Building a developer brand and simple API for email, with Zeno Rocha from Resend

Well, you're definitely in good company here. I'm Mm-hmm to chat with you about it as well. Why don't we start here? Why don't you tell me just a little bit about yourself how you got to where you are and maybe your story leading up to resend and we'll dive in from there. I. Yeah, so I'm originally from Brazil. Nowadays I live here in San Francisco with my wife and my 2-year-old daughter. And I got into programming just like most of the people in my CS class.

We just love playing games on our computer, so we're like, oh yeah, maybe that would be cool to, to do for work. And I got really into web development. Quite soon. I remember starting with flash actually, so action script three was my first language and I created all sorts of crazy things and we can even put one in the show notes. But it's this like particle thing where. This was back in, I gotta send you this because I think you, you appreciate it.

So back in 2010 or 2009, I guess we, we built this thing and then we sent to Google just to see they had like this HTML five gallery. So there's like a button on the left where you can click and it generates the particles. And this was my first time. Using JavaScript to create the weird things that I would, I would do in Flash.

And I was ama like, and the source, you can see everything on, on GitHub the source code is available and I was really amazed at the power of this new thing that was coming up called HTML five. So I. quickly transitioned from Flash to H ML five and my perspective when I started was okay, like it, and I always had this thing like, oh, I wanna be the best at whatever I do, right? So I'm like, oh, to become the best PHP programmer, it's gonna be so hard. Or the best Java programmer.

It's gonna be so hard because like there are people that are doing this for like 20 years already, but HL five is this new thing that's coming up. Maybe if I run really hard, if I like really dive into this thing, then maybe I can get to the finish line in front of others. Right? Turns out there's no finish line. You're always learning. There's like all sorts of things. Like there's no such like program is not a. Infinite game, right. Yeah, it is a infinite game.

In fact it's not a, a finite game, so you are always learning and always reinventing yourself as a programmer. But I've, yeah, I just always enjoyed building stuff, man. Yeah.

And literally the, the thing, the catalyst that set off the rest of my career as like a front end developer was I was in a room when someone walked in and asked, does anyone know what jQuery is? yeah, I've used that. And then I. Was fully, like, sucked into a gigantic project rebuilding dell.com from like the third week of my job. and the from there, like, you know, jQuery and mood tools and handlebars and all these things. And it's just like, it feels like a, a mess of things since then.

And you're absolutely right that like, it ends, it doesn't get

That was basically like deleting all my code and saying, oh, the title of the PR was something along the lines of like, delete everything. That's the only way that you can build this up again. And the the, the gif was like, actually like deleting everything. So I, I remember feeling so sad and so disappointed that I, that I've created something that maybe wasn't good for this. The guru there.

And then reaching out to people that I knew were doing interesting things like adios money and alongside with him, we like recreated that project. And then it got much better and, and much bigger later. But just goes to show that how yeah, how interesting the community can be sometimes but also. How much you've learned from from it, which is the most interesting piece for me.

So then when this was released in JavaScript, I was so excited. I. And I think the PU too has like 30,000 GitHub stars. So I've done like all sorts of side projects. I've always loved the idea of side projects. That's another thing. Just I, I've always felt that with your nine to five job, there's always a ceiling to. What I would be able to learn, and maybe that's because the, that company just uses one language, right? Or they use one particular technology stack.

So side projects, they always have been this outlet for creativity. And I've done, I, I did a little bit of product management. I, I did a little bit of, of a vp role in other places. I was a CPO at another company, but I've, I've never really. I've never really stopped loving JavaScript and open source and programming. Sure. Yeah. Okay. Dracula by the way, is everywhere. I feel like I bump into it in Mm-Hmm I expect half the time.

And that's one of things that like Gets your name out into the open, like a lot of good side projects, like people start to recognize you and a known quantity to people who've never truly actually you.

And I felt that frustration multiple times throughout my career. Like sometimes dealing with like really big customers like McDonald's, having emails going to span, I'm like, oh my gosh. Like how do we fix this problem? And just me integrating a side project so we. We just like try to come up with this idea of like, okay, what if email sending was better? Because it really feels like email sending as a whole industry got stuck in 1995 or something.

