Building WhatsApp with Jean Lee

12:47Jean Lee:Don't make the mistake that I did, which is just saying, like, I'm sorry, I'm busy. I might have been a contractor from Europe. So, like, it sounds like that was a strategy. And that was a smart strategy.

12:56The Pragmatic Engineer:Yeah, we had many contractors in Europe and they were all very experienced people. They were basically managing themselves. We had people all over the world working with us.

13:05Jean Lee:What was a tech stack like at WhatsApp? Before Gene walks us to one of the most unusual tech stacks in startup history, we're talking about eight platforms and a handful of engineers. Let's talk about keeping your code base healthy with our senior sponsor, Sonar. Sonar, the makers of Sonar Cube, helps you automate code review and verify the quality and security of your code across your entire stack so bugs don't make it to production, whether you've got one platform or eight. As agents take over the development process, Sonar has introduced the Agent-Centric Development Cycle Framework, ACDC, a new software development methodology designed for the unique scale and speed of AI-generated code.

13:43Jean Lee:It's a move towards a more intentional four-stage loop that gives agents the guardrails they actually need. The fourth phase is being guide. First, agents need to understand the canvas on which they're being asked to create so that the output fits with what the developer and organization require. Generate. The LLM-based tool generates the code it believes will achieve the desired outcome within the right context. Verify. Next, the agent is deliberately required to check its work, ensuring that it actually achieves the desired outcomes and is reliable, maintainable, and secure. Solve. Finally, any issues identified are provided to a code repair agent to fix.

14:23Jean Lee:To power this, Sonar has significantly strengthened its offering, introducing products and capabilities like Sonar context augmentation, SonarCube Agentic Analysis, SonarCube Architecture, and SonarCube Remediation Agent. Head to sonarsource.com slash pragmatic to learn more about the latest with Sonar and how it's empowering organizations to embrace the agentic era. And with this, let's get back to Jean and all the different tech stacks that WhatsApp had.

14:46The Pragmatic Engineer:We were actually pretty unique. I don't think any startup ever really does this, but we had seven different stacks. We had, I actually looked it up because it's hard to count them all. So we had, of course, everybody has iPhone and Android, but we also had BlackBerry and Windows Phone, which is also pretty common. But we also had Nokia S40, S60. We had a thing called KaiOS for a while, but not for a long time. And we had a web client. So it's actually eight.

15:16Jean Lee:So you have, of course, you know, we know that iOS is Objective-C. Android was Java back in the day. And then all of these, like the BlackBerry, the Nokia, they all had, I think Nokia was Symbian, C++. So they all had like their own different language. And then we've not talked about the backend, right?

15:32The Pragmatic Engineer:And the backend was Erlang.

15:34Jean Lee:Erlang. Can you tell us about Erlang? Because that is one of the most exotic tech stack. I've heard Erlang in telecommunications context at Ericsson, again, in Europe. It is popular with the telcos. But startup-wise, I'm not sure I heard anyone else use Erlang.

15:51The Pragmatic Engineer:You might be right. They do have an Erlang conference. I think it's called Erlang Factory. There's a really great talk by one of our engineers, Rick Reed, if you're interested in learning more about it.

16:02Jean Lee:We'll link it in the show notes below.

16:03The Pragmatic Engineer:I'm pretty sure it's still on YouTube. I haven't looked up recently. But he gave a really great talk about why they started working with Erling and it was the perfect choice. And he describes it as trying to maintain the engine of an airplane while it's flying 24-7. because if you imagine like WhatsApp is so international, we can't take a break, right? We have to continuously keep running and it's always busy. It's 8 a.m. somewhere in the world, right? And Erlang was a really robust language that was really good at concurrencies and they stumbled upon it because they were using this other tool that happened to use Erlang and decided this is the perfect language.

16:47Jean Lee:And I guess at the core of WhatsApp, what was the core engineering challenge? Was it like so many messages being kind of coming in, needing to be seeded out and sent to different platforms?

16:58The Pragmatic Engineer:Yeah, that was one of the main challenges. Like, for example, for New Year's or Christmas, because everyone's saying Happy New Year at the exact same moment. That was always our biggest challenges every year. And we would celebrate, hey, we didn't go down after New Year's.

17:14Jean Lee:so the interesting thing about the seven different mobile platforms specifically is the conventional wisdom wisdom before and after has been like look if you want to support all those platforms don't be silly do cross-platform either build your own layer that is cross-platform or use you know there's all sorts of frameworks why did whatsapp not do this do you remember the discussions of like why why hire seven including some really hard to hire people like for nokia and symbion and you mentioned the contractors in europe i mean sounds a bit of a nightmare. Why?

17:44The Pragmatic Engineer:So Jan used to always say, I want a grandma in a remote countryside to be able to use our app. So what does that mean? They may not have the newest iPhone, the shiniest phone with the biggest memory, right? In the countryside where a grandma is using it, you need the app to be lightweight. You need it to work on any kind of device. and you need the app to be simple. So those were our goals and priorities and that's the thought process that went into our decision to build seven different platforms.

18:20Jean Lee:And then inside WhatsApp, how did you get things done? Do you remember like how a project got done or what was the concept of projects and kind of what engineering processes people might have followed, especially, you know, later you worked at Meta compared to like how, you know, like more kind of, you know, standard startups work. Because I have a feeling WhatsApp was not exactly a standard startup, was it?

18:40The Pragmatic Engineer:Not really. Even Meta compared to other big tech, especially when I was at Meta, was pretty scrappy. Like not so much on writing documents, for example. The move fast and break things model kind of allowed them to be a little bit more lean in terms of their process, at least while I was there. But WhatsApp was like the ultimate lean company. By the time we were acquired, we only had 20 something engineers, so under 30 people serving 450 million monthly active users. So we didn't have code reviews. The only time I got my code reviewed was the first time I made a commit. Brian asked to take a look at it before I committed it.

19:27The Pragmatic Engineer:And he asked me a bunch of questions, which I had to think through a lot. kind of like a coding interview. But that was it. After the first time, we didn't really have a formal code review. But I mean, people read the Git commits because there's only 30 engineers who can read other people's code and they would discuss it on the WhatsApp groups.

