How Sahil Lavingia (Gumroad CEO) Learned to Code

Hello and welcome to the Learn to Code podcast here at One Month. Last week on the show, I chatted with Sahil Lavingia, and the topic of conversation was reflecting on my failure to build a billion-dollar company. This was a viral article that he wrote and published on Medium about his failure, with quotes around it, to build the site gumroad.com, which is in my eyes a very successful company, but in the eyes of Silicon Valley, because it's didn't become a billion-dollar company and all of the pressures that go along with raising money, he was considered a failure by a lot of his peers and investors and all this kind of stuff. And it's a great story, and feel free to, you know, check out that article. I'll put it in the show notes here at onemonth.com, and, and go back and listen to it if you would like. You don't need to listen to that episode to listen to this episode. They are two completely different conversations. And in this episode, what you can expect to hear is some Sahil's story of building Gumroad and learning to code. So some of the things that we will cover in this episode is how did he learn to code, how he used Python to launch Gumroad and then to raise $10 million to fund the project. We'll talk about, you know, now what is the best programming language to start with if you're just learning to code right now, as well as a really cool hack that he used to build Gumroad. And I think it's something that's often overlooked and still very useful for many different kinds of businesses. We'll chat about that a little later in the episode. So, I hope you're having a great day wherever you are and excited to join me in my conversation with Sahil Lavingia of gumroad.com.

I'm gonna start with the question that I usually ask I ask everyone that comes on the show, how did you learn to code?

Yeah, so it happened a couple times. The first time I learned to code, I learned iOS development. This is back in 2009, I believe.

That's right when the iPhone came out, like give or take a year.

Yeah. But yeah, I learned Objective-C from a Stanford course. Called CS193P on iOS development that was free on iTunes University. I spent like 3 weeks literally every day, uh, just watching the videos and doing the assignments, doing the homework, etc. And then 3 weeks later, I knew iOS development, it turns out. And so yeah, I started making my own apps, and then, you know, that ended up like a year, less than a year later, I was working at Pinterest as employee number 2 and I shipped Pinterest for iPhone, which is kind of cool. And then the second time I learned to code, I learned Python. I learned how to make a web app because iPhone apps are cool, but often if you're building something, you sort of need it to be able to talk to a website. You need to build an API or something like that. So I started learning Python. Just some tutorials I found. Google had written them for App Engine, which I think they've rebranded now to like Google Cloud or something like that. And so yeah, I learned Python in a couple weeks, started making web apps and things like that. And then, you know, fast forward maybe 5 or 6 months after that, I shipped Gumroad, which was at the time basically one Python file, like 200, I think it was around 2,700 lines of code and it worked. There was a lot of copy pasted code, but it worked. And yeah.

When people are getting started, oftentimes they'll be kind of conflicted about which programming language should I choose? It's sometimes hard to know, you know, can Python do this or that? How did you have the confidence or how did you go about choosing Python and was it like, was it everything you had hoped for? Like when you chose it?

It was the programming language of my dreams. Correct. Honestly, I don't think it, yeah, I think a lot of people do get caught up in stuff like that. And it doesn't matter that much. I think at the end of the day, it, if your app does the thing then it doesn't really matter, you know, if it was Python or Ruby or you sitting at a typewriter, like manually responding to requests. If you can do it just as fast, your users probably don't care super much. I think for me, I had the confidence because I was working at Pinterest at the time. And so I, you know, we were using Python and Django at Pinterest, but in general, like there's, you know, there's a couple options. There's Ruby on Rails. There's you know, now there's like Laravel, there's Python and Django. And now, I mean, there's all the JavaScript stuff, which I'm not familiar with really at all, but I'm, you know, React and Node and all that sort of stuff. But I think at the end of the day, like, you know, find something that you like that was built and then figure out what was used to build it is— I don't think that's a bad, sort of a bad system. To have, you know, and I saw like when, you know, Governord was initially Python and our first couple hires were Ruby engineers, really familiar with Rails. And, you know, it took maybe a couple weeks to rewrite the whole thing properly in, you know, in Rails, on Rails. So, you know, you can always revisit these decisions, right? I think the important thing is to actually like build the thing and like I don't think it's necessarily wasted time to sort of contemplate which programming language or which framework you want to be using. But at the end of the day, you can always revisit it, you know, and, and often startups do, you know, it's, it's one of those things that it might work for a year, might work for a day, might work for 6 months, might work for forever. But, you know, nothing's set in stone. You can always replace whatever you're using when it, when it comes time.

