‘The Technology Opportunity of Our Lifetimes’: Bessemer's Byron Deeter

0:05Welcome to Goldman Sachs Exchanges Great Investors. I'm Ken Hirsch, co-chairman of the Global Technology, Media, and Telecom Group and head of venture capital coverage within Goldman Sachs' global banking and markets business. Today, I have the great pleasure of speaking with Byron Dieter, a partner at Bessemer Venture Partners. Bessemer is one of the largest and best-known venture capital funds out there, and Byron himself has backed 26 companies that are currently valued over a billion dollars. Byron and I will discuss the venture landscape and the most promising developments in AI and the software industry.

0:37I'm particularly excited to hear his uniquely informed views about what tomorrow's most valuable companies are doing today.

0:48Byron, welcome to Great Investors. Always fun to be with you, Ken. Thank you. So, before you were a venture investor at Bessemer, working with CEOs of promising companies, you were a CEO of a promising software company yourself, working with the venture investors at Bessemer. Tell us about Trigo Technologies. What was the genesis of that company? I got the bug for tech. So I was actually working at TA Associates at the time back when they did venture. They've moved into private equity more now, but saw the market opportunity, went looking to invest in a business that was addressing the supply side of the e-commerce and marketplace wave, didn't find a company to invest in, and so decided to start it instead and was fortunate to get matched up with the Bessemer team as our investors at the time.

1:34built the business through the dot-com crash and the early craziness of the early 2000s. And we're fortunate to build it up, got to profitability, global scale, and then ultimately sold to IBM and joined them for a year and a day back in mid 2000s. And what motivated your decision to join Vessemer yourself? I saw their product on the other side. As a CEO and founder, we were fortunate to work with a number of great investors, but Vessemer was the firm that I thought was most authentic to the art of serving an entrepreneur and our team. They were the ones that very sincerely I'd call when I was driving home from work and things went very badly that day.

2:13Customers canceled us, employees were trying to resign, or we had issues. And the other investors were the ones I called when we had good news, when I felt like I needed to put on a show or perform or share good things. I mean, we had many of both, but Bessemer was truly the supportive, value-added, collaborative partner. And when thinking about joining venture and returning to the industry, that was the ethos and the platform I wanted to be a part of. Strong, top tier global, those things, but fundamentally serving entrepreneurs in an additive, valuable way. So let's pause and talk about Bessemer for a moment.

2:48The firm goes back more than a hundred years, tracing its roots to a co-founder of Carnegie Steel. Very much so. So not as old as Goldman Sachs, but close. Many venture firms will tell you about their most successful investments and Bessemer has many. But one thing that caught our eye on your website is the anti-portfolio. And I know you've talked about this a lot publicly, but it's a list of massively successful companies that you had the opportunity to invest in and missed out. And I love the way you say it. I think you say, we've had an unparalleled number of opportunities to completely screw up.

3:24Just the humility and humor, it's striking. But what's the thinking behind that? Is it just that? Yes, largely so. Obviously, we've had a lot of success. And so the portfolio page is one we're very proud of. But we like to have fun and we're constantly in learning mode. And the anti-portfolio is that. It's not the companies that we've invested in and have failed. Those are shared errors. We also try to learn from those. But this is very much what were the big opportunities that could have been. And in our business, the crimes of omission are much greater than the crimes of commission. And the deals that we miss and the opportunities to work with transformative companies are truly the things that keep us up at night and occupy our off-sites and occupy our headspace.

4:04And so the anti-portfolio is that. It's constantly trying to learn from the what could have been. For me, it's companies like Tesla and Atlassian, which are great generational companies and have a lot of ties to the teams and work to them in other ways, but for a combination of reasons didn't ultimately invest. And those sit high in my mind and have Proud positions on the anti-portfolio of what could have been and what should have been. Yeah. Two incredible companies. Do you often go back to investments you missed and invest in a later round? We do. And actually that's one of the reasons why we've added growth capabilities, scaled up the firm.

