Robert F. Smith’s “Worldwide Exchange” Interview With CNBC’s Frank Holland

Welcome to CNBC.com. I'm Frank Collin, anchor of Worldwide Exchange. I'm joined by Robert Smith, founder, chairman, and CEO of Vista Equity Partners, one of the largest software investors in the entire world with over $100 million in assets under management. Mr. Smith is a legend in the tech investing business. Mr. Smith, thank you for joining us to talk about tech, your career, your new book that just came out, leadership, your life, and much more as part of CNBC's coverage of Black Business Month. We really appreciate your time. Great. Thank you, Frank. Good to see you. Why don't we start with your book? You just released a book this week. It's called Lead Boldly that features your insights on seven principles of Dr. Martin Luther King, Jr. I'm going to read part of this book that you put in before the chapter on economic inclusion. You actually ask a question. You say, Dr. King reminds us of our power to make a difference in the communities we care about. How will you embrace that power to create lasting change? I'd like to ask you the same question. How are you embracing your power? Thank you for asking that. I'm quite frankly honored not only to be on this show, but when the King Center and HarperCollins asked me to collaborate. It's the first of a kind collaboration where they've done this work. Obviously, I've been honored 55 years after the death of Martin Luther King to present some thoughts and reflections on his writings, his works, as they are represented today, in many respects through my life. I grew up in Denver, Colorado, fourth generation from there. I lived in part of one of the chapters about a beloved community, a community that was very focused on bringing to that community the resources, the know-how, the wherewithal of people in the community to other people in the community. It's a beautiful experience. I call it an expression of love. As that was impressed upon me as I progressed in my life, progressed in my career, I knew it was incumbent upon me to do the same. The ability in terms of what I'm able to manage as a manager of assets in software, over $100 billion in assets, over 91 software companies. Software technology generally is an enabler. It enables productivity, enables access, enables opportunity, and being able to drive that in different parts of our community. Things like InternXL, we have now over 30,000 students on our platform and match them up with now over 1,200 corporate partners for internships. That makes a huge difference in people's lives. Leveraging companies for our Southern Communities Initiative to enable the CDFIs to actually lend more capital with a lower default rate and a higher frequency into communities that don't necessarily have banking infrastructure. Those are just two of the areas that I think are critically important that people like me and me in particular can enable because of my knowledge of software, knowledge and capability of the productivity that software brings into different ecosystems. You were mentioning one part of some of the initiatives you're looking into. It's actually helping minority deposit institutions and community development financial institutions. Why do you think it's important for you to put your resources towards those? What do you think that enables? Because going back to the book, Lee Boldy, a lot of people don't talk about this very often, but part of Dr. King's message, it wasn't just social justice, it was economic justice. It was. A lot of it was underpinned with economic justice. People have to understand that the stronger communities in America make a stronger America. What really drives opportunity in this country typically is access to capital and utilization of capital to build businesses, to create jobs, to create stability, to improve what I'll call the conditions of the communities that people live in. If you don't have access to capital, you can't necessarily improve your homes, borrow against to fund your businesses. In many cases, over 70% of the African-American community actually does not have a branch bank present in that community. In the community I grew up in, there wasn't one. In fact, we didn't have a supermarket either. As a result of that, the small business owners in my neighborhood didn't have access to capital so that they could actually grow their businesses and then hire more people. Everyone knows that in America, the vast majority of companies or employers are small to medium businesses. All of those are important components of being and participating in enabling these community development banks and minority depository institutions to lend more capital at a higher frequency. Again, the name of the book is Lead Boldly, a really interesting read. I'm going to finish it. As I admitted to you, I'm about halfway through. I wanted to go through a couple of different parts before we spoke. I want to talk about your journey as a leader. Of course, you're the leader of Vista Equity Partners. You're the chairman, the CEO, and the founder. Before that, you were just a leader of yourself out in Goldman Sachs San Francisco office. You were the first person really focused on tech, M&A. How did you become a leader? What are the things in your life that led to you being a leader? Following great leaders. I had the great fortune in some respects first growing up in my household and to see my father as a leader of the community. While he was a teacher and a principal, he also was chairman of the local YMCA. He was chairman of what was called North City Park Civic Association. He associated himself with community leaders and some