Netflix's Blueprint for Building Inspired Engineering Teams | Carol Barrett

0:00Nobody will prioritize it for you. So we try to advocate good judgment from the ground up, from the most initial ways in which you do it. How do you apply good judgment? You can ask for context among the various projects, and then you can decide where it makes sense for you. That said, you're not doing it in vacuum because you do have the list of what is important for the business and where do I want to invest my time. So under a wide construct, you can ask for that context to help improve your judgment and how you prioritize your work. And I think that has been the singular, I would say, very powerful way in which engineers feel empowered to how they spend their time, how they spend their day, and eventually how they spend their quarter.

0:39What's the impact of AI-generated code? How can you measure it? Join Linear B and ThoughtWorks Global Head of AI Software Delivery as we explore the metrics that measure the impact Gen.AI has on the software delivery process in this first-of-its-kind workshop. We'll share new data insights from our Gen.AI impact report, case studies into how teams are successfully leveraging Gen.AI, impact measurements, including adoption, benefits, and risk metrics, plus a live demo of how you can measure the impact of your Gen.AI initiative today. Register at LinearB.io slash events and join us January 25th, January 30th, or on demand to start measuring the ROI of your Gen.AI initiative today.

1:24Hey, everyone. Welcome back to Dev Interrupted. I'm here live at Lead Dev SF, and I'm your co-host, Connor Bronsden. But I'm delighted to be joined today by Carol Barrett. Carol is an engineering leader in consumer identity and access at Netflix, aka the CIA. Carol, what a title. Great to have you with us. Thank you, Connor. I'm happy to be here. And I should clarify, not the actual CIA, actually Netflix, but it's a fun acronym. And I'd love to dig into this a bit more. Can you tell us a bit about your team and what it covers, and then we can dive into the topics. Yeah, absolutely. And I have to say, you know, it's when I was introducing the organization, the CIA, we would just laugh because it's consumer identity and access.

2:04It's meant to be that we keep your secrets. We know your identity, we keep your secrets. And as the name states, we basically look after all of the concerns of our identities, the consumers that use our services in terms of their identity, their email and credentials, as well as authentication concerns and authorization concerns. So they have the ability to access both our product and our content. And we basically deal with all of the systems and technology that handles all of that at Netflix. Kind of important. Maybe crucial in what I would say. So let's open up with this. It feels like doing more with less is kind of a common trope, but is it actually true or do you believe it's an increasing problem or something?

2:49So that's an interesting, it is a common trope, but it's such a misnomer because when you say doing more with less, like what does that mean, right? And so what we're seeing is that we are, because we serve all of Netflix and we serve all of our product priorities, we constantly have, we're constantly in the middle of the action in many ways, like the CIA thrilling action as well. But it's fun because we get to get involved in all of the various product innovations. And at the same time, we also have to invest in keeping our systems healthy because our subscriber base relies on us to be able to get access to their content.

3:25And so this is constant juggling of prioritization that we normally have to do. And, you know, this the doing more with less, I'm like, you know, you can't you can't wash your laundry with less detergent. That's what about AI? And is AI going to automate everything? Maybe eventually, but it's, you know, it's not, I don't believe it's quite there yet for the kind of use cases that we would need. So, so yeah, so it's, it's, it is something that we constantly consider. One of the ways in which we have, we're trying to attempt, I would say, to solve it is through this concept of annual planning. So at Netflix, we only used to do quarterly planning.

4:02We just sort of paid attention to what we were doing in the next quarter, at most the next couple of quarters. But now because we are struggling with really understanding what's coming from product, how do we balance it with the innovation that we need to sustain our technology in engineering? How do we make sure that both product management as well as engineering is really in sync about what we can truly do? And those are strong prioritization exercises that actually happen even at a leadership level. But that's one of the ways in which we're attempting to solve this, just even at Netflix. So, I mean, planning is a great thing to talk about.

4:36What are some of the actionable insights that you may have gathered that other engineering leaders can put into practice based off of the shift that you've made at Netflix? Yeah. One thing is everyone needs to be aligned on what the topmost priorities are. I think that's the key, right? Because if you have leadership that is not aligned at the, really at the executive level to what might be important, then you're always going to be struggling with what needs to happen. And I can say that that prioritization, you can see that alignment happening because resources actually are shifting from one part of the company to another part of the company to be able to sustain this.

5:10And including there are some, there are some, I would say, product opportunities that are more well-funded than others. they're funding the other teams that need to support those product initiatives. So generally it's, you know, the money is where it is in terms of resources and how you support it. And I think that's a very actionable way of making sure that you really follow what you're talking about is you fund the priorities that you believe are the most important. Now, certainly for us as engineering leaders, we have, you know, there's limited resources and there's limited time than what you can do.

