How Whop Hit $1.2B GMV Run Rate with Just 20 Engineers

0:00You're doing 1.2 billion in GMV run rate. you've hit 5 billion in creator views you're at massive scale like instead of scaling up team like you're going to be scaling back i really enjoy working with very talented engineers right now because they can take ai and they can split up their project and get it done in five to a next other time i don't see a world in where big organizations really need junior engineers as much to do it that's why my advice to junior engineers is don't even try to fit that mold just go build your own shit and you'll probably be way more successful than you ever could have been being a junior engineer at an organization jack sharky welcome to sorcery thank you It feels like just yesterday we were on the Startup Westport panel with Dylan of Saturn and the team at Crowdvolt.

0:38How are you doing? Doing great. That was just, what, six months ago? Wow. I think so. Time has flied. Thanksgiving. Yeah. So you guys have been crushing it. You guys are at a$1.2 billion GMV run rate, and you've reached over 5 million views for your creators. Yeah. Creators are having the best time of their lives using our platform to really run their online business. um like you said we're hitting month over month highs each month last month we processed just over 108 million dollars on behalf of our creators and then uh you touched on content rewards we've done over 5 billion views generated in the last like eight ish weeks um eight weeks yeah the products were really taking off probably the biggest growth we've ever seen organically in a product okay so can you give the full WAP breakdown just start from the beginning I want to know everything yeah at its core WAP's mission is to make people a lot of money on the internet that's it um we grew up me my two co-founders Cameron and Steven and even half the people or more that worked there grew up on the internet selling things making money selling software building sass all these different things so a few years ago during COVID when everyone was home and trying to sell at-home workout equipment or baseball cards or gpu tips everyone's trying to buy a ps5 we were in that market we were those people um and we built a platform that really allowed people to um prosper on that so at its core wap is a platform that allows people to make online money most of the products people sell are digital for now um and yeah people run communities they sell software they sell courses they sell live streams they have long-winded forum posts they sell information.

2:22Pretty much whatever you can imagine is kind of sold in a while. You guys have had tremendous unlock in the last couple months, probably last year. What exactly was that? Because I remember in the beginning you guys were shoes, right? Yeah. So we started with software that facilitate people to buy sneakers or clothing or any of those goods. So like I said, we kind of grew up on the internet. I started in high school selling sneakers and clothing and hoverboards and headphones and all these things online. In order to get a lot of these items that are hard to get, you need to use something called a sneaker bot.

2:57So during COVID, when everyone was trying to get a PS5 because everyone was home and quarantined, they were selling out instantly. And they were selling out because people were using these same sneaker bots to sell them. So we found a Discord forum of people buying and selling software license keys as if like you were buying a Microsoft Word license key and someone sold it to you. Very discombobulated forum. People were scamming each other it was venmo and it was cash app it was paypal so we were like okay this is not good this is a very discombobulated market let's quickly build a quick website so we built a airbnb like website where you can buy and sell and actually rent out software for a duration of time and in our first month we processed i believe it was like fourteen thousand dollars um just simply moving that form over to our website month two we did about fifty thousand dollars and from there we just kept kept building everything that people kind of wanted and how did you and your co-founders come together you're the cto talk about the team yeah so um i'm the cto steven schwartz our ceo and cameron zab is our cgo chief growth officer cameron and steven started working together when they were very young around 13 or 14 they met in a facebook group building sneaker bots um and they kind of went through that whole trajectory too i met cameron online when i was about 16 also in a Facebook group selling sneakers we were kind of friendly at the time we never really worked together um and then a few years passed as COVID happened I was working with Cameron and Stephen on a separate business um they had a consulting company I was kind of just hacking away developing for it COVID happened we tried a few startups I believe we had like four in one year that all kind of failed four startups four startups we were just trying random shit we can find on the internet um and WAP happened to be that fourth one and it just kind of took off from there one of the things that I've been having trouble finding on the internet is how much you guys have raised in total and by who so could you share more about your funding to date and your investors yeah so we've raised about I want to say about 80 million dollars in total um series b up to so far some of our earliest investors were people like justin between much justin mateen founder tinder justin khan founder of twitch um peter teal and then we had our first series they raised led by insight And then lately our Series B was led by A-Star and Kevin Hartz, founder of Eventbrite, as well as Bain Capital.