19:47Jean Lee:So everyone was trusted, all engineers, that they just pushed their code to, they merged it into production, pushed it to production without a manager review. And it was trusted that, you know, they would ask if they were unsure or something like that.

20:00The Pragmatic Engineer:Exactly.

20:02Jean Lee:okay and it worked it worked what about the release process like if if you tell me 450 million people the first thing i'm gonna say is like okay did you do canarying did you do feature flagging did you do experiments did you do you know what kind of safety next did you have right we didn't do much of that but we were really big on dogfooding so every time we were about to do a

20:24The Pragmatic Engineer:release we would all internally use it ourselves yawn i think he might still say it on his linkedin If you look up Yeon, he said just quality engineer.

20:36Jean Lee:His title, when he messaged me because I didn't know he was CEO, it said chief QA officer.

20:41The Pragmatic Engineer:QA officer.

20:43Jean Lee:And I didn't know what that meant. I thought it was some sort of weird joke from the outside. So now it makes sense. So he was going around. He was making sure that it worked.

20:53The Pragmatic Engineer:He would try to break things as much as he can. and then if he finds a bug, he will like really try to break it and then he'll come to and say, hey, like I found this bug.

21:04Jean Lee:And you also said that Jan said no a lot.

21:08The Pragmatic Engineer:He did say no, almost as I recall, 99 % of the time he would say no, which I thought as a, again, as a young engineer, I was very confused because when you look at all these other apps, there were like dozen different messaging apps at the time. Like WeChat is notorious for having everything, right? They have so many features. And I was so confused. Like, why don't we build all these features? These are the newest, coolest things that we should have. Because at the time when I joined, we didn't have groups. We launched groups shortly after I joined. We didn't have voice calls, video calls. We didn't have any of these, no stories, you know.

21:45The Pragmatic Engineer:All the cool features were missing in my mind. But that was by design because we really wanted to prioritize, again, the quality of a grandma in a remote town being able to use our app at any given time.

21:59Jean Lee:WhatsApp held back features for years until they were absolutely sure about quality. They worked on video calling long before they shipped it. This leads us nicely to our presenting sponsor, Statsig. Today, you don't have to choose between speed and confidence. Statsig lets you see features behind flags, experiment with real users, and only roll out broadly when data shows that you're ready. Here's what it looks like in practice. You ship a change behind a feature gate and you roll out gradually, say to 1 % or 10 % of users at first. You watch what happens. Not just did it crash, but what did it do to the metrics you care about?

22:34Jean Lee:Conversion, retention, error rates, latency. If something looks off, you turn it off quickly. If it's trending the right way, you keep rolling it forward. And the key is how the measurement is part of the workflow. You're not switching between three tools and trying to match up segments and dashboards after the fact. Feature flags, experiments, and analytics are in one place, using the same underlying user assignments and data. This is why teams and companies like Notion, Brex, and Atlassian use Statsik. Statsik has a generous free tier to get started, and pro pricing for Teams starts at$150 per month.

23:05Jean Lee:To learn more and get a 30-day enterprise trial, go to statsik.com slash pragmatic. With this, let's get back to how Gene and the WhatsApp team ship quality code with close to zero formal processes. so it sounds like whatsapp had very very little process this was very very interesting because when i worked at skype at the same time as you joined whatsapp and i also joined in 2013 i joined skype and you joined whatsapp in 2012 skype was very proud that they sent everyone to scrum training i was a scrum master other people was scrum master so here we were with all the scrum all the consultants all the everything and whatsapp al competed us with like a lot smaller team and and no no scrum no tdd no agile skype 1000 engineers oh that's a lot of people yeah i mean when you have a thousand people you kind of need these yeah and and in all fairness like for example one thing that this whole scrum thing solved for a little bit is we had more more than 100 teams and everyone working on different things.

24:05Jean Lee:And because of all this organization, we had a prioritized list of which teams are the most important and those got all the support. So I guess one lesson might be that when you're just big, it's just so much harder to move fast and a small team can out-compete you.

24:19The Pragmatic Engineer:Yeah, it just takes a long time even just to communicate with everyone.

24:23Jean Lee:Being inside of WhatsApp, how did it feel to see this massive growth? Not in your team size, but in the product usage, the people, the media, the feedback.

24:34The Pragmatic Engineer:We didn't have much media. Nobody knew about WhatsApp.

24:38Jean Lee:One interesting thing you told me about the office is you had countdown displays. Can you tell me about them? What were these? What did it display?

24:44The Pragmatic Engineer:Yeah, so you asked me a lot about metrics. And I think really the only metrics we tracked, we didn't really pay too much attention to media or Skype's usage numbers or other messaging apps' usage numbers. But the one metric we counted down was number of days, like X number of days since the last outage.

25:04Jean Lee:Wow. No pressure.

25:07The Pragmatic Engineer:Well, the numbers started to go up over time. Maybe that helped to have it visibly there.

25:13Jean Lee:And when an outage happened, do you remember what happened after? Because these days in the tech industry, it's all about blameous postmortems. If an outage happens, you know, we first mitigate it, then we get together, then we write a document where we try really hard to not say who pushed, caused this, but we come up with why the system is like this and so on. How did you go about like dealing with outages and also following up and ensuring that they won't happen again?

25:39The Pragmatic Engineer:So I know they did these discussions in the server group chats, but I wasn't in the server group chats, so I can't really say for sure. I mean, for sure, we did not have documentations.

25:50Jean Lee:It sounds like a lot of things were pretty simple. You talk with people. If you have a problem, you try to fix it. Don't over-document things for no reason. And it seems to just work. And then have the key thing. I guess if you put out days since outage, people will know, like, okay, I should do what I can to not have an outage.

26:08The Pragmatic Engineer:And everybody knew exactly who was working on what. So we didn't have to blame anyone. Everyone just knew.

26:15Jean Lee:WhatsApp was a massive, massive success. What do you think made it so successful in the early years, especially for the product itself? You've seen Kakao. You were aware of some of the competing messaging apps. What did WhatsApp do that others did not?

