are the one digitizing it. You own the usage of that data, and then you can build all these AI functions that sit on top of it because you own the data. So, I do think some of the largest businesses will end up being those full stack hardware software solutions.

Alex Bridgeman:  Yeah, and a lot of these you are talking about too have access to a proprietary data set that either they own or have pretty exclusive access to. How important is that proprietary data aspect to any new investment you’d make?

Brian Yormak:  I would say, if the underlying advantage is commoditized, so if you think about like LLMs and building on top of GPT, your edge is not your algorithm, and your edge is not the data if you’re building on top of like accessible data. So, then the question from an investment perspective is like, what is your edge? And I think at the seed stage, we found it to be very difficult to make bets on things where there’s no initial moat, unless you have visibility into how you will develop a moat. So, an interesting concept is we’re going to build a managed service solution to apply LLMs to enterprise because we’re going to be focused on integrations into the enterprise to pull their data to be able to drive the best insights because really what you’re doing is you’re hacking into a proprietary data asset. You’re hacking into the enterprise’s data asset to be able to refine your algorithm in a way that most can’t. So, I think even if you’re starting with something commoditized, I do think it’s important to figure out what it is that you can tap into that becomes from proprietary over time. The companies that start from scratch with the proprietary asset, so Nanit, for example, the really interesting thing about those companies is you have to spend a lot to train the models to be useful enough to make the business look like a software business. So initially, you have humans in the loop that are expensive to waive all that data to be able to get the models to the right spot, which means that you build an inherent moat because you’ve actually trained the algorithms to optimize the data properly. And so, there’s this real friction to competing with businesses like that because they have an advantage in developing on top of an asset that nobody else has. So, we’re definitely more inclined to invest in businesses that either collect a proprietary asset and/or have a path to getting access to a proprietary asset. And so that’s generally what we focus on.

Alex Bridgeman:  Yeah, well, especially too because that company is going to be building from the beginning knowing that they’re eventually going to build another product with their data. So, user permissions, terms, and all the other things are built around that ability. Like, it’d be very hard today for Salesforce to build a data product on top of all of its customer data.

Brian Yormak:  Yeah, right. Plus, the fact that like they didn’t really think about the relevance of building on top of data. So, the structure is horrible. So, it’s like actually having to deal with that underlying data, like there’s massive complexity. We looked at a company called Flatfile a while back, and a big fat fan of the founders. The whole thing was building these data interoperable API, which is how do we normalize data from one source to another, so we’re going from Salesforce to some other CRM, the portability of that data is quite difficult when Salesforce doesn’t really want to allow it but too because the structure is pretty poor. And so how do you think about kind of like these middleware normalization layers so that data is usable? And I think we’ll see a lot of companies in that world bringing kind of like V1 cloud companies into V2 cloud companies where people recognize that you want to have the ability to leverage the underlying data, not just for visualization, but for other functions, primarily like analysis.

Alex Bridgeman:  Yeah. You’ve talked about API businesses a bunch. Would you just help by defining an API business and some characteristics of attractive API businesses to you?

Brian Yormak:  Yeah. So, the way I think about it is that there are data sources that exist. And in most markets, there are different vendors of data. So, if we take auto, for example, we invest in a company called Motorq, which is a connected vehicle API, you have all vehicles are now built with inneractive connectivity, which means that they are generating data. But the way that Ford’s vehicle generates data is different than the way that Toyota’s vehicle generates data, which is different from the way GM’s vehicle generates data. And then you have a fleet management company on the other end or a company like Avis. They have vehicles from all those different automotive companies. They want to be able to access data on that vehicle, which is where is it, what is the performance of the driver, what is the fuel level, things like that, but they have no technical ability or limited technical ability to pull that data from the respective vehicles and normalize that data so that they can see it in a dashboard across vehicle types. And so Motorq builds a connected vehicle API. They build integrations into all the automotive companies, Ford, Toyota, GM, so they can pull data from the vehicle. They agree with the OEMs on the frequency of that data, the depth of data. So like, how often am I pinging that vehicle or pulling data from the cloud, what is the package of that data. They then normalize all that information and pass it along to the end users. So that’s generally an API business. And similar for Plaid, like Plaid plugs into all different banking institutions, pulls the data, normalizes that data for an end use. And we see that across some sectors. We invest in a company called Particle Health, which does it in the healthcare space. I would say the properties that are particularly interesting about API companies, generally, all of them are interesting because they’re really just repackaging versions of the same thing over and over again. So, they’re immensely scalable businesses. Like, you have to maintain the infrastructure, but it’s actually quite easy for a fairly broad range of end users to leverage that data because it’s really just a function of like packaging and frequency, which is not that hard to tweak once you’ve built supply side integrations. So they’re really scalable businesses. I would say the thing that we’ve recognized is that the fewer integration partners that the supply side wants, the more powerful. So for Motorq, again, Ford, Toyota, GM, they don’t really have interest in servicing and supporting a bunch of partners integrated into their data. So, it’s not going to be an open ecosystem; not every company can go to Ford and say, hey, I want to access your API. What that means is that Ford is going to pick one of very few companies to actually integrate into their systems, which means that there’s only going to be a few companies like Motorq that exist because Ford doesn’t want to manage those relationships. So, you have this captive supply access, which is really powerful. And then you obviously want like a large end market of demand across use cases. But the captive supply component is actually probably the most interesting thing where if there aren’t going to be that many companies that can build those integrations, you sit at a really interesting choke point of value.