5:19Can you share anything on the valuation? The latest valuation was$800 million. That's public? I believe so. Okay, we'll check. I think so. I think we got leaked by someone on Twitter. We try to keep it – we're so heads down building all the time, and we're very not like a pompous out there company. You can even look at kind of like the brand we've portrayed over the past few months. it's very not like this is WAP and this is what we do it's very like here are the people that kind of use WAP empowering a lot of the entrepreneurs and creators that use it so our brand and our perception also kind of matches that so we weren't like raised around go immediately to TechCrunch tell everyone it was like okay we have this like keep in our back pocket for later let's just keep heads down building we almost like forgot about it and then someone on Twitter leaked it a few months ago you know where to go for your series c sorcery sorcery i want to talk more about wop's mission because you've also created a little bit of a cult following people are obsessed with wop like it's a true thing and it's real and then even with your new product creator rewards like that's creating even more wop i don't i don't know what you call whoppers you call them whoppers not yet gotta come up with the name for that but we do call them you have a wop you have a wop like if you were sold no up you have a wop okay but we didn't frame that we actually noticed that people on twitter kept calling referring to their thing their thing they sold on wop as their wop and we just kind of took it and ran so what's your mission our mission is to make people a lot of money uh on the internet we um we also come home say is that we want to make people sustainable income on the internet we believe a sustainable income is around two thousand dollars a month our product basically makes it so you don't need to worry about payments like you don't need to like build up your own payments integration you don't need to worry about like building out your own landing page or your own checkout you don't need to worry about like your back-end business dashboard you don't even worry about like connecting to different integrations to chat with your customers or to go live to your customers we have all that in a product all you really need to do is come to WAP and just provide whatever value it might be and in a lot of different people it's a lot of different forms.

7:28Or that might not be you. You might not be a creator. You might not be an entrepreneur that's ready to have a following and start taking payments. So we have other ways to earn, such as content rewards, which is the fastest growing way that we have seen earners come. They're coming on the platform and they're helping some of our biggest creators or brands or whoever it might be market and really unleash their audience. So our mission is to make people a lot of money. sorcery is brought to you by brex the financial stack trusted by more than 30 000 companies including one in three venture-backed startups in the u.s nearly 40 of startups fail because they run out of cash brex is literally built to help founders avoid that unlike traditional banks that let your money sit idle chipping away at it with fees brex is designed to help you spend smarter and move faster their all-in-one solution combines checking treasury and fdic protection into one powerful account.

8:24You can send and receive money globally at lightning speeds, get 20 times the standard FDIC coverage through their partner banks, and even high yield from day one. With same day and even same hour liquidity, access your funds anytime. Companies like Scale AI, DoorDash, Service Titan, HIMS, Anthropic, Flexport, Robinhood, and Plaid trust and use Brex. Start today at brex.com slash sorcery. that's b-r-e-x dot com slash sorcery what is the onboarding process for different cohorts of users because you just mentioned a couple yeah so um any user that onboards they have to they sign up and you actually can't even use our website anymore if you're not signed up and that was a conscious decision to try to make it a higher intent community of people on WAP so once you sign up you go through our basic onboarding you know verify your email and stuff and then you actually have to join your first WAP before you can even use the platform.

9:23And that is, it's a confusing platform. Like we know it's, there's a lot going on and there's a lot of, with a lot of functionality and features and abstractions, it's hard to actually simplify it until what we're doing. So our thesis is let's put people in one or two WAPs when they join, let them land on a populated home feed that shows you who's talking, who's commenting on long forum posts, who's even live, who's paying off for content rewards or different things and let them experience it. And then from there you might stay as a user you might join other ones from our marketplace or you might go and become a creator yourself and just hit the new WAP button the top right and create a Molly's podcast and on there you're just bringing your audience you're sending out links to your friends you're using WAP to drive our your own distribution so you can just focus on what you do best and that's it so you mentioned that you originally started as a sneaker bot went to a marketplace now it's its own platform could you break down the entirety of the different products you're offering yeah i think to break down the products we offer it's very the best way to do it is to explain how we got here so like i said back in february of 2021 we started building this platform to buy and sell software that's essentially what it was so you can come on you can pay 30 a month and once you bought you get access to a downloadable zip file and a license key and that was it and then everyone was selling the software that we called sneaker bots but no one knew how to use a sneaker bot it's like what is it it's some software made by some 16 year old developer in their basement making 100 grand a month but like they're not they don't know how to use it so what they did is they sold alongside their software a discord community a community of all their members where people would share success stories and how-to tutorials and they could ask questions with tickets and all this stuff so our creators the software creators will come to us we're like we're like hello can you please allow us to sell discord we're like okay fine so we built a quick discord integration when you pay to get out of discord server when you stop paying you got removed and so on and from there that really opened up our market to non-developers now people that were selling just like information or a fitness community or teaching people how to flip sports cards or um nfts were getting kind of big at that time we now expanded our entire market to people that were selling access to discord servers and communities and what is a discord server.

