422. How AI Will Transform the Future of Marketplaces, the Shift from AI Co-Pilots to Agents, & Frameworks for Investment Timing (Pete Flint)

422. How AI Will Transform the Future of Marketplaces, the Shift from AI Co-Pilots to Agents, & Frameworks for Investment Timing (Pete Flint)


Pete Flint of NFX joins Nate to discuss How AI Will Transform the Future of Marketplaces, the Shift from AI Co-Pilots to Agents, & Frameworks for Investment Timing. In this episode we cover:

  • Founding a Successful Tech Startup and Industry Insights
  • Founding NFX, Software Approach to Venture Firm, and Orthogonal Initiatives
  • Venture Capital Investment Strategy and Philosophy
  • Leveraging AI in Marketplaces for Efficient Operations and Novel Consumer Experiences
  • AI-Driven Career Counseling and Autopilot Technology
  • AI Value Creation and Distribution
  • B2B Marketplaces, Defensibility, Investment Criteria, and AI

Guest Links:

The hosts of The Full Ratchet are Nick Moran and Nate Pierotti of New Stack Ventures, a venture capital firm committed to investing in founders outside of the Bay Area.

Want to keep up to date with The Full Ratchet? Follow us on social.

You can learn more about New Stack Ventures by visiting our LinkedIn and Twitter.

Are you a founder looking for your next investor? Visit our free tool VC-Rank and weโ€™ll send a list of potential investors right to your inbox!

Transcribed with AI:

0:18
Our guest today is Pete Flint, co-founder and General Partner at NFX.
Pete is the only guest that I can recall who has both founded a company that exited for a $B+ and a firm that manages more than $B+ in AUM.
As an operator, Pete was on the founding teams of both lastminute.com and Trulia, both of which were acquired for over a $B+. After his days building internet marketplaces, Pete founded NFX in 2016 to invest in them. Pete, welcome to the show!
0:49
Thanks. Great to be here.
0:50
We’re gonna spend most of the time today talking about NFS and your views as an investor. But prior to diving in, I’d be remiss if I didn’t talk about your background as an operator, as you’ve had two multibillion dollar exits. So can you share with us the founding story behind Lassman? And then how that eventually led to founding Trulia? Yeah,
1:12
sure. So as you can detect from the accent, I was born in the UK, and I kind of uh, when I was at university, I, I basically was got passionate by the internet. And just the power of it. And it was, this was kind of a mid 90s. I showed my age there for the I ended up writing to like every internet company in the UK, one for internships of which there was 17 at the time, and anyway, I got a job at one and met some people there, one of which was a guy called Brent Hoberman. And we ended up sitting next to each other. And then when he started that company, lastminute.com, he asked me to join us, part of the founding team, along with Martha Linfox. And so that that’s just a crazy story from you know, and true.com Mania from like, launching and, like 1998 going public, literally 80 months later, you know, people talk about going public for 10 years afterwards, we went public 80 months later, at the peak of the NASDAQ. And it was travel marketplace. And then, and then we had to navigate September the 11th. You know, the next 18 months was basically the.com For ligaments and deliverance, which is disaster. And then I and then I left in 2003 when we’re about 2000 people profitable. And I said just a crazy journey. And just you know, I was in my early 20s And, you know, wearing many different hats, but what, what incredible ride. So I, you know, I left, really to go to Silicon Valley, I was fortunate to get accepted at Stanford for an MBA, which was, you know, just, for me a way to physics undergrad sounds like a really, and last minute was a great learning environment, but not, you know, it sort of taught me a kind of one, one type of business, a truly crazy environment. And then and then doing an MBA at Stanford just got me that got me the ticket to Silicon Valley, which was really important to me. And so and then then I then I started Trulia, it was in the second year between the first and the second year, there were to build an online real estate marketplace. And so that was another 10 year journey, so many so many lessons and journeys from, from whole that whole experience. And then and then Trulia ended up merging with Zillow 10 years after we launched it. Two public companies come together to create the world’s largest real estate marketplace. So really, really fun.
