theCUBE + NYSE Wired: Capital Series | Mark Roberge, Stage 2 Capital

theCUBE.net

Clips
More from theCUBE + NYSE Wired: Capital Series

Mark Roberge

Co-Founder

Stage 2 Capital

play_circle_outline Capital-Light $100M Fund Targeting Outsized Returns in B2B Application Layer and Vertical AI

play_circle_outline AI Hype vs Reality: Experiments, Fragile Moats and Replacement Risk

play_circle_outline Preference for mid-market and SMB verticals with regulatory complexity and low technical talent

play_circle_outline Fast-follower opportunity: copycat strategies could undercut high-priced incumbents during AI boom

play_circle_outline Leading Value Metrics: Net Dollar Retention and Instrumenting Customer Value Delivery

Info

Mark Roberge, Stage 2 Capital

Mark Roberge

Co-Founder Stage 2 Capital

search

Mark Roberge

>> Palo Alto Studio connecting, Silicon Valley and Wall Street. I'm John Furrier, co-host of theCUBE here with Dave Vellante, my co-host.

Gemma Allen

>> Welcome back to theCUBE Studio here at the New York Stock Exchange. I'm Gemma Allen, co-host of NYSE Wired. We connect Silicon Valley to Wall Street. Today we're talking all things capital investment. And joining me now for a conversation on exactly that is Mark Roberge, co-founder of Stage 2 Capital. Welcome, Mark.

Mark Roberge

>> Thanks, Gemma.

Gemma Allen

>> So Stage 2 Capital, interesting thesis, interesting investment VC fund. Break it down for us. Give us the 101.

Mark Roberge

>> Yeah, it's been a blessed experience. Been about 9 years now. I never intended to go into venture capital. I certainly benefited from it as an entrepreneur. But I was approached by a gentleman at Bessemer who felt like there was a massive need for great counsel to these young founders around the go-to-market, the sales, the marketing side that wasn't happening in the boardroom. And the vision was to assemble all of our investors, our LPs, as the best sales, marketing, customer success leaders in tech and mobilize that network to help these founders in the early stages of their own journey. So there's a number of ways that we deploy that. We've been at it for 9 years across 4 funds. Probably a bit over 100 companies under management, and we've been excited about the journey so far.

Gemma Allen

>> I want to get into the types of companies you're investing in, the bets you're making. But first, in terms of what's unique about Stage 2 Capital from the perspective of professional support and investment into these portcos, we had a founder on recently, a founder who talked about choosing between who he was going to take funding from. Right. And he mentioned that Sam Altman had actually considered investing in his round and said, you will get 15 minutes with me once a quarter. What's your response to that? And how do you think about that from the perspective of what you're building and what's unique?

Mark Roberge

>> No, amazing. the best founders out there— it's certainly a founder favorable market today, and the best founders should be thinking about building out their cap table not just for the money but for the value. And I totally agree with that. Sam Altman perspective is like, if you're that founder and you feel like Sam would be hugely useful, and he probably will if you're in anything AI, you should take that money. And so that's why we win a lot of our deals is it's— you're probably going to need to acquire customers through your journey. And we've assembled the best firm to be able to do that. We have the sales and marketing leadership from OpenAI and Anthropic and GitHub and Snowflake and Databricks and HubSpot and Salesforce and all, so no matter what you're selling, and who you're selling to. We've got folks with decades of experience who've gone through that. We mobilize them through a number of ways. We've set up over 400 advisory engagements between our LPs and our portfolio companies. We do a handful of workshops on very specific go-to-market challenges from pricing to distribution to account-based marketing with individual portfolio companies every single month. And so just make sure you've got smart capital around your cap table. And make sure it's diversified across every element you need, including go-to-market.

Gemma Allen

>> If you were advising that founder, 15 minutes is a pretty short space of time, right? You could babble for 15 minutes quite easily. What do you think that time is best spent? What advice would you give to folks who have this opportunity to talk to experts in the industry, highly connected individuals?

Mark Roberge

>> Yeah.

Gemma Allen

>> How do you spend the time?

Mark Roberge

>> It's tricky with Sam. That's what you're going to get. And he's so ingrained in the frontier that you can leverage that. But in most cases, that won't work. We joke amongst my peers that oftentimes folks are like, we can't get sales to work. Can you spend 30 minutes with that company? It's just not going to happen. In that situation, outside of the Sam example, we highly recommend pre-reads because actually these busy executives don't mind. They actually probably prefer spending 10 minutes the evening before skimming through a couple of pages of material to familiarize themselves with your context so they can dive right into the value add. And to be honest with you, realistically, you're probably better off trying to find folks who can do sort of an hour a week for some sequence to really work through the diagnosis and the fixing issues for, or the evolution of strategy issues to understand your context. So that's the reality of advisement. With Sam, I would really just like to truly understand where he's been spending his time lately on the frontier and how that overlaps with the key tensions or opportunities within the organization and come in with some pointed questions, a little bit of context to leverage that 15 minutes.