11:38It's a big chat room. It's a member directory. It's maybe some long winded posts with a bunch of crazy reactions and they can go live and so on. So we built that out. And at this point, we just kept listening to our users. They were like, Hey, I want to sell access to a Telegram group. We're like, okay, fine. Here's a Telegram integration. They're like, I want to sell access to a multi-step course. We're like, fine. Here's a course integration. I want to sell access to files here web apps here and they kept asking us all this stuff um and like our secret sauce is we just kept building it like whenever anyone asked we just kept building so i would say about a year or two years progressed we were processing about let's say 30 million a month and we kind of took a pretty big stance that the user experience kind of sucks like you're coming to wop we're driving traffic we have this beautiful one click checkout with eight different payment options you can use crypto card whatever it is but as soon as you buy now you get kicked off platform to discord and where you now need to onboard a discord verify your email do like eight captchas understand how discord works just to see something just to see what you're trying to buy just to like maybe people are selling um teaching you how to sports bet and you want to know how to use a unit system and i need to go through all these freaking hoops just to figure that out we're like okay the user experience is kind of bad.

12:54As a company, we don't really, we can't do much because all the consumption is done off platform. So we kind of took a step back and we're like, okay, let's build out our own way to provide on platform consumption. How do people chat on WAP? How do they go live on WAP? How do they engage their customers on WAP? So what we did was we took a pretty big bet, arguably too big of a bet at that time, but it's coming back now, is we've kind of built, invested in this developer ecosystem, a way where developers can come on and they can build, we call them experiences for our creators. They can build out chat.

13:26They can build out games to engage your customers. They can build out ways to solicit feedback from your biggest customers. So we kind of went deep on this around October, 2023. And we got a little over our skis, but we released something called WAP3 at this point. There's actually a crazy Apple keynote that Steven and Cam did. And over the next year, all through 2024, we kept working on our apps platform. As a result, we by accidently built chat. Like we were just kind of messing around on a weekend project and all of a sudden we now had chat. We kind of built this forums. And over the last year, we've really been working on getting a really nice chat up to par with iMessage and Discord and Telegram, the experiences you imagine, making it so you can drop a forum post and everyone gets notified and you can do a bunch of reactions.

14:12We've had people drop a post like at 10 ,000 people in a second and kind of doing that. We've built out live streaming because people want to go live to their members. That's what they do on Discord. They might host a webinar or live coaching or stuff like this. So we went really deep on this on-platform consumption. We did it through the lens of developers. We built out a lot of these tools internally. And now here we are about a year and a half later after starting that endeavor, and we have almost all of our organic creators, our signups are using our on-platform apps. We've moved a lot of our traffic over from people that are selling on external integrations onto WAP.

14:47we are finding ways to really give you more data as a creator like who's your most active engaged customer who's chatting the most who's making the most money in your WAP how do you solicit feedback you can now automate DMs so after someone joins it's like hey welcome to sorcery I would love to have you here's what sorcery does if you want to DM me you can DM me we built on platform DMs and all this stuff really to make the user experience better and also give you or whoever the creator is, more building blocks to build something fantastic. So that's kind of how the product progressed and how we kind of got here.

15:21And you've reached massive scale in a very short amount of period of time. So one of our sponsors, Brex, is all about performance. They're the modern finance stack. They help you spend smarter and ship faster. You're shipping incredibly fast and your metrics are incredible. I'd love to go deeper into your metrics. How do you look at that from a high level? and what are you really kind of observing to determine success? So at our core, once again, our mission is to make people money. So our top line metric that we really care about is how much are our users and our creators earning through our platform?