were in government and some were in private industry in ways that they could actually create real change in our neighborhoods. When you see that as an example, you realize that's frankly part of your responsibility as you grow and build your skills and build your capabilities. That's something that was an early impression on me. As I matriculated and I went to great institutions like Goldman Sachs and saw spectacular leaders there and saw how they not only led in terms of the transactions that they did, but actually led their teams and teach and trained and mentored their teams so their teams would be more effective and be able to be leaders and call it the league tables as we call it. All of those things are just a whole series of events that, as Dr. King says, build upon who you become as a person and drive ultimately your leadership style. Those are important aspects of my upbringing and my rearing, I'll say, that led to me being the leader that I am today. Do you have a philosophy when it comes to leading that's overarching over everything or do you lead in different situations in different ways? You have to lead in different situations in different ways. The world that we're in, for instance, in technology, it is constantly changing. One of the ways that we have to lead is to think about how we embrace new technology into our best practices to enable our companies, our CEOs, to become more effective. One of the more recent things we're doing, of course, is we've now built in what I call an agentic factory with a very large partner being Microsoft and saying, okay, how do we take these enterprise software companies? In the early days, we were taking on-premise software companies. We built a factory to convert them to become, in essence, what we call cloud or cloud native or SaaS companies. We did, I think, more of those conversions than I think really any institution on the planet. Well, now that there's this new technology going, you know, Gen AI, we've been investing in AI for over a decade, but now how do you take these companies, these cloud native companies in some case or hosted companies and make them agentic? Building a factory is what we've done and now we're driving companies through that. In that case, you actually have to understand the technology and lead by example in building capability and resources so you can actually drive and enable your portfolio company executives see how it's done and give them tools and resources and partnerships to be effective in that challenge. Now, you've said agentic AI is going to be a differentiator and some companies are going to have the right to exist because they use it correctly and others are just going to go by the wayside. Do you still feel that same way? You said that a couple months ago and when you're looking at companies right now and their use of agentic AI, do you believe they are using it effectively? Are you starting to see that difference between the companies that are going to make it because they've figured out how to use it and those that aren't going to make it because they don't? What we are seeing is I like to say enterprise software, enterprise software, consumer software is a little different. We'll ultimately end up in those three states. Agentic, I call it rule of 70, you know 40 is the new 70 or no right to exist. And part of that is do you have sovereignty and dominion around workflows and data sets and then can you actually create agentic engines as part of the delivery of a product or solution to your customer base? What we're finding now that we've gotten again the first company through the factory and now we have dozens more now moving through it is a few things and some great insights. AI is going to feed enterprise software into these states. What it's really going to do though is start to you know attack what I'll call the services industries because the capability you go from a license to building a team of workers, agentic workers that have a high productivity. Enterprise software already is over 625 percent ROI on the purchase element. We're actually seeing inflections of exponential proportions of the productivity that agentic solutions can deliver into an enterprise and so those are real actual you know things that we are experiencing with companies today. Our companies today as we install these systems in their customers. Well this leads me to my next question. So if you're using it in your portfolio companies you must be using it just here at Vista. We use it every. Right so I have to ask has AI changed the way that you value companies, the way that you look at software, the way that you invest, the companies that you want to invest in as opposed to the companies that agentic AI says you should invest in? I mean what's the difference? Yes it changes it insofar as you now can understand what you can change in the business. If you have a business operating with a certain margin, certain growth profile and say okay I'm going to bring the the agentic solution to code development. It can reduce the cost of that code development. It is highly effective in new code. Not as effective in what I'll call you know legacy code but we found what the proper benchmarks are so we can actually drive against that in terms of our annual operating plans. You can use it for customer support. You know exactly what sort of deflection rates you can cause and reduction in cost in customer support. For go-to-market so all the P&L aspects of running a software company we now are applying you know some form of generative AI solution set and now actually have specific KPIs that we can measure against and actually manage towards to reduce the cost