Yeah. Tell me a little bit about this decision. It's like, it sounds like a pretty difficult choice to be working at Pinterest around, I'm guessing this was around 2012-ish when they were like at their heyday. I mean, I know they're still a pretty huge company, but Pinterest was like huge. Like everybody was so excited about Pinterest and you got a job there and then you decided to leave. What was that decision like for you?

Probably not the best financial one, but, but yeah, I mean, I think for me, I sort of got hooked on building product, shipping product, and I really got into the mindset of like, I want to start a company. And when you want to do that, it's kind of hard not to do that because nothing really comes close to that, you know. Working at a company is great, being an early employee, uh, you can learn a lot, but there's nothing that really comes close, for better or worse, often worse, uh, than starting a company and like being a CEO and sort of being forced to do all the things because that's, you know, no one else's job anymore. Yeah.

What, what attracted you about, like, back then, what attracted you about being like running a company? What, what was in your mind of like what it would be like Yeah, yeah, I think, I think there are two parts to it.

One, one was this idea of like success, riches, fame, influence, power, all those sorts of things. Sort of a lot of my heroes, you know, when you think about Steve Jobs or Bill Gates or any of these sort of like canonical, like, you know, sort of almost myths really, they were all founders of companies that had a really large impact. And so that was like a thing. It was like, could I do that one day? And if not now, you know, these are all, you know, when you look at a lot of these sort of origin stories, they're college dropouts, they started their first company at 18, 19 years old, and then they were incredibly successful. And so I sort of like, was like, if I don't do this right now, I'm not gonna have my chance to be one of those, which in hindsight is kind of not the most rational thought process maybe, but But that's sort of one of the motivations. And then I think the other one was just like this idea that I felt, you know, a lot of people could do the job that I was doing at Pinterest. You know, I was having a large impact. I was making and designing Pinterest for iPhone, but a lot of people could go do that. And I felt like a lot of the innovation, even though I was early, you know, I was employee number 2, once things start working, you kind of like, you kind of should stop innovating in a way because the thing you're doing is working and you should just make things like faster, fix bugs, make sure the site stays up. Site stays up. And so I felt like building Gumroad was something only I could do, right? And so in terms of sort of like maximizing the amount of value that was being created in the world, it would be better for someone like me to go do something I wanted to go do than other people didn't want to do. And then a lot of people who wanted to be early at a startup and do iPhone development could go do that job. Right. And so I think it's sort of those two combinations of sort of the selfishness of like, I can go do this thing and maybe make it versus the selflessness of like, if I do this, like that's a great way to add value to the world in a way that is just sort of unique to me and my skillset and my sort of my worldview.

I love that. Yeah. Kevin Kelly, who is one of my inspirations, he has said something like that in his writing. That he, when he has an idea for something, he often tries to give it away in hopes that someone else will create it. And that if he can't get someone else to do it and he really wants to see it in the world, then he'll go, ah, goddammit, now I have to do this thing. Which I kind of love. I kind of love that, like, ethos.

I totally agree. I think if you really believe in an idea, it shouldn't matter who's building it, right? If you're the one— if you, if you believe in an idea, but only so much as you sort of restricting the idea to yourself, then like, I don't know, it might signal that something else.

Well, tell me about the idea that about Gumroad, you know, at that time that you had. So you launched this app, it was Python, it was just like whatever, 2,700 lines of code. What was the big— what was the big idea back then for you? What was Gumroad?