5:42So I, and this is what I'm going to talk about tomorrow in my talk, is that I think partners are a great way of also ensuring that they can support you because you may not have the resources, but maybe your partners have some kind of bandwidth that can help support you in your endeavor. We've done things like having, you know, other teams sort of code on our systems or we code on other people's systems and those kinds of sort of ways in which you can explore some of these ways in which you can continue to scale, but not necessarily have a direct set of resources that work with you. You brought up a couple of really important points there.

6:13So one prioritization, like at a leadership level, crucial to have that alignment and crucial to make sure that everyone's kind of growing in the same direction. But it's also true in smaller team units, right? It's so easy for someone to get distracted by the shiny technology they want to use or decide that, you know, their work is taking priority and not look at the holistic team approach. And I think I see that kernel in what you're doing where it's like, this is something that can be applied at leadership level on down and it needs to be aligned or else there's so much risk about resource alignment and lack of resources in key areas.

6:45I'm curious how you do resource allocation tracking at Netflix to make sure that you're actually allocating resources in the right priority order. Yeah, so that's a lot looser because engineering managers have a lot of flexibility in how we need to. And we rely on our engineering leaders as well, both technical and management, to have really strong judgment into what might need to be and where you need to invest your time. So roadmaps continue to be built by the engineers as largely as possible so that the most important aspects get highlighted. and that way when we can even tell product like, hey, I know this is important and this innovation is important, but we need to take a step back and build this infrastructure because once we build it, we can go that much faster.

7:29And that could be, that's really a conversation about having, you know, that kind of prioritization exercise with management. That said, even within our engineering teams, like we just had a two-day offsite just this past week where we talked about what are our biggest priorities for 2024. And the way we did it is what, if we only had five people, where would we invest them? And you sort of really say, okay, you know, even among your own leaders, even among the leaders of my team, we had that prioritization exercise and we stepped away thinking, we know what we need to do to support each other to meet those highest priorities.

8:03And we came to that agreement and we walked away feeling really good because we were aligned with what that is as well. And that brings up something that you brought up a moment ago, which is that partnership piece of building partners within the organization to enable sustained innovation. Can you unpack that concept a bit more? I know you're going to be speaking about it shortly. You know, I'll tell you, one of our back-end engineers, for example, we needed to build some UIs and a back-end engineer decided to step out and say, I'm going to build a UI and then give a great presentation about why back-end engineers should not build UIs.

8:36So it was great. But it was a lesson for us because we realized that as our systems are getting more complicated and more harder to, you know, they're getting a lot more features, they're getting more complicated. How do you scale those systems? How do you make them such that your engineering resources are not going to be the bottleneck and they're the only ones that know how to use the system? So you need to invest in tooling and developer infrastructure, but there's no resources to do that. So then what do you do, right? And so what we're looking at some innovative ways and some of them is, of course, our backend engineers.

9:06It's like, can we basically just have rudimentary screens? We don't have to create anything fancy. But can we do something more rudimentary that we can at least start to go down this path of what we need to do to allow a better developer? The developer experience will not be perfect, but let's start building it and then iterating on it. That's one aspect. Other avenues we're exploring is if we want to consolidate different systems into a single one, that should free up technically some bandwidth among those other teams. There are partner teams that might be able to help us in a joint project that we can do.

9:37Other things that we've done is, you know, created a squad and said, let's just put this, let's see what we can do in a quarter with this group of people to sort of see, you know, and then the squad sort of elects a leader and then the squad can decide what is, what can they do within that one quarter to move this initiative forward. So we'll look at just different, just ways in which we as engineering managers align first that this is important and then really work with the engineers too. And the engineers may align and say, well, this is something that we're putting together and we'd like to do that.

10:07And then we support them in whatever way is possible. It sounds like you're leveraging this priority exercise to then encourage adaptability within these systems and also engineering-led innovation. Would that be accurate? I think so. In our case, and this is purely because we're backend systems and we're critical systems, as you saw, we actually try and do this. It's not a strict priority. So I just wanted to be sure and clear about that because we do allow our engineers the flexibility of saying, today I need to work on this quarter, we want to be able to work on it. But across the year, we want to make sure that there's, we've spent enough time to support our product initiatives.

10:42We have spent sufficient time to support our technical initiatives, as well as what we call our run the business, which is paved path migrations. You know, you modernize your systems and manage your alerts and your on-call operations. So within the year, we anticipate that there's some percentage of fixed time that you would spend across these three. And that's actually been kind of helpful because we're like, okay, I kind of know that if I spent 75 % of my time doing product initiatives, I had no time to build engineering, the engineering infrastructure to scale for the next year, then we can have a conversation and say, how do we improve that time?

11:17What strategies or tactics are you leveraging to build trust in these partnerships so that these conversations can go smoothly and people feel heard? That's a great question. And actually the point of my talk tomorrow, which is really the way to sustain innovation is to really think about trust and adaptability across all of these dimensions of technology and, you know, operations and teams. And my belief is that, you know, partnerships are, it's harder in the remote world, I will say. You have to make a conscious effort to put time on your partner's calendar and follow those opportunities. But some things that we've done is like, we have engineering brown bags, we'll invite our partners to come and talk about what they do.

