Instagram Principal Eng (IC8) On Building IG Stories, 1 Promo Per Half, Small Teams

It actually felt in some ways like Instagram was dying. This is Ryan Olson. He grew to a principal engineer or IC8 at Instagram in just a few years. One promo per half or two halves straight. What was that like? I was like, hey, I think I got misleveled and I think you should promote me from four to six in this cycle. As the lead iOS engineer on Instagram stories, he shared what it was like behind the scenes. The CTO and co-founder of Mike Krieger, he was like, you know what? don't worry about trying to A-B test this thing. Build and ship it. Do you think that the company would be better off if you just laid off half the people? Even if it's a random selection, do things go better? And frequently we felt like, yeah, maybe. In 2011, I understand that you interviewed for Facebook and you failed. Can you talk through that experience and what that was like? I thought Facebook was super cool. The time. I wanted it so bad. And I was just incredibly nervous. I remember the interviewer, he kind of asked the question. He's like, oh, you can just do it on my machine. He turned around his laptop, pushed it over to me. I put my hands on the keyboard. You could actually hear like the keys like rattling because I was shaking so bad, which just, you know, it starts this feedback loop, um, hearts pounding. And, uh, the question that he asked was actually one that ended up getting banned at Facebook, um, which I feel some, uh, somewhat better about, but, um, what was the question? It was this Angram, uh, buckets, uh, question, which, yeah, there's a very simple solution. If you know, uh, data structures, um, where you use a dictionary, and that solution didn't come to me. So I wrote this like horrendous triple nested for loop. I mean, it was definitely like, you know, the interview went quite poorly and yeah, so he, he actually cut the interview short, which as, as an interviewer, that's like something I would never do. Even if somebody is failing, like you want to, you want to try to have a good candidate experience. And so he cut it short and he was just like, look, you, you, you're just not going to be able to make it at Facebook, which was like kind of a knife to the heart. Like I, you know, I wanted this thing so bad. And then this guy was just like, yeah, you're just not good enough for this. And it actually took me like, I was really down from that for a while. It took me a while to kind of like bounce back after that. Luckily I did bounce back. I feel like I had a pretty good career at Facebook, so I managed to prove him wrong. I never got the name of the interviewer it um so i i have no idea who it was but yeah yeah well i mean that to me shows how broken the interview process is if i mean you're an incredibly high performer and if you are not getting through then clearly it's not you know testing for the right things yeah i've always i've always struggled actually with like the delete code style um interview and and uh it was something i was always interested in trying to change while at Facebook, but, um, you know, it's the interview process. There is such a, such a machine across a big company. Um, so one of the things I enjoy about having my own company is like, you know, I don't have to do those things that I don't agree with. And so after failing that Facebook interview, I understand that you went to go work at a startup named Flipboard and you were an intern, one of four interns. What was that experience like working at Flipboard? Yeah. So, um, I'll tell a little bit more of the in-between story there. Um, I, uh, a couple months after the Facebook interview interviewed, uh, with Amazon, um, and, uh, it was a phone interview. So I was actually a lot more comfortable. Um, I did well on that. I got the offer, uh, they did the kind of exploding offer. It was like, you know, I got a call on Friday. It was like, you need to tell us by Monday if you're taking this or not. Um, And I was prepared to take it. It seemed like a cool, cool opportunity. I had an investor friend. He founded Insight Partners, which is one of the big venture firms. And he had mentioned at some point in the past, he's like, you know, at some point, maybe I'll connect you with some startups. And so I just reached out to him and I said, hey, you know, I got this intern offer at Amazon. I think I'm going to take it. and he was like, don't take it yet. Like, let me connect you with some startup people. So it was on a Sunday and he sent an email connecting me with a co-founder of Flipboard, this guy, Evan Dahl. Evan called me right away. And I actually, I didn't really want to pick up the phone because I was like, oh man, I don't want to do another interview. And I kind of just want to take this job offer that I have. But I'm very glad I picked up the phone and had a great conversation with Evan. He had worked at Apple on kind of the core of iOS. He actually wrote UI View Controller, which is, you know, like the core kind of UI class that you interact with. And he'd also taught this course at Stanford called CS193P, which they made available on iTunes. and that was how I learned iOS development. I took it a different year that Evan didn't teach but we had a really good conversation about that and so after that phone call I was like yeah okay this this makes sense like this this just seems way more interesting than whatever I'm going to do at Amazon. So I went there in the summer and the company was about 50 people at that time And I think, yeah, we had like maybe four interns or so. And it was just an extremely high density of talent. Like people were all super smart. They come from these backgrounds, you know, like writing the iOS frameworks at Apple. um the one of the other interns that i sat next to he he wasn't really an intern he had worked at apple and he was actually doing a medical degree and he just needed a job for the summer he was friends with evan so um but he had written the push notifications framework and he was talking about like how he set that up for the demo at wwc and it was like running off this is like mac mini just in Moscone where they had the conference. And, um, and another story he told was, uh, he had written like the, um, kind of clock framework, the clock app, and they had a bad time zone bug one year for, I think it was around new year's or something. And they screwed up and like people's alarms didn't go off and it was like fairly catastrophic. Um, and it's like, yeah, actually making sure people's alarms go off is pretty important. So yeah, it was just a really, really good learning environment. And I learned a ton about iOS and about building product. Two of the other interns have gone on to make multi-billion dollar companies. One is Dylan Field, who founded Figma, which went public yesterday. I think it's a $50 billion market cap today. He actually, he was an intern through the first half of the summer that I was there. And then he ended his internship and he founded Figma then. So yeah, that, that it was, it was cool. People were kind of into startups, entrepreneurial. I actually really questioned whether I should go back to finish my last year of school. I'm ultimately glad that I did, but it felt like there was just this moment happening. I was like, I can't leave. I have to stay here. And if I go back to school, I'm going to miss out on something really critical. So Dylan Field, co-founder of Figma, was one of the interns. And I saw in your writing, the other intern was Devin Fincer, who's founder of OpenSea, which is also at some point it was billion dollar. I don't know the exact valuation now, but that is absurd. Were there any traits that you noticed in those people or that intern class that kind of foreshadowed that they would be that successful? They were definitely determined to start companies. It was like, that was what they wanted to do. And I think that was a bit different than what I saw, you know, from like my university classmates, people just wanted to get a job, you know, kind of like the highest aspiration was like to work at a big tech company. and both Devin and Dylan were like, you know, we want to start companies. And there was a lot of talk about ideas and what they would do. It's interesting to think back to the early days of Figma or, you know, like talking to Dylan around that time. I think the idea that they had, they weren't totally landed on the idea, but they had this one really interesting insight, which was WebGL was this new technology that was really going to enable new things on the web. And Dylan's co-founder, Evan, was a really talented engineer and was like very early to WebGL and very skilled at that. And I think they went through some different iterations of the product that weren't really tailored toward designers at all. Like Dylan's had talked about how they were a meme generator app for a week or something like that yeah i mean so i think it took them a bit to to find uh exactly that's kind of their use case and their customer base um but they they had this vision of like this technology is really going to change things and they stuck with that and uh it's it's cool cool to see how how far they came yeah it seems like while you're working at flipboard you built a a successful open source project and you know yeah some of the first commits are in 2014. I'm curious what's the story behind that and why did you open source that? So during my internship I did some work on just kind of internal debugging tools for our iOS app and it was the first time I'd really worked deeply in Objective-C and I found the language really interesting because it's a compiled language but it has this dynamic aspect has this runtime that exposes actually quite a bit of information you know while the app is running you can inspect a lot of things about um what are the classes what are the methods on those classes you can call them dynamically you can see all the state all the properties and uh so the first kind of iteration of that was just building a tool to kind of inspect the state of the app and tweak it um on demand I think we called it tweak it was FL tweaker um and I just kind of continued iterating on that um as like a side project I mean my main you know we were a startup so we had probably between three to five people on