4:40One, because we think the industry is heading there and much like Goldman Sachs experience in financial services, this maturation where platform global scale was needed to serve your clients. We feel the same way in our industry where the global platforms are really what's important and the scale is needed, but we wanted the ability to still participate in these category-defining transformative companies, even if we'd missed them at the seed or inception stage. And so our goal is to work with them as early as we can. We will write seed checks. Twilio started with$125 ,000 seed check, but many of our investments we miss at that stage and still believe can be fund returners and material outcomes because their opportunity is so big.

5:20I read the Twilio memo that you posted on your website, and I found it interesting that it said they wanted to raise a full$3 million round, but you were recommending a portion of that for specific reasons. Isn't that unbelievable? It's just fascinating in the context of what we're living through today. Yeah, Ken, my first IPO with Goldman was Cornerstone On Demand. I think they were at 55 million run rate and IPO'd for probably 700, 750 million, traded up ultimately to several billion, but that's a series A for some companies these days. We've seen a lot over the last 20 years. We'll come back to that.

5:55So getting back to your story, you're an entrepreneur, you sell to IBM, you come to Bessemer. What was the transition like from being an entrepreneur to being a venture capital investor? And what did you find most challenging about it? The mindset does need to be different. And as an entrepreneur, day in and day out, you're solving problems and it's all about the what can be. And it's a game for optimists in a wonderful way. And in venture, what I found early on was I was constantly imagining the what could be in a lot of these businesses and taking on harder problems than I needed to because I got so excited about the potential in the teams and the what could be.

6:31Whereas in a lot of cases, the best venture deals are finding the companies and teams that are already succeeding. They're going to frankly succeed with or without you. And you can just change the slope of the line on the margin to take something that's already going to be transformative, but to truly make it an N of one outcome. And that's really what so much of the venture business is about. And the unfortunate reality is we say no 99.9 % of the times. We're constantly getting approached by really intriguing businesses and opportunities that could be. But really the peak of the opportunity is to find those businesses that are truly at the top of the pyramid and to be more selective and then concentrate the impact on those businesses specifically.

7:13Let's maybe shift to your investment process at Bessemer. Help us understand your investment process. What are the criteria that you most value, that you most look at when you're making investments or choosing not to make them? Certainly. So I think firms philosophically often fall into two camps. There's the gold nugget or the blackball. Do you, as a team, make investment decisions, but anyone can blackball something if the group consensus isn't there? or is it the empowerment and independence model of the gold nugget of, hey, I'm gonna pound the table for this one. I believe in this company, I'm gonna put my neck on the line.

7:50We're much more in the latter. We are much more in the camp of empower our investors. And if you make a bunch of bad investments sooner or later, that'll catch up to you. But we promote people, give them checkbooks and encourage them to run. And so that's very much what our process is. We do take votes and do those things for record keeping and the like, and to stop the truly outlandish perhaps, but we don't vote down deals. if a partner wants to do one, they can. And so this is all about learning an industry going deep in a process we call road mapping, where we really will have a point of view about a sector, a theme, and the types of companies we want to invest in so that when we bring those forward, we can be fast, we can be responsive, we can pay aggressive prices, and we can really look to work with the leaders because the market's moving that fast and requires it.

8:33And when you're looking at investments yourself, just given your success, are you more focused on TAM or the entrepreneur, or is it always both? So I think most investors will tell you both are critical. I have come to back the jockey more than the horse at this point, where I do think in these especially dynamic markets, AI front and center there, team matters way more than it ever has. These things are moving so fast. Look at the prices of talent. Yes. I think you're seeing that from the incumbents, recognizing that and getting in these bidding wars to recruit people back. And the opportunity speaks to the reason why they have to pay so much because the opportunity for these talented individuals in very small teams, 10 person teams can now do Herculean things and create multi-billion dollar businesses.

9:20The leverage from those individuals is so extreme that we are oftentimes backing individuals with ideas at big numbers or empowering small teams with big war chests to go out and figure things out in very open-ended ways that historically hasn't happened in venture. One of the things that you made a comment earlier about imagining scenarios for businesses when you were making that shift from entrepreneur to investor, and one of the things that I took from reading some of the investment committee memos that you post on your website is almost the conservative bias in balancing the vision and the ambition with realistic outcomes.

