EPISODE 1707

[INTRODUCTION]

[0:00:01] ANNOUNCER: GV, or Google Ventures, is a venture capital investment arm of Alphabet. Erik Nordlander is a General Partner at GV and invests across enterprise software and frontier technology, focusing on developer tools, cloud infrastructure, and machine learning. He's backed companies like Cockroach, Bowery, and Neo4j. Prior to GV, Erik was at Google and led development of the company's next-gen display and serving system, and built statistical and machine learning models for Google's ad businesses. Erik joins the podcast to talk about his work.

Gregor Vand is a security-focused technologist and is the Founder and CTO of Mail Pass. Previously, Gregor was a CTO across cybersecurity, cyber insurance, and general software engineering companies. He has been based in Asia Pacific for almost a decade and can be found via his profile at vand.hk.

[INTERVIEW]

[0:01:04] GV: Hi, Erik. Welcome to Software Engineering Daily.

[0:01:07] EN: Hi, Gregor. Thanks for having me today.

[0:01:09] GV: Erik, it is great to have you here. You are at Google Ventures, so this is a pretty exciting one for us. We don't often get to have venture capital people on the podcast and get to hear from that side of the industry. Thanks so much for making the time today. We'd love to, I think, just to kick off with what was actually your path? Google Ventures is very well-known, I think, probably to a lot of our listeners, but I'm sure there's a lot that happened before that. What was your path to becoming a VC? I don't think a lot of people maybe think actually VCs had a life before VC. Yeah, I'd love to hear your story.

[0:01:50] EN: Yeah, absolutely. I think very few people grow up thinking, "Oh, I'm going to be a VC one day." That was definitely not my path. I was always more technical. I loved math. I taught myself to program when I was relatively young. Then I ended up going to school to study CS. I did an undergrad in CS and was really excited about becoming a software engineer. I ended up sticking around for grad school. Out of grad school, I heard the call of the startup. In this case, the startup was Google. People forget that Google was a startup once upon a time. I'm showing my age.

I left Boston and moved out to the West Coast to join late in Google's evolution as a startup, but worked on what was there at the time, which was really deeper infrastructure projects. That was first in search and Gmail and then lower-level stuff that touched a bunch of different systems and products.

After a little while in California, I had the opportunity to move to Google's New York office, where I still am today, I'm still based in New York, and just continue expanding my career. This was a really exciting time for Google, the early 2000s. The company was growing really quickly, so I was leading larger teams of engineers and having a ton of fun building and shipping. Google was also growing by acquisitions. At that point, they were buying lots of companies, some of which are still core parts of the Google experience today, things like YouTube and Android and others. Then they were also just buying smaller companies just to ingest talent and technology as quickly as possible.

I was in this interesting position where there was a corp dev team that obviously, was going through all these transactions, but I was also helping look at companies in terms of tech and talent while still working as an engineer. That was in some ways, a VC 101, because I was meeting all these startups and seeing chapters 10, 11, 12 in the epilogue. I was really excited about working with entrepreneurs, and really pretty curious about what did chapters one, two and three look like, and even before that.

As luck would have it, there was a group of people thinking about, what does a venture fund inside of Google look like? Starting to put that together and super, super early in the fund's life, I just got connected with some of the early partners and started spending time there. Then at the end of 2010, I made the leap. I would say, we had a few people on that early team that had worked in some investing capacity before, but many of us were coming in as operators, builders, founders, just different sorts of experience, whether it's building stuff inside or outside of Google. And so, we really had to figure out how to be a VC and how to be a VC with Google as our backer. It was an exciting time. What I would say is after making that leap, I've still remained an engineer. I'm still a working engineer today. That's probably what brings me to this podcast.

[0:04:48] GV: Yeah. That's super interesting. I think, maybe a lot of people don't think about VC being able to work hand in hand with a company that actually still builds software. VC maybe sometimes feels, or looks a bit disconnected in that way. I think Google Ventures specifically is a super interesting one, where, as you say, you are still very much a working engineer, but also part of Google Ventures. If I was to say, there's probably some myths and stories about VC culture, or just VCs in general, what's maybe something you'd want to debunk from the outset if someone's listening to this and thinking, "Oh, I think I know what VCs about"? Is there anything you could say that maybe changes someone's mind?

[0:05:32] EN: Yeah. I mean, what I would say about VCs is just, it's one of the few jobs where there's not a single path to getting into it, or being successful inside of it. I mean, if you want to be a doctor, there's one way that you can become a doctor. Maybe that's a good thing. We don't want people figuring that out on their own. Even if I look at the GV team, or other really successful VCs, I mean, there are people who have been founders, journalists, software engineers, you name it. People who have been investing their whole career, right? That's all they've ever done. There's no one background, or one set of experiences that predicts whether you're going to be great at it, or terrible, or somewhere in between.

I think that maybe if you're looking at the industry from the outside, it can feel very homogenous. I would say, it's a really interesting group of individuals. Especially for founders, as you're thinking about, "Hey, who do I want to partner with?" It's worth learning a bit more about how people got to where they were. That may be a really good signal in terms of whether that's someone that is going to work well with you, or work against you and what that means.

I think, especially as the industry has expanded, we're seeing even more backgrounds. I think that's great, because those types of backgrounds that were great at discovering companies today may or may not be the right people to back the best and the brightest of the next class of exciting startups.