the iOS team through through most of my time there and so mostly I was building Flipboard building the app building new features. So it was kind of like a nights and weekends thing that I was working on this debugging tool. And people in the company found it really useful. It was just, yeah, a way to kind of like understand what was going on in the app under the hood. You could see all the network requests. You could see what was on the file system. I kind of asked, hey, can I open source this? I think it's like going to be useful for other people. And everyone was on board with that. So we pushed it out and, uh, yeah, it's, it's become, um, one of the more popular iOS debugging tools. I think a lot of, a lot of apps use it. It's definitely use it, uh, Facebook and Instagram. It's, it was also cool to see it, it took on kind of a life of its own. So I've been less active as a maintainer of it in the past, probably, probably really since I started at Instagram, uh, or maybe a couple of years into that. Uh, but this guy Tanner really took it over and has continued evolving it and adding features. That's awesome. I remember hearing about Flex at Instagram as well. So I wasn't aware that that was actually something you'd built before you were at the company, which is really cool. I'm curious, did building something like that and open sourcing it have some unexpected butterfly effect on your career in some way? In some ways it did. In some ways I thought maybe it would have more impact than it did. So an example of that is when I came to interview at Instagram, I had already released this open source. And so I was like, you know, here's my code. You want to see what I can do? Like, this is a thing that I built that you can go look at. And none of my interviewers had any interest in looking at it. You know, they're like, we want to know if you can do this phone number sorting thing. And so that was like pretty frustrating to me. I was like, you know, you can see my work here and you choosing not to look at it. Other people who are kind of slightly outside of the interview loop, but part of the recruiting process, like Jonathan Dan was this guy who worked at Facebook on the iOS team. He kind of I think he actually did the file new project in Xcode for the Facebook iOS app. So he'd kind of started that. And he was the one who got me to interview in the first place. And that was all through him seeing this tool and what it could do. And yeah, and I think he advocated in the candidate review or whatever, because I think I had kind of a mixed interview loop. Yeah, I mentioned before, like I was very nervous in my internship interview. Um, and, uh, I did something for my full-time interview that, um, it's maybe helpful to some people out there, uh, that get nervous in interview situations. Um, there's something called beta blockers, which, uh, they, they block adrenaline. So when you have kind of the physical effects of being nervous, like pounding heart or, you know, sweaty palms, that type of thing, you start to feel those. and it can create this feedback loop where you kind of spiral down and really starts to like block your performance. And so what a beta blocker does is it just stops that adrenaline from firing. And so you can kind of stay calm and never end up in that spiral. So I got a prescription for that for my interview, interview performance enhancing drugs, I guess. And I took that and it was very helpful for me. So I was able to stay calm. I think performers sometimes use this, you know, comedians or whatever, just to kind of like, you know, be able to stay themselves and not end up in this nervous loop. So I did that. But, you know, I was still like kind of elite code style interview loop, which I don't do that well at. So I think I had like a few like absolute confidence hire interviews and probably a few no hires. Somehow I managed to get through. But I think that also contributed to that came into to Facebook as IC for I guess one level up from new grad engineer. And I would say like I was under leveled higher. I think, you know, part of that was it came from my interview loop. And when you use the beta blockers, I'm curious, is it basically completely removed the nerves component for you or is it just kind of like helps a little bit? I've only used them a handful of times, but pretty much completely, you know, like no noticeable kind of nervous effects. I mean, I think you can still be in your head for sure about what you're saying, but it takes away a lot of those physical effects. so no pounding hard or sweating, that type of thing. After you passed the interview for full-time, you started working on the Instagram team, working on iOS, and I'm curious what the environment was like at the time. What was the size of the team? Yes, I came in, and I think there were about 10 iOS engineers working on Instagram then, which was actually the team size had come down. So the way the story has been told to me is there were maybe about 20 iOS engineers and there was kind of this internal war over the direction of the code base, like the infrastructure, how the Instagram iOS app would be built. And there was a really talented iOS engineer, Scott Goodson, who was managing the team. And he had made this framework for the Facebook paper app called Async Display Kit. And it was kind of a different approach towards how you manage writing iOS. And I guess there were kind of like warring factions. Like some people were very pro Async Display Kit. And other people were like, no, we should just stick to vanilla iOS, how Apple builds apps. The way it was told to me is like these two factions, like they fought each other and they destroyed each side, destroyed each other. Everyone just left and nobody won. So, so I came into kind of like this, this team. I mean, I got, I feel like not many engineers had been there more than, you know, six months to a year. So it was like a pretty new team. Um, and because of that, there was actually like a lot of low hanging fruit, um, a lot of places to, to have impact. One of the first things I did was just to put the app into, uh, a tool called the time profiler. Most iOS developers will, will know this in Xcode, um, or in instruments, um, and just looked at, you know, okay, what is happening on cold start? And there was a bunch in, in kind of the startup path. there was a bunch of work happening for the profile tab, which many people will never tap into. And if they do, you can kind of do that work at that time. So it was like a very easy win to shave 20% off our cold start time just by deferring that work until later. Another example was we had this networking library, AF networking, and there's something, I mean, I guess most programmers are familiar with assertions. In iOS, the way assertions are typically handled is it's something that you run when the app is in debug mode to kind of crash and alert you to a problem. But you kind of have fallback behavior for production and you don't actually crash the app. And so this networking library was being built with assertions on and it was crashing the app for like benign failures, like a network failure, right? Where the app could recover, we were just crashing. And so that cut our crash rate by 80%. And, you know, it was just like a one line change, just NS block assertions, but it was a pretty significant impact on the functioning of the app. And we, Instagram had this cool tradition where every week they would give what was called the ax. And it wasn't getting fired, which it sounds like if you got the axe, it was like you did something of outsized impact. And it was actually across all functions. Didn't have to be engineering, could be product design, whatever. And so for that 80% crash reduction, I won the axe. And it was this big physical axe that you got to carry around for a week. At first, I thought you got to take like, I was like, oh, I get this axe. Yeah. I had it at my desk for a week and I was like, oh no, there's just one axe. Somebody else is going to get it next week. But yeah. I remember the axe. What's the story behind that? Is that something that Instagram founders like had done? At some point they had been asked by like GQ or some men's magazine to do like a holiday gift guide. And I think it just kind of came up with like random gifts. And one of them was like, they found this service that was making like bespoke axes. You could get this axe. And then one of their investors like read that article and then they sent them this huge axe. And apparently Facebook security was like pretty unhappy about this axe, this weapon in the office. So they made them put it on a plaque. It was like mounted to the plaque. You couldn't take it off. but it was yeah it was a cool tradition you know later in my time there they ended the axe as like a very intentional thing they were like I guess they thought people were feeling excluded because it was like a weapon that was being given or something like that but I was kind of sad when that happened because it felt like you know at that time the founders were gone and it was like this piece of like early Instagram culture that they were kind of like sweeping under the rug so So I guess going into your first major project after all those low-hanging fruit, I understand there was a big redesign of the Instagram app called Whiteout. Can you talk about the story behind that project? Yeah, it was a couple months into my time there and I kind of got word that we were working on a new icon, which ended up being very controversial. And as part of that, we were going to do like this kind of redesign of the app, a big visual refresh. and I don't remember exactly what I did, but I remember being like, I have to work on this thing. This is like, this sounds so cool. This is what I want to do. And I, you know, I told my manager, I was like meeting the designers that were working on it. Um, so I basically just maneuvered myself in, into, to working on this project and, uh, the, the main designer on it, Joy Vincent, um, who was just like, yeah, just an incredible talent and the nicest guy ever really sad. He passed away while we were at Instagram. Um, but, um, I sat in a war room, um, you know, just like a conference room basically, uh, with him for, I think probably two or three months. And just every day we'd be like, okay, new