3:50
Yeah, well, you’re you’re being modest because it was a three and a half billion dollar exit to Zillow. But what was what was the founding moment for Trulia? Like, do you recall the insight or the collection of insights that led to founding the business? Yeah, it’s,
4:04
it’s there’s a sort of different periods of triangulation, though. And specifically, I was, I was, you know, as a two year program, in the first year, you know, you put you off in campus housing, and the second year you’re on your own. And, you know, the way that it works, typically works with like, there’s a group of people that sort of self organized and say, Yeah, do you want to live together and second year? And we said, Sure, and so there were like five or six of us. And we said, yeah, we’ll friendless live together. And, and it was born out of the six went off to like New York and London to go and do their summer internships and me and classmate of mine, Ali, we were like, we were tasked with finding somewhere to live. And I would like Okay, here we are. We’re in Santa Rosa Silicon Valley in Santa Rosa, the center Epi Center globally of technology. We need to find some deliver k right You know, we’ve got Google down the road, you know, Apple, and Facebook was just getting going. I was like, Okay, where’s the tour to find the place to live and asked around and said, Well, there’s two ways to do it. One is you go on Craigslist. And the other is you speak to a real estate agent, or just floored, that is like, this is the best you got. And I was just like, this floor, this shouldn’t be the case. And so and so I, you know, and I, you know, that was a sort of this is interesting to me. And then the other piece, because I’ve worked in travel, you know, the, the core product experience, you know, that the actual moneymaker travels, most people don’t out with hotels, not flights. And actually finding a hotel is a million miles away from finding somewhere to live, as you know, as Airbnb is found out. And so that kind of whole mindset of like, okay, how do we organize the product? Was the secondary information beyond the listing to help you make that discovery? How do you aggregate all the listings was it was a industry that, you know, it’s very different hotels versus real estate. But I could sort of wrap my head around, how do you take the client side and bring that onto the platform. And then also, I could figure out the demand side as well, because that was my job at last minute to be the sort of what was latterly called the kind of head of growth. And so I kind of knew not only is this a big problem, there’s not many incumbents, but I was quite unique, you suited to actually figure it out?
6:32
Do you? Do you think you would have recognized the opportunity for Trulia or had the success that you did with the business without lastminute.com? And that might be impossible to answer definitively. But how much did lastminute.com Prime you for the success? The recognition of the opportunity? And then the ultimate success? Of? Good
6:53
question? I think it did, I think it gave gave you it sort of like, gave me two things, quite honestly, they gave me one was just basically a a playbook for how to build a, you know, high scale high growth internet company, which I’d never seen is really my first job out of undergrad. And so it’s like, and so it gave me a playbook. And the elements of that playbook that last bit, they got really right and a bunch of stuff, they got really wrong. And so you could sort of like, okay, this is, you know, this gave me a playbook, which I think today, there is so much startup content on the internet, right? It’s like, you can spend hours listen to podcasts in incredible detail and YouTube videos will get consumed every paper, but back then there was no playbooks that you could consume. And so I found that I found it very kind of helpful from that perspective. But, you know, I think you can find substitutes for that now, I actually think the bigger one was, was one not of experience, but of like, honestly, lack of experience, I think the, you know, it gave me the confidence that yeah, I can, I can, you know, I’ve seen this, I’ve been there. I can do this, but actually is the sort of the inexperience though, naivety, that immigrant mentality of like, coming in, you know, I have no idea where Illinois is, like, you know, you know, I literally just been in the US for a year. And we’re just like this thing, and, you know, I spoke to all the professors about it at Stanford, this was going to do and they were like, No, that’s not going to work. You know, they were, you know, they would like, you know, that’s why their professors, right, because they’re sort of tenured, and they don’t take risks, but they, you know, there’s definitely a sort of a sort of a perception that this will be challenging because regulation, this is the way it’s done. It’s slow moving. And I think they actually outside in view, great naivety.
8:53
Yeah, I could ask you 20 More questions on on all this. But I There are a number of topics that I think you’re uniquely qualified to talk about today. I guess I’m I’m very curious of how you ultimately made your way to founding NFS. Because clearly, you don’t need to find a venture firm with all the success that you had as an operator. So you must really love working with founders partnering with founders. But we’d love to hear what transpired what led to founding the firm.