So Gumroad started out because I wanted to sell something that I had made. I designed this pencil icon in Photoshop. It took me a few hours and I was just sort of curious, like, could I sell this thing to my online following, my audience? And I quickly realized that it was not easy to do that. It was pretty difficult if I wanted to sort of say, hey, this is $1. It would have been easier for me to like have them like mail a dollar bill in the mail, literally to me, you know? And then me like, I don't know, put it on like a thumb drive and ship it to them. You know, it was just, you know, it just felt like, oh yeah, like there's Dropbox and there's, you know, there's all these ways to share it, but for some reason, like I don't really understand, it was not possible for me to sort of charge $1. And so the sort of the premise for Gumroad was, what if you took Bitly and added a credit card form in the middle? What sort of, what would that do? What stuff is out there that doesn't get sold because it's too painful? Or what stuff would people make if they knew there was a way to sell it? And it sort of just got stuck in my head that I just kept thinking about all of these opportunities. These musicians, these designers, these writers, these photographers, these filmmakers, painters, stand-up comedians, whoever, all these people that make stuff, make content. On their computers, on their phones, and sort of don't really have a way of like putting that out there and getting paid for that.

I love, I love that.

Yeah.

Because even from my perspective now, I would have a hard time thinking, you know, if I wanted to sell something, I can get a Shopify account and it's like a lot of setup and maybe you have to get a membership, you know, or on the low end I could get like a PayPal URL link. But I mean, it's kind of F PayPal to begin with, but also whatever, just my politics. But like there's like they're not giving you anything there. You kind of also have to like have the— there's no image or branding or like, you know, it doesn't have like a feeling.

Yeah, the way that sort of we pitched in the early days was like, there are a lot of ways to set up a storefront on the internet, but like there are not very many ways to set up a lemonade stand. Sometimes you just— and there frankly is much easier on the internet to like not even get a storefront, just like go to Target, be like, hey, could I sell my stuff here, right? Which is like a marketplace like Airbnb or iTunes or Amazon. And some people need lemonade stands, right? You might eventually need a storefront, but to start, I don't think there's any better way to say, okay, instead of investing all of this time, effort, money, stress in building a storefront, let's just put this on the sidewalk and busk, or set up a little CD shop on the corner. And see if anyone's interested before you quit your job and do that.

All right.

So now what I'm curious about is a lot of people have ideas. I get pitched on startup ideas all the time. I'm sure you've heard similar things. You had this idea to make this thing. You took one step further, which is you actually made a prototype of it using Python, and that's super cool. I'd say that's a pretty big leap from idea to making something. The next leap there though is for me, if I was doing this, I was— I'd be wondering like, does anybody want this thing? Like, is it actually— does it have legs? Like, so how did you kind of make that next leap? What were some early signals that you thought like, hey, maybe other people, maybe the market wants this, maybe some other people want this?

Yeah, so I had that idea that I designed that pencil icon Friday night, and then by Monday morning I'd built this prototype, you know, really simple CRUD app. A friend of mine had started a payments company called Stripe, and so I got a beta key.

That's a cool name drop.

Yeah, my buddy, uh, he started Stripe.

Now, yeah, now it's like a $20 billion company, but back then it was like, you know, 4 dudes or 6 people. I do think they had women, so not 4 dudes, but yeah, I think we were like user number 40 or something like that.

Oh, that's cool. We use Stripe and we are huge fans of Stripe here.

So yeah. Yeah. And yeah, it's awesome. And so, yeah, I built this thing in a weekend and launched it Monday morning. I tweeted about it, I put on Hacker News, and it did really well on Hacker News. It got up to number 1.

Nice.

Yeah.