screen and app, we'd go look at it. you'd be like, all right, I want to do this, this, this, and I would do that. And I, you know, pass the phone over to him and just back and forth like that. Um, you know, we, we did kind of a whole new color pop for the app, new icons. Um, and one of the goals was to really make all of the focus in Instagram on the content, on the photos and the videos, and to take all the color out of the Chrome And it basically you know how Instagram looks today But at that time you know we had this kind of dark blue and black Chrome everywhere in the app and things were a lot heavier. And yeah, so we just tried to really simplify it down and make all the color come from the content itself. Was it a gated launch or was it a just launch it all at once kind of project? So AB testing was like pretty new at Instagram at that point. Um, you know, Facebook was, was doing a lot of it and Instagram was kind of more of the like, Oh, we know what's good. So we'll just ship it. And, uh, when I started working on, on this project, the, the CTO and co-founder of my Krieger, he, he was like, you know what, don't, don't worry about trying to AB test this thing. He's like, just, just ship it, like build and ship it. Um, and actually at that time, like they were still sort of using branching as like a way to, to build big features, but I had seen some long lived branches, uh, like the, the, um, direct messaging feature had been built as a, as a long live branch. And it was a nightmare merging it back in. Cause they branched off for like three months and weren't really rebasing. And then they, you know, merge this thing back in. It was, it was, I think pretty awful. And this redesign touched literally every surface of the app. So I was kind of like, well, this is not going to be a good way to build it. So I put it behind kind of a feature flag. I was constantly merging it and had to build abstractions. Like all the colors became semantic colors. So, you know, it's like instead of saying that icon is black, it's icon color or something like that, or it's disabled icon color. And then you could switch inside of that function for, are we on the new design or the old design? And so that ended up being just kind of a much easier way to build it incrementally. And then it also actually opened up the possibility of testing it. And so we went to ship it, and it was going to be just like a 2% holdout. So like, you know, we were going to ship the new icon. 98% of people were going to have the new design. And then we were just going to have like 2% of people, you know, with the old design just to make sure that like nothing was broken or changed drastically. And I had submitted the app to Apple. I dropped in the new icon. You know, we were ready to go. People had champagne ready for the launch party. And it was like the night before we were supposed to launch this thing. and um someone kind of in the facebook executive executive team above uh instagram came in and was like no you guys are not doing this um i guess they facebook had had a redesign that went poorly um and they just saw this as the same thing and uh and so they were like you need to run tests of this upfront. Um, so I, you know, I was up at like 1am just like backing things out and trying to like resubmit the app to Apple. And, uh, we, we tested it and actually tested really well. Um, it was a little bit sad cause it kind of crushed like the launch that we, we had planned. Um, you know, cause then there are all these articles about this new UI that Instagram's maybe going to do. Um, but, uh, but yeah, it, I mean, it all worked out. Um, but I, I did have a, a takeaway from that, which is like, um, I mean, testing definitely has a place. It's extremely useful. Even in our startup, you know, we test things where you sometimes get counterintuitive results, um, things in our onboarding flow, how people convert on things. Um, it's extremely useful to run a b test for that i think for like high level product direction and like what you want the thing to be i prefer to come in with a stronger opinion and not just kind of like um look at the look at the data and only let the the data guide the decisions um so i i think there's a risk if you do too much experimentation that you get trapped in kind of incrementalism so it's you know, easy to get those 1% wins, but you're never going to get that 50% jump. On the major redesign case, though, if you just went for it, though, too, there's also the other side risk, though, right? Which is your product taste might be off and people hate it. And then it's a pain to come back. Yeah, yeah, for sure. I think we had enough confidence, I guess, in using it. I mean, it felt really good internally. And I think at least on the kind of redesigning the app, we thought that would be good. The icon was definitely a little bit more controversial. And then when we shipped that, people hated it. I mean, it was like, I say it's the last time I looked at Twitter after a product launch because I went on and, you know, it was supposed to be the celebratory day. We had worked so hard on this thing. It went out and we were just getting raked on Twitter. and uh yeah i mean it was it was kind of sad but i think you know change is hard um especially i think people feel ownership over their home screens and and all of a sudden this thing just changed on them um and it was kind of like a a little bit more of a bleeding edge design direction too like i think the the flat icon and the gradient you know today feels very at home, but at that time, a lot of the icons were pretty skeuomorphic in 3D. Yeah. So, so it was a bit rough. The, the funny other thing from that, you know, we could see from the data actually that, uh, this icon change, uh, it materially improved how many people a day open the app. Um, and we've actually seen this in, in, um, the app of my new company called retro, but actually like the icon, can impact whether people open the app, like how visible it is to them on the screen. So it became more noticeable, I think, amongst the other icons and people actually tapped it more often. It reminds me of, I think some of Thomas Dimson's work on the ranking versus chronological, like the public perception is so different from how actually people are voting with their usage. Like people are using it a lot more when it's, when it's ranked, but you know, for some reason there's this like very vocal minority that is saying, I absolutely hate this. Totally. Yeah. That, that actually, um, the feed ranking, um, shipped about the same time as, as this redesign. And, uh, I mean, I was actually skeptical of it as well. Um, the, the metric that changed my mind on the ranking, which I think the story might be a little bit different today. But at the time, one of the things that moved was actually how often people shared to Instagram. So people that had this feed ranking experience were actually creating more themselves. They were sharing more. And it was partly because the feed ranking allowed Instagram to show them their friends more often. And so you saw content from people like you, and I think you wanted to have that mutual connection so you would actually share more yourself and so that kind of was an insight that flipped it in my mind where you know because people would say oh I'm just using the app more because like you're not showing me the things I want so I'm scrolling further or whatever but I was kind of like okay they're actually like creating more content that that's kind of hard to dispute like that's that's probably a good thing um I I really liked kind of the mission, the original mission of Instagram was like capture and share the world's moments. And I really liked it as a way to kind of encourage people to be creative and see beauty in the world. And that was just a mission I felt like I could get behind. And so something like this where people were actually creating more, I was like, okay, that's probably a good thing. I think at this leg of your career, I think one thing that you wrote about is that, you know, finding an amazing designer was like a big part of it for you. And so I'm curious, how did you find the amazing designer that you did work with and what makes a great designer great? Yeah, I've always just tried to, you know, be the engineer that like the top designers want to work with. And yeah, it's just an incredible opportunity if you work at these types of companies where you can just have their work kind of flow through you, be a part of it. So that for product engineers, that's one piece of advice I often give is like your impact can be multiplied so much by finding a good designer and creating a really good working relationship with them. Is there a reason why you say specifically the design function versus the engineering function? Like imagine I identify some very senior engineer that has some engineering design and I just devote myself to it and kind of attach to them. What's the difference between doing that versus the really talented designer? So I should make a bit of a distinction between like, if you're in sort of a product engineering function where you're working on like the features for users, the interface, kind of the front end, then I think like, you know, that's where you really want to pair with a designer. If you're in a more infrastructure role, maybe the analogy is like finding that really talented, super senior engineer and being the engineer they want to work with, you know, they've got great ideas and you can kind of implement that. So yeah, probably different depending on kind of what role you have. And so this Whiteout project was in 2016. I'm kind of surprised in the same year, you also built stories with Tiger Squad of some very famous people I'm aware of. Can you tell me the story about, you know, building stories for Instagram? The redesign wrapped up and I had actually been kind of kicked off my previous team. Right after I joined, they said, we're moving the team to New York. Do you want to move to New York? And I was like, no, I kind of, I'm okay here in California. They're like, okay, that's fine. well, you just have to find a new team. I was like, okay, I kind of wanted to be on this team. So I kind of like delayed it. I worked on this redesign project. And then I joined the search and explore team. And