Gemma Allen

>> Talk to me about the sorts of companies and investments you're making and Stage 2 Capital is making. It's been an interesting year for SaaS. Your career started, I believe, in the SaaS space. I know you were at HubSpot. You were early there, correct?

Mark Roberge

>> Yeah. So I was the fourth employee at HubSpot and founding CRO, taking them through from zero to IPO over 9 years. Correct.

Gemma Allen

>> And if you think about how that SaaS journey has evolved and what's been happening in 2026 in terms of the narrative around the SaaSpocalypse and all of those things, how does it frame what you invest in? Are you all in on AI? Give us some thoughts on where you see the money flowing and the moats remaining.

Mark Roberge

>> Yeah, there's a lot of dimensions to that particular question and it's actually evolved every 2 or 3 years. part of you, you have to be set up for the value you want to create, but you also have to be aware of the macro around you to understand where you have conviction against the consensus. So the focus for us has been in the B2B application layer for the most part. We like to be a $100 million fund because you have the option to generate a 10, 20, 50x return potentially if you really crush it, that a multibillion-dollar fund just mathematically can't return. And we believe that's the point of our asset class in venture is to provide that risk reward. So we like that fund size. And because you're in that fund size, you really can't do heavy CapEx businesses in the physical infrastructure layer. And furthermore, we've built our network from the companies I mentioned, which largely play in the B2B software arena. So that's been a focus for us. And we do everything. A lot of people think we only do go-to-market. That's only about 10% of our investments. We'll do every single category, from cyber to infrastructure, a lot of vertical AI. And yes, this year, as of late, it's every software company has an AI component and we're heavily weighted in that direction. I would say, over the years, we went into COVID and we felt like it was an overheated market. Turns out that it was. So we were very focused at that time on not believing the hype and making sure that we were investing through a recession lens, which we didn't go into a full recession, but it was definitely a correction in the tech community. And then more recently with AI, you have to be really aware of moat and durability. We think that's probably a theme right now is the market has heavily rewarded revenue growth at all costs. And there's not a lot of diligence being done on the lifetime value of the install base. How many of these implementations are experimentation and not in production? We're still seeing strong evidence that there's a lot of experimentation. And the other piece is the durability, the moat, the barrier to entry. Those are all terms to say the same thing, which is, will this last the long term or could someone swap you out for a better model or better app fairly quickly? And we think a lot of the players today lack that and we invest heavily through that lens. I'll give you a couple examples of hypotheses that we've been investing in that check those boxes. The first one is in vertical AI software in end markets that are mature and perhaps tech laggards. If I'm a tech company with a fairly sophisticated CIO or CTO suite, I could probably pull off a build versus buy on my tech stack, meaning I'm now going to replace my sales tech, finance tech, HR tech, product development tech with a general foundational model on top of, say, a Databricks. But if you are a mid-market bank, regional bank, or if you are a hospital or if you are a mid-market construction company, they just don't have the technical talent around them to be able to do a build. And there's an opportunity for specialist vendors to become the entire tech operating system for those end markets. Some of those end markets haven't been big enough to justify a $5 billion outcome for venture. But in the last generation of vertical software, even, say, Toast, which is a massive winner in the restaurant scene, you walk into a restaurant on Toast. That's not all the tech they have. They have other tech running. And because AI has accelerated the product development cycles and because AI benefits from a singular brain, there's an opportunity for the future vertical AI softwares to be the entire operating system for these businesses so they can capture the entire tech spend, making the ACV per customer much larger and expanding that market. The other nice thing about that is I just don't see Claude going after those markets. There's two— the more regulatory unique requirements there are for those end markets and the more they cap out at, say, $5 or $10 billion, which is a massive return for us but doesn't move the needle for the tier 1 VCs or, you know, Anthropic.

Gemma Allen

>> I don't know if you followed the kind of media speculation this week that there was a comment that Anthropic might be the world's only private company 10, 15 years from now. Right. It was like, a couple of outlets are running it as a headline. Apparently somebody very, very senior at Anthropic said that. But, there certainly seems to be this belief that Anthropic is owning the enterprise. And I want to use that to go back to a comment you made around these vertical application layer bets. Right. So we don't hear a lot. We have a lot of founders come on this show that are building in the vertical space. Right. We hear vertically integrated, horizontally aligned. That tends to be the line because it gives you a level of flex. Yeah, I don't want to straw man, but what I want to ask you is we don't hear a lot about selling into mid-cap, small-cap SMB, selling into— we hear about selling to JPMorgan, not a slew of credit unions. Right. That tends to be like the bigger bet is that enterprise bet. And you would assume that if you're competing in the enterprise space, there's going to be far more likely competition from an Anthropic or an OpenAI or a Gemini than there is going to be in that laggard, right? That laggard layer. And there is a lot of money to be made in that industry. How do you think about that? When you have companies pitching to you, are you looking at how they're profiling audience and buyers? Is that a

Mark Roberge

>> big—Totally.