Alex Bridgeman:  Do you have other examples you can think of of that kind of constrained supply that you just mentioned?

Brian Yormak:  I look for them. So, I would say one that is not so much this, but they are their own dynamics, so we invest in a company called Particle Health, it’s an API for electronic health record data. So you go to the doctor, they input information into their digital system, that’s called an EHR. The interesting thing with EHR is that historically, they don’t speak to each other. So, you go to a hospital for a special surgery, you put data in there, or a doctor puts data in there, you then go to Mount Sinai. The doctor at Mount Sinai doesn’t know what happened at the hospital for special surgery, which is why it’s so annoying you have to provide all of your data every time, which is like crazy. There was something called the 2020 Cures Act, which basically mandated that if a patient requests access to their data, then the EHR systems have to provide that data. And so, because of that, there was now this opportunity for an interoperable layer to emerge, which is to plug into Epic, Cerner, All Scripts, all these EHR companies then pull the data from those systems. And so, what will be great is over the next five to ten years, when you go to the doctor, they’re actually going to have your data, which will be really nice. The interesting thing in that market is that consortiums form. So, one is called Care Quality, the other one’s called Commonwealth, basically groupings, nonprofit groupings of these EHR systems that Particle Health integrates into. There’s broader access there. The interesting thing within that market is the complexity of data is where you can build differentiation. So, if you think about automotive data, it’s incredibly structured. So it’s easier to streamline in terms of how that data is used. For EHR data, there are doctor’s notes, there’s different structures as to how doctors input information. And so a lot of their work is around the normalization and leveraging algorithms to normalize that data so that it can be usable. And so there’s less defensibility on the supply side; there’s more defensibility on the know-how of how to process that data and make it usable for an end user. So that’s just like a different dynamic. In terms of supply, we haven’t seen another one like Motorq where there’s a real constraint on access to supply. There are interesting opportunities in real estate, a lot of other like built world things. But we haven’t actually found one where there’s like a very natural reason as to why there’s a supply constraint on the number of vendors that will sit there.

Alex Bridgeman:  Yeah, it’s interesting, too, that you mentioned that Particle Health is- part of the catalyst for that business was that legislation that mandates you can have access to your data, and that kind of paved the way for this to become a more interesting opportunity. Do you see that as a kind of continuous momentum from regulators and governments who over time want, generally want data to be more and more available overtime, thinking even of like SEC rule around ESG requirements becoming more and more common, this kind of like ever going push for more transparency and more open data? Do you see that as a trend and something that’s going to continue potentially?