of the product and in the agentic solution in terms of the value that it brings to our customers. The key now is how to price against that. If we're deflecting you know tens of millions of dollars or in some cases maybe hundreds of millions of dollars in cost for our customers, how do we capture our fair share with our agentic solution? So that's the real challenge today. All right so that's the challenge right now. So I do have to ask how come we haven't seen big changes in the portfolio? I mean so agentic AI sees things the way you see it because these are the companies that you and your team thought should be in the portfolio. Agentic AI hasn't said not this one has to go or this one should be in. It is more complex than the words. Okay. Okay because you have series of agents that actually perform the tasks and then you actually have to now build a whole series of administrative agents to manage and monitor those agents in ways that they are actually giving you and producing solutions at a high precision. Enterprise is very different than consumer. Okay and the consumer you and I can deal with a certain level of tolerance or hallucinations as to how it's characterized in a consumer. In the enterprise we can't have that that that that that fuzziness of answer. Okay and those randomness of outcomes because you would be upset if it would if it gave you a 95% probability that your you know your wire transfer goes to you and a 5% probability goes where this agent might think it may be a good place. Right. I think I would. Yeah. Yeah. Okay but you know in a consumer context you're probably a little you know more indifferent about oh you know this restaurant versus that one and it might and you're okay with that call it that fuzzy answer but in an enterprise you need a lot more precision in it and so you actually have to build and construct the infrastructure with a with a higher degree of certainty of outcome. Just about everybody I talk to says hallucinations even in enterprise AI applications are still a very real thing. We see it sometimes in the consumer side but also you know on the on the business side where you know maybe it's Alphabet's Gemini system doesn't quite work the way it should or we see it with Grok coming up with answers that I don't think Elon Musk wants it to come up with I would assume. I'm going to assume they don't. When do we get to the point where that's no longer a problem and is it simply a matter of building up what you're seeing to be calling like an army of different agents to monitor different things or is there just a timeline when we get rid of those? In the enterprise part of what we do is we constrain what are the outcomes ultimately for things like underwriting an insurance claim or an insurance policy or managing you know a loan syndication. You actually know what the likely you know improbable outcomes should be and you can actually eliminate in some respects things that aren't probabilistically true. So that's a function of the architecture that's a function of the way you actually build out the capacity of your of your agentic solution. That's why the enterprise is so different. So now the installation of those and the configuration of this these agents activities in the enterprise is what we are doing today. So it's not going to be a mass proliferation but as it proliferates it creates massive implications in terms of the value that these agents bring to the customers we serve. So it is it is occurring it's just you know as it occurs you actually have to get it exactly right as the kind of close and so that's what we're focused on. All right there is a mass proliferation of AI infrastructure. People are just building it up. I mean seeing billions in spending announcements in just a couple last months. Other also big spending plans whether it's Alphabet, Microsoft saying they're going to spend way more than people thought they were. I'm going to add some other names. Meta, Core, Weave just to name a few. How does all this spending on infrastructure shape your view of software? So again for us we will need the inference capabilities of what these models deliver. But in the enterprise again it will be my view I think ultimately will be compute more at the edge. Because there's only so many ways you're going to underwrite a certain you know activity which might be again it might be again insurance claim. It's only so many outcomes so you actually don't need call it the power that a lot of these models deliver. All right in that environment you might need it for other applications might be diagnostics or you know the development of a new drug that's a different that's a different use case than the sort of use cases that we use. So we actually think we're going to be mass beneficiaries or beneficiaries of the mass proliferation of that compute because that compute cost will come down pretty dramatically. And then as you actually build out capabilities at the edge we think we're going to even gather you know more economic opportunity in that context. Again the enterprise is very different than consumer and depending upon the use case as to you know how much of this this compute power do you need. And at the edge of course typically it's going to be lower power, lower power consumption, lower cost, and higher frequency, lower latency in terms of the work that that you need for those particular applications. So with that thought in mind is the infrastructure build up just getting to be too much? Is there some people that believe it's just becoming too much? Other people say they're going to keep building because they feel like the demand and the capacity are going to meet sometime