Gemma Allen

>> So break it down for me a bit because you're the first person to actually say, no, there's mid-cap money too, right?

Mark Roberge

>> Oh yeah. and again, for, for our outcomes where we don't need a trillion-dollar outcome, we— if we, if we get to a $5 or $10 billion outcome, we're going to be a top decile, fund just the way we've set up our fund math. So that allows us to fly below the competitive pressures of some of the tier 1 $5 billion funds, as well as where OpenAI and Anthropic are trying to compete. You know, they already have gone into coding. They've already gone into a lot of product development and design. They might go into go-to-market. They've gone into healthcare in the bigger systems pretty aggressively. I just don't see them going into mid-market, construction and banking, etc. Now, your question, the dilemma is, do you go after the big accounts out of the gate or do you start with SMB? Over a 20-year cycle, if I fast-forwarded 10 or 20 years, I felt most founders incorrectly went for the big enterprise out of the gate, thinking that if they closed them, the rest of the market would follow. They completely underestimated how long it takes to close an enterprise account and all the requirements around SOX compliance and HIPAA compliance or whatever. And so typically the advice was to start in the SMB to get some fast learning cycles, build up a brand, make sure the product works. And over— once you hit $10 million, $20 million, now you can go to the enterprise and that creates a really durable revenue cycle for you. There's arguments that that's not fully true these days. Harvey's a classic example. That's a vertical software company in the legal space, and they've made a great name for themselves by just building some of the— getting some of the big law firms up front, literally camping out with forward deployed engineers, which is an important concept today, to not only deploy the software, but more importantly, move through the change management. The change management— if I choose between an A+ product and a B- change management services program accompanying it, versus an A+ change management program and a B- product. I think in a lot of industries the latter wins today because it's not just about having an awesome AI product, it's getting that 20-year veteran to actually adopt it. And that's something that Harvey did well, is they deployed in there with the big brand and the rest of the folks followed. So there are some contextual decisions there as to do you start with big and catch the rest of the market or do you start SMB mid-market and move up?

Gemma Allen

>> It's an interesting example though, right? Because you have Am Law 10, their budgets are huge, right? From a technical spend perspective, top 100, probably still very sizable. Harvey's licensing is very expensive, right? It's a big upfront line on your P&L that's new, right? And I don't know if it's replacing associates. People say maybe it is, maybe it's not. I'm sure it's like securing budgets in other parts of the business to cross-subsidize it. But it's still an interesting dilemma because there's a whole plethora of law firms out there that again probably won't be able to afford Harvey or integrate Harvey into their P&L in the next 5 years.

Mark Roberge

>> All right. So you are building on the second hypothesis that we're really early on in exploring, but I don't know if Stage 2 can exploit it. I hope we can. But I do think this will unfold over the next few years, which is I think this is the best time in the history of tech to be a fast follower, a copycat. Because of the point that you said, we have seen some of the highest valuation multiples from a P/S ratio, the valuation to the top line revenue in early stage like pre-IPO tech in the history of tech because of a speculation that we are entering such a historic time with this AI movement. Perhaps it will pan out and I've made some arguments that it will. And I've also seen a lot of arguments that it won't. And to your point, these first movers have to sustain a fairly high ACV with their customers to grow into these massive valuations. And the combination of rapid product development cycles and this high valuation, what I call ACV jail, where you can't reduce price. If I were two MIT CS undergrads right now, I would literally just pick a category and copy the product and sell it for 70% off. You can make a lot of money if you study the archives of a company out in Europe, Rocket Internet, that was based in Germany. They ran this fast follower copycat model in the beginning of the century as companies in the US took off on the internet and they just copied it for the European market and made billions. I think that opportunity exists today. And over time, we'll settle into a much lower cost for all of this software for the end customer, which is really the promise of AI. But it can't be delivered with these high valuations.

Gemma Allen

>> I love someone who takes a bold position, Mark. So I love that. So back to these opportunities, these pitches that are coming across your desk, right? We've also heard a lot in the VC space around ARR. Does that actually mean anything anymore? You know how these metrics are changing. What, 10 years ago these rounds were very, very— like, seed rounds were small, right? Now they're like— we spoke to one yesterday, their seed round was $40 million. Like, that's pretty huge.