Like, to do a, a rounded corner is extremely difficult still. So we're like, no, like may like this, this whole idea of my email not rendering the same. On Outlook and Gmail and Yahoo Mayo and superhuman like that. Just for me, like in 2023 that didn't, like, my brain wouldn't process that idea. I'm like, we gotta fix this.

So we started building like an API for sending emails, but then as we were building, we're like, oh, the API is super cool, but there's this problem right before sending an email, which is the actual email template. Like putting together that HTML and people don't wanna use tables they don't like it is just like so outdated. But what are they using now? They're using React. They're using XJS, they're using Tailwind.

So we came up with this idea of a project called React email, or you could use that modern stack to create your emails so you can have. Type definition using type script and JSX. So no more like, hey, empty space, comma, right? Like that type of stuff. And that was super helpful and we released React email in December, 2022. And then in January, 2023 is when we announced recent.

You know, for example, I don't know if you think back far enough, you probably don't want an email with a Marques scrolling across it or like blinking images or dynamic stuff loading. So it's just a different problem to solve. And, and I have tried to send good looking emails in the days long before React email existed and essentially every time gave up. It was just like, good enough and you know, move forward from there. And this is something that your team attacked head on, right?

And, and I feel like that's a really interesting inroads to then the API, which becomes interesting too and was a really good hook to get people interested. So, so you launched React to email and then, I mean, it was, I guess I just a matter of weeks before resend launched.

That's the first step of the journey of sending an email, right? You get. This beautiful Figma file from the designer. And when you start implementing, you're like, wow, this is so much different than what I'm used to in the web. So we wanted to fill that gap and the answer was react email. But after the email was done, I need a way to send it. And again, thinking about primarily that no JS stack, how does that fit into that world?

And from a DX perspective I feel like no one was really giving the love that email deserved. So like this thing is so crucial for the web, but no one is investing time to build a, a proper API or to have a open API spec to have a postman collection. Like everything that surrounds, like building and distributing an API. So that's where recent came up about like. Once you have the email, you need a way to send it.

And it's more than sending, because the moment you send, now you have to worry about was the email delivered or not? Did it bounce, was it marked as spam? Where is it? Is it on the primary box? Is it on the, the spam folder? So there's this whole observability on top of email that is. Taken for granted. So that's why we were like, no, we gotta build a dashboard. We gotta build the stripe of email. We gotta build the of email. And that was the whole exercise from day one. How can we do that?

So it's gonna require amazing docs, it's gonna require a lot of attention to detail. It's gonna require speed. Like you can't just navigate on that dashboard and it's extremely slow, like. Other competitors are. So we were just like trying to tick those boxes to really make something that people cared. And I feel like there's something when you go to a, a website and like you understand if, if there are s STKs for all the languages, you're like, oh, cool.

Like you start gaining trust from developers by, by putting those building blocks in place.

And often one of the discussions I have with people that is hardest to convince people is that the story is as important as the product may be more important than the product in your early days. And that even if you're building something where the first releases are flawed. Hard to Your early adopters are the people who love the product because they love you or they love your team, or they love the brand and the story that you're telling. And that's a hard thing to get across.

So how, how did you land on that as your path for getting this out into the wild? Was this something you had gained along the way?

And after I run it, like it takes a long time from the moment I sign up and get the API key to the moment I see the value of the product. Just getting to those conclusions myself as a user and then trying to avoid that as much as possible, like.

One thing in the email industry is that when you sign up for all these other services, you got a first request to access and they put you on this verification list for three days, and then they tell you, oh yeah, now you're approved As a developer, I hate that with my whole heart. Like that's so against everything that I believe that level of friction.

Like, I wanna be able to experience the API in 30 seconds because that's the time I have on a Saturday afternoon while my, my daughter is running around. And if I like that experience, then maybe I'll bring it over at work Monday morning, I'll talk to my coworker and say, guys, you gotta check out this API. Look how cool it is. So there's definitely, I. For me, my angle is I'm not building a company, I'm not building a team. I'm building a brand. And to sustain that brand, I need a stellar team.

And to sustain that team, I need a stellar company, a stellar machine, to be able to sustain that. So. Not, not many people think like that because building a brand takes 10 years, right? Or 15 years. It's a long time. But I, I definitely spend a lot of time thinking about this rather than, oh yeah, what's the next feature we, we need to ship.

