Guillermo Rauch at Founders in Arms Live: Simplicity, Focus, and the Bet That Built Vercel

Being a founder can be so lonely because it's not like you have a whole team of coaches that are helping you not burn out. So a lot of it is what I think Ben Horowitz says, like managing your own psychology. Because in reality, whether you burn out or not, you could totally go off the rails and just burn out, but also you can be your own coach in the sense of like, all right, it's time for me to unplug. It's time for me to go for a run. It's time for me to invest more in my social life. And I know that it was difficult for me to figure all that out because I was by myself coming from another country. So I appreciate the concern over those kind of things, but also part of what one needs to master.

I met a lot of customers today. So thank you for putting this group together.

I hope so. We should definitely do.

Yeah, the way we talk about it is that we have a web stack, which is all of the tools and infrastructure that you need to build for the web, and then you have an agent stack, which is all the tools and infrastructure you need to build agents. What I've learned over the past year is that you kind of always need both. So people really like that Vercel is becoming the everything cloud. You can come in because you're building a site, or you can come in because you're building an agent. You can come in because this is your first time ever building software, which is what we're seeing with v0, with the Vercel plugin for Claude Code and Codex. A lot of people, some of the people might be in this audience, had never even thought about creating software before and now they can create and deploy on Vercel. So what we've seen over the past I would say probably six months is this massive new onslaught of users who are new to software engineering. So the platform has become the place where you have an idea and you bring it to reality online.

Yeah, interestingly it's the second company that is a serious company. I think for many of us in this room we're serial entrepreneurs, right? So we've been having companies since... I probably had my first company was like 10, selling CDs in my school during recess.

Kind of. The teachers didn't love it. But this is the nature of being an entrepreneur. Like sometimes you rub people the wrong way. But essentially this is my second startup. So I came to Silicon Valley. I was initially working for another startup and one of the things that I noticed is I love building stuff. I don't love working for a boss, so I want to work for myself again. And I also noticed that in Silicon Valley, it's really easy to start things. In many ways, I think in this day and age, it's extremely easy to start things, but the journey has also gotten really difficult and competitive. So, I started a startup, then I sold it to the company behind WordPress. So it was my first exit after quite a few years of iterating, pivoting, trying different things. And then when I was at WordPress, I realized that the world is going to change because of different big waves. One big wave was the cloud. At WordPress, we had our own data centers. And I thought, man, I don't think the next billion companies are going to have their own data centers.

This is 2013.

Yeah. So I started Vercel in 2015, end of 2015, 2016, and I had spent two years at WordPress. And by the way, WordPress had to create their own data centers. They were scaling really quickly and they needed to keep up with their scale. And so what the cloud brought to the world was this idea of elastic computing, that you could have a quick idea. An anecdote that I like to share is in this room you might find the next multi-billion dollar success that everyone's going to be talking about in 10 years. For me, it was the case when I went to my first co-working space and I met the Instagram guys that were my office mates and they had built on AWS. And what I noticed is that as they kept growing, obviously AWS was able to keep up with their growth. So they became quickly one of S3's biggest customers because people kept uploading photos and then they would put them into S3. So that was one of the big things, is that the cloud has this ability to scale with you. And I noticed, I bet on that wave. I didn't know exactly what I was going to do frankly when I left WordPress but I knew okay that's going to be a very important wave. Another one was front end. So when I was at WordPress we'd been really successful with the PHP stack but it was all this monolithic thing. You had to put all of your eggs in one basket in the same runtime in the same codebase with the same sort of limitations. And I noticed that the world was going more towards more interactive rich client experiences, so React was happening as well. And so when I left I didn't know exactly what I was going to do but I had a bunch of intuitions. I knew that there was going to be a potential for something like Ruby on Rails but for this new generation. I also knew that the cloud was really powerful but it was insanely hard. So I think noticing these asymmetries about the world is really important for entrepreneurs.

What I would do is because I was always obsessed with starting something new or coming up with new ideas. I would go out to dinner with people and I would try ideas on them. I remember I was with a couple of friends in Finland and I was like, 'What do you think about this framework idea?' And I go back and that might have been the first time that I articulated verbally what would become Next.js, right?