9:20
So So I came out of Trulia, Zillow in 2015. And in I was, it was a 10 year math and I was kind of burnt out at the end is like a whole bunch of antitrust regulatory review, kind of bring the company together. Combined, the business is like, Okay, I’m gonna take a year off and just like in a very young kids, and so we’re like, Okay, I’m just going to sort of recharge and I did a lot of kind of investing and advising, you know, you just get poured in because you have free time and so little advising. And so and I was thinking about should I become another just start a company should I join a company should I do go into investing. And it was during that time I reconnected with James courier and giggy Lev Iwai. So I’ve met previously. And, you know, if you the track of sort of my career has really been about in my investment thesis was like, really early, because I’m kind of like uniquely suited to the kind of the product market fit the founding stage. I can do spreadsheets quite well, but I’m like, I’m just, you know, I just I have a sort of a superpower that really early stage. And I love network effect businesses, you know, like marketplaces, last minute, Trulia, Zillow, they were just like this. And once you were in, like a company with a lot of network effects, you just realized, like, there was no other interesting, sort of like, Oh, you, you don’t want to be in any other type of business, you just want network effects. Because you know, the power of it is like, if you can get that initial velocity, then scale you do great. And James, and giggy, had exactly the same point of view founders as well, operators, investors, and a passion for network effects. And so they were getting sort of accelerated going, and then and then I joined to help to turn it into the institutional fund it is today. So we, you know, and it’s actually this perfect situation of, for me, at least of of investing, and supporting founders. And obviously, we kind of like out of the operations of those companies. But we get to run a firm, and we choose the culture, the product, people kind of all that boring stuff. Which is really fun.
11:38
Yeah, in the firm does not look like most venture firms. I mean, you have a head of engineering had a product ahead of content. Can you explain that the strategy behind these initiatives that you undertake, at NSX, and specifically the ones that are orthogonal to investing activities?
11:59
Yeah, I think that this sort of observation that I’ve found raising money from, you know, the best, in a classically the best venture firms in Silicon Valley, we’re fortunate at Sequoia and excel as lead investors at Trulia, and then on the board with Zillow with benchmark and many others. And it’s, you know, it felt that the perspective and organization of these firms is incredibly analog. And yet, they’re advising all these companies to be incredibly digital. And it was almost like, there’s no technology, and most of these firms, and we felt that, okay, to, to if we really want to practice what we preach. And we really think we could do these difference differently. We want to augment the human component, which we think is necessary. And with a strong backbone, and infrastructure and data, and software. And so we started the firm, in 2017, with the head of product were the head of engineering, in a building a bunch of internal products to help us with our management of the deal flow with our evaluation with diligence collaboration, or the rest of it, and a bunch of external software tools that we we have to kind of help to build the ecosystem, just to and help to provide free free services to founders. So we literally think about it like a software company to give us that advantage. And it’s, and this was the approach back in the sort of inception. And so we run it, you know, we literally run like a software company, we have KPIs, we have teams, we have metrics, we’re building proprietary advantages out of their software. And it’s and it’s what is really interesting is that that thesis that we had is back into them 70 software and data can enable a an advantage, institution advantage, a venture firm, where we bring it up to today. And we’ve seen AI start to kind of really transform that every industry. And we think that’s sort of the backbone of kind of data and institution knowledge that we built for email conversations, or the index that we’ve seen, or the diligence calls we’ve done that we’ve kind of, you know, having an internal CRM, you know, that that institution knowledge is not in the heads of me or James. It is, of course, but it’s, but it’s also kind of in a digital format as well. And we think that kind of the opportunity of the next few years is gonna, how do we leverage that and scale that and do interesting things on top of that?
14:36
Do you think it’ll be paramount that most venture firms follow a similar strategy? Like of course, I think we’ll always have the benchmarks of the world where there’s a select few partners that were very intimately with founders, but for the majority of venture firms, do you think they’ll have to adopt technology or do you think they’ll be left behind? What do you what do you see for the future of venture capital over The next five to seven years.
15:01
Thank you, I think you can do really well, in many different flavors, I think I think it’s, you know, the very small bands or kind of super angels or, or, or groups together just a really good deal flow, they, you know, I’m a small lp in some of them where they run an air table, this is the deal flow. You know, they outsource love the back office, and they spend time kind of meeting with founders and helping them. And I think that worked just great. You know, that said, I think that the kind of larger scale firms are building and should be voting on a mod this mod, there’s partly because that the, the ecosystem is so large today, like, you know, when you, you know, back in the day, that’s sort of like, okay, there’s only a couple of interesting companies that are unicorns that have found it every year. And now, there’s dozens and dozens, all over the world. And it’s just it’s hard to keep track of all this stuff, and so many different companies. And so I do think that there’ll be advantages it. I don’t think it’s, I think it’s helpful, but it’s, we think that software, we think venture firms with a backbone of data that don’t have exceptional partners, will fail. Just because I think it’s sort of it’s, it’s an accelerator to what the people do. I don’t think it’s an utter substitute of what they do. Yeah,
16:30
doesn’t doesn’t replace good deal, acumen or good judgment work exactly as being able to advise them, you know. So we’ve talked a lot about the structure of artifacts, whether it’s the software that you provide to the ecosystem, you have your own internal tools, the content, specifically from an investment philosophy, can you share? Can you share the firm’s investment philosophy? And are you thinking about making investments?