10:00And the one that really stood out to me, if I read this correctly, with Shopify, because I think it says that in the upside scenario, we think this could be a$400 million exit. And it's today a$200 billion company. So it does seem to be inclusive of some conservatism. In a culture where you have the golden nugget, does that encourage a bit of a more conservative view among the partner who's pounding the table? How does that dynamic work? Yeah. I love that you raised it because I think it speaks to two things. One is the scale that we've witnessed over these 20 years. So investments like Shopify and Pinterest and Twilio started at the seed stage where companies used to go public at, as I said, 700 million and a billion was the home run.

10:45Now that's a fueling stop, usually in a private round. And so the second thing it speaks to is horizons of business building. And this is something that we spend a lot of time with our entrepreneurs on is thinking about the businesses and just how you can grow TAM an opportunity. In the Shopify case, Commerce and Payments opened that up. And it's interesting the way we actually brought those learnings to our vertical SaaS businesses, you know, Toast and Service Titan and a number of others have implemented a financial opportunity that's doubled their TAM and unlocked it. You see this in AI right now as these businesses are addressing successive markets.

11:19And these TAMs, I often use a comment, use the phrase of adding a zero, often to our valuations and the investments we make, but certainly for the exits and the outcomes. These markets are so much larger when you can think of the horizons that you can address. And that's been the part of the journey that we want to be rational in terms of how we scale the business each step, but always go after the maximum opportunity. And that's how those two worlds can coexist. A rational conservatism in terms of how you think about scaling in the phases of the business while going out to build a hundred billion.

11:49And now, you know, in the anthropic case and perplexity and Canva and these businesses, we're going out to build a trillion dollar company. Right. We're going to come back to that. I want to just touch on one thing. You mentioned roadmapping. And another Bessemer story that jumped out to me was, I'll call it Twilio, SendGrid, both companies that you invested in, both went public, both combined. Is that just the ultimate embodiment of a successful roadmap? That was a wild experience where I was on the board of two multi-billion dollar independent public companies that then decided to merge. And so I'll say it was both the most stressful kind of emotional cycle for me to go through and the easiest deal I've ever done because I had to recuse myself from both.

12:30But we don't invest in competitive companies. And so you have the scenario where companies often will get adjacent. Usually they're helping each other. And that's one of the really fun parts of our portfolio. When you get folks together for events and portfolio things, the CEOs truly are lifting each other up and they succeed together and they're collaborating and they're encouraging each other. And in this case, Twilio made the decision they wanted to go into the email business and they wanted to do some things that the SendGrid platform offered. And they were either going to buy SendGrid, buy a competitor or build.

12:57And the probability was mathematically, they're probably going to do something else. And so it was incredibly fortuitous. And I think they made the decision they wanted to acquire the leader, which SendGrid was, but it was a fantastic outcome. With that came a lot of the SendGrid talent. It was a win, certainly for both sides, but a roller coaster because of the likelihood that it was going to go another direction and they were going to end up competing. A win for Twilio, a win for Bessemer in that you got another great investing partner in the same way that you came to Bessemer. That's right.

13:30You're speaking to Samir, who is our SendGrid CEO, stayed at Twilio for a while and has now joined us as an investor. And we do have several former Bessemer operators on the team and also several Bessemer employees that are now operators in our portfolio. And so it's a really fun back and forth. And I do think there's skill sets that crossover and learnings that are relevant. So one more on the investment process in the context of the era of AI that we're now entering and living through. How does that investment process change in a world where capital needs are that much larger, where ownership stakes are that much smaller, and the need to continuously underwrite at higher values just given the rate at which these companies are scaling is really unprecedented.

14:17How do you evolve the investment process to meet a totally unique situation? Well, and I would add one thing even on the front end of that, which is early on, you can't look at business models and unit economics. And that's where a conservative financial investor would never do early rounds in these AI companies, given the magnitude of an investment, the risk you need to take in terms of building out early capabilities. And so it's both that, which is you have to have high conviction, be visionary and buy in combined within what become extremely deep capital commitments for businesses that are operating near the infrastructure layer that are doing foundation model work that are at the infrastructure phase of this build out.