Thinking out loud and seeking feedback. I think I've been always pretty outspoken about what I want to build, which is sometimes hard to do as an entrepreneur because you think everyone's going to steal your ideas. Like I've been there, done that. But it's really helpful to notice how people react to your ideas or to the things that you're selling them on a Zoom call or whatever.

I think everything is feedback. Even when they notice that, and by the way, I've gotten better at this over the years, but even noticing someone is not following your train of thought.

I kind of upset an engineering manager at Vercel today because I was like, 'Hey, on Friday we're going to have this customer. They're going to come to us and they're going to tell us everything that's broken. And by the way, a lot of what's broken is relating to your area of work.' So it's like setting up expectations. It's like don't feel bad. This is how we're going to improve. But I try to get all of the negative stuff out of someone if I can because one lesson that I learned is that people sometimes might not volunteer all of the bad things about their product experience. Sometimes they might even be a little oblivious about what's painful about their day-to-day. They might tell you like, 'Oh yeah, I love everything about ChatGPT,' but then you start digging in and like, oh yeah, you just got a bad answer and like, oh, that didn't work. Oh, and that was really slow. And so you start identifying opportunities commensurate to your ability to dig.

Yeah. Like I try to be really annoyed about everything at Mercury. Like I think you can get used to things and then you just take things for granted. But if you're just very annoyed about the world, it's actually a useful way of learning and improving.

There's a yin and yang. I think I'm a huge spiritual believer in that, that I have to give you enough of what's bad and also I have to celebrate the wins and I have to bring out the joy of our accomplishments but not so much that we just live in this permanent idealized state of oh everything is perfect. So I think it's also how you give the feedback. I think if you make it very much about like, hey, I don't like this. Maybe if we did it like this, and you make it more about we're creating something together. I don't think people mind that. I think if it's purely critical without proposing a solution or if it's about someone like, hey, why did you do a shitty job of this again? Like no one likes that. One thing that I adopted as a mental model that changed my perception of everything is Jony Ive has this thing about opinions versus ideas. So opinion, by the way, and I preface this by saying all feedback is a gift for me. I have enough thick skin that you can tell me an opinion. I still kind of value it. But there's a superior form of feedback which is an idea. So the example here would be if there's something you don't like about an app, you can say ah this app sucks, screenshot, or you can say this app sucks because so and so and here's my idea on how to improve it. So I think the final form of feedback and the most dynamic and productive workplace would be one that is a free exchange of ideas that are high quality ideas and the more concrete the better.

And I sometimes correct people when they tell me, 'Oh, that's an awesome thing you guys invented.' It's like, 'No, that was invented by HashiCorp, who was a customer for Vercel in the early days, or that was invented by TripAdvisor.' I actually memorized some of the big wins that we've had that are attributable to customers.

Yeah. One of the ways that we try to suss it out as early as possible is by actually building. So can you ship it as an experiment? Can you ship it as a v0? Can you actually show me your idea in action? Obviously, as long as the cost is not too dramatic, by actually giving it shape, people realize pretty quickly that their idea is bad. There's also ideas that we all convince ourselves are good ideas. And we have to face the real world. And so I try to always use the voice of the customer or data to actually contrast the hypothesis with what actually happened. And I really try to use data, you know, I'm not the crazy data science guy, but I try to constantly orient myself around data. So I have all of these really cool Slack workflows where now an agent will give me every week a download of metrics and an analysis of the metrics. And so those are ways in which you can be reoriented. And by the way, sometimes it's not that it's a bad idea, is that there's this other much greater idea that you should be feeding your attention. And so this is actually where data can be really helpful. Obviously, I'm speaking about a company now that has multiple product lines and we're at a certain scale right now. You're probably in the trenches of your zero to one idea and there's different techniques for that.