Brian Yormak:  Yeah, I would say I think there’s two trends. And I’ll reserve my opinion on whether or not I think it’s good or bad, but these are trends that I see. One is all around open data. So, we’re seeing- so in the EU, there’s already open banking standards. I think we’ll see the penetration of that into the US ecosystem. And for the most part, the push is around consumer control of data, which is the ability for a consumer to control data. But in order to do that, you do have to create better data interoperability functions because a lot of these companies don’t even have a mechanism to share that data. So within healthcare, for example, something called Fire, which is like the fast healthcare interoperability, R, I don’t know what the R stands for. But that is now like the new common language that they have to work towards so that there can be interoperability. So I think that’s a really interesting trend because you have these common languages that start to form. And then I think ESG is another one. I just spoke with a company which I won’t mention, but super interesting, where because of the regulations and because of kind of like subsidization that’s kind of been tied to regulations, which is kind of like a carrot and stick dynamic, like if you don’t comply, you’re going to get fined. If you do comply, there are tax credits and things like that. There are definitely some really interesting businesses that are going to form that are like make money by doing good, which the government is definitely incentivizing right now, which is, hey, we can save you money if you are compliant with ESG standards and climate standards. And I think it will create some really interesting business opportunities where it’s like, hey, I’m going to sell a commercial product that helps your business, but the result of that will be that you have a lower carbon footprint. I think there’s some really powerful dynamics there that will be interesting to track.

Alex Bridgeman:  So if you’re a data investor, like yourself, and you want to be on the frontlines of any future opportunities, is upcoming regulations or bills in the process or any legislation coming out, is that something that you’re paying close attention to that’s important to be on top of?

Brian Yormak:  Yeah, I mean, I’d say like when we talk about what we invest in, we say that the why now is incredibly important. Like there needs to be a why now, and there either needs to be a technical why now or regulatory why now. And I would say the vast majority of the investments we’ve made have been technical why nows, but Particle Health is an example of a regulatory why now, legislation around onshoring. We invest in a company called Via Photon; they assemble and manufacture fiber optic cable that plugs into data centers, 5g towers. A lot of the thesis was that there’s going to be subsidization for onshoring of critical infrastructure. Plus, we’re obviously bullish on physical infrastructure if we believe in the proliferation of data. But the regulatory components definitely create interesting tailwinds. Candidly, we don’t have- I would honestly want to spend more time, like a lot of what we learn is kind of reactionary, which is that somebody tells us we’re building it because of this reason. I’m like, oh, that’s really smart. Now, the nice thing is when a founder tells me, I’m like you’re the type of founder that I want to back because you’ve done that work. But we definitely, I definitely think it is worthwhile to explore, and there are lots of really good businesses to be built. Like SOC 2 is another one. Like when that happened, you had companies like Vanta that blew up quickly because there was just a regulation that came down that required that companies get SOC 2 compliance, and large companies were pushing that mandate. And so you had these really quickly scaling businesses in that category. I think it’s a really good trend to try to jump onto.

Alex Bridgeman:  If you think of the ongoing research that you’re doing around data opportunities, it sounds like it’s very network driven. And you’re looking for founders who are seeking out these opportunities. Does it happen the opposite way where you spot a trend, see something interesting, see new technology, new legislation, what have you, and you go find the right founder to go start that business?

Brian Yormak:  Yeah. So I’d say that there’s research driven firms, folks that we think super highly of that are putting out content around trends, soliciting founders for those trends. I would say that we are thesis driven but not proactive about trends for the most part, which means that like once we see something, we will often latch on, and we are probably learning about what’s happening in the ecosystem because of the types of founders we speak with before most. But we are not defining something that we absolutely want to find. We have toyed around with incubation and building businesses, especially in healthcare because of the complexity of the category. But for the most part, I wouldn’t put us in that like, I would say we build a good brand so that we get a lot of inbound for folks that are building data driven businesses. And so, I think that’s a really good sourcing function for us. But we are not leveraging marketing to really put a stamp on like research centric approach to say we want find X, Y or Z business.

Alex Bridgeman:  And is that a function of size in that if you had multiple funds or a larger set of resources, you could do more of that work, and it’s just the early days, and it’s hard to get there at this point, or is that like a strategic decision as well?