14:57And it's changed the methodology in a lot of ways. One, the teams are always thinking about layers of financing, layers of access to capital over time. They'll walk down the capital stack to non-venture providers as well. But as a firm, we've also needed to scale up to meet those demands. Our goal is to be able to fully fund our companies in most cases, in the extreme cases that you see in the quarter trillion dollar rounds, those become harder, but to be able to meet the needs and then to be collaborative with partners here. And so what we're doing more and more is syndicating these deals. We're working with multiple firms.

15:30We're pulling in our community. One of the benefits of having 150 IPOs and counting is a lot of folks have worked with us on a lot of successful businesses. And now we're actually enjoying the chance to partner again. Whereas in the early stages, these syndicates have been tough in venture. And a lot of people have sharp elbows and want to do these rounds solo and do fight for the ownership. In these later rounds, there is an opportunity to work with great folks again. And frankly, that's a positive all of this. I know you're involved in several of the top AI companies out there. You just referenced, I think, indirectly Anthropic, but Canva, Perplexity, AI native, AI powered.

16:07Just spend a moment on how you see the potential for AI to change the economy and the world. And has there ever been a moment like the one we're in? So we are 10 figures deep into pure play AI bets. Like you mentioned, Anthropic, Perplexity, verticals like Abridge and MaintainX, as well as a number of our incumbent challenger companies. So a company like a Canva or an intercom that's already at large scale, but is reinventing themselves through AI. So, you know, investments in both sides. And I think the legitimate answer is no, there's never been an opportunity like this. We collectively live through many transitions in software.

16:45I do think the closest analogy is cloud wave one. And what we saw as with the hyperscalers analogous to the foundation models, what we saw then with the pass and SAS layers, what is going to start happening in the app layer side of this. But you add a zero to these outcomes. You're pulling in services dollars. You're transforming these businesses in entirely unforeseen and different ways, which will be the technology opportunity of our lifetimes. Where do you see the most attractive opportunities for firms like Bessemer? So much of the value to date has accrued to the incumbents. And where will venture backed companies have the greatest right to win?

17:24Yeah. So let me give you two answers to that. The first is a direct answer on the investing side. And when you look at NVIDIA, which today is plus or minus 4 trillion market cap, you look at what's happened with RODCOM, you look at the hyperscalers, it's playing out similar to how the cloud wave did, which is the infrastructure layer is scaling and monetizing in many ways first. The foundation models are following. And so for public investors, it's been incredibly frustrating because there's been really only hardware ways to take advantage of this, whereas the venture market is starting to see it working from the bottom up.

17:55But I do think it's going to follow the trend where the middle and app layers are going to follow this chain. And either you believe that those trillions of dollars that are being monetized on the infrastructure side will carry over through software value capture or this whole thing is a hoax and it's going to collapse. I believe in the former. I believe that the fundamentals do hold it together. And that, so it will inevitably roll to the end user through anonymization, which the apps will largely capture, and then it'll roll through shared with the foundation models as the next gen of hyperscalers.

18:26We certainly saw that in mobile. Very much so. In terms of the value creation on top of the platforms. And just given the way that AI investing has evolved, you referenced this indirectly earlier, but how does it affect your strategy when you think about the ability to serve the companies at all stages, seed, early stage, and growth, and the way that you run and operate Bessemer? Yeah. So we think much more as a full service platform today to catch those needs. And I'll also admit as a former entrepreneur, the Hippocratic Oath of Venture definitely applies, which is do no harm, first of all. So where things are working, job one is just stay out of the way and don't screw that up.

19:07Job two, three, and four and five is help where you can, which is whatever they need. If it's capital sources, as we noted previously, then lining those up. If it's helping with talent, if it's recruiting, if it's partnerships, it's strategic advice, etc. So we have teams and over half of our firm is now in these areas, platform and operational support, all sorts of expert advisors that are specialists that can be called in a moment's notice, whether it's the absolute chaos. I've been woken up at 2 a.m. for an API business with a DDoS attack that could have literally taken the company out if that wasn't resolved by midday.