For the day, which was great. And you know, the first day, I don't know, maybe I think around 50,000 people or something saw it. But really, and honestly, it's funny because people are like, whoa, that's crazy. Like you're successful. Like now all you gotta do is like, Listen to users and like, you know, keep going. And it's like, honestly, like just cause 50,000 people saw it doesn't mean anyone really used it. You know? It sort of like got all this traction and sort of mindshare and then died, you know, like no one really used it, but I couldn't stop thinking about it. And I couldn't really forget about this idea and sort of the same concept around like they're all there. And I think that's why 50,000 people saw it the first time around, even though they might not have used it is they're like, I can't use this, but I know people that could. And so that was enough for me to say, you know, I want to work on this thing more. And I couldn't at the time, I was at Pinterest. So I left and started working on it more, raised money for it and all of that stuff happened pretty quickly after that. And I, you know, and then it was just telling people about it. You know, it was like email, literally cold emailing people that I thought were cool on the internet that I might've followed for a long period of time. And said, hey, like you make music or you sell videos or you do this or you do that. Would it, would you consider, you know, trying to sell some content using this tool that I built? And I did that hundreds, if not, I mean, probably almost definitely thousands of times.

You emailed thousands of people just like, hey, you should try this.

Yeah. Yeah. I think people, people sometimes don't realize like how manual some of this stuff is.

Yeah. That's amazing.

Even getting press, you know, like sometimes I'll get press for something or other and then people will ask me like, how did that happen? And I'm like, I just sent them an email. I mean, you know, people are busy with their lives and they're very unlikely to kind of stumble upon you and your product and go use it and definitely write an article about it. But you know, if I make it really easy for someone, hey, I built this thing, this is like why it's interesting. "Here are some screenshots, like, let me know what you think. It would be cool, you know, I think your audience would appreciate this." You know, 1 out of every 30 emails probably led to some article, you know, which is funny because now Gumroad is big enough that I kind of forget to do that, but I should probably do that more often. When people are surprised, what it means is people don't do it enough, which is why it's a surprise to them and why it works for me.

Well, there's like, there's like a whole kind of two sides to this because on one hand, spam is not surprising. Everybody's familiar with it. You get a bunch of it. You get these cold LinkedIn emails. Like, it's kind of like a dodgy— it feels not good, you know. But on the other hand, to hear that like the founder of a company is like actually one-on-one reaching out to artists, that's not something that you necessarily hear so much of because it— there's like a bit of heart. There's like a lot of heart in that, you know. And I think People are— I'm a little more like, even just hearing, I'm like receptive to like, yeah, that's, that's really cool.

Right. Because the minute it's very, it would be very easy for me to, to say, okay, copy paste and send it to 50 people. But when you do that, you're basically guaranteed to have no result. And sometimes I would just, you know, I would copy paste the link or like maybe one sentence about Gumroad, but then I would like write a new email every single time. It would just be slightly different. It might be slightly more catered to that person.. And it would take 3 hours to send 50 emails instead of 15 minutes or something like that. But, you know, including finding the email and doing the whois and everything like that. But your response rate goes up dramatically when you do stuff like that. And, you know, you probably get emails, I get many emails, and it's very easy even just in the list on Gmail to be like, okay, these are the 40 emails that I'm not gonna read because they're all obviously mass market emails. And then these are like the 7 that are from human beings that need feedback on their product or that want Gumroad to use it or think they want me to invest in or whatever the reasons are. But it's like really easy and it certainly helps because once you've been on the other side of it, you can kind of be like, okay, these are the type of emails I'd like to see. It's kind of like the golden email of the golden rule of emails, right? Which is just write the email you want to read. And if you do that and you do that enough, you're going to start building an intuition around what's going to start working for you and what's not.

Before we, before we finish up here, I would just love to know kind of what's going on at Gumroad today. And, you know, what are some things maybe that you're looking forward to or that you'd love to share with us about Gumroad?

Yeah, Gumroad's in a pretty exciting place right now. I think partly because the essay, no, like there's never been a time where more people were thinking about it and it's sort of wanting it to succeed even more than it already has.