Like there are many people that bought Teslas because, you know, they had a, an association with Elon Musk four years ago. Like a lot of people love for sell because they see Guillermo very involved. So there's, there's these things that play out and I definitely feel like the personal branch. That like I, I, I'm always thinking how does my personal brand follows recent and how does one support the other? If you, there's some silly things like, oh, you go to the website of that person, right?

And then they have a personal website live. That's the first thing do, do they have or or they don't. And then you go, it's all of like outdated blog posts and stuff like that. So you're like, oh, okay. Or it's extremely ugly and it doesn't match. The values that so resend is very polished, right? And then you go to my personal website and it's super like me. It's just there.

Like no, it needs to communicate that because then it's a cohesive message that, hey, if you care about details, this will be for you. If you don't, that's totally fine. There are all these other competitors you can do, you can go and use. But if that's something that you, that's important to you. Here's another person that also cares just much as you do.

We've, we've had, I've had investors in, in our company mention blog posts of mine from years ago,

We sometimes go to other APIs and we see like a lot of information and, and then it's hard to translate that to the different use cases that users have. So for us, we're like. How can we make an API to send just one email? Just the easiest interface possible. And then there are more advanced use cases later. Like, oh, I wanna replace a template variable with this content. Okay, let's not care about that right now. Oh, I wanna send 100 emails in one API call versus 100 API calls for 100 emails.

Okay, let's not care about that. So there's this tension when you're building of. You gotta build fast and you've gotta build high quality, right? And then what people tend to do is just, they're like, oh, I can't have both, so I'm just gonna pick one. And if you pick just quality, then it's gonna take two years for you to ship your API. If you just pick speed now, you are gonna introduce tons of breaking changes in the first year that you launched, right? And we know how hard it is for.

To, to migrate APIs. It's extremely difficult process. I feel like the answer is you gotta do both, but then the way you do it is you gotta cut scope. So you gotta make your API as easy as po Like maybe you're not gonna cover all use cases and that's fine, but you need the, the, that minimum lovable product of an API is is not just the functionality is everything surrounding it. And I feel like that's where most people fall apart because they're like, oh yeah, the API is running. Here it is.

Here's an endpoint. No, let me check your docs. Like, do you have all that surrounding ecosystem that people need to get started? I very much believe that we're living in a SDK first world, not an API first world. I think an API first world was what we were living in 2015. I. In 2024, we are living in SDK First World because that's how people consume APIs.

Like if I want to use, let's say Firebase, I haven't even started using, I, I, I didn't even sign up, but I already know that, oh man, if I use their SDK, I'm sure they will handle offline. Sink, and I'm sure they'll have something already, like if I'm using Stripe, I'm sure they'll have some sort of like pagination built in. Like there's all these things that you know, that you expect that it'll make your life easier.

So I don't think people start from rest purely like they start from, I, I love no js. I needed a no JS dk, I love Ruby. I needed Ruby s dk. So we are very intentional about that too. Like starting day one with a lot of coverage for, for different SDKs, but it's extremely hard to do that. There's a lot that we can dive into that one particular subject, but I think, I feel like there's a significant shift and you gotta think about everything in between.

Not only like what's the body and what's the response.

Do you do, do you have a gigantic backlog of things and are you reprioritizing them or do you have a pretty good like, roadmap ahead of you that you're building towards?

So if that's the case, then at all times we are re revisiting scope. We might come from a user interview and then we learn all these things and we start implementing. Then we're like, oh, but we gotta ship it fast, but we can't ship crappy stuff. So we're like, okay, we gotta remove this feature, remove this other feature, remove this other feature. I feel like that's a very interesting, like some people are afraid of cutting scope or they feel like they can't push back.

So it's important to foster that culture of like, yeah, saying no, like, no, we're not gonna do this for this reason, and I. We can. Yeah, totally right. It's super hard, man. Super hard.

So what I do is I, I first like, let me go to their source code. Like this is a, an opportunity where. They, it's actually available. So I can take a look and then I can see every single API call they make and I can see like, what are the parameters? How do they send those things? Like do they transform the value before sending? What is the type of transformation can we do the transformation on, on the SDK like.