15:55So like I said, last month was about 108 million, but that's cute, but like, are we really making a difference? And one of the biggest things that we look at is how many people are earning on our platform? How many earners are we having each month? How many people are coming to WAP and actually seeing a difference made in their lives? I think last month we were up to about 28 ,000 earners, which is an all-time high every month is getting obviously an all-time high after each one. And one of the biggest things that we're trying to do that we deem success is when you come to the platform, are you earning in your first seven days?

16:27We know that if you come to a WAP and you earn your first seven days, you are 3x likely, you are 3x more retentive to stay than someone who does not. So it's in our best interest to from the moment you sign up to get you earning money. We've been able to do that very efficiently lately through content rewards because we have thousands of probably over a million dollars in budgets now that are waiting to be depleted from brands that are either offering you money to clip their content or create native UGC for them. What's really cool with content rewards is it's very democratized. So you can be someone from Northern Africa that has a relatively modest internet connection and a low computer and in a poor region, but you can come on to WAP and you can actually find some aspect of financial freedom because you can come on.

17:18We have companies like Windsurf, for example, that are offering real cash for people to come on and tell them how they vibe-coded their software. And I've seen people from like Nigeria come on and make UGC content about how they vibe-coded their app through Windsurf and have now made over two, three,$400 just through that one video. So the way it works is they make their UGC, they put it on TikTok or Instagram or Twitter, whatever the creator is looking for, and they just submit a link and we automatically pay them out for the reviews. So there's no like back and forth as if you're working with an influencer.

17:52There's no hassle. All they do is they put in a budget and then we forget to plead and it gets depleted by real people. And you spoke about growth. Where we're seeing the most growth right now is because our goal is to make earners and it's so easy to earn with modern day content rewards, we are seeing a massive explosion in global growth. Sure, WAP works well for US and Western world capitalists that are used to this, but it's even more powerful for the global economy, people that don't have an easy way to earn income where they're at. So we've seen massive growth both from platforms like India and Pakistan and Morocco and South Africa, which is really, really cool because our mission is actually becoming something that's actually important now.

18:40How are you reaching those countries? How did you end up going global? Funny enough, we put most of our marketing budget into content rewards. We've spent over, I don't know,$150 ,000,$200 ,000 to generate probably over 200 million views by now telling people that they can earn on our platform using Content Awards. So we use our own – one of our keys to success, and I credit Steven, our CEO, to this, we dog food our product more than – or at least we try to more than anyone else out there. What does that actually mean? I told you about a year and a half ago we built our own chat by accident. We actually stumbled into it.

19:22But what did that mean for us? We were on Slack at the time and Stephen was like, nope, we are moving to WAP. And that was crazy because the first three months, like we couldn't even communicate. Like our chat was so bad, it was broken. But that started like we don't have Slack anymore. All of our communication company-wide is on WAP. There's a WAP engineering WAP. There's an internal WAP. There's a growth WAP. That kind of speaks to the level of which we're trying to do dogfooding. So you asked how we're going global. We put all of our money into content reward because we know that product works.

19:53And as a result, we are spreading the message that, hey, come on to WAP and we'll help you earn. And we have our own funnels. We have WAP Clips or WAP University, our own maintained WAPs that have thousands and thousands of people that we sponsor and we run that teach you how to make a living on WAP and how to earn on WAP, how to be power users. And we've really gone on to that. So I would say that's one of the big things. And yeah, we can talk about brand a bit later, but I would say a brand is also a big one. Would you say you have an infinite TAM now? I would say our TAM is 8 billion plus. Wow, that's great.

20:27The entire world, which is – it's funny because I told you how our platform progressed. And I think I'd be lying if I sat here and I said we had this vision four years ago. I think visions are kind of bullshit, especially from early stage startups. Like no one's actually sitting there. If they're sitting there before they have any users or like, I want to change the world and like, here's my crazy vision, they're probably not going to do anything. I think the most successful startups just kept iterating and iterating and iterating and building lists to their customers, which is one of our early success stories.

20:57But you asked about TAM, and I say that also because when we were first starting back in April 2021, we actually applied to Y Combinator to try to get funding. And they denied us because they told us our market was too small, which in their eyes, like, fine. We were just a sneaker about rental marketplace. But as we've progressed, our TAM is now$8 billion plus. Can we get the entire world on WAP transacting, consuming, being social? It's a way more, it's so much of a better way to spend your time. Like we believe that, we say internally that financial like freedom, like why does it actually matter to make money?