26:31The Pragmatic Engineer:There is a little bit of the networking effect. The thing about messaging app is that if you use it, you need your friends to use it. And if your friends use it, you need to use it. And WhatsApp was the first to be on the market. that certainly helped. But there was a lot of competition. But again, I think a lot of other apps and messaging apps were chasing features, thinking about adding the shiniest, newest features. Whereas WhatsApp was very intentional. They actually worked on video calling for a very long time. We were probably working on it by the time you joined Skype when your founder said we have video.

27:10The Pragmatic Engineer:We were working on it, but we just didn't launch it until much later when we were actually like 100 % sure about the quality of the feature. So we often held onto features until we felt really sure before launching them.

27:26Jean Lee:Interesting, because that is a little bit of a different than the conventional advice, which is if you're a startup, launch early, get feedback, improve it, and iterate it. It sounds like you did the opposite. It's like polish it and then do when you have full conviction.

27:39The Pragmatic Engineer:Yeah, we did use it internally. internally we use the voice and the video calling features with our families so we had like a list okay like i have family members these are all my my parents and my brother and sister's numbers let's enable it for this beta group and we used it for a very long time before we launched it with the public two years into working at whatsapp in 2014 facebook announces their biggest ever

28:03Jean Lee:acquisition whatsapp for 19 billion dollars what do you remember of this time how unexpected was it and what kind of feelings, what kind of emotions went through you and the team around you?

28:14The Pragmatic Engineer:I actually journaled soon after the acquisition. So I looked up my journal around this time 2014. So it's been over 10 years. But I looked at my journal and I remember I was coding. I had this Spotify playlist with noise cancelling headphones. I had this playlist called Let Me Think. It's the one I listen to when I want to focus. and again like we were in a pretty small office where i can see everything i was sitting in pretty central location so i i could see people bustling and hustling which was a little bit weird but i tried to tune it out so i can code but then from the side i saw um niraj who was the head of business at the time he was just like waving his arms he was a pretty tall guy so i could see it he's like like stop whatever you're working on right now come into the um we had one meeting room come into the meeting room and i was like what is happening like we never have meetings like we we so you didn't have meetings i mean we we have scheduled meetings every now and then but we rarely have like we have never had unscheduled meetings and we rarely have meetings at all So I was confused and I dropped whatever I was working on and I went into the conference room and then they asked, turn off your phones.

29:38Jean Lee:WhatsApp, turn off your phone. That's kind of weird, right?

29:41The Pragmatic Engineer:And I thought, oh my gosh, what's happening? Like, did we go out of business? I was one thought. I thought, are we getting another raise of funds, like round of funding, like a new investor coming on board? It can't be that we sold the company because Jan used to say he will never sell the company. He used to actually say, selling your company is like selling your baby. And I remember we were waiting for quite a while because there was one person missing. And it turns out she was getting her eyebrows done.

30:16Jean Lee:With her phone tucked away.

30:18The Pragmatic Engineer:Yeah, she came after the announcement. But the news was about to hit the public and they wanted to tell us before the news hit. And I noticed that Yeon and Brian were making this face and I couldn't tell what it was. And then they made the announcement, WhatsApp has been acquired by Facebook for$19 billion. And I realized, oh, that was them trying to hide their excitement.

30:46Jean Lee:That was a face?

30:47The Pragmatic Engineer:Kind of smiley, but not smiley. And that was a really exciting moment. And I kind of joned out for a little bit because I was trying to remember, hey, like how many shares did I get? Again, it was my first startup ever. I didn't even negotiate my equity. And honestly, I couldn't remember how much equity I had. And I was trying to think how much is a billion dollars? That seems like a lot of money. And how much is like 1 % of 19 billion? I couldn't do the math. And I remember sitting there thinking, like trying to do the math. And then I thought, you know, no matter how the math works, I think one thing is clear.

31:25The Pragmatic Engineer:I'm going to be rich. And then Zuckerberg walked in.

31:30Jean Lee:Zuckerberg walked in to the meeting. Yeah. Wow. And then you had like a Q &A or something.

31:36The Pragmatic Engineer:We did. We did. Yeah.

31:37Jean Lee:What kind of questions can you ask at this point? Or what kind of questions did people ask?

31:41The Pragmatic Engineer:There was a mix of excitement and nervousness, right? Are we going to have to change everything? Because I think a lot of the engineers were more experienced and they talked about how when Yahoo acquired companies, they changed 100 % and lost the essence of the business. So there were a lot of questions around that. And Mark is actually very charismatic in person. And I thought he had great answers at the time. he made sure everyone feels assured that nothing's going to change and he will try to maintain it as much as possible at least that was the messaging at the time clearly this this was an amazing exit

32:27Jean Lee:and to this date it's not really been repeated maybe a few companies might have come close but definitely not with such a small team how did you and and your colleagues deal with the fact that wow, you've just got an amazing financial exit. But I guess the company kind of continues inside of Meta. Like it seems like, you know, two things at the same time. Like, okay, I have this like amazing financial exit, but there's also work. How do I balance? How did you balance? How do you decide what next?

32:56The Pragmatic Engineer:That's two folds. So the finance side, in terms of that aspect, we actually got a lot of support. our business person organized many meetings with like the accountants or even a financial advisor. We invited a professor who was the founder of Wealthfront and he gave us an hour of finance advice and he recommended books. I read The Random Walk Down Wall Street, which is a great book. I recommend people read it if you're interested in financial management. And I read several other books to really educate myself to be able to manage this new wealth that I came across as a young 29 year old.

33:44Jean Lee:Yeah. What changed the day to day once you officially became part of Facebook? Did you have to move offices that, you know, did you get a new title added to like the meta org chart, that kind of stuff?

33:56The Pragmatic Engineer:The changes were very slow in the beginning. We didn't even move into the meta, where at the time it was called Facebook headquarters, Menlo Park, until at least a couple years after the acquisition. So in the beginning, everything was same as usual. We still had our old office, or we did actually move to a little bit nicer office, a slightly bigger office. But other than that, it was business as usual. It was Jan and Brian, and we were hiring, but not, at our similar, so steady pace. And I think not until when we actually moved into the Facebook office, we started seeing a little bit more cultural influence and merging.