Mark Roberge

>> That's actually cheap these days, but 10 years ago—

Gemma Allen

>> oh yeah, insane, right? So things are changing so fast. What are you truly measuring and looking for though? Like, what metrics are not changing in your mind as you consider investing in some of these early-stage companies?

Mark Roberge

>> Yeah, sure. So I just came out with my second book in February called The Science of Scaling, which helps companies understand when and how fast they can scale. Half the entrepreneurs I meet scale too early and too fast. Half the entrepreneurs I meet scale too early, too fast versus too late and too slow. And it's because they don't have the same quantitative approach to their scale and understand the durability and readiness to scale. And by the way, I'm donating all the proceeds to mental health just so folks know. It's a very important cause to me. But so at our stage, we don't really have a ton of revenue and customers to evaluate. So we look at leading indicators of customer retention. This LIR concept is what a lot of folks set an initial North Star of their business on revenue growth. But great sales teams can sell ice to Eskimos and Eskimos don't need ice. That's the classic historic example. And so we really set the North Star around value creation. The sales funnel doesn't end at the signed contract and the wired money. The sales funnel ends when you've delivered the value you've promised your customers. And there's ways to make that super instrumented and programmatic. That's a key measure today. And I think even if you're valued over a couple billion dollars, take a hard look at that measurement for yourself. And yeah, we're looking a lot more at a combination of ARR and net dollar retention. TVPI, which is a sum of the valuations of your companies, is almost worthless these days because we're not sure how healthy these valuations are. There's such a variance between how private markets are valuing companies today and how public markets are valuing them. The judge and jury is going to be when some of these companies need to go public to deliver money back to their investors. And I don't know if public markets are going to be as bullish on some of these native AI companies or whether they're going to see some durability risk that we're seeing as well.

Gemma Allen

>> So before we finish, I want to talk about another string to your bow, and that is that you are also a professor at Harvard Business School.

Mark Roberge

>> Yes.

Gemma Allen

>> You lecture on a number of programs and you can break it down for us. But I'm interested in your thoughts on this next generation of young, enthusiastic business-minded founders. It seems as though there's a lot of trepidation out there in the market, especially in tech, around what it will take to— or what a computer software engineer will look like 10 years from now, whether these investments are worth it. We hear a lot about You mentioned mental health as a cause close to your heart. People feeling a whole lot of ambiguity. The generation that is kind of a little bit lost in space in the world of 2026. What are you seeing and hearing? What gives you hope? Yeah.

Mark Roberge

>> Okay.

Gemma Allen

>> And I worry, I guess, about this next gen?

Mark Roberge

>> I teach an entrepreneurial sales class there, which is very popular. I also teach a field trip class to Silicon Valley in January. We would get behind closed doors with all these amazing people out there and 50 students. It's amazing how my students point out to me the imbalance of how much capital, talent, and time is going into building AI and not helping society come along. And that's very worrisome. I think a lot of folks in tech, whether you're on the operational side or investment side, those with a good moral compass are doing a lot of thinking about that. And I am as well. It would be a whole nother episode if you want to go there and more of a philosophical one. But bottom line, and that's part of the reason for the proceeds to mental health, is We who are in the AI epicenter, we need to make sure we internalize that of how much are we putting into building and how much we invest in helping society come along and do our little things. For the MBAs in particular, over a 15-year horizon, we've seen a massive shift from going into banking and consulting to going into entrepreneurship and tech and founders, which is really exciting. I happen to think they're probably one of the best talent pools and overlooked talent pools right now, specifically in tech, like the computer science departments. They've been hurt a little bit more recently. And for example, when I see a software company today, 10 years ago, they used to spend maybe 10 to 20% of the time figuring out what to build and 80 to 90% building because it's just what was necessary to build really strong code. That's flipped substantially. The best businesses who are AI-enabled in their tech community or tech departments, they can build code so rapidly they're spending a large portion of their time figuring out what to build. And MBAs are very well suited to do that, especially those of us with a tech background. So those are some of the themes I'm seeing in the Harvard and general MBA community.

Gemma Allen

>> Well, Mark, before you go, where can folks find your book?

Mark Roberge

>> Yeah, it's on Amazon. The Science of Scaling by Mark Roberge. 100% of the proceeds are donated to mental health, just for two quick reasons. One, it's a big part of my life. I've been a caregiver. I've also been a patient. And a lot of people can't bravely say that publicly because of the stigma associated with it still. So I'm blessed to be able to have the resume to be able to say that bravely. And secondarily, everything I talked about with where society is going, this is going to be a massive shift for society, as was fire and electricity and the internet. And each one of those We evolved to a better species, but it came with scars. If we don't adjust, we'll have massive scars here. And everyone needs to do their little thing. This is my little thing right now.