I was only on the team for a couple of weeks. And a friend of mine, who was also an iOS engineer at the company decided to quit. And he had been working on what was called the creation team and leading the project that would become stories. And, uh, my manager, um, was like, Hey, you know, this is actually a really important effort for the company. Like, I don't necessarily want you to leave my team, but like, I think this is a good opportunity for you. And, um, I really believed in, in what this team was trying to do. Um, at that time, it it's sort of surprising in hindsight, given how big Instagram is today, but it actually felt in some ways like Instagram was dying because a lot of the sort of everyday sharing, uh, from people you knew normal people was evaporating. It was kind of being replaced by creators and influencers. And, uh, some of that everyday sharing was going to Snapchat, which had this ephemeral format. Um, and so we were tasked with just, you know, how do we get kind of normal people to feel comfortable sharing Instagram again? So I went over to, to lead the iOS team on stories. And, uh, one of the first things we did after I joined the team is we actually cut the team size significantly. So there had been a lot of people working on it. It'd been pretty churny. Um, they had tried different product directions. Uh, a lot of them didn't really feel that great. They weren't working out. And in some ways they were working on things to have work for people to do. Um, or, or there was just like, not, um, not enough space for the people that were there. And we just decided, Hey, we can actually move a lot faster if we go down to a smaller team. So, uh, it was myself and one other iOS engineer was kind of like the core team. And then we'd get some help from other iOS engineers to Android engineers. And we didn't even have a dedicated server engineer. It was the infrastructure team where like you can have half a person, um, it's half their time, uh, which, you know, for what stories has become, it's, it seems kind of crazy. Uh, but it, it really allowed us to, to move quickly. you had ownership over the whole thing. So it was never like a question of, am I working on something that someone else is working on? Am I going to step on their toes? You know, if there's a bug, it's like, okay, that's my bug. I got to go fix it. And there was less discussion around decisions. We could just make them more quickly. I say like, if you want to go fast, go small. And I'm a strong believer in small teams as like really the best way to operate. It's definitely not the only way to operate, but it's my preferred way. And yeah, so we went through that and we built it just over two to three months. It was pretty quick. I never worked so hard in my life. I was working, yes, like 16, 18 hour days, seven days a week in the office every weekend. And yeah, I would like leave at like one or 2 a.m. to go home. I was driving back and forth from San Francisco. I was really determined to not sleep in the office. I was like, you know what? I'm always going to go home, see my girlfriend. And it was kind of silly because I was spending this extra time driving. I really should have just slept in the office. But yeah, it was intense, but it was fun. It felt like we were building something really important. And we were using the product ourselves. We were really enjoying it. It also was like this bonding experience amongst this small team. So our PM on the project is now my co-founder at my company. And, you know, I'm very close still with the other folks that worked on it. And it also, it felt, it's funny because in some ways we were kind of the incumbent to Snapchat, but in other ways they were very much winning in terms of this kind of like everyday sharing. And so we actually kind of felt like the underdog. We had a poster up, a Ghostbusters poster up in our war room. And yeah, it felt like we were kind of in this war against them. And I thought, you know, maybe we could win. What was it in your opinion that like when you look at those two products, what was it that the Instagram version was doing so much better than the Snapchat one? Yeah, I think we had a sort of unfair advantage in that people already had their friend graft on Instagram and the people that they wanted to share with. But we just kind of weren't giving them the right outlet to actually do that sharing. And so as soon as we had that, I think it really stopped the outward flow to go to Snapchat for that type of sharing. Um, you know, I think we also did push the, the format forward. Um, certainly Snapchat deserves all the credit for coming up with this container that was a lot more comfortable. Um, you know, it's kind of 24 hour, the content doesn't stick around forever. Um, full screen, immersive, all those things. Uh, but at the time there was like a lot, uh, that exists in both stories products today that was not there in Snapchat. So, you know, even being able to navigate backwards by tapping left on the screen like that didn't exist. So I think we brought a lot of nice touches like that. The hold to pause, something I mentioned in the career notes. So, you know, I was using this early version that we had built and these things were auto playing and they were kind of going by too quickly. And just like intuitively put my thumb on the screen and wanted to pause it and it didn't pause. So I was like, oh, I can just do that. So then I built the hold to pause and, you know, now that just feels like a core part of the navigation and the format. So, yeah, that's another thing. Like when I mentor engineers, I mean, you just have this power as an engineer to just build your ideas. Like if you are using something and you think it should work a different way, you can just go do that. And that's such an awesome thing. And so I encourage, you know, for sure product engineers, it's like, if you have ideas, just build them and put them out and have people try them. And it's such a cool thing to be able to do. It's insane that you were commuting from San Francisco to Menlo Park. So that's like almost an hour in each direction. And you said you were working 16 to 18 hours a day, about seven days a week. I can't imagine what that was like. One of the learnings from your note is that you should find work that you care about deeply. I still wonder, were there times where you were thinking, you care about the work, but that just, as a human, I can't imagine surviving that kind of work schedule. I mean, it probably wasn't 16 to 18 every day, but it was working a lot for sure. and I definitely sacrificed other parts of my life. Like I used to be a very serious rock climber. I was on the US team. I competed in World Cups and I basically gave up that part of my life, you know, for sure in that time. But it was also such a unique opportunity to like be able to build this thing that was just going to go out to hundreds of millions of people. And, you know, one of the things I loved most when I was working on Instagram is I'd be like riding the Caltrain and I'd look over someone's shoulder and I'd see them using the thing I had just built like a week before. And like, that was such a cool experience. So, uh, yeah, it was like, you know, you sacrifice certain things, but I was kind of happy to do so. Uh, definitely not a sustainable model. So I wouldn't, you know, I wouldn't recommend doing that over a long period of time, but in, in stints, it can, it can make sense. And I think by doing that work with that intensity, it paid dividends later on. So the Instagram Stories product came out. It was much better executed, I think, than some of the other Stories efforts at the company. There were ones happening in Facebook and in Messenger. I think part of that was just the care that we put into it. It contributed to its success. I think the reception on launch was very positive. I actually was really worried about it. I thought the whole thing was just going to crash and burn just because I could see all these cracks in it. You know, you, you see the crash reports, you see all the bugs, like you're kind of your own harshest critic. But then when it went out, people were like, wow, this is so polished. And I was like, really, are we using the same thing? But yeah, it was, it was nice to have that reception. And so I think it helped the product. And then, you know, personally for my career, having kind of demonstrated that I could deliver this project on a, on a tight timeline with this level of craft helped me get onto, you know, the most interesting projects going forward. I saw at the end of the notes as you started at IC4 at Instagram and you got to IC6 by the end of, you know, after this gauntlet of these projects of whiteout and stories, that's incredible. Like one promo per half or two halves straight. What was that like? There's a bit of a funny story. So the promotion cycle happened in July and we shipped stories at the beginning of August. So calibrations are kind of happening in July. And so we're in the middle of this like very intense project. and I think like the there was a lot of recognition of how hard we were working I think the managers like they were being cautious around us and I knew so I felt fairly confident that I'd been misleveled on hire I just wasn't that familiar with levels when I came in and I kind of looked around and I was like you know I'm executing a lot higher than IC4 um my android counterpart who was an incredible engineer will bailey he was sort of the tech lead for the whole stories uh project and he really like set the example for me of like how you could be a successful product engineer and he was incredible at getting me into the meetings and you know being in the room with mike and kevin as we made all the decisions so he was an ic8 so i kind of like, you know, we, he certainly had a bigger role on the project, but we had some somewhat similar roles and I was like, okay, I'm here at IC4, you're at IC8. Um, and so that kind of opened my eyes to, to some of the leveling stuff. So I told my manager at the time, Eddie, uh, he, he was, he was the director. So I was reporting the director just, um, because, uh, you know, it was kind of a unique project at the company. And I was like, Hey, I think I got misleveled. And I think you should promote me from four to six in this