34:44We started using their HR services

34:49The Pragmatic Engineer:or recruiting services and things like that. But it was a very gradual change over time.

34:54Jean Lee:And then when WhatsApp became part of Facebook, as I understand it it still is even to the state its own organization like inside of Facebook I understand there's organizations like Messenger or like there's the Facebook group etc so like did WhatsApp remain its own kind of organization a little bit shielded from the rest of Facebook?

35:12The Pragmatic Engineer:We had our own area before WhatsApp and in the beginning we even had like our own chairs and our own whatever like walls and decorations that we were using we brought them all over but But over time, you know, there was more and more mixing.

35:29Jean Lee:After the acquisition, how did you start to hire more people? How did the projects change? Did things become more ambitious? Did you start to add more features? Because clearly, like, you were about 30 of you. And then in a few years, there was hundreds of people working on WhatsApp. These days, it must be thousands of people. And like with those people, like, what new work came up? Because, again, originally, WhatsApp was so minimalist, right? And kind of so scrappy.

35:52The Pragmatic Engineer:I guess we were choosing to be small. Not that there was not enough work for us to do, right? So one of the reasons why we also tried to remain small was actually Brian and Jan did not want to raise too much money. And it actually costs a lot of money to serve so many users. You have to pay for the servers. You have to pay for the SMS registration codes. Every year, Jan and Brian would do all-hands meetings. So we did have meetings.

36:20Jean Lee:Once a year?

36:22The Pragmatic Engineer:and Brian was very transparent. He will walk through our earnings and expenses. Oh, interesting. Yeah, I had a lot of information around this. So the three main buckets of our spending was server cost, was about a third, and then about a third on salaries for the engineers mostly. And then a third, the rest was for the SMS fee. When you try to register, you get that code and we have to pay that 10 cents or whatever, how much it costs to send up international messaging. Those numbers, I mean, they add up when you have millions of people using your app. So they actually didn't want to grow too fast because it gets very expensive.

37:07The Pragmatic Engineer:WhatsApp was free for the first year. And then after that, WhatsApp was charging$1 for every year. But they were only using it in certain countries really to suppress growth because they didn't want to grow too fast.

37:22Jean Lee:Fascinating. Because I remember in Europe and in the U.S., there was this$1 cost, which I think people were like, yeah, well, whatever. I don't think we realized that this was a growth suppression tactic. Fascinating. And then when Facebook acquired, I guess they got rid of it.

37:37The Pragmatic Engineer:Yeah, Facebook said, we don't need the dollar. We can grow as much as we can because they had the funding for it.

37:43Jean Lee:And then growth just, did it speed up? Do you remember?

37:46The Pragmatic Engineer:It did, yeah.

37:47Jean Lee:Incredible detail. Using payment to slow down growth.

37:51The Pragmatic Engineer:The lesser known detail about the$1 is that that$1 was enough to pay for all of these. The server cost, the salaries, and the SMS code.

38:01Jean Lee:Per year. So you were roughly break-even.

38:04The Pragmatic Engineer:Break-even. We did have funding from Sequoia, but we never touched that money.

38:10Jean Lee:Incredible.

38:10The Pragmatic Engineer:Yeah. Brian explained it as how his dad was a business owner, and they would wake up in the middle of the night worried, what if I cannot pay the salaries for the employees tomorrow? And he explained that he took the funding from Sequoia as like a backup.

38:29Jean Lee:And I think it was$8 million of funding, if I recall, if I looked at that backup.

38:33The Pragmatic Engineer:Yeah, so we never touched that money. The$1 paid for everything.

38:37Jean Lee:And it slowed down growth enough to be manageable. Yeah. When you joined Facebook, what title did you get? and how did your career change?

38:45The Pragmatic Engineer:So the thing about Facebook is that everyone's actually a software engineer. I'm pretty sure they still don't have titles.

38:53Jean Lee:They don't have titles, but they have levels. What level did you come in at?

38:56The Pragmatic Engineer:So being one of the five youngest people, I got leveled as a junior engineer.

39:02Jean Lee:No, you did not. Yeah. L3 or L4?

39:06The Pragmatic Engineer:L3, L3, yeah.

39:08Jean Lee:No.

39:08The Pragmatic Engineer:I had to climb all over again.

39:11Jean Lee:Oh my gosh, that must have been a bit awkward.

39:14The Pragmatic Engineer:I was not too happy about it, but what's the alternative? Do I want to give up fasting the rest of the shares? And eventually I got promoted.

39:22Jean Lee:But it was within WhatsApp, so you got promoted pretty quickly. How many times did you get promoted there?

39:27The Pragmatic Engineer:A few times. I mean, I eventually became an engineering manager.

39:31Jean Lee:And then as you became an engineering manager, at some point you decided to help and start a new office in London. How did that decision come and how did you go about it? that was actually an ask from facebook headquarters so they said hey like we're

39:48The Pragmatic Engineer:actually running out of space in menlo park and also whatsapp is so big in europe so why not have a presence there it'll be much easier to hire engineers because everybody actually uses whatsapp so let's let's start a new office there and we didn't have that many engineering managers right I was very lucky because I got asked to go along with a couple other engineering managers. And all three of us actually became managers around the same time. We actually even trained together. We were relatively new managers when we got asked to go there. But I think we were the only ones who could go because, you know, people have children and they have to think about school and they couldn't go.

40:27The Pragmatic Engineer:I remember one, the director that I was working with, he couldn't go because his wife says she doesn't want to move with the children.

40:34Jean Lee:It makes perfect sense. You arrived in London, you landed with these two or three other engineering managers. How did you start to grow the office? From a practical perspective, what can I imagine? Like, you know, like how did you start hiring or leasing space? Or what are the other things that you had to do that, you know, like were maybe a little bit unexpected for you?