[0:06:54] GV: Yeah. I think you hit the nail on the head there. I, not of any relation to the date we were recording, I happened to have finished the book, The Power Law, which goes into the whole history of the VC industry. Highly recommend it to anybody who just wants to maybe learn about the history of, I would say, the valley, or just how a lot of ventures have taken off. I think one of the big things that came out of reading that was just every VC outfit has a very different personality, very different way of looking at things, and that's fantastic. If you're a software engineer, or a company looking for any venture backing, as you said, Erik, go and meet people and get to understand them. It's very much a two-way street, which is maybe not reported as much.

Yeah. Given this is Software Engineering Daily, most of our listeners are software developers. I think a lot of listeners out there, they're either in the thick of creating their startup, or maybe just thinking about making the leap to doing their own thing. You've seen it from both sides as being an engineer, and then very much also, looking from the other side and saying, "What are we going to back as Google Ventures as well?" What would you say are maybe some of the most common mistakes, or pitfalls you maybe see the very early-stage software startups make? I think, software is maybe the key here where we could talk about all ventures, but maybe let's talk about software.

[0:08:22] EN: Yeah, it's a really good question. By the way, Power of Law is a wonderful book. Two thumbs up on that recommendation. A lot of times I'm working with founders, they're often very technical. There's maybe multiple technical founders. Sometimes I think people believe that if they just build the right software and put it out into the world, people will come, right? This is independent of whether it's for consumers, or enterprises, or something in between. The mistake that I see a lot of these companies make very early on is not thinking at all about any of the marketing, or sales aspect, or how that's going to fit in.

Certainly, as you're designing, you're writing a design doc, or you're putting pen to paper on the early builds, that's not going to be top of mind. It's a very crowded space for software, right? There's all kinds of opportunities and there's all kinds of competition. Understanding from a fairly early stage, how you're going to promote it, how you're going to distribute it, how you're going to get people excited about it, it matters, because it's going to influence your design, it's going to influence the team that you build around it. All of these things can obviously be added on later on. Many times, that's the journey that we go on is that we build the product and then we, along the way, figure out what's the best way to distribute it.

The earlier that you can think about that and you can get feedback in terms of who the right customer is, or who the right set of initial users are going to be, it's just going to make your life that much easier. Because then you can start building that flywheel of feedback early in the company's life. That's, I feel like for most software companies, at least in that seed stage, that is really the most valuable thing is making sure that you're building the right product, you have the right technology, you're solving the right problems. If you do that too much in isolation, one of the pitfalls is going off the wrong path and losing track of where you need to be.

[0:10:13] GV: Yeah, very much. Can you just confirm again in terms of Google Ventures, what kind of stages does Google Ventures tend to come in at, and what opportunities do you tend to evaluate? Like, what stage?

[0:10:26] EN: Yeah. GV is definitely a multi-stage firm. We can write a seed check, we can lead a traditional venture round, like a series A and B, and we can even write a growth check. I would say, the bread and butter is seed to be. These are checks that are start at a few million dollars, up to maybe 25 or 30. We have a lot of capital to deploy, so we'll often double down on portfolio companies. I mean, I have a company invested in almost 10 years ago and we've been in every single round since their first financing. Sometimes we'll miss something. We'll miss the A or B, and we'll just jump in a little bit later. There's a lot of flexibility.

I mean, I think that's one of the great things about having this single LP being Alphabet, is they give us a lot of trust and flexibility in terms of how we get the capital out there. I have to say for myself personally, I mean, my joke is I'll go as early as they'll let me. That can often be at inception, but typically a bigger seed round where people are raising several million dollars. If it's under a million, the math doesn't make sense for our fund given its size these days.

[0:11:29] GV: Yeah. I think one of the, especially again, to the listener base, we have probably one of the, I would say, most underreported parts of maybe this evaluation is actually the technical due diligence, or DD as might be said. How do you evaluate the strength of an opportunity? Or maybe, let's just back up for a second. At what point would you actually get to the point and saying, "Okay. Actually, we're going to do some technical DD on this company"? Then maybe just talk through what might be some differences between say, stage, if you're looking at a seed versus say, a series A, are there differences in terms of the level of detail that you'd go into in those? Then, what would you actually be doing?

[0:12:11] EN: It's a good point. Because it does change as things move along. Normally, we're going to start our engagement with the company, probably like everyone else. We're going to hear the pitch. We're going to go through the materials. We'll ask a bunch of questions. Then step two will be us double clicking on all the different areas. Obviously, tech is definitely going to be one of them. I feel super comfortable personally just sitting down with the CTO, or technical co-founder, whoever it might be and just getting into it.

I mean, part of it is just evaluating that individual, and their ability to grow and lead a team and how would they think about the technology and the roadmap. One of the nice things about being at those early days of Google is I was a person that had to do lots of interviews. I mean, I've probably interviewed hundreds, if not thousands of software engineers at this point. Certainly, I don't make anybody code on a whiteboard. That's not part of it. But just understanding how they think about building technology. It's a really important part of the process.

What I would say is early on in my time at GV, I maybe was a little bit too concerned in terms of what the stack was and how the individual technology choices would impact what they were able to do. You quickly learned that a startup is about solving problems. Sometimes the tools are just in means to an end and things can change very quickly. I'll just give you one fun example here in New York. We backed a company in the e-commerce space. I went to start the tech DD. This thing was a complete .NET shop, right? They were using all Microsoft all day long. I was like, "Hey, you're a startup. Do you really want to be paying all this money to Microsoft? This seems really - here's all this stuff going on in open source."