Yeah, if I were to name the chapters of the company from your question, I would say chapter one was finding focus. That was the whole arc of the first few years of the company. It wasn't finding ideas. I think right off the bat I had some pretty good ideas. So Next.js is obviously a pretty good idea. Maybe I had too many ideas. I was starting to build a lot of other open source projects at the same time. Like some might know, we built a terminal called Hyper and we built a microservice framework and a bunch of other things. So finding focus and narrowing focus, first identifying that from all the portfolio of open source things that we're working on, Next.js was like, 'Holy shit, that one is getting adopted by enterprises and individuals alike. That was going to have a lot of traction.' But the other one is on the infrastructure side, narrowing our focus to do this particular kind of deployment really really well and focusing more on the CDN and the capabilities of distributing these workloads globally. So that required a focus because my prerogative, my intuition was I'm going to build infrastructure for everything. Even this is so crazy because I was watching these two movies of wow Next.js keeps taking off and everybody in the world is using it and it's like the infrastructure can do Next.js but it can also do kind of everything. So it wasn't focused. So it was a process of trimming the forest and finding the thing that would stick.

Okay. So in retrospect, the things that have always worked the best and I teach the company still is the things that were solving our own pain points and had been proven before they kind of got out into the market. So Next.js we built for ourselves. So when I left WordPress, I realized I'm going to use React for everything. And yet React was kind of a pain to even just build one page, one website. You had to almost take a PhD. And so I built it for myself. And then I started noticing that a lot of other people were running into that same problem. It wasn't immediately obvious to me that my problem was relatable. I almost discovered this gem. And by talking to other people and paying attention realize okay like this thing someone was asking me earlier how do you get into a market that's seemingly crowded with a lot of competitors and actually I try not to think about competitors because I would have never shipped anything if I was motivated by competitors. Like think about how many ways there were to create a website. There were actually probably like 20 React frameworks before we nailed the one that had this minimalism and zero configuration and server-side rendering and all of these things. So I think focusing on that pain point but then noticing that that pain point was so relatable and was going to give rise to a gigantic market opportunity. So that's the other thing I always give advice to people is you really have to play in enormous markets. I think the intuition that entrepreneurs sometimes have is that well if I get like 5% of this market, 10% of this market, doesn't matter what the market size is. But what I think we really benefited from is that we entered the, you know again from a wedge, Next.js became a wedge into what has now become the entirety of the cloud for us and so we can keep dreaming about.

Some people said, many people said.

Yeah, it's tricky. There were a ton of very skeptical venture capitalists. So yeah, I do think by the way, and this is potentially bad news for some, but like what you do cannot fully be consensus. If it's fully consensus, if it's like you're listening to the idea on every podcast about agents and AI, by the time you roll up your sleeves and you're working on it, it might already exist. It might not be a good idea after all. And so there was a certain amount of skepticism about the market, about the category. And I'll explain why. So there were appliances that you could buy to do all of these things that Vercel does. It's a little bit like when I was a kid in Argentina, I assembled my computer by buying all of the parts and then I put them all together. At the time, it seemed like a good idea. My dad thought that the way to have a computer at home was buy a bunch of parts and accessories and toss them all together. And later came along Apple and they said, 'Here's a MacBook.' And I remember when I got my first MacBook, I was just in disbelief that computing could be so simple. And I think there was a version of this for the cloud because I think rightfully a lot of VCs would tell me, 'But isn't that solved by CDN?' And I would say, well, but it's really easy to build and deploy and configure it. 'Well, isn't that solved by CircleCI that gives you the CI part?' And I would say, well, yeah, but you also have to host the thing. 'Well, isn't that solved by Heroku?' And so I kept getting these intellectually good objections. But the thing I guess my north star was I kept dreaming of that simplicity that I could give to the world. And I do remind people that there's a lot of advice that I may give you that is maybe good, bad, whatever. But I do think that so much of the world is won by the people that bring simplicity and ease of use. Is that simple?

It's not necessarily about like okay the market is excited about AI don't do AI but even that is provocative right? AI might not be the solution to every kind of problem. The thing that I try to do when I speak with entrepreneurs is talk about their customers. Talk about especially when they don't understand the idea like I'm always giving them a freebie like all right I blacked out for the last two minutes can we start over, but tell me about it from the point of view of a customer. And I think at the end of the day, there's going to be more and more opportunities for legal AI. I'm sure not everyone has the clarity of how to do that.