40:54The Pragmatic Engineer:A lot of the logistical part was taken care of for us because Facebook already had an office there. So we kind of moved in. we got our own section and it wasn't big because at the time again we had a lot of contractors in Europe so we had one contractor already in England so we turned we converted them full-time and then we had one in Scotland we also converted him full-time so he would commute from Scotland every now and then so we had two engineers plus three managers and we started hiring there I think the hiring part was something that took longer to set up. We worked very closely with the Facebook hiring team, which was really great that we already had people who are familiar with the local recruiting logistics there.

41:42The Pragmatic Engineer:So one thing we focused on a lot was really letting engineers know, hey, WhatsApp is hiring in Europe now. Come apply because we were hiring from all over Europe and also a lot from India. Yeah.

41:54Jean Lee:Do you feel it was easier to hire for WhatsApp in Europe just because people knew about it? Did you get more excitement, more applicants?

42:00The Pragmatic Engineer:A hundred percent. You wouldn't believe like I used to do a lot of university recruiting. And when I used to go to Stanford, maybe 2013, like anytime before the acquisition, I would say, hey, like the people will come up to the booth and I would say, hey, do you want to give me a resume? And they would be like, tell me about your company first. Because they have never heard of WhatsApp. What is this company? I'm not even going to give you my resume. I'm not going to give you my resume.

42:29Jean Lee:I have only 20 of these.

42:32The Pragmatic Engineer:Versus in Europe, people were actually excited to talk to us.

42:36Jean Lee:What were the good and bad things of working in what basically is a remote office? Like, yes, London was a big office, but HQ was in California, Menlo Park. That's eight hours of time zone difference, a lot less overlap. There's probably some good things about this and downsides.

42:52The Pragmatic Engineer:It helped because the three of us were from Menlo Park and we actually had great relationships with other teams and other engineers and other managers. And we also traveled back to Menlo Park every quarter. And then we had the leadership from Menlo Park also travel to London almost every quarter. So there was a lot of back and forth to really strengthen the relationship in the beginning.

43:19Jean Lee:Your growth went to like being, I guess, one of the most junior people in WhatsApp, which is crazy to say because you were experienced as well. But then you were also L3 and Facebook, which I still cannot believe. But you went and became a manager. What pushed you to actually say, I actually want to try to manage people?

43:37The Pragmatic Engineer:I actually never asked for it myself. Someone on my team begged my manager, hey, can I please report to Jean? And that's how I became a manager.

43:48Jean Lee:Wow. Okay. What do you think this person saw in you? They wanted to report you when you were not a manager.

43:53The Pragmatic Engineer:I was the tech lead, so I was already managing the project. So it was sort of a natural transition for me.

43:59Jean Lee:And when you became a manager, what parts of the job came naturally to you and what parts were hard that you had to learn or get mentorship for?

44:07The Pragmatic Engineer:You know, I started reading books. I love reading books. So whenever there's a new challenge, I like to read, learn, and research. There actually at the time weren't a lot of courses on how to become a manager and not a lot of books. Like I still don't think there are too many books about how to become a manager.

44:25Jean Lee:There's a little bit more now. There's like three or four good ones, but they all came out after like 2015 or 2016.

44:31The Pragmatic Engineer:Yeah, the resources were pretty limited, but I did what I can to read as much as I can about leadership. And I think I've read actually a lot about communication and psychology. There's several books. Like I love the book Surrounded by Idiots. Have you read that one? It talks about the disc personality, the different types of personalities. And I try to really understand like what motivates people, how do you communicate with people in a way that makes sense to the other person. And also I reflected personally, like what were some good managers and bad manager in my experience? Because you hear the saying that people don't leave companies, they leave managers, right?

45:17The Pragmatic Engineer:Your manager can really break or make your career. They can make your life miserable if you're matched with someone you don't fight with.

45:25Jean Lee:What are the traits that you found? As you recalled, what were things you said like, I think this makes a good manager. I want to do more of that. And I think these were terrible managers or bad managers. And I want to avoid doing that. Do you remember some things that stuck out?

45:38The Pragmatic Engineer:Yeah, I try to really understand each individual person. So for example, like one person that I had on my team really loves going deep into problems or debugging and finding out how to improve things, right? Whereas another person really loves building new features. And you cannot ask this person who loves to build new features to go debug 10 bugs. That person will go nuts, right? And then one person who was really good at building new features was not so great at mentoring new colleagues. So I try to really look for their strengths. And of course, you also want to set them up for challenges so they can learn as well.

46:21The Pragmatic Engineer:But you want to balance them out. So I try to really understand by asking them a lot of questions to understand, like, how do they want to be challenged? When do they feel excited about their work? Or what are the things that they're really good at? What are the things they want to improve on? So I spent a lot of time really talking to them.

46:39Jean Lee:As a manager, you were part of calibration meetings, right? Now that you're not at WhatsApp, not at Meta, can we talk honestly about what are those meetings like? you know what are maybe the good things about them how can you prepare and what's the kind of reality because I feel outside of a small group of managers who are in there it's not many people

47:01The Pragmatic Engineer:know like how these things go so people number one biggest mistake people make is they think your manager is the one giving you a promotion or a salary boost like as a manager middle manager I have no authority to give you a promotion.

47:19Jean Lee:You have no budget. Typically directors have a discretionary budget sometimes to be able to give a reward, but not even a promotion they cannot give, right?

47:28The Pragmatic Engineer:Right. And the bonuses are tied to your performance review, right? So at Meta, for every level, there's exact percentages lined up by the comp team. I have no control over it. The only control I have is I think of myself as the lawyer representing my clients.

47:46Jean Lee:Wow. Yeah.

47:47The Pragmatic Engineer:It's true. I'm making a case for them. Yeah. Why they deserve to get a certain performance review rating or a promotion. And obviously, I want my clients to do well. I want my team to get the recognition that they deserve because I know they worked hard. But it's not up to me. All the other managers also have to agree. That is the nature of performance reviews.

48:13Jean Lee:And being specific, on a performance review, who were the people that you saw, the engineers, who got these high performance reviews from this committee? What kind of tactics did you see? Were there things where like, well, some managers kind of like, you know, politics, where they kind of like, they're calling in favors for each other and pushing someone up? Or was it mostly meritocracy, meaning this engineer was actually doing great work that a lot of managers saw. And they just naturally agreed that, you know, this person who's on Gene's team is actually, they should be above my great person.