He's like, "Oh, yeah. No, no, no. I get that. I'm aware of all that stuff. We're just, we're going with Microsoft." I was a little surprised, because I'll be honest, it's not something you see a lot in early-stage company. Very quickly, I learned it was actually a recruiting act. He was here in New York, pulling people out of the financial service industry, who, there are a ton of people who'd just grown up with Microsoft. That's all they'd done. They would love to go to a startup, but nobody wanted their skill set. He had full reign over pulling and all these people. He'd go to all this .NET meetups, would be the only startup there, and everyone was like, "Please, give me a job. Get me out of this corporate hierarchy," or whatever they were in.

Yeah, sometimes the actual stack can be valuable in ways that have nothing to do with product building or anything else. It was just a great way for him to recruit. I'm very open minded about the individual technology choices. It just has to make sense. I think that as long as you have a founder, or CTO, or whoever that can communicate it in a great way and we don't think that they're putting on crazy amounts of tech debt, or doing stuff that's going to bite them down the road, that's generally how we get to a point of it being excited about the tech. Obviously, the vision there matters a lot in terms of us being aligned that this is technology that it's going to be exciting in several years.

Because when we're investing in these companies, especially at seed or A, that could be a 10-year relationship. We're totally fine being patient investors, continuing to invest into technology as long as the promise is there. Often, if you're building something really complicated, it can take several years before you even get a dollar of revenue, or have lots of users, or anything else. Yeah, we need to be pretty confident that the technology and product is going to be valuable once you finish the build.

[0:15:34] GV: That was a great anecdote about why .NET for that specific company. Certainly, has been a past life as a CTO, a larger outfit than I am now, and having to defend why we were picking certain stacks and actually saying, this is a hiring thing. Yeah, we're not going to be able to find the engineers, or we're going to have to pay outsized amounts to find engineers who know the latest and greatest. When actually, if we pick something just a bit safer, so to speak. Yeah, I think it's a great call out.

I mean, just a little sidebar. I mean, I'm curious what you've seen. Have you seen there being quite a lot more startups coming with these, I guess, slightly more evangelized languages now go and rust basically come to mind where everyone's saying, "Oh, this is where we should be going, especially if we're we've got a green field right here. We can just start from anything." Is that something you're seeing? Do you think there's any merit to running with them just as languages, because you can start from scratch? Or, yeah. I'm just curious what your thoughts are on that.

[0:16:35] EN: It's funny. I have two of my portfolio companies are pure go shops and one is a pure rush shop. Cockroach Labs, which is the company behind Cockroach DB, they've been go from the start and go is a great choice for them, right? They are building a really complex distributed system. That's the whole point of go. It's the perfect fit for that.

Yeah. They'll do things around the edges in different languages. Again, in terms of the recruiting thing, which is really at the earliest stages, that's probably the hardest part for technical teams is getting the right talent in there. Go is great for distributed systems people. I think there are a lot of folks coming in from C++ and other areas that were delighted to switch to go. That was a real asset to be someone new. By the way, they started with this back in 2015. It's been a while. In that point, Go is really the new shiny object.

I'm also involved with Warp, which is this modern command line terminal. They tried out a bunch of different technologies. I mean, they actually built prototypes in the earliest days on different stacks. For performance reasons, ended up in Rust. Again, Rust was definitely new. I would almost say, it was maybe too new right when they get started, but they've been a huge beneficiary of the ecosystem exploding. Again, maybe it's the opposite of my .NET story, but in this case, folks are really excited about the language. They're excited about learning something new. I don't think most of the people coming in there have a lot of Rust experience, but it's actually pretty easy to pick up. If you're an application developer, there's so many just nice properties about it that I think if you're pulling someone out of, I don't know, a Java place, or something else and you say, "Hey, by the way, you get to write 100% Rust," that's actually a really good carrot to bring in smart, young engineers.

[0:18:19] GV: Yeah, that's also a great point. Yeah. exactly. They just, there really are so many factors, do language choice, and yeah, what might attract engineers to your startup, for one language might be a completely different reason for another. Yeah, it's a great thing to consider. Going back to some other things that you'd be doing diligence on, or asking questions about, what kind of - I think, again, maybe as a myth, or just something that's never clear to people, when they're pitching, or just generally talking to VCs is which metrics or milestones. Everyone says, "Oh, you have to have revenue." Then the other one say, "No, it's ridiculous. But you don't have to have revenue." Or if you do, how much revenue?

I think so many numbers and things get bandied around. Like, if you go in Reddit, you can find all sorts of nonsense, basically. Great to hear from you. What do you look at in terms of metrics, or milestones? Yeah, how do you look at that?

[0:19:15] EN: Yeah, great question. I think the revenue one is always interesting, right? Because if you have lots of revenue, people can invest in your company without knowing anything about your business, right? They can plug whatever you're doing into some spreadsheet and pop out a number and write a check and maybe understand nothing about what you're doing. As companies get much later in their life, that's how it works, right? I mean, you think about the public markets and things like that. You look at what they do in the quarter and then the stock goes up or down. I mean, I think to me, the whole point of venture is investing before these things are super obvious. When we're getting involved, again, we are thinking about what happens if this goes right? What is this going to look like five years from now, 10 years from now, even further into the future?

Obviously, we want to invest in something where we think the opportunity is big enough, right? I mean, that's a big part of venture. It's like, the prize has to be worth it. If we're going to work on this thing for 10 years and put in lots of capital, yeah, it's got to be really big at the end of it. That doesn't mean that in year one or year two, you have to have X amount of revenue, otherwise it's not going to work. In fact, I think many of our best bets, it took many years of zero, or it's called marginal revenue before things really pick off. That's because you're building this technology moat.