Yeah. I mean, this is where you also have to bet on the founder to figure it out and figure out the nuances. I don't think it's possible for investors to understand all the nuances. It's not possible. They're looking at too many deals. They're looking at too much stuff. But a good founder will figure out the nuance. So, I think you were mentioning the stages of a company. The first one was Next.js. What was the next, did that story complete?

Market with that made the case for Vercel really easy to understand to a business buyer. I've always been really good at creating developer experiences and convincing developers and agents to use a certain technology. It's like in the DNA of the company. It's happening organically. You might have seen some of the things that Vercel Labs have been putting out, like we launch projects and they get really good adoption, but sometimes it's not obvious to a business buyer. People have asked me, and they weren't trolling. They were like, 'Why should I make the lives of my developers better?' Like CIOs, CTOs have asked me this. To me it's so obvious. If you make the lives of your developers better, you increase productivity. You create higher quality products. Those products make you more money. But they wanted me to spell it out. They wanted, okay, four jumps in logic. Okay, like you was happy, more money. Can you like draw it out for me? And so e-commerce gave me that. I was like, all right, Next.js cost this much, for sale cost much, improvement in conversion rates, improvement in obviously development cost, total cost of ownership, and then profits like increased topline retail sales, etc., etc. And sometimes we would, the most beautiful thing that gave me the repeatability is I started marketing through what I would hear from customers. This is the easiest way to do marketing in the world. No one wants to hear, by the way, I think we all kind of navigate the world with shields up with some kind of BS detector. For example, I'll tell you one is a pet peeve of mine. I go to a website and I see a banner of logos and I immediately think none of these logos use this company because it's so easy to throw in an SVG. Now Claude can look it up for you. It's like, oh yeah, cool, Nike uses me and this and this and that. And so you need substance. Okay. Sometimes I ask entrepreneurs, just tell me about one happy customer, just one. And let's go from there. And so when I started hearing from our ecom customers, okay, this is what you did, this is what we measured, this was the ROI to the business, I was like, okay, awesome. I'll run with that. So that started giving me repeatability. And then I found a bunch of other use cases and it started scaling the company revenue-wise because basically what I was describing with the finding focus arc was like basically zero to one million, and then one million to five million was like the next. A lot of it was ecom marketing. We're doing a lot with obviously startups, has always been a very good segment for us, like leaning into like if you have an idea and you need to get it to market really, really, really fast. We became synonymous with that solution for crypto, for agents, now for AI, for a bunch of other things. Yeah. So, that was chapter two.

Well, interestingly, right, like you realize at some point that there's a couple things in your trajectory. I'm sure you guys are also in a similar journey. So, one is that you always have to keep growing, right? If you're in the venture game, you're always growing. So, like we went from like, we 5x revenue again. And we went from 5 to 25. We started going into areas that we were less comfortable I would say or like were less natural. So the classic one is enterprise. So we've always been an awesome like product-led growth machine where people know about our product, they recommend it to their friends, they invite their friends and so everything grows bottoms up. Then we layered more top down. Top down is really important. Top down pushes you into places where you're less comfortable, adds requirements to the product and things that you normally don't hear because like I think we all can get trapped in local maximas even with really fast growing businesses. Things seem to be going amazing, there was a sense of like, okay, are we getting trapped in like what's a known quantity for us? So we layered more top down sales. For me like the next chapter I guess would be like scaling the company as an organization. I always have loved this idea of turning Vercel into an innovation machine. It's not really about Next. It's not really about AI SDK. Maybe this is the third chapter, right? Which is the product is the company itself.

I would agree with that, right? Like going zero to one is the hardest. Is that why Peter Thiel wrote Zero to One? I haven't read the book. Zero is the hardest thing.

I think, okay, so I think we've established some consensus. I think it's not as hard as zero to one. That's the hardest thing in the world. And I agree distribution advantages are awesome. But you've also probably seen, for example, the company that's so annoying about growing their new product that they advertise and vomit all over their existing successful product.