16:56
Yeah, so we’re, we’re stage specific sector agnostic. So we really love this kind of early stage, early stage engagement, you know, optimally to be the first institutional investor. And we think we’re quite uniquely suited because we’ve collectively founded 10 companies, which have been worth more than 10 billion. So which I think it’s more than pretty much any firm of any strategy. So it’s like, so we love that early stage being really hands on. And we’re currently investing out of a $450 million bond. So we’re pretty active in terms of the companies that we work with. And so we’re, you know, so that really is leading pre seed and seed, seed rounds of founders. And then and then sectors that we invest in, we really look for sectors, which we have both an internal expertise around, as well as just massive market opportunity. So you know, the those, there’s an that covers, kind of 80% of stuff. And then there’s 20% of stuff, which we just didn’t this bound is just amazing. And there’s no real precedent. But it’s, it’s unusual, it’s interesting, but we just have to back them. For the sectors that were typically active in like things like marketplaces, FinTech, prop tech gaming, and then tech bio. Omri, one of our partners who joined us a couple of years ago, he’s just phenomenal at this, the intersection of technology and bio. And I think where you probably is that in the common language today, you’d call it longevity. So it’s just the massive healthcare so as less making drugs and things that’s really this sort of, so let’s making sort of specific drug discovery, and making drugs that sort of traditional traditional biotech, and we really call it tech buyer. So there’s so many interesting things happening now you’ve got this combination between computation and biology. So that covers like 80%. And there’s 20% of stuff, which is, we just never really anticipated and doesn’t have network effects. So an example might be we’ve, we’ve invested in mortgage loan, a couple of deals in space, which have done really well. And we just did this sort of massive industry that we’ve been tracking. And we’ve had a couple of companies over the last couple of years in that area, which we, which is sort of perhaps outside of our kind of core areas, but we just love the founders and we’d love the opportunity we we want to support them to build their these amazing businesses.
19:36
I want to go a little bit deeper into marketplaces. I know it’s an area that you have a lot of experience with both operating and investing in. In recently. You said it, it’s an exciting moment to be building any company, but it’s especially exciting for marketplace founders. Why do you feel that marketplaces are uniquely positioned to leverage AI?
19:58
Well, I think there’s I mean, plus The Marketplace business model is just, it’s just uniquely attractive, in that you kind of have these incredible network effects. You think of this sort of the remarkable businesses over the last few years from Airbnb to Uber to others. It’s, there’s a ton of opportunities, kind of in that area. I think, what we, what we think specifically about the opportunity with AI, is it that we think it can really unlock different elements to spin up new marketplaces. And I’d say there’s, you know, there’s kind of a couple of different ways that manifests itself. One is that sort of it makes the internal operations more efficient. If you can, you can bring on supply and bring on demand more efficiently using AI, whether that’s automated decision that just makes the the business itself more economically attractive to is that you may be able to in areas where there’s constrained supply, use AI to start to kind of spin up new areas within that, can you create new supply, that sort of classic marketplace? Model is like there’s proven demand, you just need to create new supply? And are there ways that AI can, you know, can it make lawyers more efficient, for instance, and in which case, the AI combined with lawyers can make it more efficient, then you can spin up this more supply, which kind of gives you this increased liquidity, which makes the marketplace more valuable. And then there’s what we call AI first marketplaces, where we think that okay, these are product experiences, which just couldn’t exist, without without AI. And I don’t know exactly what that looks like. But you’ve got this, it’s, it’s kind of fascinating, right? Now, you’ve got this kind of new content creation, text, video imagery, you’ve got this sort of new form factor, whether it’s chat or you know, that’s word spoken word. And so I think you can, it’s quite likely, we’ll start to see some new different new different marketplaces and networks built both b2c and b2b, I didn’t know exactly what they look like. But it’s going to happen. And if you were to take a parallel, perhaps with what happened in the mobile internet, you know, mobile phones came out, and it took a couple of years for founders and designers to really understand, okay, this is the form factor, this is the technology, the camera, the GPS, etc, to like, Okay, this is, this is what we can do with it. And out of that came consumer apps, from WhatsApp to Uber to others, and to Instagram. And I think with what we think with with AI is going to have, there’s gonna be a bunch of novel consumer experiences, and potentially b2b experiences, which little more predictable, but the b2c is like, very unpredictable, the people are going to come out with these with these user experiences, which have kind of really navigating the unique capabilities of this platform. And, and I think the opportunity is there to build multibillion dollar companies, the 10s of billions of dollars of companies. And we’ve seen how quickly check GPT has grown. And I think that sort of shows that there’s interest in kind of many other businesses. So we’re far we’re far from done. And this enabling technology will open up a whole bunch of opportunities. Do
23:38
you think it’s likely that we see more consumer matching marketplaces versus search marketplaces, so search, meaning, we have to, we have to search the entire category of supply, we need to find the best solution for ourselves versus matching like Uber where you’re matched with the driver, because in that case, the supply is very homogenous. But with AI, if it can truly predict the best unit of supply for yourself, it feels that that could be a shift in terms of the consumer experience where there’d be more matching marketplaces, even if the trust needs to be built up to get there. I’m curious if if you have if one, maybe if you agree or two, if you think that that’s a phenomenon that we may see.