I'll go back to Cockroach. I mean, I think they did several rounds of venture funding before they had any serious customers. That's because building a big global distributed database that people are going to trust with all their data, that doesn't happen overnight. It takes many years. Of course, look, along the way, you're going to hit all these milestones that are measured both internally and externally in terms of the technology, in terms of its reliability and performance and all these other things. All of that matters, right? I mean, again, once you have revenue, maybe some of those things are slightly less important, but early on, that's what we pay attention to.

I think, because many of us at GV can speak that language, we can go in and build a lot of conviction maybe before others, or before people who would just prefer to see the revenue. I feel like, hey, revenue is great. Obviously, that means that you're on to something and customers are excited and it's not something to be ignored. The same time, if you build something and you turn on revenue two days later, what does that mean? It probably means someone else can do that and you're going to have a lot of competition, too. Striking that balance is really important in terms of building both a big business and one that's going to remain defensible for a while.

[0:21:38] GV: Yeah, that's a great point in terms of just what you said at the end there, that there is very much a potential trade-off between getting revenue in early, but exactly, your moat, if you want to call it, that is pretty small, if that's the case. Whereas, if actually you're building up a product and in theory, building up that moat on the product side effectively and then able to turn on revenue. Slight sidebar again, Cockroach Labs, we've had a couple of episodes with them on. Curious what attracted GV to Cockroach Labs?

[0:22:08] EN: We knew the founding team really well. Spencer, Peter and Ben were all early Google employees. They'd done lots of great work there. I actually personally worked closely with Spencer right when I started. On top of that, they were building something that was a little bit inspired by a Google technology, which was Spanner. Spanner was an internal database that was distributed, transactional, all these other great things.

Unfortunately for Google, fortunately for Cockroach, it was only running on the Google Cloud, right? One of the big things that Cockroach was going after was companies that wanted to be multi-cloud, they wanted to build on top of open source. These guys were famous for creating GIMP and GDK. They'd been open source from the beginning. We were really excited about that. It was very clear that that type of database was going to have a huge market opportunity, but also, it was something that was our earlier discussion, it was going to take a while to build, right? This is not something that's done overnight.

I mean, it took Google, obviously, minimum engineers over a while to get Spanner. We knew that doing the outside would probably be even more difficult, right? Because they didn't have the other supporting infrastructure. Yeah, we were very excited to back that team. I remember reading the first white paper for it in early 2014, so 10 years ago now and being very excited about it, but also, understanding that this was a big technical build. Those are great projects for GV. We see an enormous opportunity, but it requires building lots of complex, meaty technology in order to get there.

[0:23:40] GV: Yeah. That maybe segues into given, GV seems to backing these companies that, as you say, are potentially quite meaty builds, how hands-on would you say GV is with companies, whether it's technical, specifically, or product strategy, or do you ever bring in anyone in-house? I mean, yourself, you're a practicing engineering. Is that you engineering effectively on GV portfolio companies? Or how does that look? Cause again, I'm aware that VCs can have very different approaches here in terms of the hands-on this and why they have that approach.

[0:24:18] EN: Totally. We're very lucky at GV that we have a whole team that we call the ops team that is really world-class people that we've pulled in from startups, Google, other places who work hand-in-hand with our startups. They're working across a bunch of different areas. That could be finding great talent, thinking about press and marketing, thinking about UX and design. We have some amazing designers and UX researchers on the team. Really, it's about different inflection points in the business. I mean, again, these are almost consultants that you would tap into when you're hitting, "Hey, I really need to get into the press, or I need to get out of the press. I need to launch this new product. Can you help us test it out and see if this is the right moment, or right strategy?"

Then beyond that, there's obviously lots of people at Google with they don't work directly for GV, but they can be really helpful to us in terms of just checking in on different areas. Yeah, there's a ton of resources we can bring to bear. I mean, our goal is certainly not to build your company for you. I mean, we're backing founders for a reason and we know that they have incredible talents and we want to really just supercharge that at the right moment.

For me, again, you probably don't want me at this point writing any code for your company, but I'm always happy to be a sounding board and resource as you think about, again, different technology, different strategies, things like that and other people that I've worked with other CTOs. Just plugging into that community can also be very, very valuable from our end.

[0:25:41] GV: To me, that just sounds hugely, hugely valuable to have, especially a technical sounding board, someone effectively that you can trust and just lay out all the code and be like, "What do you think? Am I going in the right direction here?" I think I think a lot of engineers struggle to probably have someone that they can call upon to do that with. Yeah, that sounds super valuable.

[0:26:00] EN: I mean, again, it can be super lonely as a founder in the earlier stages. Again, especially if you're coming from a bigger place where you're working with many other engineers and you just had that support system. Yeah, bridging that gap can be really valuable in the earlier stages.

[0:26:14] GV: Yeah. You touched upon, obviously, founders. You need founders for companies. I think a lot of people have YC, for example, is quite well-known in terms of some of the info they put out in terms of what we tend to back X, Y, Z kinds of companies. I think that filters through to most - a lot of people's thinking is this is just how it is. One of those is they talk pretty strongly about co-founder companies. It's not that they don't back solo founder companies, but co-founder companies, they laid out all their data to say why, etc.

Google Ventures, GV is coming in sometimes, well, it sounds like, slightly later stage. How do you, I guess, look at that? Does it matter at all at that stage, whether it's being co-founded, or solo founded, or how do you look at that?