21:42It's really because you get your freedom and time back. And time is your most valuable resource in life. And I've seen, like I've been growing up building startups my whole life. You're building your own thing right now. You might work harder than your average nine to five, but that's because you're so much more passionate about it. And as a result, like you find people that are like you, you spend more time on it, you're way more invested and you're probably gonna make more money because of it. We kind of believe that everyone should be that way. I mean, we're in New York. You see how many people hate their job here, right?

22:10Like they're in investment banking. They can't wait to get out of five and go hit the bars and get shit face to come back the next day. We don't really relate to that. Like our team doesn't really relate to that. We grew up starting our own shit. Like we sell our own things and we want to spread that to the world and we believe we really can. I want to talk about the technical sides of this. You are the CTO. I'm sure there's like incredible engineering challenges, but you're also scaling incredibly fast and maybe breaking a lot of things. So like what's the process for you and how have you started to scale up the product suite?

22:44Yeah. So that's a great question. So building these products is no joke. Like we have many many startups within WAP almost like our live streaming feature is a feature yet you have platforms like Twitch and kick that are built off that are Instagram live we have chat which is a feature but you see platforms like discord and telegram and Facebook messenger and whatsapp whatever we have other things like our notifications like we send out over 40 million notifications a day sometimes so these are no joke systems and we've started to really realize the difficulty and honestly the awesomeness like we love solving problems in scaling some of these problems so payments i would say is the first major challenge we did how can we make it so people can come on they can drop their link to a twitter following of 100 000 plus and the website not go down that was a challenge the website has gone down several several times um but we've been able to kind of progress that and now we can handle flash drops of over a million or two million even at one point.

23:48I would say over the last year, we've gone into some crazy technical challenges with how to build chat. We've built chat four times now. We've built our notification system three times now. We've built our real-time feed where we serve you content three times now. We're starting a new one next week. This kind of all speaks to the problems that we have to solve, but it also speaks to our team. Like we're not the team that's gonna sit back and whiteboard for four weeks and figure out like what is the ideal solution? because frankly, like you don't know the ideal solution until you try it. So the way we kind of operate is we move really fast.

24:21We ship a lot of code to production. It has its drawbacks. Like some things might be broken sometimes, but we care a lot. We work hard. We'll fix them very quickly. And I would say we kind of took that mentality from early on and now we're supplying that to bigger and better challenges. How's the team structured on the engineering side? Our team is about 15 to 20 people, a mixture of full stack web ios android engineers we don't really believe in kind of middle management or rewarding staff engineers with scope and junior engineers need to do tickets it's very much a come and take it mentality we encourage all of our engineers to be very product focused like we're not going to spoon feed you prds or sometimes even designs which we're working on but we expect all of our engineers be able to take a project a product an idea and turn it into something beautiful and that's why we also really push on people to be be a creator be a user of wop so you can really put yourself in those shoes and our team actually company-wide i would say over 50 60 of our team we found on the internet um we found them on discord servers they were early users um they were we dm'd them on twitter and said hey move across the country like i can name of our founding 10 team one person was a customer support specialist that dropped out of school at Duke two people moved across the world from Australia someone moved across the world from Spain and like they didn't go through traditional interview technical interviews they were like you get our product you get our customer you want to build it come along and we've kind of really really tried to embody that mindset so to answer your question directly our team's about 15 to 20 to make sure of a bunch of different types of engineers at all different specialties and anyone could come in and lead a project at any given time we reward success and so on do you have a platform team?

26:06We are actually just building out our platform team now. So this has been my focus for the past one or two months. I'm obsessed with platform right now. I'm obsessed. My goal as our CTO is to make it so our engineers can ship at the highest quality possible, at the highest velocity possible. People think quality and velocity is a trade-off. I think they're both requirements. So with the platform team, the goal there is to empower our developers through that. So we're investing heavy into local development. We're investing heavy into our monorepo. We're help investing heavy into AI and agents to come in and be able to one-shot features or help onboard new engineers, making it easy so you can push to production within five minutes.