in the very near future. In your mind and I want a real question I really answer out of you on this one. Is the AI infrastructure build out is it becoming too much? Are companies in your mind spending too much on data centers? Are they spending too much on chips? Spending too much money to try to get chips that haven't even been made yet to book them in the future? It's it's it's hard to say that that question of too much. You know there is. What can you say though? Again I don't think all the math is completely done okay because again I said I don't think there's enough calculus on how the enterprise will be effective ultimately because I think you can actually deliver a lot of the compute in a much smaller format at a much lower cost with much less power consumption ultimately. So then that leads me to think you think the build but again that's of the enterprise. There are other aspects and in in terms of how this is going to be used that may exceed that demand of current capacity which might be you know we we may decide that it's important for us to have you know certain experiences that only AI can deliver you know to a consumer market that they're willing to pay for. So again it depends on what assumptions if you're thinking about consumer assumptions. I live in a world of the enterprise and I have a good sense for what computers required for the businesses that we own today. Well in all fairness a lot of this build out is focused on enterprise functions especially the companies I named whether it's an Alphabet or Microsoft or Cori. That's really focused on enterprise functions and also I want to go back to what you said you have a bunch of agents. Yeah well they are enterprises a lot of their functions are doing are that capacity is for consumer activities. They're not necessarily driving consumption and saying oh this is going to be used for insurance okay specifically they're saying in other aspects of the business of other aspects of business that it will be used more frequently. All right but just going back to your your agents but your agents are telling you on the edge low latency less power that's where it's going to go. Yeah in our in our space for sure. All right I want to go to your companies. What companies in your portfolio are you excited about? What companies do you see taking advantage of a genetic AI or having the best ability to take advantage of a genetic AI in the near future? Yeah so things you know companies like Gainsight, companies like Avalara, companies like Sales Loft. These are all businesses that as you agentify the solution set they bring massive massive return on investment for their customer base. And so not surprisingly you know companies like Resolink, companies like Gnosis Freak. These are businesses that because of how they sit in their ecosystem agentic solutions provide again kind of a massive return on investment for their customers. And now the question in our mind is how do we capture our fair share of that economic rent that's created from the highly efficient agentic solutions that we're delivering to our customer base. So with those companies in your portfolio are they going to stay in your portfolio with them now being agentic AI powered? Are you going to possibly look to put them in the public markets? Because according to your thesis they're going to have an advantage over a lot of their competitors. Yeah well it depends. Certain ones will go public. Certain ones will you know we will find wonderful trade sales. Others will put in like a continuation vehicle like you know CSG Cloud Software Group. I mean those are businesses that have unique positions in their environment. You know CSG of course is a virtual machine dynamic. If you think about a world where agents have to live somewhere, living in a virtual machine is actually a good place for them to live. So we think that's actually one of those businesses that you know built properly, configured appropriately with the right partners can actually have a very you know long tail of opportunity associated with it. So take us basically around the corner to give us an insight into what's going to happen that we can't just see in a straight line. You take a lot of companies private. Are there any sectors or areas in the world of tech right now that you're really focused on that you think are undervalued and there's a lot of opportunity to take them private, infuse them with agentic AI and then either keep them in portfolio or again turn them public. And I tell you that and then the prices go up and then I can't buy it Frank. You know that's a bad question to ask me. Again what we think about in our underwriting is what are the things we call them critical factors for success under our control. We have a whole new set of knowledge and capacity because how we have built out this agentic factory that enables us to underwrite differently than we did even six months ago. And so that's what we're applying in our analysis and oh these are the businesses that operate this sort of margin. You create an agentic solution it operates at this sort of a margin. If you're actually able to reduce costs in the production it operates at this sort of a margin and it actually has sustainable infrastructure or moats around it that are protected from I'll call it you know agentic solutions attacking that industry particularly. So that's the the new insights that we're bringing forward in our underwriting. All right if I can't ask you directly let me ask you about some different areas. Cybersecurity. We see a deal pending. Palo Alto Network's looking to buy CyberArk. A lot of people talked about an increased need for cybersecurity