[0:27:00] EN: Yeah, it's a great point. I mean, I think we try not to have too many hard and fast rules about what we'll invest in other than things that are ethically not aligned with us. I've backed many companies with multiple co-founders and I've backed many solo founders as well. I think, again, you're really just evaluating what that person, or group of people can do. It's clearly a lot more work as a solo founder, but often then they're making hires to fill the gaps early on. I mean, I've backed solo technical founders and then they're hiring people who help with product and UX, maybe much earlier than they would if they had someone else on the co-founding team, who was really great at that.

I think as long as the founding team has the ability to recruit, to fill in the gaps in terms of their skill set, I'm totally fine with it. I think there's maybe just a little bit of safety with investing in more people. Like, "Oh. Well, if so-and-so leaves, then we still have a company and things like that," which I get into earliest stages. That's probably more of a risk once things are series A, series B, there's just more company there. Really to me, no number of co-founders other than, I don't know, we have 10 co-founders or something crazy is a flag, as far as I'm concerned.

[0:28:21] GV: Yeah. Great point on just the risk piece there is maybe what some founders aren't thinking about in terms of, well, if you look at it from the other side, what it was, it's not that someone's saying, we don't think that you specifically are not capable of building this up. At the same time, we've got to look at it. We are putting money into you and the company and what was the risk profile on that, if that's a, just a you company and not maybe to all of you. Yeah.

[0:28:49] EN: Yeah. I think, again, if you're a solo founder, it's just important to build that support network. Again, talking about it's a lonely early on, make sure - I mean, something I've done is help connect founders, these other pods of founders, they can chat with each other, get feedback, talk about different strategies and things like that. Which again, if you have a co-founder, those are all the conversations you're going to be having internally. A lot of that, again, if you just have a good support network through investors, other founders, just other people in the industry, you can get a lot of that. It's more work, obviously, but then you're the sole founder, you have more ownership of the company. It all balances out.

[0:29:27] GV: Yeah, absolutely. Getting to the end of the looking at how, I say, GV would evaluate companies the, I guess, sometimes the beginning and the end, its devaluation, which is always the bit that - the real numbers. What's your opinion at the moment? Or like, how valuations are looking? We definitely went through a very frothy period two, three years ago. Then, there's quite a lot of noise, or depending, I would say, geographically, it's quite dependent on people saying, "Oh there's no funding anywhere, etc." I sit in Singapore and actually, I don't see that, if I'm being honest. I see there was quite a lot of opportunity, certainly in this part of the world, at least at the moment.

You're sitting in New York and I'm not sure geographically if you have any leaning in terms of GV and where you tend to invest. Yeah, how are you seeing valuations and just the environment at the moment?

[0:30:23] EN: Yeah. I mean, it's really easy to talk about valuations in the past. I mean, as you pointed out, there was this COVID bubble, where really all asset prices shot up, just because of all the money that was flowing into the system and startups weren't immune to that. I think a lot of these things matter a lot in the short term, right? If you were someone that was about to sell your company, let's say, 2021 would have been an amazing time to do it, because the prices were super high. Almost anything that you'd been doing in the past may or may not have mattered, because the market was approaching a peak and you could have had a great exit, regardless, right? There are lots of specs and all these other things that happened.

I think when we're investing, again, you're coming in at the earliest stages, seed series A, you're planning for a journey that could be five, 10 years, maybe more. Whatever those conditions are in that moment are probably going to look a lot less relevant when you're looking backwards. We may pay a little bit more, pay a little bit less, of course, right? There's always going to be those perturbations. Really to me, what's always been true, if you look at venture going back to the earliest days, if you're in the best companies, the top 10%, the top 25% of companies, there's always going to be great returns for them, regardless of whether things are at the peak, or you're in a recession, or whatever it might be. That's really how we think about it.

We're thinking like, is this going to be transformative in the future? If so, we can't really control what the macro situation is going to be then when that company is having an IPO, or whatever that the outcome is going to be. We have to pay attention to it. I mean, one of the other things I do at GV is I run something we call the macro group, where we actually have a bunch of speakers that come in and talk to us about different things that are going on in macro. I think what I've realized over the last five or six years is macro is starting to affect venture more and more. We could care less what's happening quarter to quarter in terms of most of our companies. Really, it's about, can we be in the best companies and what's their full potential? Then we'll let the macro fall where it may.

[0:32:37] GV: Yeah. I think that's a great point there. Yeah, just that people might get a bit lost, thinking that the macro does affect the venture side, maybe a bit more than is true.

[0:32:49] EN: It changes the amount of venture, right? Again, if people are raising lots of money for venture funds, there's just going to be more money out there. Probably means more startups gets funded and leaner times, there'll be less of it. Again, I think my point is that if you're building one of the best companies, there's always going to be venture money for it. We've never been in a point where venture goes to zero.

Again, we're very lucky to have Alphabet as our partner. They have lots of capital and they're super, super patient. That's the approach we take. Again, I think that that means that when things are hot in the market, we're not doubling and tripling our fund size. When things are a bit leaner, we're also not cutting it in half. We have more of this steady approach, which I think if you're doing long-term investing, that's what you have to do.

I wanted to touch on the piece about geography. I mean, I do think geography does affect it a bit, although, obviously post pandemic entrepreneurship is everywhere. You can start companies in almost any place on the planet and it's going to work. New York is where I sit. I feel really great about the ecosystem here, especially post-pandemic. I mean, it's becoming a great place for young, smart people to move to, founders, engineers, things like that. I mean, there's a great ecosystem of both universities, big tech companies, startups. We feel pretty bullish about what's possible in New York, just because of all of the work that's been put into the ecosystem over the last 10-plus years.