19:40We've had PR crises, we've had opportunities that are unprecedented. And so it's that combination that now they look to for a venture partner. I also remember I promised you two answers to the prior question. So let me just add in there. I do believe that the benefits for society around things like education and healthcare and where we start to see AI roll through and transform things and pull forward scientific discovery that wasn't possible will also be one of these lasting enduring benefits that this tech wave will bring. And it complements how we think about our business. And it also makes it fun to be in this industry.

20:12As we've talked about many times, It's so damn exciting to be in the middle of tech right now and to be in the middle of this transformative phase, which we will talk about for generations. Our grandkids will talk about the stories of when AI was early. You're perhaps best known for your expertise and your many successful investments in cloud computing leading up to this era of AI. And I believe you've now published your 11th version of the 10 laws of cloud computing. and you've made a lot of predictions over the years around how to win in cloud. When you look back with the benefit of hindsight, what did you get most right?

20:51What did you get wrong? I'm guessing not much on that front in your predictions over the years. Oh, that's a great question because a ton of learnings, and I think the core of it was around underestimating TAMs with these horizons. Payment's probably the most obvious one that we saw that early for the industry, but we could have rolled it through much more in the portfolio. And frankly, we should have done three times more investments than we did at the time. When we look back, just about every company that we brought in for a third meeting or a partner presentation, we should have invested in.

21:24That there was just such an explosion of opportunity and market transformation. And those thoughts do sit heavily with me now in this window when people look at it and say, oh, these valuations are crazy and it's overvalued and people need to pull back. And I look at it and say, this is that opportunity again. This is, whether you call it horizon two or three of cloud, this is the next phase of cloud. This is only possible because of all of these GPUs daisy chained together with the centrally provisioned architecture and recurring business models. This is the extension of cloud and it is way larger than the first version has been.

21:57There's definitely a lot of debate about the net new revenues and profits and companies that are going to need to be created to make all of this investment successful from a returns perspective. But maybe uniquely you can answer this question. Can you share what you think the most successful venture-backed companies, the companies that are startups now, but will be household names in 20 years, what are they doing today? Where are those companies? And I know it's a hard question, but if anybody can answer it, you can. Yeah. I do think that this next phase is going to be the app layer, the consumer side of this, and rethinking what it means to have a successful product in the sense that we used to measure things like minutes in application and engagement and data throughput, where in many ways, the next measure will be how much app time did you save the user and what happened behind the scenes.

22:51And I think great institutions like Google search business are now completely under attack and available to be redistributed. And when we type in a search query and we want an answer, we got a page of links and you had to kind of find it versus this idea of answer engines and this idea of true business productivity in the sense of supercharge me so that things can disappear from what I have to engage with. And so I think it's all around this next extension of productivity, engagement, connectivity in a re-imagined application sense. And productization. Yes. Yeah. I think it's something that OpenAI has just been remarkably good at, right?

23:31Oh, that chat GPT moment that we all lived through. And now you look at what perplexity is doing and how you can see the chain of reasoning unfold and how you can introspect these answers and how you can interact with it to do thoughtful follow-ups and the deep research capabilities and things like cloud code and the ability to spin up applications. that business users can now create product. And this is, again, going back to the idea of 10-person teams are going to create massive businesses because it really is at your fingertips now, the ability to stand things up and the ability to iterate and experiment.

24:07And prototypes are going to become the new PowerPoint, where when you're talking about, oh, I'm thinking about new features, you're going to be able to spin those up pretty quickly and interact with it. And it'll completely change how people think of computer science and how people think about architecture and how they think about product. And you're seeing it already, right? Like your recent cloud 100 benchmarks and just the revenue per employee that you're seeing at these companies is orders of magnitude higher. Yeah. It's steeper and deeper. The slopes of these curves. I'm a skier, so I love powder.