24:26
Yeah, I absolutely think so. I think if you, you know, a lot of the, you know, the a lot of the more mature areas in AI, today are in specifically the ad matching models that have been sort of prevalent within matter and, and Google and so and that’s, you know, that people don’t search for ads, right? You kind of just match it up a ton of data. And so if you do have specific marketplaces with the ability to have deep profile or or inferred data around, particularly individuals, I think you’re going to start to kind of move into this kind of beyond the search box environment where you’ll be making more of these matches and recommendations. And it’s sort of, you know, it’s, you know, I go onto LinkedIn and like, and I look at LinkedIn, LinkedIn, good be doing this really well. It’s got, like, this whole my professional context. It knows my entire career history and knows everyone else’s career history, in terms of where they started and where they’ve been and where they’re going. And yeah, I look at the ads and the recommended jobs that show me and that awful, you know, they’re like, they’re really, really, really bad. And I’m like, Why? Why doesn’t like LinkedIn? And I’m sure they will, but like, why isn’t LinkedIn give you a sort of proper career counseling recommendation service that is not so much ad driven, it’s kind of more experienced, driven. And so I think it, you know, I think it speaks a couple of things. One is that there is very much this kind of like, you know, there will be this recommendation based system, which is going to create a lot of matching, but to is going to be the companies that are sitting on a lot of these data are probably going to be the biggest beneficiaries around this. And and it’s, you know, LinkedIn is like, it should win at the kind of AI enabled career coach, and maybe, maybe that breaks his business model around recruiting, I don’t know. So maybe there is an opportunity for startups to get into it. But they, but the data they’re sitting on is so vast, and it doesn’t intuitively feel that hard to build something like
26:38
this. Yeah. It reminds me the article that you wrote around how the search, searching for whatever the matches you’re looking for marketplace really starts with intent. It’s not I have the problem, but really the intent. And it feels that large companies, whether it’s LinkedIn, I feel like Apple could be another one that is primed to leverage Siri and really use if we can build up the trust the autopilot experience down the road. But yeah, it’s gonna be interesting to see the way this evolved. Are you? Are you seeing anything on the autopilot? And right now at no effects? Are our founders coming and pitching you autopilot experiences? Or where do you think we’re at right now in the progression towards adopting autopilot?