[0:34:13] GV: I was back in New York - I say, back. I lived there a while ago and I've tended to go most years for various reasons. Yeah, I was back there in January and I couldn't agree more. Just the way things look in the heck ecosystem there at the moment, pretty awesome. A lot of my friends - I lived in Hong Kong for a long time and actually a lot of friends have moved from there to New York. A lot of people are congregating, I would say, in New York now, which is super interesting to see and just a bunch of really interesting companies in New York. Yeah.

[0:34:46] EN: That's cool to hear. I mean, one of the things New York has that I think San Francisco has in spades and maybe some of the other cities are not quite there yet is just this convergence of talent and ability to switch jobs. The joke in Silicon Valley is always like, hey, you can change where you work without changing where you park. New York has big engineering offices from Google, Microsoft, Meta, Amazon. I mean, the list goes on. They're all here, sometimes hiring thousands, if not more of engineers.

If you leave one of those places, or you come to New York and you start a company and it doesn't work out, you might be able to just go and work at one of those areas and you not have to uproot your family, or move across country, or things like that. I think that safety net, that was really important for Silicon Valley. I think New York is one of the few places that really has that, too.

Some other cities, if you went there and you started to start up and the startup had to be there and it didn't work out, okay, you might then have to make a shift. Also, there's tons of great universities. There's this influx of talent. As you said, people are excited to move there from other big cities in different places. I think that that is always the cornerstone of a good tech ecosystem. I feel really positive about where things are today and where things are headed for the city.

[0:36:10] GV: I think that the real human network should never - can never be understated. There was probably a phase where I would, especially during pandemic, people thought, "Oh, it doesn't actually matter. We can be wherever and we're still connected, etc." Yeah, I've always been someone that has deliberately lived in bigger cities for that reason. I think it pays off so much. Yes, they're are expensive places to live, especially if you're maybe a younger engineer trying to get your startup off the ground. The value you get back in who you meet and like you've just said, Erik, being able to almost have a fall back. If it actually doesn't work out, you probably have already met someone that's ready to employ you, more than just if you're sitting off in a different country somewhere trying to ping with your Ellington somewhere. That's a great advice.

Looking at how Google Ventures - well, I want to just touch on this. Is Google Ventures, even though mothership, I guess is now called Alphabet. I'm curious, was there any chance it was going to become Alphabet Ventures? Or is Google a specific name now in terms of Google Ventures?

[0:37:16] EN: I think we go by two names now. Both GV and Google Ventures. Look, we restarted in 2009. We technically predate the Alphabet, even though we are always set up independently. Our structure is one where we're measured on financial return, ultimately. I mean, we want to have great impact, but it's about generating real returns. That keeps us really aligned with the companies we're backing. We can back people that are good for Google, bad for Google, something in between. I mean, almost half of our dollars go to life science companies. Very little overlap with what's going on at the core Google business. We've backed competitors to Google products, things like Slack.

I've had big exits to Facebook. I guess, now Meta. That's great. Their money is green. We're happy to take it. In terms of how we work, I mean, we work like any other VC, which means we're paid carry, or a portion of the profits when our funds return. We don't get any Google stock. I don't have any Google stock from GV, as nice as that might have been over the last couple of years. It keeps us really, really aligned with the founders we're backing. I think that's the most important thing. I'd feel awful if I go and write somebody a check and then, "Hey, you have to use these Google products, or else," right? No. You want to use AWS. You want to use whatever. That's totally fine. Obviously, if they do want to do things with Google, we're happy to make those connections, but that's not a - it's a non-factor in terms of how we think about who to back.

[0:38:45] GV: That's a, yeah, I think a great clarification point. Just, I think naturally, people see the name Google there and probably think there is more connection actually, than as you've just described. Again, it makes complete sense. As a fund, you want to have that independence to - you have to be able to make those decisions. If we look at which areas you - you just said, I think, life science is actually is about 50%. Any other areas that are outsized as well?

[0:39:14] EN: The area that I spend most of my time in, which is what we call enterprise. That's everything around infrastructure, developer tools, classic SaaS. That's probably the other really big one. We also do lots of consumer investing. We do frontier stuff, which involves hardware, climate, semiconductors, things like that. It is really broad and really diverse. I mean, at the core of it, unsurprisingly is different types of software.

[0:39:40] GV: Yeah. Let's talk just about any of the port codes. That's the shorting portfolio companies. What would you say? We've touched on a couple, like you mentioned, Warp. We've talked a bit about Cockroach Labs. Yeah, I'd love to hear about what are some others that come to mind in terms of just some really impressive transformative products and companies that GV's invested in, especially with yourself being involved specifically? Yeah, I'd love to hear about those.

[0:40:10] EN: Sure. I mean, I'd love to just touch on Warp, because you only mentioned for a second. I mean, it's a really fun one. Most of us, whenever we got started with programming, we had to use a CLI, right? I mean, I'm old enough that my first exposure to computing was around using a CLI, whether I wanted to play a game, or do something productive, or anything else. Now it's been relegated to the world of the programmer. Honestly, it's not something that's changed in 40 years, maybe more. It's the same thing.

I mean, I remember getting my first Mac laptop. I think terminal.app is pretty much exactly the same as it is on a 2023 MacBook Pro. Zach came to me pretty early on and was like, "Hey, what if we completely rebuilt this from the ground up?" At first, I wasn't - I didn't get it. I was like, "Oh, that's just not that interesting." then I started realizing, this is probably the most widely used developer tool. It hasn't been improved. It has no UX improvements. There's no AI. All these other things that are showing up in every other aspect of the development workflow, we can have this least common denominator in terms of the terminal and the CLI.