Gemma Allen

>> Well, Mark, you're certainly a fantastic role model. Thank you so much for joining us on theCUBE and NYSE Wired.

Mark Roberge

>> Thanks, Gemma.

Gemma Allen

>> I'm Gemma Allen here at theCUBE studio at the New York Stock Exchange with NYSE Wired. This is our Capital Series. Thanks for watching.

Mark Roberge, Stage 2 Capital

search

Mark Roberge

>> Palo Alto Studio connecting, Silicon Valley and Wall Street. I'm John Furrier, co-host of theCUBE here with Dave Vellante, my co-host.

Gemma Allen

>> Welcome back to theCUBE Studio here at the New York Stock Exchange. I'm Gemma Allen, co-host of NYSE Wired. We connect Silicon Valley to Wall Street. Today we're talking all things capital investment. And joining me now for a conversation on exactly that is Mark Roberge, co-founder of Stage 2 Capital. Welcome, Mark.

Mark Roberge

>> Thanks, Gemma.

Gemma Allen

>> So Stage 2 Capital, interesting thesis, interesting investment VC fund. Break it down for us. Give us the 101.

Mark Roberge

>> Yeah, it's been a blessed experience. Been about 9 years now. I never intended to go into venture capital. I certainly benefited from it as an entrepreneur. But I was approached by a gentleman at Bessemer who felt like there was a massive need for great counsel to these young founders around the go-to-market, the sales, the marketing side that wasn't happening in the boardroom. And the vision was to assemble all of our investors, our LPs, as the best sales, marketing, customer success leaders in tech and mobilize that network to help these founders in the early stages of their own journey. So there's a number of ways that we deploy that. We've been at it for 9 years across 4 funds. Probably a bit over 100 companies under management, and we've been excited about the journey so far.

Gemma Allen

>> I want to get into the types of companies you're investing in, the bets you're making. But first, in terms of what's unique about Stage 2 Capital from the perspective of professional support and investment into these portcos, we had a founder on recently, a founder who talked about choosing between who he was going to take funding from. Right. And he mentioned that Sam Altman had actually considered investing in his round and said, you will get 15 minutes with me once a quarter. What's your response to that? And how do you think about that from the perspective of what you're building and what's unique?

Mark Roberge

>> No, amazing. the best founders out there— it's certainly a founder favorable market today, and the best founders should be thinking about building out their cap table not just for the money but for the value. And I totally agree with that. Sam Altman perspective is like, if you're that founder and you feel like Sam would be hugely useful, and he probably will if you're in anything AI, you should take that money. And so that's why we win a lot of our deals is it's— you're probably going to need to acquire customers through your journey. And we've assembled the best firm to be able to do that. We have the sales and marketing leadership from OpenAI and Anthropic and GitHub and Snowflake and Databricks and HubSpot and Salesforce and all, so no matter what you're selling, and who you're selling to. We've got folks with decades of experience who've gone through that. We mobilize them through a number of ways. We've set up over 400 advisory engagements between our LPs and our portfolio companies. We do a handful of workshops on very specific go-to-market challenges from pricing to distribution to account-based marketing with individual portfolio companies every single month. And so just make sure you've got smart capital around your cap table. And make sure it's diversified across every element you need, including go-to-market.

Gemma Allen

>> If you were advising that founder, 15 minutes is a pretty short space of time, right? You could babble for 15 minutes quite easily. What do you think that time is best spent? What advice would you give to folks who have this opportunity to talk to experts in the industry, highly connected individuals?

Mark Roberge

>> Yeah.

Gemma Allen

>> How do you spend the time?

Mark Roberge

>> It's tricky with Sam. That's what you're going to get. And he's so ingrained in the frontier that you can leverage that. But in most cases, that won't work. We joke amongst my peers that oftentimes folks are like, we can't get sales to work. Can you spend 30 minutes with that company? It's just not going to happen. In that situation, outside of the Sam example, we highly recommend pre-reads because actually these busy executives don't mind. They actually probably prefer spending 10 minutes the evening before skimming through a couple of pages of material to familiarize themselves with your context so they can dive right into the value add. And to be honest with you, realistically, you're probably better off trying to find folks who can do sort of an hour a week for some sequence to really work through the diagnosis and the fixing issues for, or the evolution of strategy issues to understand your context. So that's the reality of advisement. With Sam, I would really just like to truly understand where he's been spending his time lately on the frontier and how that overlaps with the key tensions or opportunities within the organization and come in with some pointed questions, a little bit of context to leverage that 15 minutes.

Gemma Allen

>> Talk to me about the sorts of companies and investments you're making and Stage 2 Capital is making. It's been an interesting year for SaaS. Your career started, I believe, in the SaaS space. I know you were at HubSpot. You were early there, correct?