27:26
Yeah, I think we’re definitely biased to the auto pilot world with, I think it’s, I think they’re, we’re in this sort of transitionary transition phase right now, where we’ve got people that are using AI, to supervise what they do. But then quite quickly, I think, mentally, people are going to be like, well, this is so good, I’m actually going to trust it more than I’m going to trust myself. And I think that’s the is the companies which are leaning more into that autopilot phase, it doesn’t mean they’re not supervised. And there’s a ton of industries when supervision is necessary. But I think it’s where you start to automate the work to be done is, is often a far superior economic model. And just a different way of thinking about the product experience is like, you know, it’s sort of, you know, the analogy is like, taking a newspaper, and putting the onto on the internet is sort of translating an old school approach to a new school, and it didn’t really work. Right, it was okay. But actually, you know, thinking of the, what is the technology and device native experience in the way that should be in the next five to 10 years. And if you think about it, in that context, you’re probably gonna end up with an autopilot and a bunch of this stuff. You know, it’s like a car is a, you know, is a horse drawn carriage without horses, just like you’ve just got to think about it, like what’s the optimal use case? And then kind of work backwards? How do you make it so these autopilots are really interesting, we’ve, we’ve invested in a couple of companies in that area, one of them does a bit further on, it’s called even up in the personal injury attorney area, which essentially automates the role of a personal injury attorney to help the claimant to maximize their benefits their settlement and get the most dollar for what they deserve for the, the situation that we’re in. And so that really automates the, the demand after a lot of the modeling around that. And we let the seed and that’s gone on to do incredible things. And so that, you know that and that’s automating a whole bunch of the processes, but there’s very much a human element within there, but we think just that the data and the processing is just, you know, much of it is better suited by using technology and as we know it can the technology can do remarkable things. It
29:57
talking about this may makes me think about the element of timing. And I’m curious if your thinking in the way that you evaluated timing at all for the AI paradigm shift is different versus past paradigm shifts that we’ve seen in the past and an example. So as I listen to Brad Gerstner talk the other day, and he was bringing up, which this feels somewhat trite to even bring up, but the example of Google and how the majority of search value for Internet companies or internet search companies has accrued to Google. So if you invested early in 98, through 2002, and Lycos, a, you know, pick your favorite search AltaVista, you missed out on 97% of the returns. I’m curious with this major paradigm shift that we all agree, a lot of value is going to come out of it. How do you decipher the signal from the noise in the element of timing for some of these spaces that will yield very large outcomes? But getting the timing right is just as important as getting the horse you’re betting on? Correct?
31:04
Yeah, but I think I mean, timing in general, is, I guess, two things. One is the timing in general for for venture investing, you know, being too late, is or being too early, it’s a real problem. So many companies have failed. You know, and I think the, you know, the, the way that we look at it is really the, the catalyst for the why now is often one is this technology change, which we’re seeing right now, which opens up opportunities that weren’t even there before. You know, secondly, they can often be a sort of economic gain something that were expensive, become cheap or cheap, become expensive. And that gives you that sort of the fuel in your rocket to be successful. And the third is the, you know, the sort of cultural zeitgeist loaded with that sort of, like, you know, stepping on other people’s couches, like, going into strangers cars, there’s like, you know, that, you know, what goes from unacceptable to acceptable. So, you know, we, we look at it through that lens of kind of very much the why now, and it can be a number of those different, different things. Specifically, we’d like, how we think about, you know, the sort of why now, I think what you’re getting at is, is, how, where’s the value going to be created in AI, in this kind of next evolution? How much of it is going to be at the sort of, you know, in the sort of different layers, that sort of chip level, the foundation model, that kind of infrastructure, falling layer, and then the app level itself? And I think the, you know, you know, I think the chip level is just really hard, right, and I think there will be a bunch of companies beyond Nvidia that do well in this area. But it’s like, not, it’s unlikely to be a big startup opportunity, I think the LLM the financial model area, I think it’s, it’s, most of the benefit will be going to incumbents. And you know, whether that is going to be Microsoft slash open AI, whether it’s Google, whether it’s matter, and maybe a couple of others, but I think clearly most of that is going to go to the incumbents at this stage. I think the the tooling layer, I think is probably going to be subsumed within the, the tools that the large incumbents provide, they’ll incorporate a bunch of tooling to get people to use their infrastructure, and pull that demand over and a lot and a lot of it may be free or open source, and then it becomes the app ecosystem. And I think that is where that there’s going to be a very large number of successful companies, sure, there may not have the trillion dollar exits, that the you potentially could see the trillion dollar value creation that we could see at the, at the foundation or model. But I think you’re gonna see dozens, if not hundreds, of kind of multi billion dollar companies that are creating that use case, and there’s going to be a bunch of them, we’re going to be accelerated, they’ve already encumbered and they have data distribution, there’s gonna be a bunch of new companies going to be going to be brought out. But it’s that bit that if you’re a startup founder or a startup investor, then that’s going to be the interesting things where they’re going to be the startups that, okay, AI is an accelerant to the business, and there’s gonna be a bunch of them to existing business, that’s going to essentially just massively improve that margin profile, because they don’t need 1000s of salespeople, customer service people, they’re going to like, you know, automate a bunch of that, or they can provide services that are expensive, that are that are cheap, or there’s entirely new use cases that can be can be built up. Yeah.