in the AI world because not only would companies have AI so with potential hackers or other threat actors out there. So what's your view on cybersecurity right now? Which in the public markets is actually underperformed in recent months surprisingly. It's interesting you know you and I as individuals have what I call a surface area of attack that all these wonderful cyber security companies like know before are like one of our companies to to protect. As you and I start to manage agents not surprisingly the surface area starts to increase exponentially. And so you've got to think about how do you protect them. Do you protect individual agents or do you protect the environments in which these agents live. Okay virtual machines are great places for those agents to live and then protecting those you have the ability in many cases to to keep the same level of call it an attack surface area and manage that in a way that as you proliferate agents it actually doesn't proliferate the the amount of the call it the attack vectors that then can be exploited. Interesting. Another part energy. A lot of talk about nuclear energy possibly powering these data centers. Just the need for energy in general and also energy prices moving higher. What's just your view on stocks or companies that aren't necessarily stocks. Focus specifically on energy on providing energy and also infrastructure for data centers and also empowering AI. Yeah energy again you've got generation, you've got transmission, you've got distribution, you've got utilization and those are that's the complex factors that that go into that question. And if you look at the current assumptions that we've looked at you know there's going to be a massive demand for it if we continue to build out not in all aspects of that. Of course now you're seeing people say I'm going to put data centers right next to energy. We're of course looking for more efficient ways to generate power and energy to support this. But at some point in time if you actually believe the thesis that a lot of the compute will be done at the edge with a fraction of the power consumption which it can be in some cases you're going to go through that same sort of an overbuild cycle. So we'll have to see who's right there but that's a dynamic that we think ultimately there's going to be that that dynamic is going to come into play as to you know what the energy consumption will be. So just to be clear data center build out not clear too soon to tell but you're saying it seems like we might be on the precipice of an energy build out that's going a bit too far because you think the direction is moving towards the energy build out again you know our transmission and distribution systems actually can use some augmentation in any in any respect. Do you need the generative power right now based on calculations? Yes but ultimately you might not if you actually have different forms of compute that have you know a fraction of the consumption to deliver the inference capabilities. All right. Tech's your wheelhouse but people just want to hear your general views on investing. So I know you watch CNBC. We have Rick Reader on our air. Come on you know you watch. We have Rick Reader on our air. He said this is one of the best investment environments maybe ever but certainly in recent years. He was largely talking about public equities public bonds. I want to get your take whether it's public equities public bonds just in general. How are you looking at the investment environment? The investment environment creates opportunity because now what I'll call the volatility. There are a lot of things changing all at once and so being able to participate in that is how you can actually make money. You know you introduce a new technology in this case Gen AI. Guess what it's going to affect every single thing. It is a general purpose technology like electricity right. So it is going to affect every single industry in one way shape form or another and so what you have to do as an investor is to think about where are there where is there an arbitrage of opportunity and where can you create an arbitrage of opportunity if your value if your investors like us who are valued and we bring a value creation dynamic to it. That's how we think about it and so when we look at the different sectors still believe enterprise software if man if you have the Sovereign Dominion Workflows and Data Sets is a wonderful sector to enable that enterprise software to become agentic because it actually delivers solutions that have massive implications on the services sector. You need fewer and fewer services capabilities in terms of you know people doing their work that you can actually inure to the benefit of those who are utilizing that technology. So it sounds like you agree but you you're saying because of the volatility. The volatility is creating opportunities. Volatility really just being the change okay. New changes in markets of course we've got tariffs. We've got all sorts of changes in the equilibrium systems that we lived in you know two years ago and three years ago. Anytime there are those sort of new introductions of changes that's when you can actually start to look for opportunity sets. All right so also a couple months ago you said things are kind of expensive right now. So with that thesis in mind and you say it seems like you're agreeing with Rick Reeder. It's a pretty attractive environment to invest right now. Are you thinking about putting some money to work? We know Warren Buffett or at least Berkshire Hathaway is putting about five billion. Not a 348 billion. Yeah but still looking to put about five million to work possibly in