Brian Yormak:  I think it’s both. So, I’d say I generally describe venture firms, there’s a scale from investors to marketers. And what I would say is, the more you are a marketer, the more you are likely motivated by expansion of AUM, so like assets under management, the size of your fund. The reality is I generally believe there’s like an inverse correlation to fund size and your ability to drive returns, like to a point, there’s a certain saturation where this is the right fund size to drive the best returns. And so for our strategy, for example, I think the ceiling on our strategy is $150 million. I think above that, you would actually dilute your returns because you can’t deploy enough capital through the strategy properly. I think marketing firms will lean into visibility because it allows you to raise capital to deploy into strategies, which is completely fine and viable for a lot of reasons. We definitely lean into the kind of investment side of this, which is we’re very focused on driving best in class returns. What that means is that the time we spend is focused on optimization of returns. So that’s time spent with a portfolio, that is time spent on the BD side, that is time spent on refining our diligence process to make sure we’re selecting the right company. I would love at some point to have time to be a little bit more prescriptive about what we’re looking for. We just haven’t prioritized and given what we thought was most important to drive for returns. The only caveat I’ll say to that is I do believe there are research specific firms that are making investment decisions based off of their research, and I think that’s a really good strategy. So there’s folks like Compound or Equal Ventures where they do a lot of research that I think actually really helps them to make a decision. I think it is particularly helpful in areas where the research is kind of a front run for your diligence process. So, let’s say you’re looking into a new LIDAR sensor, there’s a certain amount of homework you have to do to have an opinion on the best LIDAR sensor. You’re either doing that reactively in a diligence process, or you’re getting ahead of it to say hey, I want this type of team. And so for like R&D centric investors, true R&D centric investors, I think the research component can make sense because you’re kind of front running a diligence process anyways.

Alex Bridgeman:  We talked earlier about the right founders to back, especially the ones that can not just go from zero to one but one to n and beyond. But can you describe, like maybe it’s helpful to frame like the various types of founders that, generally speaking, you partner with and are good in certain parts of the business or certain growth stages versus others.

Brian Yormak:  Yeah. So I think the first very important dynamic to understand is we are looking for venture returns, and venture returns necessitates a far narrower subset of founder type. I think people have talked about this, around this, which is like venture is a very specific type of game that is driving towards a very specific type of result, which is the largest possible outcome possible because you’re selling a portion of your business to create a larger pie. And that does not fit for most people. So when I talk about founders, I’m only talking about the profile of venture founders that I think works really well. For those folks, I think there’s like a few core underlying traits specifically for the types of businesses we invest in. Like, I don’t have a strong opinion on a consumer founder or other types we don’t do but for like b2b complicated businesses, the things that we focus on are mental acuity, just how sharp are  they. Like we think that you need to have a certain level of natural intellectual horsepower to be able to deal with the complexity of problem solving within the b2b universe. I would say communication skills, which is not only how do you articulate things, but how do you motivate the people around you to communicate. You need to be able to raise- in order to raise capital, in order to hire a team, in order to motivate people, you need to be able to communicate well. And there are lots of different styles of communication, but you need to be able to communicate well and with clear intent. I would say that there’s a certain level of obsession for venture founders in particular. Like a joke I make that is kind of a joke is I will sometimes ask founders do you have a hobby, and if they have a hobby, we will pass on the investment. And it’s kind of a joke, but there is a real aspect, which is that the founders we’ve seen build the best businesses are maniacally obsessed. They are not balanced people. And again, I support balance, but for the founders that we’ve seen be most successful that build really large organizations, they are naturally obsessive people, and their obsession is towards their business and whatever needs to happen there. And so we definitely look for kind of like obsessive qualities. And then I would say we’re generally looking for some level of industry knowledge/strong commercial understanding. We’ve invested in first time founders, really talented, we will do that still. But ideally, you’ve been in the category, you have an opinion around a unique insight and the workings of that category because you’ve been in it long enough. And so ideally, you have somebody that has that experience, where they don’t have to learn, but they actually just know what they’re trying to execute towards.

Alex Bridgeman:  You also touched on the various stages of a company as it grows and that early on, it’s very problem focused, customer focused, and over time you’re managing this larger growing organization. What are some maybe characteristics or traits that work well in both?