Mark Roberge

>> Yeah. So I was the fourth employee at HubSpot and founding CRO, taking them through from zero to IPO over 9 years. Correct.

Gemma Allen

>> And if you think about how that SaaS journey has evolved and what's been happening in 2026 in terms of the narrative around the SaaSpocalypse and all of those things, how does it frame what you invest in? Are you all in on AI? Give us some thoughts on where you see the money flowing and the moats remaining.

Mark Roberge

>> Yeah, there's a lot of dimensions to that particular question and it's actually evolved every 2 or 3 years. part of you, you have to be set up for the value you want to create, but you also have to be aware of the macro around you to understand where you have conviction against the consensus. So the focus for us has been in the B2B application layer for the most part. We like to be a $100 million fund because you have the option to generate a 10, 20, 50x return potentially if you really crush it, that a multibillion-dollar fund just mathematically can't return. And we believe that's the point of our asset class in venture is to provide that risk reward. So we like that fund size. And because you're in that fund size, you really can't do heavy CapEx businesses in the physical infrastructure layer. And furthermore, we've built our network from the companies I mentioned, which largely play in the B2B software arena. So that's been a focus for us. And we do everything. A lot of people think we only do go-to-market. That's only about 10% of our investments. We'll do every single category, from cyber to infrastructure, a lot of vertical AI. And yes, this year, as of late, it's every software company has an AI component and we're heavily weighted in that direction. I would say, over the years, we went into COVID and we felt like it was an overheated market. Turns out that it was. So we were very focused at that time on not believing the hype and making sure that we were investing through a recession lens, which we didn't go into a full recession, but it was definitely a correction in the tech community. And then more recently with AI, you have to be really aware of moat and durability. We think that's probably a theme right now is the market has heavily rewarded revenue growth at all costs. And there's not a lot of diligence being done on the lifetime value of the install base. How many of these implementations are experimentation and not in production? We're still seeing strong evidence that there's a lot of experimentation. And the other piece is the durability, the moat, the barrier to entry. Those are all terms to say the same thing, which is, will this last the long term or could someone swap you out for a better model or better app fairly quickly? And we think a lot of the players today lack that and we invest heavily through that lens. I'll give you a couple examples of hypotheses that we've been investing in that check those boxes. The first one is in vertical AI software in end markets that are mature and perhaps tech laggards. If I'm a tech company with a fairly sophisticated CIO or CTO suite, I could probably pull off a build versus buy on my tech stack, meaning I'm now going to replace my sales tech, finance tech, HR tech, product development tech with a general foundational model on top of, say, a Databricks. But if you are a mid-market bank, regional bank, or if you are a hospital or if you are a mid-market construction company, they just don't have the technical talent around them to be able to do a build. And there's an opportunity for specialist vendors to become the entire tech operating system for those end markets. Some of those end markets haven't been big enough to justify a $5 billion outcome for venture. But in the last generation of vertical software, even, say, Toast, which is a massive winner in the restaurant scene, you walk into a restaurant on Toast. That's not all the tech they have. They have other tech running. And because AI has accelerated the product development cycles and because AI benefits from a singular brain, there's an opportunity for the future vertical AI softwares to be the entire operating system for these businesses so they can capture the entire tech spend, making the ACV per customer much larger and expanding that market. The other nice thing about that is I just don't see Claude going after those markets. There's two— the more regulatory unique requirements there are for those end markets and the more they cap out at, say, $5 or $10 billion, which is a massive return for us but doesn't move the needle for the tier 1 VCs or, you know, Anthropic.

Gemma Allen

>> I don't know if you followed the kind of media speculation this week that there was a comment that Anthropic might be the world's only private company 10, 15 years from now. Right. It was like, a couple of outlets are running it as a headline. Apparently somebody very, very senior at Anthropic said that. But, there certainly seems to be this belief that Anthropic is owning the enterprise. And I want to use that to go back to a comment you made around these vertical application layer bets. Right. So we don't hear a lot. We have a lot of founders come on this show that are building in the vertical space. Right. We hear vertically integrated, horizontally aligned. That tends to be the line because it gives you a level of flex. Yeah, I don't want to straw man, but what I want to ask you is we don't hear a lot about selling into mid-cap, small-cap SMB, selling into— we hear about selling to JPMorgan, not a slew of credit unions. Right. That tends to be like the bigger bet is that enterprise bet. And you would assume that if you're competing in the enterprise space, there's going to be far more likely competition from an Anthropic or an OpenAI or a Gemini than there is going to be in that laggard, right? That laggard layer. And there is a lot of money to be made in that industry. How do you think about that? When you have companies pitching to you, are you looking at how they're profiling audience and buyers? Is that a

Mark Roberge

>> big—Totally.