They really want you to adopt the new product, right? Like so they deal with difficulties around like are we ruining the existing good product, are we sabotaging it or are we inflicting self-damage in promoting the second product. There's that risk although I don't think it's particularly high risk. Another risk that is real is your second product might be a mediocre success. And so that is also unacceptable because like if you're looking at like public markets, they're expecting that to your point about distribution advantages now the bar got higher. So if it took you several years to get your first million ARR, the new product should get it in a day, right? And because you have the distribution advantages. I think the cure to this is that I've always believed in companies need to build a very strong core and this is why zero to one is really hard because it's going to be the main substance of your company and then the way to succeed is by building really smart adjacencies to that strong core. Now to make this advice relatable to the people that are just getting started, be careful in not jumping too quickly to that second thing because the main thing, so the metaphor I like to give the team is iPhone and AirPods. I think we've all heard the sales figures from AirPods is astonishing. Tens of billions of dollars a year is just maybe a lot of that coming from me because I lose them a lot and they break and whatever. I have like 20 pairs. But even excluding me, it's been such a successful product. But it goes like bread and butter with the core product and that product would have never been successful if the iPhone wasn't this good. So your goal, your job is to create the iPhone. No big deal.

Serious alpha by the way. I violently agree with these two things like you can get, I mean this is like imagine like you're working extremely hard 20-hour days and like I'm telling you about you can visualize this amazing future in which you don't even have to innovate anymore. Your product is so successful that people pull features and products out of you like your main thing has already gotten so much adoption is in a critical path that they're inventing on your behalf. They're saying, 'Oh man, all my Bluetooth earbuds are BS, Apple, couldn't you just make one that always works and then you invent the AirPods?' And so, a lot of subsequent innovation, if you pay attention, can directly come from customers. And so, I was giving some examples like Trip Advisor, etc. Where they invented features that have now become, you know, hundred million dollar revenue businesses for us. And obviously one remembers that first insight or moment where it was the customer that said, 'Yeah, I need that thing.' Or, 'No, no, no. What you just said, can you change it slightly? I need that other thing.' And they're both giving you the product and then they're giving you the marketing story. So the example with Hashi was they had grown too big to do static websites, but they really like the snappiness of static websites. So they came to me with a conundrum. They were like, 'Okay, I don't know how to solve this, but this is what I want. I want every click to be instantaneous, but I want everyone in the team to create new data for everything to be dynamic, personalized, etc. You guys go invent.' So that was the dance. It's like, and then it's like, okay, if I can come up with a solution, they're going to remove the existing provider. They're going to use more of Vercel. So it's highly, highly motivated. And that's what like a true good design partnership sounds like.

There's a mental model, I've heard it from Satya from Microsoft, he calls it brand permission. Sometimes you might have the right team and you might have the technological ability but you don't have brand permission. And so I think with Google they fell for this trap quite a few times where people didn't want them to be a social network, it didn't fit in the conception of how you think about that company. And but you have the best engineers in the world and the best technology and a great development platform. Just because you can doesn't mean you should. And this is advice that I'm finding myself giving to everybody today with AI. Just because you can prompt and yes, code is seemingly free because you can turn anything, any prompt with enough tokens into a piece of software. It doesn't mean you should. It doesn't mean you'll be able to afford to maintain it. It doesn't mean you'll be able to market it. It doesn't mean you have to, right? And so I think saying no to a lot of things, to the right things, and to a lot of things is also part of the job. Which is so ironic because a lot of people sometimes praise us with, 'Oh, you guys ship so fast.' But if you saw just how many times every day I'm blocking things or saying no or asking for refinement, you would think that we would never ship anything. And so I don't know, you know, like the formula for that is, but like you do have to have a fair amount of no in your own head even like before you verbalize an idea. And also before you decide to greenlight a product or ship a feature.