48:47Jean Lee:And I kind of agree with that. Because there's bucketing, right? Let's be curious, bucketing. You're going to have buckets and you need to put like, I don't know, X people in the top bucket, middle bucket, bottom bucket, and so on.

48:58The Pragmatic Engineer:Yeah. Yeah. When I was coaching engineers, I learned that different companies have different ways of self-promotion. So like, for example, I heard some companies use emails like they send mass emails every time they do a new release or launch. Or like at WhatsApp, we use WhatsApp groups for everything. But at Facebook, they used Facebook Workplace, which is like Facebook groups where you have a group for team or your org and you're like everything has a different group. And I noticed as I'm representing my clients during performance reviews, the people who post the most often, who have the most visibility, usually get the easiest consensus.

49:45The Pragmatic Engineer:Because it's just like all very natural. If I have no clue what you worked on and your manager tells me you're great, maybe, but how would I know? I don't know anything about you, so I'm less likely to be inclined to agree with your manager. Maybe your manager is right, but I don't know. Whereas if you have been actively posting and telling me indirectly or directly what type of work you have done and what type of impact that has made and what are the lessons that you've learned and what type of people you work with, then I already know. Oh, OK. Like when your manager tells me you're ready, then I say, yeah.

50:24Jean Lee:And then in terms of WordPress, this was actually like it's more than just groups. It was like this Facebook feed where, you know, like it's a bit like LinkedIn, right? just to make it so you see these posts come across the company and sometimes you will put hit like and what you're saying is like if you've seen this post from this engineer on some other team saying oh we've launched this feature here's an interesting thing we've learned that we're using for facebook and i hit like uh i now remember it and then when performance review comes like oh

50:48The Pragmatic Engineer:i remember that person they wrote that exactly and i might even have some questions right maybe like if your manager tells me i might be like well what about this what about that but if you make a post I can just ask you directly through the comments, right? There's a lot of engagement happening in the comments. I might ask, have you thought about this other thing? Have you thought about this thing? And you might give me answers and I think, oh, okay, yeah, he's thought about it. He's really good.

51:12Jean Lee:It's amusing because it sounds like simplifying a little bit, but to be successful at Facebook, you need to also be good inside of the Facebook app and do interesting work and not hide it, actually make it visible. That's interesting. Now, stepping up a step back, Ian, you were a manager at Facebook. you saw a lot of engineers. Outside of the performance and people posting about it, what traits did the best engineers that you remember share? Like what made them so good?

51:39The Pragmatic Engineer:I struggle with this question a little bit because there's a difference between like how do you measure skill? How do you measure what a good engineer is? Is a good engineer someone who can bang out new features? Is a good engineer someone who can design a complicated system? Is a good engineer someone who can communicate all of this and explain it to non-technical people? I struggle a little bit with the definition of a good engineer because I can have a definition of a good engineer, but it may be different for every culture. A different company might have different definitions.

52:16Jean Lee:That's a good one. At Facebook, what was the definition? I remember that a lot of it went down to just a very simple characteristic impact, right?

52:25The Pragmatic Engineer:Definitely. And I think there are many ways to measure impact. And definitely at Facebook, their way of measuring impact was through these posts. If I know about your work and you tell me you have impact, and I agree, that's impact.

52:41Jean Lee:So going back to when you were in London office and started to grow, at what point did the London office start to feel less of a startup, a scrappy startup, and more of a big tech?

52:51The Pragmatic Engineer:I remember a time after about a year and a half or so, I realized I don't know who that person is or I don't know their name. That was a turning point.

53:03Jean Lee:And at what point did you actually start to think of leaving Facebook?

53:08The Pragmatic Engineer:I think I really enjoyed the intimate environment. So I appreciate being able to like at WhatsApp with 30 engineers. I knew everyone's names. I knew where everybody lived. I knew their spouses and their children and their dogs' names, right? I really like that type of intimate environment. We still hang out. We have a pretty strong bond. And I feel like when I don't even know this person's name, I just feel less connected.

53:39Jean Lee:Yeah. So was this the point where you decided that maybe it's time for you to leave and do something else?

53:44The Pragmatic Engineer:Oh, so, okay. I was in London on a contract. So I had a two-year contract. They said, hey, like, go start this office. And then once the contract ended, I had the option to either stay there to continue working in the London office or I could come back to Menlo Park. But then at that point, I had been working there for eight years. And honestly, I think I was pretty burned out. I'm the type of personality who likes to get like A plus on everything I do every single time. so it was pretty tiring after eight years i needed a break yeah and when you left whatsapp

54:24Jean Lee:what did you decide to do what i say whatsapp but it was facebook at that point yeah um i actually

54:31The Pragmatic Engineer:because i know my personality i don't take breaks so i actually had a goal it's simple but i said i will do nothing for the next six months i'm going to challenge myself to do nothing for six months.

54:47Jean Lee:Did you manage?

54:48The Pragmatic Engineer:I did it. I did it. I did read a lot. I exercised. I went on long walks. I did multiple meditation retreats. But that was my challenge to myself to not work for six months.

55:00Jean Lee:So after six months of successfully doing nothing, after setting yourself that goal, what did you do to figure out what next?

55:07The Pragmatic Engineer:So initially, I thought maybe I want to go start a new company or join another startup because I like working. I love building things. So I decided, okay, I'm going to start talking to other founders or people who are hiring or people who are looking to start a new company. So I actually talked to 100 founders. I have a spreadsheet to really see like, is there any interesting opportunities that I might feel passionate about joining or building? And then after talking to 100 startups, I realized I wasn't really passionate about joining any of them. And I thought, what would I feel more passionate about?

55:45The Pragmatic Engineer:And what was the thing that I liked the most about working at WhatsApp for the past eight years? And I realized I actually really liked being a manager because I felt like I was creating a culture of support so that other people can really be learning and thriving and, you know, be able to do things freely without people bringing down your neck. There are many things that make for a happy career, but I found it really gratifying to be able to find that from each person and really try to help them out and create whatever that is. It might be different for different people and trying to unblock them so they can really flourish.