34:36
Yeah, I was I was on that point of where value is going to occur. I was listening to a different podcast, and I believe it was Tom Tom goose that said this, the the and he was talking about where value accrued for the inferred cloud infrastructure providers versus the cloud applications. And they had about equivalent market caps, but the issue was That was a trillion dollars worth of infrastructure, it went to three players versus the application layer was distributed over hundreds of public companies. So it feels that we may see a similar situation play out at the model layer versus application layer with AI.
35:19
Yeah, and it’s not. It’s there’s just there’s very much a core question right now about how much do the models become? commoditized? Yeah. And I think, and then the sort of perception is that commodities become zero cost, if the digital commodity lead sending an email is essentially zero. But and so the price will come down, you know, that said, you know, the cloud businesses of Amazon and Microsoft and others a diverse and as you’re like, a very profitable businesses, there, they are perceived commodities, but they’re, they’re pretty nice businesses that are growing very quickly. So there’s not it’s not bad to be in that kind of business. The second question, I think, is, is there’s there was a, there was a perception back in the day when Google came out that search is a commodity. You know, Yahoo, famously lightens Google, because there’s like, Well, we, it’s just a kind of like a feature. And I think there will be in and these models are both b2b infrastructure plays, but also b2c consumer experiences in the former chair, GPT, or Gemini. And I think they’ll be particularly consumer side, there’ll be ways to build strong network effects, brand locking and distribution lock in that there will be non commoditized aspects to it. And such, you know, while the cord sort of technologies is commoditize, in the form of, you know, Google, or Bing or others, actually, there’s, these businesses are massively defensible. And I think we’re going to, they’re going to figure out, so I think they, the technology is going to be very, very valuable. The third, the third piece is that it’s, this is a exponential technology, not obviously an asymptote in and show if I want to talk to help me to do my homework, you know, my 14 year old math homework is like, that’s gonna be like asymptotics is going to be, you know, there’s tools out there that do that very well right now, and they’re not necessarily going to get much better. But if I want a tool to help me to navigate, you know, the next breakthrough in quantum physics, that stuff is not going to be awesome to team or other breakthroughs, which we just can’t really sort of imagine this point. And so I don’t see this stuff. I think that innovation will push the edges of commoditization that the value will be continuously created for many of the leading models.
37:56
On the topic of defensibility, has your thinking evolved on defensibility of the marketplaces, you’re investing in an earlier you mentioned, the ability to create more supply using AI. And when you’re saying that my mind goes to a situation in which we used to procure a graphic designer or tutor therapists, what have you, and thinking about five to 10 years out, I mean, some of this is already happening today, but being able to procure a artificial agent of the graphic designer or be tutored by an AI or potentially, you know, have a conversation with a non real person through a therapist. And I’m curious if that dynamic by artificially being able to create supply almost infinite supply are tailored directly to the individual, if that changes the way that you think about certain marketplaces and which ones make it attractive today to invest in or if there’s just inherently a different playbook that these startups need to follow if they’re going to create durable value. Yeah,
39:05
it’s a very interesting question. And we look at it a lot in marketplace, I think, just the sort of, if you take marketplaces, they’re sort of abstract of AI. We actually haven’t seen that many big consumer marketplaces over the last five, six years. And part of that is that the matching technology, which is in the ad networks, which is Google and Facebook, combined with the establishment of a lot of sort of the marketplaces that have captured a lot of the economic value, you know, whether that’s Amazon or whether that’s Uber or Airbnb is like, they get a pretty good job, you know, they do a pretty good job in matching stuff. And so it’s kind of hard to get this liquidity going. We and so we haven’t seen that many big, big marketplace opportunities over the last couple of years. It’s not clearly not done, but I think it’s just it’s the exact same platforms, whether they’re Instagram or Facebook ads are as do a good job in that. On the b2b stuff like, I think there is more interesting kind of whitespace to be done. And we definitely much focused on what is the, what is the engagement process on both supply and the demand focus. So matching is no longer really a meaningful source of defensibility. Just matching supply and demand, the classic model of marketplaces is just weak it because there’s so many options out there for businesses and consumers, we really what is the embedded software tool that helps one preferably both sides of that marketplace that helps them to run their business more effectively. And then the matching is really a value piece on top of that. So you know, you could say this is comfort the tool stay for the network, or we have similar thesis called the market network, where we can extend our SAS work to workflow with marketplace and network. And so I think the so where we’re focused today in terms of defensive ality is this embedding. In we one area we’ve been quite active is in labor marketplaces. And you speak to a recruiter? And like, what, where do you go to kind of find applicants, wherever there are applicants, like, they’ll use whatever works? And then similarly, if you speak to a job seeker, where do you go to find work, whoever has availability and plays the most, like they’re very promiscuous, in a good way, like to kind of find it. And so you really want to build a sort of economic value that you don’t fragment the liquidity, you really need a bunch of value added services, meaning they get more than just matching at your platform, then helps them something that helps them to run their business.