They have completely rebuilt it. I mean, if you're a programmer out there, you have to download it and give it a shot. It's on Mac and Linux right now. They're expanding to different platforms throughout this year. It's really a total rebuild. It's just a joy in terms of what they can do. If you've ever been frustrated in terms of knowing, "I need to expand this and tar dash, whatever," if you haven't committed all that to memory, this is the perfect tool for you, because there's so many more ways that even - it's almost called relatively simple technology in terms of sharing commands and using AI to infer what you do. It just makes you so much more productive.

I love that, because this is a - it's a company built by developers, for developers. Hopefully, making a lot of people's just regular day-to-day experience, that much more productive by improving this tool that we all have to use, but it doesn't have to be so painful.

[0:42:10] GV: Yeah. I want to try and get them on something.

[0:42:13] EN: Absolutely. Zach, who's the founder, is incredible. I'll put in a plug for making him a future guest, for sure.

[0:42:20] GV: Yeah, fantastic. I need to download it. Admittedly, I caught on to it fairly recently, but yeah, I'm going to download it. Because as you see, I'm in terminal every single day. Yeah, it hasn't changed. I still have to use the same shortcuts that still aren't that great and etc., etc. I'm sure, yeah, Warp just sounds awesome. Any others that come to mind?

[0:42:39] EN: Yeah. Another one more recent addition is this company, FlutterFlow. FlutterFlow is an app creation platform. It's based on the Flutter language, which if you don't know it, it's really cool. It allows for cross platform development. If you want to build something that works on iOS, Android, mobile web, you can write the code once and deploy to all of those areas. The thing about Flutter that I think people forget is there are so many businesses, and even people that are just building an app themselves that have these parallel code bases between iOS and Android, and it's really a waste, right? You'd really love to just write something once and deploy to the widest set of users possible.

You also want to be able to use platform specific stuff. Again, that's pretty easy to do here. FlutterFlow is yet a layer on top of that, which makes designing and building these apps super, super easy for someone who's technical and wants to jump in and out of the code, or for someone who's just comfortable building UIs, but doesn't necessarily want to get into all the code. What's really happening with them is they are democratizing the ability to build software and also, allowing people who are really technical and people who are, we call it medium, or less technical to collaborate in the same app, right? I think that that's really a model for how software is going to continue being built into the future, where AI has a bigger influence.

There's just other people who understand how to build software, but maybe haven't been trained as a computer scientist. We need platforms that allows everyone to contribute. Because frankly, we just need to build a lot more software and we need to break down as many barriers. The mission of this company is really to go from ideation to ship production code as quickly as possible, focusing on mobile, obviously. I think they've done a great job in terms of what you can do, whether you're starting with, or just a simple UI design and then turning that into a working app.

I mean, what's really, really cool about this company is I was at a board meeting there and we have paying users in almost every country you can imagine. I mean, it's really, really cool for a young series A company. This thing has global impact. I think it's because the desire to build and ship code is universal. It's not like we've done a bunch of internationalization work. There's more to do there. They've really struck a chord with a bunch of people who are interested in building apps. I really love working with that team and seeing just where it goes, because I do think the future of building software is just about bringing more people into it.

Despite all the kind of, "Oh, AI is going to replace everyone," I still think we have way too few software engineers. Really, it's about making more people - giving more people the ability to create software.

[0:45:30] GV: Yeah. No, I couldn't agree more. Always hear effectively non-technical people saying, "Oh, AI is going to come and take your job." I'm like, "Ah, it's not exactly how I think it's going to work." It's just that AI and as you say, platforms that democratize the level at which someone can come in at to build things is always changing. I think every software developer needs to be always aware of that and just never sit still. One company that did that, I think really well, there's a ton of reasons you might not want to use this product, but I still think Webflow did a great job of actually changing that bar, where you suddenly didn't need a developer for a lot of things. You actually design it, and people that called themselves designers effectively became web designers, and I love that.

I used to employ someone who was, came in as a pure UX person. Then now he runs his own shop purely through Webflow. It sounds like, flutterFlow is the same approach in that way of changing that bar, which I think is such a huge thing to be able to do.

[0:46:29] EN: Totally. I think if you just look at how software engineers have moved up the stack, so to speak, over the last several years, I actually feel that this is in some ways, more the same. I mean, I remember when I was a young programmer and I was writing little games and I would have to go in it and mess with assembly, and learn about registers and all this other stuff. I haven't thought about that in years, other than someone who's doing very low level, or systems programs, very few of us do. That's a good thing, right? Think about how many parts of building software are unpleasant. Writing lots of unit tests, fixing API deprecations, or migrating off old software.

I mean, I think about how much of the global financial system is still being run on mainframes with COBOL and things like that. There's no software near engineer out there who thinks, transitioning that code or rewriting it is a fun thing to do. I'm really excited about AI getting rid of all of the not fun parts of software engineering, and letting us just focus on really, the joy of building software, which is the creation piece. I think that all these other LLM-based tools will just accelerate that. That's obviously a space that we're spending a lot of time looking at as well.

[0:47:45] GV: Yeah. I mean, I'd love to just touch on AI. We have to, effectively.

[0:47:48] EN: It's a rule now.