Gemma Allen

>> So break it down for me a bit because you're the first person to actually say, no, there's mid-cap money too, right?

Mark Roberge

>> Oh yeah. and again, for, for our outcomes where we don't need a trillion-dollar outcome, we— if we, if we get to a $5 or $10 billion outcome, we're going to be a top decile, fund just the way we've set up our fund math. So that allows us to fly below the competitive pressures of some of the tier 1 $5 billion funds, as well as where OpenAI and Anthropic are trying to compete. You know, they already have gone into coding. They've already gone into a lot of product development and design. They might go into go-to-market. They've gone into healthcare in the bigger systems pretty aggressively. I just don't see them going into mid-market, construction and banking, etc. Now, your question, the dilemma is, do you go after the big accounts out of the gate or do you start with SMB? Over a 20-year cycle, if I fast-forwarded 10 or 20 years, I felt most founders incorrectly went for the big enterprise out of the gate, thinking that if they closed them, the rest of the market would follow. They completely underestimated how long it takes to close an enterprise account and all the requirements around SOX compliance and HIPAA compliance or whatever. And so typically the advice was to start in the SMB to get some fast learning cycles, build up a brand, make sure the product works. And over— once you hit $10 million, $20 million, now you can go to the enterprise and that creates a really durable revenue cycle for you. There's arguments that that's not fully true these days. Harvey's a classic example. That's a vertical software company in the legal space, and they've made a great name for themselves by just building some of the— getting some of the big law firms up front, literally camping out with forward deployed engineers, which is an important concept today, to not only deploy the software, but more importantly, move through the change management. The change management— if I choose between an A+ product and a B- change management services program accompanying it, versus an A+ change management program and a B- product. I think in a lot of industries the latter wins today because it's not just about having an awesome AI product, it's getting that 20-year veteran to actually adopt it. And that's something that Harvey did well, is they deployed in there with the big brand and the rest of the folks followed. So there are some contextual decisions there as to do you start with big and catch the rest of the market or do you start SMB mid-market and move up?

Gemma Allen

>> It's an interesting example though, right? Because you have Am Law 10, their budgets are huge, right? From a technical spend perspective, top 100, probably still very sizable. Harvey's licensing is very expensive, right? It's a big upfront line on your P&L that's new, right? And I don't know if it's replacing associates. People say maybe it is, maybe it's not. I'm sure it's like securing budgets in other parts of the business to cross-subsidize it. But it's still an interesting dilemma because there's a whole plethora of law firms out there that again probably won't be able to afford Harvey or integrate Harvey into their P&L in the next 5 years.

Mark Roberge

>> All right. So you are building on the second hypothesis that we're really early on in exploring, but I don't know if Stage 2 can exploit it. I hope we can. But I do think this will unfold over the next few years, which is I think this is the best time in the history of tech to be a fast follower, a copycat. Because of the point that you said, we have seen some of the highest valuation multiples from a P/S ratio, the valuation to the top line revenue in early stage like pre-IPO tech in the history of tech because of a speculation that we are entering such a historic time with this AI movement. Perhaps it will pan out and I've made some arguments that it will. And I've also seen a lot of arguments that it won't. And to your point, these first movers have to sustain a fairly high ACV with their customers to grow into these massive valuations. And the combination of rapid product development cycles and this high valuation, what I call ACV jail, where you can't reduce price. If I were two MIT CS undergrads right now, I would literally just pick a category and copy the product and sell it for 70% off. You can make a lot of money if you study the archives of a company out in Europe, Rocket Internet, that was based in Germany. They ran this fast follower copycat model in the beginning of the century as companies in the US took off on the internet and they just copied it for the European market and made billions. I think that opportunity exists today. And over time, we'll settle into a much lower cost for all of this software for the end customer, which is really the promise of AI. But it can't be delivered with these high valuations.

Gemma Allen

>> I love someone who takes a bold position, Mark. So I love that. So back to these opportunities, these pitches that are coming across your desk, right? We've also heard a lot in the VC space around ARR. Does that actually mean anything anymore? You know how these metrics are changing. What, 10 years ago these rounds were very, very— like, seed rounds were small, right? Now they're like— we spoke to one yesterday, their seed round was $40 million. Like, that's pretty huge.

Mark Roberge

>> That's actually cheap these days, but 10 years ago—

Gemma Allen

>> oh yeah, insane, right? So things are changing so fast. What are you truly measuring and looking for though? Like, what metrics are not changing in your mind as you consider investing in some of these early-stage companies?