Brian Yormak:  So I think, to your point, everything I outlined, that applies in both stages. There is one clear distinction we’ve seen, which I’ll define as like zero to one versus one to n. And zero to one is like going from nothing to initial product market fit. And then one to n is going from product market fit to scaling of business. The big distinction we’ve seen is that a lot of founders- founders need to be obsessive. They can be obsessive about one system or many systems. And a lot of founders are obsessive about one system, which is that figuring out product market fit, small team, very action oriented, which is how do I test as much as possible to figure out the right way to start scaling this product, what channel am I selling it through, or how am I building a direct sales motion? What exactly is the product message? What exactly is the product? That is all defined in kind of that zero to one phase. I think what we’ve seen is that one to n requires a shift in what system you’re obsessed with, and the system is actually organization building, which is okay, now that we have product market fit, how do we build our sales team, how do we build our marketing team, how do we build our product team, tech team, etc., how do we build the right structure so that those teams are communicating, how do we build the KPIs to measure that performance, and you definitely shift more into a management function versus like a doing function. And a lot of founders don’t particularly like that system. So I’d say the founders that have the ability to go from zero to N are just obsessed with systems and obsessed with winning the game of building their company, which means that as soon as they get to product market fit, they shift in terms of what system they’re focused on, and so they can continue to build the best systems on a go forward basis. The founders that decide to step away or are replaced at a certain point are often the founders that realize they really just want to do that zero to one, which can be immensely talented people but just kind of staying focused within a narrower lane of where they want to spend their time.

Alex Bridgeman:  So if you want to be a founder that continues evolving your style and role and what you focus on and the problems you solve in the company, how do you continuously evaluate what stage you’re at and what skills you need to start improving on or start adapting and maybe others that are leaving behind?

Brian Yormak:  The way I think about it is like capitalism is a defined game. Like the game of business building is how much cash can you throw off of that business. Like the best businesses, Apple, Google, etc., they are valued where they are because they are cash printing machines at this point. And so, you have a very defined goal, which is how do I throw money off of the asset that I have created. I think the progression is, first, you need to find something that the market wants, and that requires you figuring out what that product is, then you need to build the support to continue to sell that in. And so you kind of shift into a management role. And then it’s a function of continuing to layer product and then layering process on top of that. So I think there’s not like a defined set of skills, other than always focusing on what the game is and then reprioritizing what is most important to the business to win that game and looking at your business and saying, where are the gaps, like what risks exist that would result in some failure of this system continuing to grow? So I guess if anything, it’s flexibility, like flexibility and analytical ability, which is to say based off of the where this company is currently, can I recognize the problems? Am I strong analytically? And then am I flexible enough to say, okay, I’m going to shift my time to go solve that problem? So I think those are probably the two things that you need to be able to continue to shift your role to whatever the business needs.

Alex Bridgeman:  Have you found most founders are just naturally good at adapting to new systems? Or is there a certain element of coaching and training or learning from peers that overtime can get a founder from only being focused on the product to now being effective also at the organizational level, too?

Brian Yormak:  Yeah, I mean, our experience, which could be right or wrong, is that it’s fairly innate, which is that the founders that can do zero to n always could do zero to n. It’s hard to identify, but you see over time that they just are more flexible, and they’re willing to jump system to system. The founders that are not have not proven to be particularly good at learning that skill. I think the best founders- the best founders that don’t want to jump from system to system, I think they learn to be really good at hiring, which is I am good at a specific function, I can bring people in to help with the other things. And so, I can bring in a COO that functionally runs operations for the organization. I think you’ll see like founders that move into like an evangelist role, I think that’s often the case. They realize they shouldn’t be running the day to day because it’s very process centric. They can drive a lot of value in evangelizing the brand, speaking to the product, etc. But they’re not going to drive the most value in running the organization. And so I think it’s that self awareness component, which is like where do you fit and what do you want to do, and then proving that you can hire because you’re aware of the gaps that you have.

Alex Bridgeman:  What do you feel like non venture CEOs could learn from some of the founders you’ve worked with?

Brian Yormak:  I think to build a billion dollar plus organization, you have to be obsessive about doing everything to the best of your ability. I think that for a lot of folks, good enough works. And I think good enough can work in a lot of instances. I think the best founders are never settle. Like there is no concept of good enough. It is I have to continue to optimize. And every time I think about what needs to be done, it is not coming from a point of I’m tired but a point of there is something to improve. And I think the best founders, there’s some version of ego tied into it. Like, what do I want to leave behind, and you need to want to leave such a large impact behind to want to build a large venture business because the scale of what you’re trying to create is so large, but in order to do that, you have to be obsessive about consistently pushing the system. There’s no right or wrong. I think if you want to build something massive, just it takes more work and time and effort. And that’s for some people, and it’s not for others.

Alex Bridgeman:  Yeah, a lot of this podcast focuses on search funds, which are non-founding CEOs that come in, acquire, and run this company. And you’ve talked about hired CEOs having a different view and perspective on the business they’re running because they’re not a founder, they don’t have that same degree of attachment or maybe ego tied to the business. How have you seen them behave or run companies differently than founders?