26:45And if it fails, it rolls back or you can roll it back. You have the tools you need. So we're currently building on our platform team. We're trying to hire like the best, the best people that know what it's like to build out tooling for other developers, care deeply about developer experience and patterns and naming and all these type stuff. And this is kind of currently what i'm obsessed with i can't help but catch you mentioned ai and this is a really hot topic but how do you think about the future of engineering with ai coming in vibe coding yeah automating everything what's your view i love ai and i've seen it's a very polarizing kind of place to be in engineering because some people think they're better than ai which is naive some people are intimidated by it and some people rely on way too heavy so here's what my takes are i have found that ai i think i'm a good engineer i've been doing it for a long time i always have room to grow but like i like to think of myself as like i was a pre-ai engineer so i understand kind of how to build systems it makes someone like me 5 10x more productive because i I can just go back and forth with a robot and like explain my ideas and watch it kind of get built out.

28:04That's the same way I used to work with maybe more junior engineers on our team, but it's way more laborious on me when I have to explain tasks to someone and they might not get it and I need to go back and forth. And with a robot, if you point me in the right direction and I can break down a problem well, they're going to do it pretty damn well. So I have found that it has enabled me to ship five to 10 X faster. I have found that AI at first I thought was just a tool. I thought it was just like the tab button you hit to write your code. I've learned recently it's way more than that. It's a, it's another, it's almost like, it's almost like another person in your organization.

28:39Like they can answer questions for you. They can look up information. They can use your code base. So how we've been leveraging internally is we built like this monorepo, which basically takes all of our code and puts it into one big folder. And we added something called Claude Code and some other agents and now engineers are coming on and in their first day they're able to solve tickets or solve problems that would have taken weeks or would have taken time from other engineers on the team to answer questions so i can say hey how do i someone yesterday was like i didn't know how to end a live stream in the code because they were trying to make it so when you end a live stream it's a notification so they just went to cloud code the robot and they're like how do i end the live stream and within five minutes they had a pr up that's incredibly powerful so i'm banking on that.

29:21And then my other hot take with AI, as you asked, is I don't think there's a world really for junior engineers at big organizations. I think there's a major split here on what should happen is if you're a junior engineer, you should not be trying to get an engineering role at a company. You should be building your own software right now. You should be using the millions of tools on the internet that allow you to make designs and code. And you should just trying to build as much shit as possible because it's never been easy easier and that's also like why i'm so excited about wap is because we also are building out a developer platform that allows anyone to basically build out an agent or build out a sas tool and easily throw a paywall on top of it easily make a community of your members but why that's important is the fact that i i really enjoy working with very talented engineers right now because they can take ai and they can split up their project and get it done in five to ten extra time i don't see a world and where big organizations really need junior engineers as much to do it that's why my advice to junior engineers is don't even try to fit that mold just go build your own shit and you'll probably be way more successful than you ever could have been being a junior engineer at an organization wow so what does that mean in terms of org structure going forward what percentage of the engineering team do you think will be ai uh that's a funny question i want to keep our team lean like i think we can 10x our volume go up to i don't know 10 billion plus gmv run rate with 10 to 20 people like i don't think you need many more people at this stage i think the days of tech companies scaling to thousands of engineers are behind us because what i found is a smaller more compact team of very talented people that understand how to break down problems and naturally use AI to go faster, they won't need that much help.

31:12So I'm pretty bullish on a very lean, very smart staff plus level team of engineers that have solved these problems at scale that want to do it again and really keeping that team small. So that's kind of how we operate internally. I wish we could do, we always want to do more with less, and especially in a modern day with the modern day tools we have. Yeah. I ask that because the, like one of the leading questions that a lot of people have is, can you build a billion dollar business with one person and just using lots of tools? You already have that and you're doing 1.2 billion in GMV run rate.