41:52
Yeah, it were so early in b2b marketplaces as well. I believe Harvard Business School wrote its first b2b marketplace case study two years ago, I actually think it was an NF X portfolio company that wrote Are there are other key attributes for b2b marketplaces that you typically look for versus b2c Because they are more margin sensitive buyers. Right? But I’m curious as you think about the must haves for these b2b marketplaces. What are the shortlist call it three must have in order for NSX to make an investment in the b2b marketplace?
42:28
Yeah, that’s there’s one of the I don’t know if it’s sort of like there’s no sort of specific kind of playbook. But I think a couple of the things that come to mind that we look for one is that is this, is this a meaningfully better solution than their existing service today. And if you’re less, you know, the paper mill that’s buying wood from someone like, you probably kind of like have a pre existing system that works pretty well, to buy that word. So why would you change the system that you’re using? And so just changing the supply chain is kind of not necessarily sufficient to them? So what is the real value that you’re that you’re providing, above and beyond what the substitutes are? Another area would be just the level of promiscuity in that marketplace that like, you know, take, you know, hotel, you know, Airbnb, like, you don’t want to stay in the same place all the time. And so you’re naturally multi shopping. And so you’re looking at different stuff. And so if there are marketplaces where you’re, you’re literally trying different stuff all the time, then the margin profile on that gets significantly greater and avoids this disintermediation. And I guess the third piece specifically within AI today is like, what is the kind of hard coordination problem that you can do? And I think it is, it’s interesting with materials marketplaces, or sort of complex manufacturing, the combination of kind of all these different materials together, you know, that could be physical materials, it could be people materials, and things like construction, the coordination problem is like, is a really hard technical problem to do whether this cement is with this type of brick, whether this person is available at this particular time, whether this kind of fluid goes with it, there’s like there’s and computers are really good at that kind of like matching different products. And I think that that breakthrough technology, driven by like significant amount of data, which which it might be in the heads of the scientists are doing these things within that that technology breakthrough is really interesting as well. Interesting.
44:33
Pete, if we could feature anyone on the show, who should we interview and what topic would you like to hear them speak about?
44:40
So many people buy probably I probably pick my former co founder, Trulia semi in London, and it’s he went from on real estate with me to running a company called which is in reversing type two diabetes. And he’s just a remarkable individual who kind of is a incredible athletes as well as business leader, and that and it’s a very interesting time in that kind of space right now.
45:07
Absolutely. People book article or video, would you recommend to our listeners, either something in recent memory that you found informative or inspiring or one of your classical favorites?
45:18
Gosh, what am I reading right now? I’m actually like, I’m a bit of a podcast. I wish I read more. But I’m like, kind of an A. So I’ve just been like, I’ve been going through the back catalogue of the acquired podcast. Now they’re granted. So it’s like, it’s sort of like, you know, in commutes, or in running are like that sort of like these case studies. And they’re great. So I’m enjoying that kind of like, avoiding these avoiding the kind of like, hose of like, what’s going on with AI today, but just these like deep retrospectives of amazing business? Yeah,
45:53
yeah. One of the one of the best podcasts out there. And then last but not least, Pete, what is the best way for listeners to connect with you and NFS? So
46:02
the best way is, we’re, we’re very visible. So if you go to nfs.com, and find myself or my partners, we’re very accessible email addresses and socials on there. So and effects a common they sign up for our newsletter, as well. And you can keep abreast of everything we’re up to. Yeah.
46:19
Plus, it’s one of the few venture firms that actually puts everything, all the must haves, all the criteria that must be present in order for you guys to make an investment. So not all firms are so forthcoming around those criteria. But anyway, Pete, this, this was a blast. Thanks again for coming on. And yeah,
46:38
great chat. Nice to see.
46:46
All right, that’ll wrap up today’s interview. If you enjoyed the episode or a previous one, let the guests know about it. Share your thoughts on social or shoot him an email. Let them know what particularly resonated with you. I can’t tell you how much I appreciate that some of the smartest folks in venture are willing to take the time and share their insights with us. If you feel the same, a compliment goes a long way. Okay, that’s a wrap for today. Until next time, remember to over prepare, choose carefully and invest confidently thanks so much for listening