Brian Yormak:  I think the good CEOs that step in are very good, have a lot of similar strengths, really strong analytically, understand the business, but then they’re often really good at process and management. Where it’s like, alright, there’s already an asset that’s here. And so then the question is like how do you optimize that asset, how do you grow that asset, how do you bring in the right people, how do you build the right structures to support those folks? Like the ones that I’ve seen do the best job are really good managers because you have a lot of the infrastructure in place. So now it’s really about optimizing that system. And so that’s where we’ve seen a lot of folks come in. I would say the unknown to me still is, can a CEO that comes in build something as large as a founder? Because again, there’s like a certain level of hunger that comes with starting a business versus stepping into it, and a certain amount of emotional tie to that thing. And there’s, I’m sure, plenty of instances of CEOs that have stepped in that have built really large businesses. But I wouldn’t be surprised if on average, a CEO that steps in to run a business deal can build a really good business but not the same scale of business as a founder because of the amount of obsession that is tied to that thing for a founder. And it’s why we often try to focus on zero to n founders because if you look at, historically, the largest businesses over time, they were founder led for a very, very long time.

Alex Bridgeman:  What have we not talked about or discussed with data that we need to discuss and you think is relevant?

Brian Yormak:  The data around people, like going back to like business building. I think one of the things you see with even data driven founders is that there’s kind of like an acceptance that people can’t be measured within an organization. And I think, generally, we’ve seen the best founders measure everything. So they measure their product, and they see their performance and responses from their end customers, but they also measure their people internally, which is what- first, how do we define what success is for every person within our organization? And then how do we structure KPIs or general guidelines that actually measure that? Because the reality is like a business’s success at a large scale is probably as much if not more about the people than it is about the products. Like you can get to a pretty good size with just a really good product. But to build something massive, the people really are the most relevant system. Kind of the measurement of people, which businesses have started to do more, still feels like an underexplored, underappreciated category.

Alex Bridgeman:  Do you think there’s a data business to be built around measuring employee performance and fit?

Brian Yormak:  I think it would be difficult to build a data business because it feels so bespoke, like it’s so difficult to quantify. I think it’s very hard to have like a one size fits all. I do think that you could have a data driven consulting firm; I imagine there are some out there, which is like, I think old consulting of old private equity were very ops centric, which is like, alright, we’re going to come in, and we’re going to bring people in to kind of know this thing, and they’re going to institute it to drive better performance, to drive better margins, whatever it is. I do believe we’re entering into a period of time where tech tooling can be used to drive kind of like third party support, so consulting support, where it’s like, no, we actually measure everything, we have dashboards, we quantify it, and we can hand it over to you, where it’s not like you always need us as a consulting firm. Like we can build the structure you need based off of your organizational type to start figuring out how to quantify what it is you’re looking to do. And so I think there’s something interesting at the intersection of people supporting plus technology that I think can emerge.

Alex Bridgeman:  Yeah, there’s a friend of mine who bought an interior sensors business. I believe you backed a hospital one that tracks kind of movement of doctors and intel data from patients as well. It feels like a really interesting, like not necessarily performance based, but kind of to your point around measuring people, that’s definitely an interesting example.

Brian Yormak:  Yeah, it’s actually a really good point, which is like as we have more sensors in the real world, performance is also able to be measured. And I think it really applies for blue collar work, or let’s say like non desktop work. So Inspiring, that company you mentioned, the big thing is there’s a nurse labor shortage that’s occurring. But the system hasn’t been optimized. Like nurses don’t know where to go. So they’re just on set calendars when they should be in one room five times the amount as another, but they have no mechanism to understand it because there’s no data being collected from that environment. And so, things like that where you can really optimize your labor force because you have the information is powerful. Another example is we invest in a company called Phood, PHOOD. It’s a computer vision enabled scale that measures waste in retail grocer, so companies like Whole Foods, and the concept is Whole Foods has no idea how much they’re wasting in any given day. They produce brussels sprouts, chicken, but they don’t know how much they’re producing or how much they’re throwing out at the end of the day. What Phood’s product is you can take the brussels sprouts, you put it onto the scale, it auto recognizes its brussels sprouts, auto logs the weight, and so you now have visibility into how much you’re producing and wasting. Now, interestingly, Whole Foods tasked their workers with putting the brussels sprouts onto a scale, then auto logging that it was brussels sprouts, and auto logging the weight. And the reality is they just had no compliance. And so in that instance, you just weren’t going to be able to get the worker to do that thing, so you have to automate away some of those tasks. And now that you’ve done that, you have a lot more digital information, which means that Phood, PHOOD the company, can plug into the supply chain to change ordering behavior, which is we need fewer brussels sprouts on Wednesdays because of purchasing behavior. So you can start to automate a lot of the work that is currently done manually by digitizing real world assets in some way.