Mark Roberge

>> Yeah, sure. So I just came out with my second book in February called The Science of Scaling, which helps companies understand when and how fast they can scale. Half the entrepreneurs I meet scale too early and too fast. Half the entrepreneurs I meet scale too early, too fast versus too late and too slow. And it's because they don't have the same quantitative approach to their scale and understand the durability and readiness to scale. And by the way, I'm donating all the proceeds to mental health just so folks know. It's a very important cause to me. But so at our stage, we don't really have a ton of revenue and customers to evaluate. So we look at leading indicators of customer retention. This LIR concept is what a lot of folks set an initial North Star of their business on revenue growth. But great sales teams can sell ice to Eskimos and Eskimos don't need ice. That's the classic historic example. And so we really set the North Star around value creation. The sales funnel doesn't end at the signed contract and the wired money. The sales funnel ends when you've delivered the value you've promised your customers. And there's ways to make that super instrumented and programmatic. That's a key measure today. And I think even if you're valued over a couple billion dollars, take a hard look at that measurement for yourself. And yeah, we're looking a lot more at a combination of ARR and net dollar retention. TVPI, which is a sum of the valuations of your companies, is almost worthless these days because we're not sure how healthy these valuations are. There's such a variance between how private markets are valuing companies today and how public markets are valuing them. The judge and jury is going to be when some of these companies need to go public to deliver money back to their investors. And I don't know if public markets are going to be as bullish on some of these native AI companies or whether they're going to see some durability risk that we're seeing as well.

Gemma Allen

>> So before we finish, I want to talk about another string to your bow, and that is that you are also a professor at Harvard Business School.

Mark Roberge

>> Yes.

Gemma Allen

>> You lecture on a number of programs and you can break it down for us. But I'm interested in your thoughts on this next generation of young, enthusiastic business-minded founders. It seems as though there's a lot of trepidation out there in the market, especially in tech, around what it will take to— or what a computer software engineer will look like 10 years from now, whether these investments are worth it. We hear a lot about You mentioned mental health as a cause close to your heart. People feeling a whole lot of ambiguity. The generation that is kind of a little bit lost in space in the world of 2026. What are you seeing and hearing? What gives you hope? Yeah.

Mark Roberge

>> Okay.

Gemma Allen

>> And I worry, I guess, about this next gen?

Mark Roberge

>> I teach an entrepreneurial sales class there, which is very popular. I also teach a field trip class to Silicon Valley in January. We would get behind closed doors with all these amazing people out there and 50 students. It's amazing how my students point out to me the imbalance of how much capital, talent, and time is going into building AI and not helping society come along. And that's very worrisome. I think a lot of folks in tech, whether you're on the operational side or investment side, those with a good moral compass are doing a lot of thinking about that. And I am as well. It would be a whole nother episode if you want to go there and more of a philosophical one. But bottom line, and that's part of the reason for the proceeds to mental health, is We who are in the AI epicenter, we need to make sure we internalize that of how much are we putting into building and how much we invest in helping society come along and do our little things. For the MBAs in particular, over a 15-year horizon, we've seen a massive shift from going into banking and consulting to going into entrepreneurship and tech and founders, which is really exciting. I happen to think they're probably one of the best talent pools and overlooked talent pools right now, specifically in tech, like the computer science departments. They've been hurt a little bit more recently. And for example, when I see a software company today, 10 years ago, they used to spend maybe 10 to 20% of the time figuring out what to build and 80 to 90% building because it's just what was necessary to build really strong code. That's flipped substantially. The best businesses who are AI-enabled in their tech community or tech departments, they can build code so rapidly they're spending a large portion of their time figuring out what to build. And MBAs are very well suited to do that, especially those of us with a tech background. So those are some of the themes I'm seeing in the Harvard and general MBA community.

Gemma Allen

>> Well, Mark, before you go, where can folks find your book?

Mark Roberge

>> Yeah, it's on Amazon. The Science of Scaling by Mark Roberge. 100% of the proceeds are donated to mental health, just for two quick reasons. One, it's a big part of my life. I've been a caregiver. I've also been a patient. And a lot of people can't bravely say that publicly because of the stigma associated with it still. So I'm blessed to be able to have the resume to be able to say that bravely. And secondarily, everything I talked about with where society is going, this is going to be a massive shift for society, as was fire and electricity and the internet. And each one of those We evolved to a better species, but it came with scars. If we don't adjust, we'll have massive scars here. And everyone needs to do their little thing. This is my little thing right now.

Gemma Allen

>> Well, Mark, you're certainly a fantastic role model. Thank you so much for joining us on theCUBE and NYSE Wired.

Mark Roberge

>> Thanks, Gemma.

Gemma Allen

>> I'm Gemma Allen here at theCUBE studio at the New York Stock Exchange with NYSE Wired. This is our Capital Series. Thanks for watching.