56:24The Pragmatic Engineer:And I thought, well, if that's what I really want to do, I don't have to start a new company. I'll just do that part. But I started exploring, like mentoring people. I did a little bit of coaching I don't do anymore. and making videos on YouTube, writing, all of that to see how would I find the best way to support other people.

56:46Jean Lee:And on YouTube and on LinkedIn, you have been sharing a lot of your learnings, your observations. What pushed you to just start sharing way more than before? I think you started to do this publicly after you left Facebook.

56:59The Pragmatic Engineer:I was actually writing a blog about this. So I actually just hit 100K subscribers on YouTube. like last week. Thank you. And I was reflecting, I almost gave up doing YouTube because I was really not comfortable being seen in public. And I've been thinking a lot about this, like my grandma's from North Korea. She escaped during the war. And in that culture, like you are, you do not speak publicly. You don't want to be seen because it's dangerous. And I think there's generations of that still kind of installed in me. The fear of speaking up is real. I felt really uncomfortable. So I almost stopped doing YouTube.

57:46The Pragmatic Engineer:Once one of my videos went viral from early on, and I feel really uncomfortable. But luckily, I was talking to a mentor of mine, and she said, hey, it's okay to do something that you enjoy doing. Just give it a shot. So then I stuck with that I'm so glad I did.

58:03Jean Lee:Speaking of the thing that is happening, of course, right now, AI, you spoke about this on your YouTube channel as well. But from your vantage point, how is AI changing how engineers work, how managers work?

58:16The Pragmatic Engineer:I do find it really interesting how with AI, we're seeing smaller teams emerge. I know that a lot of teams are saying, well, we're small because of AI. but I wonder if it's independent from AI when you're small you're just more efficient because WhatsApp did not use AI but we were efficient because we were small

58:42Jean Lee:and I almost feel that even today I cannot really point to too many teams that are as small as WhatsApp and have that kind of impact and traffic might come to mind but I think even they're bigger So I wonder if there is a maybe just going back to basics with all of this, maybe AI allows to do the way most companies would have wished they operated.

59:07The Pragmatic Engineer:Yeah, and I think there's also a shift in the mindset. Like I remember back in the days, people, when you go to networking events, people would brag about, oh, like we've hired like a thousand new engineers or we're growing X times bigger. And that was a point of brag. And investors also thought that was a good thing. You need to grow. You need to hire more engineers. That was a sign of healthy engineering environment. Whereas nowadays, investors actually think smaller is better, right? Like they don't necessarily push you to hire more people. And I think as a byproduct of hiring less people and staying lean, they have found this new found efficiency and they happen to equate it with AI.

59:52Jean Lee:Although AI, I think it's clear it makes engineers a lot more efficient. Well, we think it makes them efficient because it can generate a lot of code. You can work on more things parallel as happening with agents. How are you seeing the role of software engineers change and also the role of engineering managers?

1:00:08The Pragmatic Engineer:Yeah, I mean, I love AI tools. I use it every day. As a thought partner, I often ask Chachapiti, hey, like, be my executive coach or be a hardware trained futurist and help me find the next trends. Or, you know, there are various ways of really using AI to its full potential. I feel like engineering management is less affected by AI because it requires a lot of like people to people like asking questions and learning about your engineers. AI can maybe help you with that, but I don't see AI replacing that part. But again, because the teams are much smaller, if you were the type of engineering manager who was doing a lot of these like OKR and process and writing documentation, a lot of that part is going to be gone.

1:00:58The Pragmatic Engineer:And I'm kind of glad it will be gone because I don't think it's really necessary.

1:01:03Jean Lee:Yeah, for example, a lot of performance management of, you know, gathering the impact, it can probably be done by asking agents to gather all these things. I remember as an engineer manager, I used to go through gathering all the work that my engineers have done. So on the calibration meeting, I could fairly represent them. And then it turns out that the managers who showed up without doing that, I had an advantage. But that was not fair for the engineers, by the way, right? Maybe I will get rid of this advantage.

1:01:27The Pragmatic Engineer:Yeah, AI will do a lot of the grunt work, more tedious work that maybe engineering managers or even software engineers had to do manually back in the days. Like we had an engineer who was just there to add comments. I think that is something AI can do really well.

1:01:44Jean Lee:If you had to give career advice to a new grad who says, I would like to build a durable career in software engineering in this kind of AI native world, what would you suggest they focus on?

1:01:55The Pragmatic Engineer:I say foundations. You know, tools come and go, languages come and go, but foundations don't go anywhere.

1:02:03Jean Lee:We mentioned WhatsApp. WhatsApp was very small, very efficient. What do you think today's AI or AI-native startups could still learn from WhatsApp that made WhatsApp successful and could probably help them as well?

1:02:18The Pragmatic Engineer:I think of AI. So we went through several trends. Like when I first got my first internship ever, it was a video sharing website. And I've seen how there were dozens of video sharing websites and how the ecosystem changed over time. And then I saw with WhatsApp, there were dozens of other messaging app competitors and how that kind of settled down over time. I think we're living through something similar. There are so many new AI startups and new tools and so easy to get distracted by all the different options. and it can feel quite overwhelming. There are too many options and you can feel the decision paralysis.

1:02:56The Pragmatic Engineer:But really, again, go back to the core foundation. Think about like, if you're a builder, think about what you're building, why you're building. If you're learning, think about why you're learning, what you want to learn. And if you have clear goals of where you want to go, it will really ground you because otherwise you're just going to be all over the place and you might work really hard and end up nowhere.

1:03:16Jean Lee:Do I understand correctly that you're saying that WhatsApp was successful because the goal was clear. Jan said no distractions and all the ideas, but it was very, very marching. Whereas all the other competitors, even all the messaging apps, they got distracted building, oh, let's do like, oh, this cool video feature. Let's do stories. Let's do all of these things. They saw traction and they did a lot of these things. But WhatsApp was very good at doing the core thing well and then slowly adding things that were value added. Is that a fair summary?