Alex Bridgeman:  Yeah, speaking of Whole Foods, there are these carts now where you can start scanning your items directly from your cart, and then you just check out there on your phone. I feel like there’s potentially data businesses there to be built too.

Brian Yormak:  Yeah, I mean, we looked at a company called Caper, which was bought by Instacart for a large amount. And that’s what Caper was building, connected shopping carts. We ended up passing because the market is super low margin, and it’s a very hardware intensive product. The CEO had an awesome outcome, and he’s exceptional, but there’s definitely going to be- as much as can be automated will be automated. Like it’s better for business. And I think it’s an interesting thing that’s not discussed enough, which is like capitalism definitely incentivizes automation. And so that will occur. Now, I think it creates a ton of opportunity, so I’m generally bullish on it in terms of like it’d being a net good for people, but it is going to happen, there will be displacement. So I’m very bullish on the trend. But I do think that we also need kind of like failsafes in the short to medium term because I do think people- we will continue to see dislocation of work as we continue to layer in automation.

Alex Bridgeman:  Moving to closing questions, what strongly held belief have you switched your mind on?

Brian Yormak:  I used to believe that every organization should do the best with the people that they have. But it was kind of like a reactive approach, which is okay, you hired somebody, and now how do you help that person to do the best within that environment. And I still generally believe you should do that if that’s where you’re at. But I think it is far more important to be proactive about hiring, which is to be very clear about culture. And so, I think the reality is like there are certain businesses where it’s going to be incredibly competitive and incredibly oriented around output, where to be successful, there’s a certain profile that that requires, and I think the best businesses are very clear about the types of people they want, so that the people that come into the door are aligned with that culture. I think where we’ve seen a lot of folks struggle is in not being clear about that, where the CEO wants to drive harder but hasn’t been clear enough with prospective hires, where they want more balance or whatever it is. And I think people are often afraid to be that direct about what they need within the organization. But I end up- I think it ends up being a disservice to both the leadership and the employee because they’re not a fit, they should have been told that they weren’t a fit, and there are places they will be a fit. Like there are environments that do value work life balance, but I think that mismatch I’ve grown to really appreciate over time because it happens often.

Alex Bridgeman:  What’s the best business you’ve ever seen?

Brian Yormak:  I wanted to come up with something clever to this, but I think it’s like Apple’s App Store just like has to be. Like it’s, they have a monopoly on the hardware. And there is only a single point to transact for applications. It is the absolute best business you can imagine. You have a large portion of the world that is locked into your system. And you are the gatekeeper to all the functionality of that system. I mean, I think it’s why it’ll probably be one of the first businesses that’s regulated as a monopoly because it’s a monopoly. Like they charge 30% in the app store, they could charge 60%. It wouldn’t matter. People would continue to do it because there’s so much money through that channel. So I would say anything that lends towards a monopoly is likely some version of the best business. And it’s probably like the best monopoly that has ever been created.

Alex Bridgeman:  Yeah, pretty close. There was a fun chart I saw yesterday of iOS or mobile OS market share over the last 10 years or so. And Android and Apple are, of course, at the top, and you see like BlackBerry kind of flatlining pretty quickly. But then, Android and Apple move in opposite of each other. So one’s up, the others down. And they recently flipped in exactly like March 2020, like so at the start of COVID, iOS and Android flipped. So iOS went from second place to first place and has been like gaining share from like 45 to 55% share. So that is pretty remarkable. I’ll send that chart to you. You might enjoy just flipping through it. But Brian, yeah, thank you so much for coming on the podcast. It’s super interesting to chat with you and excited to hopefully see you in New York in a couple months here.

Brian Yormak:  Yeah, this was awesome. Thanks so much for having me.