31:48You've hit 5 billion in creator views. Like you're at massive scale, like instead of scaling up team, like scaling, you're going to be scaling back. That's the goal. Yeah. Wow. Yeah. It's just easier to do more with less, less contact sharing, less communication overhead, less. I mean, it's beautiful, right? Like, and especially if you can, if the way that AI is progressing, the way that agents are progressing, how much smarter they're getting, how much more context you can feed them, how many tools are being built to write code, review code, all these things. if you can leverage that correctly and you can bank on making it so every time you have a feature idea if you could say you don't stop can i prompt this out and see where it gets stuck and then instead of just fixing it solving the root problem so like giving it a prompt to solve it we can get pretty damn far and then like my end goal is like what if i can be like hey why is the website slow and it can query our performance dashboards and look at slow queries and look at watch sessions of users and be like you know what it's slow because this pr was merged two weeks ago that adds this amount of time to loading this page and it's causing these users in india to have a 50 degraded experience like that's not that far off it's just making systems talk to each other so that's why i kind of think you can do more with less these days so there was a big migration not too long ago toms of rome one of sorcery's favorite fans had a question about this how did you do it and what takeaways do you have for other teams going through transitions and migrations like this?

33:25For context, we recently moved our database provider. So we run a database called Postgres and we move it to planet scale. Why is that important? Well, first of all, we have a fuck ton of data. It was like 800 gigabytes of data. Any downtime for our company means lost dollars for all of our creators. Or even if it's not lost dollars, it's now all the customers that just bought something can't access what they bought. So downtime wasn't really an option here. So I'll kind of tell you how we did it. And then I'll kind of tell you my philosophy on how I think migrations are. Actually, I'm gonna start the latter.

33:56Let me tell you about our philosophy. We don't really believe in refactoring over rebuilding. What that means is like a lot of companies might always try to refactor their code and make it nicer and prettier and make little tweaks. We don't do that. We're like, you know what? Like this system's here. It was built upon assumptions that were made six months ago. We have massive new revelations. let's just throw that out and rebuild it again so we've always been on the mindset of rebuild over refactor which has enabled us to kind of get this mindset of like running through a wall like you kind of it's almost like we've adopted a little bit of this yolo mindset for over the past few years we're like we kind of just got to send it and what that really means for this example is our database was it was breaking like we we came back from december break um so like january 2025 and we had incidents every single day the website was going down every single day um i remember i was actually it was like saturday night 9 p.m i was on a date and i got a call like the website's down i had to leave the date midway to like go fix this yeah it was crazy um and after that moment when like there's something you have to give and the database for us is the most complex piece of software there is like that's the full backstop like if the database isn't working the rest of your shit isn't working so we wanted to move to a more scalable reliable platform so we chose to go to planet scale um from like a technical standpoint that means like you're moving from one database which was called postgres to another one called my sequel which is like not one-to-one you can't just like export import you have to do a lot of crazy transformations midway so we kind of worked on it over the so we had to basically make it so we can move between two different databases between like postgres and my sequel For anyone that doesn't know, it's like moving data from your Windows to your Mac computer.

35:42We need to do it without any downtime, without any flawless, with fully flawlessness, while also allowing your team to continue to ship code at the exact same velocity, if not higher than you want. So how we did it is our backend is a Ruby on Rails code base, which is another hotly contested debate, but I personally love Ruby on Rails. And what that allowed for us is it was very easy for us to swap out basically Postgres to MySQL. There's a thing called Active Record, which is basically the layer on how you query your database. And what made that nice for us is we basically spent the first month making our code base run on two different databases.

36:17So in production on WAP.com, you can either plug in a WAP Postgres database or a MySQL database, and the code would have worked. So we spent our first month basically refactoring everything, making it so all of our tests passed with MySQL. Once that was done and that first chunk was over, the next step was basically getting all of our data to flow from this Postgres database to this PlanetScale database. Like I said, it's two different types of databases in Postgres and MySQL. So we needed to basically build this intermediate layer that would every time like a user signed up on our website or a payment was processed, it would take this payment.

36:52It would take this row in the database and push it through to the other database. that we basically spent like a month on that getting both of our databases stood up fully replicated all the data in there because you obviously can't have any data loss otherwise like i mean it's obvious you can't people be like where's my money um so we spent about a month and a half there that was pretty very very challenging to get working and once we got that we basically tried to do the migration like three times and the first two times we had to cancel it because we found issues but the third time was about 6 a.m on a monday um we got everyone actually it's like 5 a.m everyone got to the office and we basically conduct like a high level almost feels like you're defusing a bomb there's like a procedure you're like i need you to do this turning this off scaling this up we have an issue um but we ended up doing it from like 6 to 6 30 it went pretty smoothly there was no downtime we had no issues to the website everyone kept processing um and And that's how we moved about 800 gigabytes of data with zero downtime from Postgres to PlanetScale.