cycle. And, uh, to his credit, like he didn't just totally brush me off and be like, you know, just dismiss it. He, he said he actually went and talked to HR about it and they came back and they were like, we've done this twice in the history of the company. In both cases, it worked out terribly. The, the people left, very quickly. So he's like, we're not going to do it. But I kind of wanted to seed a little bit that probably the correct level for me is a higher IC6 And I knew it was unlikely that they were just going to do that one step But I had a guess that like if I was just on that normal cycle, you know, it would come up the next time and they'd be like, well, he just got promoted last half. Like, you know, we can wait. Uh, and I think it's good for people to recognize that ultimately these levels and promotions, they're an incentive system. I think in the long run, like there's really an effort to make it fair. Um, but there's also an element of like, it's the carrot that's being dangled in front of you. Right. And so it can actually make sense. Even if somebody is performing at a higher level that like the promotion gets delayed, just so that they're not happening in quick succession. So I kind of wanted to like seed the idea that like, you should probably get me to six pretty soon here. So yeah, the first half on Whiteout, the redesign, and some of the infrastructure stuff, I got the IC5 promotion and then the six came for stories. It's interesting that you say like your high performance. I mean, you know, of course the promotions are one thing, but one of the things that you took away personally was that it gave you freedom to work on whatever projects you wanted. How does that play out? Yeah, I think it's just like demonstrated that like, you know, if you put me on a project, I'm going to do a good job with it. And so when there is a new effort at the company, you know, it's just kind of a natural like, okay, let's take the people that have done well on the new efforts before. And yeah, so I mean, I was very lucky to have the opportunity, but then like executing well on it set me up well for the future. Yeah, you already left before Instagram threads, but I noticed a similar model, which is you pull together all the top Instagram people and you got this small kind of similar like the stories team. And, you know, all those people have proven themselves repeatedly. So yeah, and it's a very fun environment to be a part of that kind of team. The next promotion actually came from IGTV, which my understanding, it's like a, I mean, I was at the company at the time. So it's kind of like a YouTube clone almost or like the vertical video first and Instagram. Could you talk about that project that got you promoted to IC7 or senior staff? So it was very much like came from Kevin, the CEO, him and Ian Silber, who's another amazing designer. They, before like an end of your all hands had kind of designed this thing together, laid out this vision for Instagram getting into kind of longer form video was the pitch. and it was going to be mobile native. So it was going to be vertical the way that you hold your phone. And, uh, so it came actually with like a fairly developed vision to the all hands and kind of surprise presented it to everybody. It's like, here's what we're going to do. Yeah. I saw that. And again, I was like, I got to work on this thing. This looks cool. It ended up being a bit of a weird project in the beginning because, um, yeah, it was like the surprise reveal. If you read through some of the recent antitrust, uh, litigation with Facebook, there's some insights into tensions between Facebook and Instagram during this time. Like we were told not to work on it for, for a while was kind of like the end of the founder's time at Instagram. IGTV was kind of like the last thing that they were last major thing they were there for. Um, but it was again, a really cool group, very small team myself will bailey thomas dimson um and uh you know kind of like ios android and um server leads and then we each had like one or two other uh people per platform and uh yeah just like super talented group um great engineers great designers uh we moved super quickly i mean the the product doesn't exist anymore today. It didn't, it didn't end up doing that well. And it's kind of interesting to reflect on why I think like it had some things that, uh, ended up becoming the future, like vertical videos, very much a thing today. Um, but it was kind of trying to mix like this long form YouTube style of content into this vertical format. And maybe that was a bit of a mismatch. Uh, people just weren't producing high quality, long form in the vertical format. We tried some interesting AI techniques to take landscape content and reformat it for vertical. I've got a patent on that. That's kind of one of my more interesting patents. But we showed that to video creators and they were like horrified. They're like, you're destroying my content. Like why I've spent so long making this nice landscape video. now you've just ruined it. So we tried to push people to produce this original long form vertical video. And I think the inventory just never really showed up. And there was maybe a little bit of hubris. It was like, you know, Instagram can just change the industry. We can just, people will just start making this because we have this platform, this audience, and that didn't totally materialize. So I think it probably did inform a lot of like, you know what ultimately shipped as as reels um but uh yeah igtv itself didn't didn't totally work out yeah i was because i was working on the like video infrastructure team at the time so actually that was one of the things that i got plugged into at some point i remember the energy in the san francisco office and the war rooms etc yeah totally you mentioned antitrust and what's the antitrust component you're talking about? A lot of the sort of internal communication has now become public through this antitrust lawsuit where the Justice Department, I think, is suing Facebook to unwind the Instagram acquisition. And so certain things that were kind of confusing to me at the time, like why were certain things happening, are less confusing now that I read some of the internal communications that was previously private, including private to me. I wasn't exposed to it. But yeah, I mean, I think basically Instagram had been acquired. It was smaller, certainly small relative to Facebook, and then had gone through several years of just incredible growth. And now was kind of more of like a peer to Facebook and was kind of like the new popular kid in terms of like, you know, the Instagram stories was received better than the stories in Facebook. And so I think that created some tension between the Instagram leadership and the Facebook leadership. And there was concerns about Instagram cannibalizing or taking users from Facebook, which, yeah, just kind of guided some of the strategy decisions, I guess. I remember that too because I was I mean I was also on the Instagram team and I remember there was some confusion about you know why are we turning off the I guess somehow like followers are being forwarded from Facebook to Instagram or it was like cross-sharing or something like the direction from Facebook to Instagram was getting turned off or something like that and yeah I think I read the same thing you read because I was like oh this makes so much sense It's like Mark Zuckerberg's internal memo on all of that and all the tensions between the Instagram founders, Mark. He wants to keep them and they're great at building products, but he's trying to navigate his side of the equation. Super interesting. It'd be really interesting to know how he thinks about it today. That was now seven years ago. Mark or the Instagram founders? Mark, um, because I sense in those communications, like something of a very understandable emotional attachment to the Facebook product, right? This was like kind of the thing that he created in Instagram. He deserves all the credit for acquiring it, but it was a little bit less his thing. Um, and yeah, I just wonder, uh, does he view that differently today than he did now? Uh, you know, seven years on, um, he owns both of these things. Um, I always had the view that like the best, the way to get the best outcome for the overall company was actually to have these things compete with each other because it's like you own both. they're going to make each other better. And, you know, one of them will win versus like, if you don't, if you've kind of stifled that competition, somebody from the outside is like more likely to come up with something better and, and take over. Um, but, uh, yeah, I don't know. I mean, he's the one running a multi-trillion dollar company and I'm not so armchair CEOing over here. Yeah. Maybe one day if this podcast scales up and I ever have Mark on, I'll ask him that. And then lastly at Instagram, I know you started a group called IG Labs. I'm curious the story behind you starting that group. And I also understand this is your promo to IC8 or like director equivalent at Instagram. So yeah, what's the story behind IG labs? After IGTV, um, definitely spent some time like Thomas Dimson has this phrase, like wandering the impact desert, like just looking for, for things to do and, and not necessarily finding anything great. Um, and I was actually pretty close to, to leaving the company at that time. Um, and kind of the new effort became reels and I didn't feel good about working on like passive video consumption. Um, my manager was, was over the reels org. So it, you know, would have made sense for me to work on it, but I was, I was fairly certain. I, I, I didn't want to work on that. I helped the team a bit, but like, um, yeah, That wasn't going to be my project through that time. There was a lot of like culture change. The founders had left. Um, you know, the leadership kind of came over from Facebook, uh, the new leadership, um, a lot of kind of like the old Instagram culture, it felt like was being kind of pushed out. And I was talking with like the, the new head of engineering, um, about. Trying to bring back some of that energy that I think had made product development at Instagram special, small teams, attention to craft, and also really expanding kind of the scope of what we worked on. So, you