My take is when I first came to Silicon Valley, I remember people saying all kinds of BS that I just turned off. Focus on my product, focus on my customers. And I wish I'd done that even more aggressively. Like this narrative of company X will destroy company Y. I think Steve Jobs had a thing about like I never read the Wall Street Journal to outsource my opinions on the world, the markets, the tech. Now that I'm getting closer to all these people that invest in public companies and whatnot, like I realize that you are there to provide the substance and everybody has the shallow cheap opinions, the cheap seats. And so X is this double-edged sword. You can source FUD, anxiety, fears, negativity or it's the most wonderful marketing channel and like freedom of speech platform of the entire history of humanity. So every successful technology has these two sides and I try to, I've always been just insanely optimistic and so I've always believed if I create something great the rest of the world sorts itself out. I actually remember one of my, the co-founder of my previous company used to say product market fit solves all kinds of problems like just focus relentlessly on getting product market fit. You can have problems, you can have funding problems, you can have margin problems. I'm not advocating for like sell dollars for a penny, right? But there's so many problems that are solved once you get to that beautiful feeling of like customers are pulling product out of you. And I try to turn off the noise of, by the way, like sadly like venture capitalists main job is to be the expert non-expert commentator of the world like and increasingly they're realizing this because they spend a lot of time in media like Andreessen Horowitz is a media company, they're open about it, right? Like media, media, media, media like think about your life, your time should be product, product, customer, code, code, not media, media, media. The media is going to be a double-edged sword.

So that's what I'm saying. I'm saying like this gentrification kind of has happened all the time.

And social media is way more prevalent. So everything that is happening is being transcribed in real time and narrated in real time to the world which is a whole other can of worms. Like it's amazing how many conversations I have with journalists where like they come to me and say can you tell me more about this thing that happened last week in San Francisco like and Wall Street Journal story comes out, you know what I mean? So there's a lot more eye of Sauron in SF. You just have to make sure it doesn't get into your head and you keep stay focused on.

You know over the years I've become more and more interested in the story of the gold rush and it seems like one could say well 10, 15 years ago people were like doing the gold rush in San Francisco because of social media, networking, web 2.0 etc. But another point of view is that this has been going on for a century, right? And that there are people that come into the city because they're drawn in by the ability to build, the very unique ability to build that it gives you. And then there's people that are attracted by like, oh my god, like look at the stocks go up kind of mindset. So yeah, to your point, I think there's people that read stuff like that and it's like this is the best place in the world to build with like-minded people. And there's other people that are like, 'Wow.' Like there's nothing I would dislike more than a life of just being obsessed about chasing the next big thing and innovation and whatnot.

I will say there's always been like an ambient stress in Silicon Valley. I've always perceived that because it's a very competitive place and people talk a lot about work and people talk a lot about rounds and people talk a lot of performance. So that I think comes with the territory of like, you know, I always ask myself the question of like how much will entrepreneurship and knowledge work over the fullness of time resemble being an athlete? Because you have to put in the hours. You have to out compete others. You have to be always on and you obviously hear a lot about what, something I've always loved about SF is people are always optimizing. They're optimizing their company, they're optimizing their agents, they're optimizing their life like Ryan Johnson and like all like it all comes in this bundle of competition so sometimes it kind of reminds me of yeah like being in some kind of sports competition.

Well, and this is the other reality of this game is that I think being a founder can be so lonely because it's not like you have a whole team of coaches that are helping you not burn out. So a lot of it is what I think is Ben Horowitz says like managing your own psychology because in reality whether you burn out or not like there is, you could totally go off the rails and just burn out but also you can be your own coach in the sense of like all right it's time for me to unplug. It's time for me to go for a run. It's time for me to like invest more in my social life. And I know that it was difficult for me to figure all that out because I was by myself coming from another country. So like I appreciate the concern over those kind of things. But also part of what one is to master.

Yeah, there's periods of pain. But I think it goes back for me to like the yin and yang thing is like pain pleasure balance system. That over time should tip more towards pleasure. But I truly do think maybe this is even more so in the early days. I've been in extremely painful situations, right? Like there's so much uncertainty that comes with building a company that puts you in situations that you're just simply not prepared for.

Man, there's so many. I think, well, we talked about like in the early days being misunderstood or trying to go against the grain and like not getting good results. I think just like hitting a wall. I think the time pressure like if you raise money like you have to produce results by like, it's interesting. I think raising money is such a good example of the pain pleasure balance because everyone celebrates raising money. But it's a little bit like I just, I'm celebrating that I was on the treadmill and I just went from 10 to 14.

And there's another guy next to me and he's also running and like so.