[0:47:50] GV: Exactly. However, not turning the whole episode into another AI episode. Yeah, we have to touch on it for a good reason. How are you and GV looking at AI from an, I guess, a opportunities and investment point of view? There's a lot of noise. There's an AI for something popping up every day. Like a Gemini 4X, or GPT 4X, etc., again, popping up virtually every day, how are you cutting through this noise and how are you evaluating AI opportunities just as GV?

[0:48:23] EN: Yeah, absolutely. I mean, I think the goodness is we've been looking at things that you could call AI for years. I mean, probably almost since the inception of GV, even before this new LLM-based craze has taken over the industry. I think that's been really helpful. We think about things in two buckets. Whether it's for AI, so this could be infrastructure. I'm involved in a company called Lightmatter. They do optical semiconductor stuff. Turns out, this is actually super helpful in terms of increasing bandwidth between chips, which is really useful in doing large training, or inference workloads.

Then sometimes we also have companies that are with AI. Using AI, this could be something like Typeface, or Synesthesia. They're obviously the LLMs and other models are being used to drive some consumer enterprise facing application. We're looking up and down the stack. I think we have, again, the experience of not being new to it, of having seen these things for a while and just understanding where the leverage points are.

Again, for me, as someone who's really excited about infrastructure and developers, there's a lot of exciting things going on in that space. We're spending time with those founders and those companies and figuring out where we want to place our bets. Again, I would say, even though the last two years or so have been really, really exciting, it's still to me, feels quite early. I think that it's going to take time for a lot of these applications to really percolate through in terms of different consumer experiences, different things on the enterprise. Certainly, then for developers has just been talking about how your day-to-day flow will change.

I mean, lots of people are using co-pilot and similar tools. As we know, just typing out the code is just one part of being a software engineer. There's so many other aspects to it in terms of production, design, testing, integration, all of these things will get touched by AI at some point, and we're really excited to sit down with the people that are reimagining all of that and hopefully, make more bets into young founders thinking about it.

[0:50:34] GV: Yeah, that's super exciting. I love that GV is really looking at a lot of these developer-focused areas. I think that's super exciting. Just to round out, again, we've got a lot of people listening in the midst of their startup right now, or thinking about making that leap. You have just a couple of anecdotes of maybe successes and even failures, maybe that have come up in the last couple of years, that might just have helped give some takeaways to some listeners?

[0:51:02] EN: Yeah. I mean, I think in terms of successes, I think it's never a straight line, even with some of the companies I'm involved with that are clearly leaders in their respective fields today. I mean, I just mentioned Lightmatter, that's in the semiconductor space. They're what I would call a six, or seven-year overnight success. They had to do a lot of really hard technical work. Then they've gotten to this place, where now the demand for AI workloads is obviously going parabolic and they have a solution that is really well suited to building out the data centers of the future that will be doing all of our inference workloads going forward.

Again, when the company was founded, I was lucky enough to be there as they launched in an entrepreneurship competition. None of that stuff was clear, right? It was very uncertain when and how the market would come to them. Sometimes if you have a vision of the future, you have to say super, super focused on that and control what you can control and not worry so much about - obviously not ignore, but sometimes the market will just take a little bit longer to come to you. If you feel like you have something that's really going to serve a wide set of users, or people, it's okay to be heads down until that comes to you.

In terms of things that don't work, I'll go back to it. One of the first things we talked about here, which was not paying attention to feedback early on, if you're building really interesting technology, but you're doing it without talking to anyone who might use it, that's generally how you go down the wrong path, especially early on. I always - if you're building something for other developers, talk to as many of them as you can early on. Go talk to different sized companies, startups, big companies, medium sized companies, different verticals, that feedback is really the lifeblood of an early-stage startup. Again, I think for more technical people, it's very easy to put on the blinders and ignore it in the earliest days. I feel like, the sooner you can get to that feedback fly wheel, the better things will go.

[0:53:12] GV: Yeah, I think both of those are great advice. Yeah, I can turn the first one into a soundbite, I guess, of stay on the bus. I don't know if you've heard of the Helsinki bus station story. Listeners, Google it and you'll get the gist. The idea is just stay on the bus. If you have a vision for something, all startups start from the same place and it'll look quite samey to - you might be comparing yourself to some other outfits and say, "Oh, well. They're there today and I'm not." But that's not the point. It depends on your journey and actually, yeah, stay on the bus and feedback. Yeah, couldn't agree more there as well.

As a technical founder, yeah, it is difficult. It is difficult sometimes to get your head in that space of one minute you're coding, the next minute, you have to line up calls, or meetings. Yeah, you don't always want to actually hear what people are saying in the sense of that thing, but I couldn't agree more. You just be humble and be able to take people's opinions and then apply that back to what you're building and that's great advice.

[0:54:10] EN: It's not a natural thing, right? If you're just coming from a pure software background, you might be doing design reviews and things like that. Often, you're not the one catalyzing it. I think that's what happens when you shift to being a founder. You have to catalyze all these things and it may not be the most natural, or comfortable thing, but the more you do it, the better you're going to get at building tech and products.

[0:54:30] GV: Absolutely. Well, we are out of time today, unfortunately, but I really appreciate that you're coming on today, Erik, and giving lots of little nuggets of wisdom there in terms of building companies and just some really fun stories there from what GV has been doing and investing in. Erik, where can listeners find you?

[0:54:53] EN: Yeah, so I'm on X, formerly Twitter, as @ejn. Find me there and I'm usually sharing stuff about technology in our portfolio companies.

[0:55:02] GV: Awesome. Well, again, thank you so much for coming on and hope we get a chance to catch up again in the future.

[0:55:09] EN: All right. Thanks for having me, Gregor. This is fun.

[END]