know, we were very focused on like the existing Instagram app and the existing features within Instagram and just kind of like iterative features on that very like incremental improvements. And I kind of made this pitch that like we have this great brand. We could do other things under the Instagram brand. Like, let's let's go try that. Let's, you know, explore like location ideas, maps, ideas, places. And again, paired with a really awesome designer, Vivian Wong. And we made the pitch to start this team. It was just her and myself in the beginning. And the idea was just to have kind of this like Delta Force, like small group, very high talent density that would work on new product initiatives and things that just didn't slot cleanly into the org. Like Instagram had grown to such a size where you really needed like a lot of structure in the org just to keep things sane. But that meant that projects that like didn't slot cleanly into one of those orgs were probably under-invested in. And so part of the idea of this team was that we would span across, you know, we wouldn't have a focus area. We could kind of work, work across many things. Um, we, we tried a lot. I mean, a lot of it didn't ship or tested, um, probably one of the more lasting impactful things that it seems small, but I think it actually gets used quite a bit as the collaborative post feature where, uh, you can have multiple authors on a post. Um, it's another of the more interesting patents I have is, is on that one. Um, and, uh, and so, yeah, I think, I think we found good impact. And then we were also this concentration of talent that like when there was a new important initiative ultimately ended up being threads, um, you had this group that, that could go work on it. And, uh, and yeah, just trying to, I don't know, kind of encourage innovation, trying new things and push that at the company. The last thing on your Instagram journey was that you tried management at some point or as an ICA, you switched to, I guess, TLD or tech lead director. Well, what was your thinking behind that and how'd it go? It's a very unusual or rare role within Facebook. Even like TLM, I think it's quite rare or it was at the time. Um, then like tech lead director, probably even, even more so. Um, it kind of happened mostly because I had started this group and at some point it just made a lot more sense for me to manage the people in that group rather than my manager who was over the reels org and, um, just less connected to, to their work. You know, I could probably represent them better in calibrations. I mean, I was in the calibrations anyways, so it was kind of like doing a lot of this work. So, um, yeah, in some ways it was just kind of like a formal recognition of like what I was doing already, um, to, to have these people report to me. Um, Facebook is, is very much of the school of thought of like individual contributors and management should be separate and I don't subscribe to that. Um, other companies work differently. Like my wife is a senior staff engineer at Tesla and like, um, they're very flexible. It's like, I see, is how people report to them all the time. Um, managers are expected to be like pretty competent, uh, technical contributors and like doing individual contributions. Um, and I think I prefer that model. Um, and so it was interesting for me to, to try it. I think like another one of my sort of controversial opinions is like, I think senior engineers should be involved in coding. Um, there's overlap between that thought and like the thought that like, um, managers should be still somewhat involved. There's an author, I really like Nassim Taleb. And he talks about having skin in the game. And yeah, I think like if you're a senior IC who has to do some coding, like you have more skin in the game, like you're not going to come up with some architecture that like you just hand off to somebody else and it's their problem now, right? Like you're going to be involved, you're going to see more hands on what the issues are. And I think like similarly, if you're managing a team and you're much more like with them in the day-to-day stuff, I think that you will operate better. So yeah, it's, you know, maybe a bit against the grain at Facebook, but it was a philosophy I had and I wanted to try it out. And then there was an upside to it as well in a company that grew so much and was so big where at Facebook, the levels are private, right? So you're just a software engineer through your whole IC time. And in some ways, I like that for like engineering discussions that there's no like pulling of rank like, hey, I'm more senior, you know, just take my idea. Like it's a little bit more meritocratic perhaps. But in cross-functional situations, say you're working with like a PM, like I'm trying to ship this collaborative posts thing. And I have to like meet with PM, all these different PMs on these different teams. There was an element where having a little bit higher of a title, I think just made those conversations easier. Like I had a higher baseline where they were like, okay, this person, like maybe they like know something. they've been here a bit, like they're not just, um, you know, totally new and, um, you know, you, you build up some reputation in a company, but when it gets so big and there's new people joining all the time, like you're actually having to like reestablish that a lot with, with folks. So, um, yeah, I think like it, it, it was somewhat helpful to have that title. There would be discussions in the, in the senior engineering group about like, should, should there be some form of, of public levels, um, within the company. And I was supportive of like, maybe like a two, maybe not like the full level is, is public, but like, there's like a senior designation or something. Um, just, uh, yeah, I think largely for like when you're working with people outside of your normal working group so that they start from a slightly better baseline on, you know, whether, whether you know what you're talking about. The number one thing I hear people say is that this is not optimal because it's, you're doing two jobs at once and you're going to drown and your career will not flourish in either direction. What do you say to that? I think it's probably correct. Like, um, at that point I was really not optimizing for career. Um, you know, it was never my intent to stay long-term. Uh, I mentioned this in the career note, like Facebook does these internal employee sentiment surveys every six months, I think, or year. And one of the questions on there is how long do you intend to stay with the company? I never put more than one year. Like it was always like, okay, this is my last year. I'm I'm going to, I'm going to get out soon. Um, and so, yeah, I think like I didn't have aspirations to get to IC nine. Um, and so it kind of freed me up where it was like, you know, um, well, I don't want to get fired for performance, but I kind of knew I wasn't going to get fired for performance. Um, and so, yeah, there was just a little bit less pressure on that. And, um, yeah, so I could just kind of like do the thing that made sense for our group, our team. and worry a little bit less about the career stuff. But yeah, absolutely. I think like I saw firsthand, like I think my ratings were probably lower than they would have been if I was just an IC maybe. It was like, okay, you're exceeding as an IC, but you're meeting as a manager. So you're just meeting. That's kind of like the full journey from start to finish of Instagram. And I know you've since left and you're working on Retro, which is a popular social media app with a lot of the polish that I kind of recognize in Instagram as well. I'm curious, you know, what made you want to leave to start that? What's the story behind creating Retro? So around the last few months of my time there, I was talking with Nathan, who's my co-founder. We had worked on Stories together. he had gone over to the Facebook side. He started Facebook dating. And then he was working on some of the virtual reality products. And through our whole time, we were close friends. We would meet up at this bar sometimes in the mission called Lone Palm. And we would talk about how we just saw a different way to build products, to structure a company. And we really wanted to be in the driver's seat of that. And, uh, we would talk about starting a company, but it was always like the timing wasn't right for me. I was on a project that I was engaged on and then I'd be like, okay, I think I'm ready. And he was on something new. Um, and so finally our timing kind of aligned, um, where we were both ready to leave. And I had a conversation with, um, this investor friend, the same one that connected me with Flipboard, who's, he's kind of a mentor of mine. And, um, he was like, it's time. You should, you should really take a bet on yourself. And that stuck with me. I was like, yeah, I should like take a bet on myself. Um, and so we left to start the company and actually our goal with the company is really to create this world-class product studio and retro is our first product, but it's not necessarily like, the reason for the company existing. And we, yeah, we just want to kind of create this environment where great product builders can do the best work of their lives and create products that are good for people that they feel good after using. And, and also to create a successful business. Uh, so retro is our, is our first product and it's a social app, but it's actually quite different from what are called the social apps today, I would say, um, in that it's actually social. So you, you, you see people, you know, on it, uh, friends it's, it's all focused around friends. And, um, you know, I'd say like the traditional social media is maybe, well, the derogatory word I would say is like brain, brain rot media. Um, entertainment is maybe like a more favorable, uh, term, but if I open Instagram today, I don't see a lot of the people I know. Um even if I actually go to seek that out I tend to end up down a rabbit hole of uh kind of short form video entertaining content And um definitely entertaining but it kind of hijacks my attention I get sucked into it And so the idea behind retro is we creating a space that's all about connecting with the people that you actually know, staying up to