1:03:46The Pragmatic Engineer:Yeah. Yeah. And also I noticed this when I started advising and coaching startup founders as well. And also for any engineers who want to join new startups, this is a great way to evaluate new founders. Like some founders, if you're the opposite of Yan, saying no to things, I call it removing distractions, right? You're prioritizing ruthlessly. If you're the opposite of that, imagine what type of startup you end up. You say yes to everything. Maybe it might feel really nice as a 20-something-year-old if I were to go back in time and I go to the founder with all my great ideas. And he says, that's a great idea, Gene.

1:04:23The Pragmatic Engineer:Let's build it. But imagine, like he said that to every single idea that I had, the company will be all over the place. In terms of the long-term growth, it's not a very ideal situation.

1:04:34Jean Lee:Looking back, what are some kind of like pre-AI or not as modern practices that you did at WhatsApp that were really good that today is very modern AI native teams or whoever could benefit from?

1:04:47The Pragmatic Engineer:Yeah, several things come to my mind. I think one of it is by having lean teams, you get several benefits. You get to remove a lot of the distractions and process. And through that, you get two really incredible benefits, which is ownership and the freedom to build things. right because Jan was always like Jan and Brian were always very specific about what we're building but how we're building it was up to debate right I mentioned earlier that the only time we did a actual code review was the first time I made my git commit uh Brian reviewed my code and asked me a bunch of questions so Jan and Brian were both like so technically adept they were really excellent at doing this they would ask we're trying to achieve this like what is the problem here or what is the best way to solve this issue what are like different ways we can approach

1:05:47Jean Lee:this tell me so do i understand correctly that of course the small teams help with a lot of things but then having the founders push people they hire especially early on they almost like push them to excellence right is it fair to say that by brian doing that super detailed code review with you the first time, it just upped your game. And later he didn't even have to do anything, right? Yeah.

1:06:11The Pragmatic Engineer:And there's like multifold, right? Like one is to really challenge me to think critically. And then I took, I learned a lot just from that conversation. And then also like from then on, he never checked my code again. So I know I am responsible, right? And I do believe when you give responsibilities to people, people will step up. I mean, not everyone, but most people will.

1:06:37Jean Lee:But I think this might be a bit underrated. I wonder if we've had a little bit of too over babying of engineers. I remember for a long time, there was this talk in the past five to 10 years as engineering managers, like, well, I have a new grad. It will take the months to onboard. I need to sign up a mentor for at least six months, maybe even a year. And were we over they being these very capable adults you know they're adults right even if even if they're 18 but they're typically 20 something because they came out of college and they're hungry and they're ambitious and maybe we don't need to do as much of it always yeah i think as long as you hire smart

1:07:14The Pragmatic Engineer:people it's kind of like a mold right if you make a mold too small that's that's only the limit of

1:07:21Jean Lee:how far they will grow yeah if the mold is too small you have to throw away a lot of things that could have made excellent material. Finally, you're a reader. What are some books that you would recommend for software engineers or people wanting to grow professionally or in a personal sense?

1:07:37The Pragmatic Engineer:I love reading books. I did. So while I challenged myself to do nothing, I actually read. I actually took a year, but I did read 100 books during that time. That was my doing nothing. Anyways, it kind of depends of what your goals are, but you gave me some specific things like for your career, I think for me, what was really helpful was What Color Is Your Parachute? That helped me really understand my strengths and my goals and priorities in my career and life. I mentioned the book Surrounded by Idiots. I know the title is kind of funny, but it's an excellent book if you want to learn more about how to really communicate and work with different people.

1:08:17The Pragmatic Engineer:If you want to understand finance, I mentioned earlier The Random Walk Down Wall Street. It's a great book for understanding how to manage your money. Yeah, I would recommend those books to start with.

1:08:26Jean Lee:And any fiction books?

1:08:27The Pragmatic Engineer:Hunger Games was one of my favorite books. I read the whole series.

1:08:31Jean Lee:I read it as well. And I almost liked it. I liked the movies as well, but I loved the books.

1:08:37The Pragmatic Engineer:Yeah, yeah. I loved the story of this woman overcoming her challenges.

1:08:42Jean Lee:And everyone else and winning in the end. Yes. Several times.

1:08:45The Pragmatic Engineer:Yes.

1:08:46Jean Lee:Well, Jean, thank you so much.

1:08:48The Pragmatic Engineer:Yeah, thank you. Thank you for having me on the channel.

1:08:50Jean Lee:This was a great conversation. Yeah. I hope you enjoyed this rare conversation with Gene. One thing that stuck with me was Gene's point about why WhatsApp had almost no process and why it worked. Processes exist for audits, for accountability, and for tracking who did what. But when you have 30 people and everyone can see what everyone else is working on, you don't really need a paper trail. You just walk over and talk. This is a good reminder that most processes exist to solve problems that are created by scale and not by the work itself. I also found the Skype contrast really surprising. A thousand engineers, scrum certifications, two-week sprints, and a dedicated scrum master for every team.

1:09:27Jean Lee:I was one of them at Skype. And WhatsApp with 30 people and zero form of methodology was shipping faster and growing faster on every metric that mattered. This is a much-needed reminder that organizational discipline and actual shipping speed are just not the same thing. And I was in the middle of this at Skype, and Gene was at the middle of it in WhatsApp. Finally, it was interesting as a former manager to hear how Jean described performance reviews as a manager herself. She described herself as a lawyer representing her clients, as in she doesn't control the promotion, she just makes the case.

1:09:58Jean Lee:And the engineers who had the easiest time getting component were not necessarily the best engineers, they were the ones who made their work visible. They posted about their launches in the internal Facebook workplace, they engaged in comments, answered questions publicly, and the managers in those calibration rooms are making decisions about people that they might have never worked with directly. So visibility is not just vanity. It's how the system inside larger companies actually works. This is an uncomfortable truth, but I think every engineer at a big company needs to hear it. If you've enjoyed this podcast, please do subscribe on your favorite podcast platform and on YouTube.

1:10:32Jean Lee:A special thank you if you also leave a rating on the show. Thanks and see you in the next one.