Silly example, like, oh, let's say you're talking about sending attachments and I can send the whole string, or I can do the base 64 conversion already built in. Like, so there are small things that you can try to do. But yeah, we just try to get as much information as possible. And of course, most times the code is not open source, of course. And then when that's the situation, we, we try to get it like, yeah, let's screen share. Like, whatever you can show me, please do.

Because then we can make it better, right? Like we talk to a lot of our friends in the beginning of recent, like, let me understand your email implementation. What are the, the ramifications here? Is there anything that we're missing? And cover the most important 20% first and make that exceptional before getting to the, to the other 80.

You, you, you gotta need that combination. And I think developers appreciate that as well. So we try to be very. Proactive about releasing things and communicating even if there's more. But then from time to time we do an event like that where like, okay, let's just ship a bunch of things in, in that same week and really create momentum. That's something that's really important for any product that you're doing. And the whole. Scope of that week was like, how can we make the API better?

Like I mentioned that example of sending 500 or 100 emails in one API request. That's something that we shipped during that launch week because we heard from our users like, yeah, I wish I didn't have to send those 100 requests. Just one. And then you do the queue on your end. We're like, oh, okay. Yeah. So. It's one piece left less of infrastructure that you need to build. It's not easy to manage a queue. You gotta deal with that letter cues and retries and, and it's, it's painful.

So, okay, let's deal with that. But we also launched marketing emails, which was something that we, I think by now listeners already understood that I take a lot of inspiration from Stripe, for example, or, or, or other dev tools. And we're very much following that Stripe trajectory of like, Hey, you have. The rest API for like payments. And then you have stripe elements.

So those are like components that you can embed in your app and then you have Stripe checkout where it's like a whole standalone thing that you can just point to see, name to. We're very much following that. Idea of different abstraction layers. So this one abstraction layer that we introduced now is just an editor, or what you see, what you got editor, that you can send emails without needing to, you know, configure everything. Cool. So that's one piece that we're doing.

And there are others, you know I would love for you. Like no one wants to build a unsubscribe page. It's the, the most boring thing in the world. I am now looking into like, how can we build. A radix or a shed cn ui, similar solution to that, where you're just building the building blocks or the whole thing depending on, on, on what you prefer. So I think that's a another way of looking into this, like regardless of what we, we announced that week, like how do you.

Think about our, the, the different abstraction layers that your own API can empower.

And you know, marching towards more productivity for your users probably means more value and more money transparently for, for the business as well. Yeah, I, I will make sure there's a link in the show notes to the launches from recent forward. I'd imagine that was a pretty hectic week. Was it an all hands on deck situation for your team? How, how

So. It's an interesting balance, but launch weeks are, are these things that if you wanna do that, you gotta understand the impact that this will cause because it's definitely a lot of stress, a lot of attention that you get in return. But you are pushing the team extremely hard for entire week and. You gotta know the consequences of that.

Like people will be extremely tired on, on the following week, they won't be as productive or like, there's all these things that you need to take into account. So I try to be mindful of those consequences. If I'm doing it, then, okay, what are the, the trade offs

And I mean, even email lately has had new regulations passed. I think I saw something a requirement for one click on subscribe,

Like sending SMS is getting harder and harder, and I think the, the magic and the beauty of Twilio that we knew from 2013 or 12 and 11 isn't really there today. So we could explore that path and start creating APIs for sending SFS, sending push notifications, sending all these other things. But we can also just double down on what our users are telling us, which right now is like inbound emails. They want to be able to receive an email. Okay, let's, let's get that figured out.

You know, and we have a lot of challenges around scaling. We're sending dozens of millions of emails every month, and we gotta be creative. And like, whenever I, I re like, I talk to my co-founder, like, oh yeah, let's think about this new architecture. And then we have this discussion of like, oh yeah, how long do you think this is gonna last? Like, oh yeah, maybe nine months.

And then in three months we gotta revisit the whole thing because now, like we got so much more load than we were expecting. So it's an endless game. But I I, there there's some really interesting paths that we could take from here.

You are, please consider yourself more than welcome to come back and hang out anytime you have any launches or

We, we've, we've got a passionate group and if you're out there, please do give it a shot. And, and chase down Gino Zino with your feedback. Zino, thanks so much for joining. I really appreciate it, man. I hope a, a great afternoon and we'll catch you soon.