date with them, and also appreciating your own life by looking back on your photos and kind of picking out the highlights. And in some ways it's a throwback to kind of the earlier times in social media, what Instagram used to be all about creation. Almost half of the people that use retro on a daily basis actually post something, which, you know, I know from working on Instagram, like that number is quite a bit lower. And I take it as a really good sign that we have created a product where people show up to create almost as much as they show up to consume. So we've been working on that for the past few years. It's become very popular in Taiwan. Oh, interesting. It's not something we expected, but social networks are so much about the network, whether your friends get on it. And we managed to hit the critical mass in Taiwan and sort of all of the usage patterns end up looking different when, when you hit that critical mass. Um, so we were number one on the app store there for a period last, uh, last year. Um, and we still are at the top of the, of the photo category. We were number two when I looked this morning. Um, so we've kind of shown that like it, when it works, it works. Um, and we had just have this challenge of making at work in more places. And, um, you know, it was very challenging to get people to try a new app. Um, but if you miss seeing your friends and people, you know, on social media and staying connected with them, I, you know, very much encourage you to, uh, to try out retro and, uh, and yeah, it's, I, I think it's a really delightful app. I love seeing photos from my family on there and we want it to feel good when you open it and feel good when you close it so you don't feel like your attention was hijacked, that you feel like this was a good use of time, you got caught up with the people you care about and then you can go on with your life. You can get out and enjoy the world. The design philosophy you went for with this app, is it more challenging to get people to use it because it's less addictive? So one of the key decisions that we've designed around is that we don't have an ad based business model. Um, and I actually, I don't think like ads are evil, but I, I do think that ads when combined with an algorithmic feed are a little bit evil, um, because you have this misaligned incentive. Um, you know, I think the app is incentivized for you to just consume more and more. It just wants more and more of your time and attention. Um, even past, you know, what you wanted to get out of the app, I show up to see what a friend is doing. It's going to try to keep me there for the next hour, two hours, whatever. Uh, and so I, I think ads fundamentally kind of like when combined with an algorithmic feed, uh, make products that aren't super aligned with people. Uh, but there, there is definitely a challenge, uh, in terms of like growing retro. So, you know, we don't have, uh, creators, the, the model is mutual friending. So, um, everyone is private. You can only see people that you've, uh, accepted as friends and it has to go both ways. It's not like a following model. and we have a limit of 250 friends. And so, um, you can't build an audience on this platform. And so a lot of the ways that like Instagram grew was through creators that were financially incentivized to spread this app and get more followers. And so those, those pathways to growth are not open to us. Um, but we've, we've seen examples of like, you know, be real, uh, managed to get a decent growth. They had kind of a nice viral, um, mechanic built in with this notification that everyone got, that was weird. It's time to be real, you know, that created a lot of conversations. So that certainly helped. Um, but I think it is possible to, to grow a friends only, uh, social network, but in many ways it is, it is harder. And, and also the business model is, is harder. Um, that's something that, that we're very open about. Like this is a challenge. Um, there would be easier ways to do things, but we think they would lead to bad outcomes, outcomes we don't want in the longterm. We're very much taking an opinion on what this product should be and what we want to see in the world. Um, you know, some people say just, just build what people want. And, you know, maybe you could say like in Instagram, people are demonstrating that it's what they want because they spend more of their time there. Um, and we're taking a little bit more of an opinionated stance saying, no, we think like, actually that's not great. And this is better. And even if it's more challenging, it's something we strongly believe should exist in the world and should be an option there for, for people that want it. Um, I think we also actually coexist really nicely alongside Instagram. Like I still have it. I still open it. But I'm getting sucked into it a lot less and it's helped me kind of break this like phone addiction that yeah, I just didn't feel good about. How do you monetize if you're not using ads? We have a subscription, a premium subscription. And so it's only a small percentage of users that are subscribers. And it's just kind of like an extra tier of features, particularly some of the things that cost us a lot to provide as a service. So video, as you know, from video infrastructure, like the storage and bandwidth for that is quite a bit more than, than photos. Um, and so you have to subscribe if you want to share out video. Uh, but the free version of the app is like actually great. Um, you know, uh, you can share photos, you can remember your own life. Um, and it's like 93% of what people share on retro as photos anyways. Um, and, uh, and so, um, yeah, if you want to kind of like support the mission and, and get those extra features, you can become a subscriber. That's awesome. I love like the intention behind it and like the mission. It's really cool. It's definitely a product that's fun to work on. It feels good. Like, yeah, it's, it's very, very feel good. We had a tagline for a while, like feel good social media and, um, Yeah, it's very much that. Coming from big tech and starting your own company, I'm curious, looking back on across the various career axes of maybe learning, satisfaction, compensation, the things that people look at for their career, how has it been so far working on your own thing? Yeah. Compensation is quite a bit worse. we we pay nathan and i pay ourselves less than all of our employees and we have more ownership in the in the company which i think is the right way it's how it should be um uh but that means that yeah i've you know kind of um certainly put off that short-term compensation we hope to become a very successful company and that that equity is going to be worth a lot both for us and for our employees it was always the thing i wanted to do like i i was a bit surprised actually that ended up in, in, in big tech, you know, talking about like those early days in, in Flipboard and, um, kind of this like entrepreneurial culture, like it was always exciting to me. I would read tech crunch all the time, um, about the new startups. And, uh, yeah, I just always wanted to kind of build a company, you know, be able to, to bring the best people together, work on the things that we thought were cool. Um, do the things, do things in the way that we wanted to do them. And, uh, yeah, it's been, it's been awesome for that. Um, and, uh, and definitely like have learned a ton. I think the amount of time that I get to spend on interesting work is much higher than it was, especially in my later years at Instagram. Like a lot of my work became kind of like convincing, uh, you know, layers above me that we should do something. I, I would say there were probably more people working on that app than there needed to be. And so actually there was a lot of gatekeeping, um, where it was like, is your team allowed to ship? So you would do this work. And then there was like, you have to convince somebody that like, this is a good enough thing for, to ship in the product, which is understandable because, you know, it becomes very complex if you just let everybody go, but it means that there's a lot of work in just kind of like that political wrangling. And like, how do I, how do I make this person feel like this is their idea and not my idea? Um, and, and I'd spend none of my time on that now. Um, you know, we, we do a one hour standup, uh, every day. That's basically the only meeting on my calendar. So I get almost the whole day to just build, think about, you know, what people want and try to make it. So, yeah, I'd say like satisfaction is definitely higher. Learning is definitely higher. Compensation is definitely lower. I think throughout the conversation, you mentioned that, you know, smaller teams can move faster. do you think that the company would be better off if you just like laid off half the people i would say yes we we used to talk about this a lot when the company was much smaller um like uh this thought experiment of like you know the thanos like you just snap your fingers like half of people are fired even if it's a random selection um do things go better and And frequently we felt like, yeah, maybe I think they would. So, yeah. I mean, it's tough because with a business like Instagram, you know, if you can make a 0.1% improvement, it's actually like hundreds of millions of dollars to the business. And so sometimes, you know, that incremental person, maybe they're able to find those things. Uh, but I think it's less appreciated the cost of each person you add. And there's sort of like maybe the more obvious, just organizational communication overhead, you know, you, you just have more stakeholders, more meetings, more coordination, but, um, even just like writing more code, like having more code in the app. I remember right before I left, they had this tool where you could see how much time you were spending compiling the app. And I was spending more than four hours a day waiting for the app to build just because there was so much code. And, you know, the tooling just had not kept up with basically the velocity at which people were writing code. And, uh, so that's a huge cost. That's kind of, you know, it's not noticeable with each incremental person you have, but now you have this, you have a thousand engineers and all of a sudden, like everyone's just like so slowed