an industrial. Are you looking to put some money to work actively right now? So if you look at you know the world that we live in there has been a massive deflation of the prices of the companies that we typically buy. Okay in 2021, 2022 some of these companies were trading at 17, 18, 20 times. They're now trading closer to seven and eight times. So it's a great opportunity which is why you've been seeing us doing a very consistent set of acquisitions and a consistent set of realizations across our portfolio. So we've actually maintained the same sort of a balance and we're looking now to continue to be aggressive in our acquisition you know environment to go after companies that we think now can become magentic in our world. All right we're almost out of time but I do have to ask you just more broadly about the public markets. I know the private market's your wheelhouse but as you look at the public markets the S&P and the Nasdaq coming off record highs 17th and 20th respectively for the year. What's your view on public equities from now and to the end of the year? Also the fact that we're not really clear on what the Fed's going to do. Seems to be a lot of confidence the Fed cuts coming up in September. That's when we started the year. We started the year with a lot of confidence. So just your view on the public markets overall from now to the end of the year? You know again a lot of the indices are weighted to just a few stocks as opposed to you know general market call it perspectives. And so like all investors you can look at those that you know have outdo influence on those indices or you look at those ones that actually aren't being affected that people do not understand in terms of how agentic solutions are going to improve their margins or improve their market conditions. So I think is it a good time to invest? Absolutely. We'll see what happens in the bond market you know over the next really couple of weeks is my sense. But we have found in our markets and again private markets in fixed income that there are a lot of software companies or a lot of businesses because the VCs have now gone and invested in you know AI companies that these great companies actually need capital. So our credit business is actually going through a pretty interesting boom. All right so it sounds like you're saying it's a good time to invest in the public markets. I want to be clear because a lot of our viewers also can invest in the private markets as well. Do you see the public and private markets as being equally as attractive? It's hard to say. Again when we invest in the public markets it's to take them private. So that's how we think there's an opportunity set then to then you know perform some operational transformation and then drive them to a higher value. Your story is amazing. Just simply amazing. Coming from Denver, going to school, you're an engineer. You're an engineer. I'm sure you didn't even think about getting to this point or maybe you did. Maybe that's how you get here. Can I ask you just a broad question for more of our broad audience? What's the biggest money mistake you ever made and what's the biggest you know correct money decision that you ever made? Yeah that's it. So you know when I think about things that have gone well, you know listening to what I'll call mentors, building on my own capability, betting on myself and understanding that you know if you're skilled, you believe in your capabilities and you're Spartan arbitriers that you actually go put capital resources and of course your energy and intention behind them. Those are the things I think about from a positive side. I think some things that you always have to correct you know you always have to think about people and personnel and is this the right industry and certain times you will invest in a business and we're like oh man that one didn't work out because something changed or you know we missed something but you know fortunately that hasn't happened all that often but those are the things I think about every day. So the key is I think just having a view of continuous improvement and having a system that embraces continuous improvement as a core tenet and not just something you do once you once a mistake is made. So that's actually business. I mean just you personally like what's your what's your big do you have a money mistake that you've ever made? I wouldn't I don't know the good news is I have ADHD on that sort of stuff so I forget about it. Are you optimistic? Just looking forward I don't know if it's a year from now just going forward just your general view on what's going on in the United States and the world a lot of these changes that you were just talking about. Yeah I'm very optimistic actually I think you know these new technologies create new opportunities of course there's going to be some you know creative disruption in some cases on things that we thought would last forever but I see the productivity now in the work that we've done the factory that we built or the products that our companies are now able to deliver for their customers that and I again you know I've never seen this sort of you know kind of return dynamic and now you know proof of the pudding is always in the eating so you have to now go get these things you know done done not just done right and get them installed and then demonstrate the value to the customer base. So I'm very optimistic about that. That's a great place to leave things. Robert Smith, Founder, Chairman and CEO of Vista Equity Partners. Thank you. Thank you for your time. It's always a pleasure and thank all of you for watching CNBC.com. I'm Frank Holland, Anchor of Worldwide Exchange.