And so I love it.

Yeah, it's kind of this weird, uh, I did— was seriously not, you know, it was, it wasn't meant to be content marketing or anything like that, but it turned out to be a great piece of it. But yeah, I'm sort of like re-excited about it. We're sort of growing the product team again. We're gonna stay profitable. I think for me, the question of raising money is probably not right now. But yeah, we're adding a couple folks to the team as we can afford to, and we're really investing in just building a better product. And really sort of, sort of really as a way to say thank you, that's the way that I think about it, is that we You know, our creators, most of them did not leave when they heard about the layoffs. They stayed with us. And we have a lot of product and technical debt that we need to make up now because we were a tiny team for a long time. And so that's sort of my job now is, okay, like, let's ship the marketplace, let's ship ratings, let's ship, you know, a way to do membership businesses on Gumroad. Let's ship posts and public posts and commenting and all these things that people really have wanted for a long period of time. And now it's my time to say, "Okay, fine." Like, you get what you deserve. And thank you.

That is exciting. How many people are working on Gumroad right now? And the team's remote, right?

I'm guessing. Yeah, yeah, we're all remote. An office is frankly just too expensive. So we have around 12 people working on the product right now. A couple former Gumroad employees have rejoined the company, which is pretty cool. To see for very different reasons. You know, there's no promise of a billion-dollar company anymore, but there is still a promise that you get to work on something that lots of people like and use and find value in. And so it's pretty cool. Yeah, we have around 12 people. A lot of people are part-time, which is kind of a new thing we're trying to experiment with. It's sort of similar in the vein of our creators. Like, they are trying to make things work and do multiple things and spend time doing what they love. I wanna allow our employees to do the same thing and have that option as well. And yeah, things are gonna be— 2019 and 2020 are gonna be interesting for Gumroad for sure. You know, Gumroad's growing, we're bigger than we've ever been. We're processing around $6 million a month, we're growing 35% year over year. And last year we were 25% year over year, and the year before we were 15% year over year, so we're actually accelerating in our growth, which is— Kind of crazy. So yeah, we'll see how things go.

That's phenomenal. Well, yeah, anyone listening, I would say check out gumroad.com. It's awesome. We here at One Month have used it many times over the past few years. And yeah, really great to chat with you today. Is there someplace that people can go to find out more about you? Maybe on Twitter or something?

Yeah, Twitter is the best place to sort of follow me and my thoughts and see what I'm up to. My handle is @shl. Uh, so you can find me and all my weird thoughts there.

All right.

I hope you enjoyed my conversation with Sahil Lavingia. If you would like to learn more about Gumroad, just please go to gumroad.com and check that out. If you're interested in learning to code, you can do that with me at onemonth.com. I teach a course called One Month JavaScript, One Month HTML, Programming for Non-Programmers, a bunch of different courses that you can learn, uh, and get get coding in, surprise, less than 1 month. As part of our method at One Month in One Month, we divide each of the courses into 4 weeks, and every week you will make a project from start to finish. So by the end of the JavaScript course, you will have built and launched 4 websites using JavaScript. 4 websites that are cool, that you could show your friends, that are interactive, that are using data and APIs. It's super sweet. Yeah, so JavaScript, Python, and HTML are the 3 languages that we recommend as the core courses that you should start with. I would say start with HTML if you've never done any of it before, but to be honest, you could probably start with any of the 3 first if you were hungry to just kind of jump in. Yeah, so check that out. All that is at onemonth.com where you can learn with me and my amazing co-founder Matan Crippled. We have courses up there. And if you have any questions about the courses or about me or anything, you can email me anytime. Just go to, or just go to your email and then type teachers@onemonth.com and just fire away with any questions you have. I hope you enjoyed this episode of the Learn to Code podcast here at One Month. And man, you look great today. Go have an amazing day. You got this. You freaking got this. All right, go do it. Have a great day.

Talk to you later.

Bye.