down because, um, yeah, because of this overhead. So, yeah, with a smaller team, you know, like when we were 10 people, the app was tiny. It was like so much easier to get things done. Like a project like Whiteout, you know, would be like a massive project for Instagram today because there's so many more services. There's so much more code. There's so many more people. It was actually like a lot easier at the time that I did it. Yeah, I guess, you know, Twitter is kind of an interesting case study of that because they kind of went through that and the app's still operating. I do get the sense that there's a lot more breakages, but I'd be curious like on the product development side, you know, is it, how has that been being on those teams? Totally. Yeah. I mean, I think there were projections when Elon came in and I think he cut like 80% of the staff or people left as well. Right. That like, oh, this thing was just going to 100% fall over. Um, and I mean, my experience is somewhat buggy. There's definitely some issues. Um, but I mean, you know, it's very much continuing to, to run. I think also I saw like the economics of the business have actually become much better. Um, it's actually profitable now. Uh, whereas, you know, through almost its entire history, it was operating at quite a loss. And so, um, yeah, somebody coming in and with a machete cutting things, like maybe it's okay. Towards the end of the conversation, I just want to wrap up with a few career reflections. I think first thing is you, throughout your projects, you worked with like a lot of very talented people. I think you mentioned, you know, Thomas Dimson, Will Bailey, a few incredible designers. Also sounds like you worked with the founders of Instagram. And I'm curious, do you have any stories that you'd share that illustrate what made them exceptional? Mike Krieger is definitely something of a hero of mine and it was such a privilege to get to work with him. They say, don't meet your heroes because you'll be disappointed, but I think you should meet Mike Krieger. He's actually an awesome human. And a lot of the ways that I think about engineering leadership and engineering philosophy came from him. So I mentioned earlier the simple thing first, you know, that's really stuck with me. I think the attention to detail and craft. But another thing is he had this sort of, I don't know if it was intentional or just the way he is, but it was like this lead from the front, or I call it like lead from the front way of operating where he never put himself above the team. And he would just jump in to whatever team if an effort needed help. So at the end of stories, I had brought in a friend of mine from Flipboard, actually, who was then working at Facebook, not at Instagram, very different team to do the drawing tools just because he had worked on a drawing app before. And I was like, Hey, we need drawing. Like, can you work on this? Um, it's kind of a, an interesting insight into like how Facebook operates. So that I could just like go to this random, you know, person on a very different part of the Oregon and pull him into this project. Um, so he worked on it for a few weeks and got it reasonably far, but, uh, then kind of got pulled back by his team. Um, so we were in, we were in a bit of a pickle and, um, Mike just came in and was like, okay, I'll finish it. Like, um, so he, you know, he's in there coding the neon brush and, uh, you know, we were working those insane hours. He was there working with us. Um, he's up at 2 AM reviewing my diffs. Um, and, uh, yeah, just seeing that and how that felt as a team, uh, really has informed like the way I think about leading teams. Um, and, and so, yeah, that he he's, he's awesome. Um, some of the other folks you mentioned, Will Bailey, I talked about him a little bit before he has just this product vision that he sticks to very strongly and frequently it's quite good. And he, at the kind of like, I would say like dark days of development of the stories project, when it was spinning in circles, it was taking different forms. None of them felt very good. He just kind of outlined this full vision in a document for like how this product should work, you know, the things that we could do with it going forward. And I looked back on it five years later and I was like, wow, this was like the five-year roadmap for what we ended up doing. Like he basically nailed it all from the start. Um, and, uh, and yeah, he was such a good example for like how you could succeed as a product focused engineer, um, which actually was fairly calm. It was much more of a thing in Instagram than it was in Facebook. I remember being in bootcamp and people were saying, okay, if you're junior, you go work on a product team. And then if you're senior, you work on an infrastructure team. And that was kind of the culture was like, you know, product is easy, but like, you know, real engineers, they work on infrastructure and, uh, and Instagram actually prioritized more in some ways the, the product side. And I think that was reflected in, you know, the various versions of stories, for example. And so Will was such a great kind of mentor and example of like how you could have the successful career as a as a product engineer, how you could influence product direction as an engineer. And then Thomas is probably just like the smartest person I've ever met. Like he, he, he's incredibly smart, but not in like a savant kind of way. Like he's also like, you know, uh, got great social still skills and understands people. Um, and, uh, so he did the feed ranking or, you know, basically started all the ranking at Instagram. Um, he started his own company, open AI, open AI bought his company. He's a researcher there now. Maybe Zuck's coming to him with a million dollar offer. I don't know. Um, but, uh, but yeah, I, again, like, you know, he actually had like this skill on the infrastructure side, um, you know, built a lot of the infrastructure for the ranking stuff, but always approached it with a product mindset. And I think would probably consider himself like a product person. Um, I have always been interested in technology as like a means to create things for people. Um, I've never been that interested in like the technology itself. Um, it's like only interesting to me, to the extent that like I can create interesting experiences for people. Uh, so yeah, I think like, um, Thomas and, and will, um, you know, kind of had maybe a similar approach, very like human oriented, um, thinking about the psychology, how people use these things and how it makes them feel. And then your last question is if you could go back to yourself after all this experience that you have and talk to yourself when you were just graduating college and give yourself some career advice, what would you say? Like just invest in the tools of your time to You build things for people, like build products for people. And I think if you do that, you'll be pretty flexible and adaptable. And in many ways, you have a unique opportunity as a new grad because you aren't sort of stuck on this path of like, I'm an iOS person. I know this technology really well. It's all new to you. And so you can pick up the latest thing, right? And so I think like with these AI tools that are coming out, you can do incredible things with them and you're probably going to be more skilled at those than the senior engineers because it's just not the thing that they, you know, learned natively that they've invested their time into. and I had that to some extent with iOS where it was a newer thing. And so the senior engineers were writing Java or something. They were writing our backend and I could actually stand out and excel on this thing because it was the first thing I learned and I was eager or hungry. And I think there's something to that with this new set of AI tools. So yeah, I think my advice for new grads is just learn the tools of your time and get good at them and make good things for people. And I think that will work out. I imagine it's a bit scary probably because it's hard to say how it's going to play out, but it seems like there's reducing demand for junior software engineers because of some of these AI tools. But I think if you can yourself get really good at using those tools, there will definitely be roles for you. Awesome. Well, yeah, thank you so much for your time, Ryan. And that's really, you're a legend at Instagram and I worked there for so long. Super excited to talk to you. At this point, is there anything you want to redirect the audience's attention to? If the ideas behind Retro sound interesting to you, definitely try it out. It's a great place to reconnect with people that you want to see into their lives, but maybe they've stopped posting on social media. So get them on too. If you're a designer, you're an engineer, definitely reach out. I think we have some interesting ideas in how we're building this company. I mentioned before, our goal is to become a world-class product studio, multiple products. We have Retro, we have two more in the pipeline right now. And I think we have some interesting ways we're structuring revenue share with employees. So as these apps start to make money, even if the company's ambition is to kind of like take that and reinvest it into growth, you can see some of that upside right away. And we're building infrastructure that helps support products across the portfolio. So we're building growth tools, analytics tools, all of that, that I think are going to really help us Excel as a product studio. So if that's interesting to you, you know, reach out for a conversation. And yeah, it's just been a pleasure chatting and thanks for having me on. Thanks so much, Brian. Really appreciate your time. Thanks for listening to the podcast. I don't sell anything or do sponsorships, but if you want to help out with the podcast, you can support by engaging with the content on YouTube or on Spotify. If you want to drop a review, that'll be super helpful. and if there's any guests that you want to bring on to please let me know i feel like sourcing very senior ic's there's no well studied list out there on google that i can just search this up so if there's someone in your org or at your company who you really look up to and you want to hear their career story let me know and i'll reach out to them