12:02And just, you know, we might have an icebreaker that says, what's your favorite ice cream? And it's surprising how people, you know, never really think of connecting outside will suddenly have an opportunity to say, oh, yeah, you know, this is what we talk about. So you have something in common. So just bringing partners in, having much more of that interaction. And then that can happen both at the engineering level, at the team level, at the leadership level, really across all levels. That's one aspect. Other things we do is when we're on site, we try and make time for partner lunches and the team is having a lunch anyway and we'll invite our partners and we'll say, hey, you know, we're having lunch.

12:34Do you want to just come and drop by? And that starts a whole bunch of casual conversations. And in fact, I believe that that mechanism of just getting people to meet and connect on a non-project work, non-work level was actually instrumental in healing some of our partnerships that perhaps were a little bit difficult in the past. And we were able to really create some really interesting initiatives, joint initiatives going forward with that as well. So yeah, it's interesting you bring up the kind of org design piece of remote work where there are huge advantages to remote work and recruitment and some of the focus time abilities and a lot of employees say they're happier in those situations or seeking those.

13:12However, you have to be really intentional about how do you actually encourage interaction, set up org design, and encourage partnerships, to your point. So I appreciate the intentionality you're bringing to this social circuitry development to say, okay, like, let's facilitate these connections. Let's build these up. What are other ways that Netflix is adapting to the current world of remote hybrid? Some of the things that we're doing is introducing the idea of weeks of work, weeks of a quarter, where the entire engineering team comes on campus for what we call high bandwidth programming. and we experimented with one week, a quarter, everyone comes in.

13:51We experimented with a lot of structure, with no structure, with limited structure, but we found that the best things that we can do is in that week is make those opportunities for our engineers to connect, for engineers to connect with leaders. So we have what is called an IC day, an ICs program that IC as an informed individual contributor. Totally. We have informed captains at Netflix, too. Thank you for the clarification. Yeah. Yeah. So our engineers basically programmed the day to the way that they wanted. One time they picked a hackathon and just multiple groups got together and did a hackathon.

14:27Another time they picked, you know, hey, let's do these group exercises. This most recent time they had war stories and they created a panel where senior engineers came together and shared war stories about how Netflix grew and the kind of challenges they had. Another thing that they did is, you know, we were three large groups coming together in this over 100 person organization. And the engineer said, we don't know where what our systems are and where they are. So they put together a series of whiteboards and every team came together and drew a system architecture on the whiteboard. This is that's so creative because every every one of these weeks across the quarters is different.

15:03And it's so it's very interesting in how that's programmed. And then we end up having like a lot of fun with, you know, casino tables and all this kind of other stuff as well. So we found that doing hybrid well is about, first of all, ensuring that there's a time when people really have enough time to just connect one on one. There are some there is some structure, but enough of inflexible enough of flexibility in the schedule to allow for that. But also in the downtime. So we actually have a document that talks about how to how to do well in meetings, being inclusive. the kinds of things that, you know, turn your video on, you know, let people see you.

15:39And we do it actually quite intentionally. And it's really helped to continue to keep that connection, even outside of not being physically there in the workplace. It's always interesting for me to hear about how Netflix is adapting to changing circumstances within your engineering team, because honestly, I think a lot of folks in the industry look up to how Netflix approaches engineering. And so this experimentation approach, this adaptability, as you brought up, I think is really crucial. Can you dig into some more examples, maybe a specific one around how you've enabled adaptability within your teams?

16:10So one of the things like I mentioned is we don't have a lot of structure in our planning process. And part of it is because we do, we have at Netflix, we call ourselves stunning colleagues and stunning colleagues support stunning colleagues. To do that, you have to allow the engineers to decide what's best and where they want to invest their time at that moment in time. You can't really structure it for them. They have to structure it themselves. So all the way from our least tenured engineers, we encourage them to say, if you don't know what to work on, then ask your colleagues for what might be important.

16:45But you make the decision about what you believe is most important. Nobody will prioritize it for you. So we try to advocate good judgment from the ground up, from the most initial ways in which you do it. How do you apply good judgment? You can ask for context among the various projects and then you can decide where it makes sense for you. That said, you're not doing it in vacuum because you do have the list of what is important for the business and where do I want to invest my time. So under a wide construct, you can ask for that context to help improve your judgment and how you prioritize your work.

17:17And I think that has been the singular, I would say, very powerful way in which engineers feel empowered to how they spend their time, how they spend their day, and eventually how they spend their quarter. Maybe one engineer might pick, hey, I really want to invest in spending doing product work this year, this quarter, but I will spend a lot of time doing technical, you know, migrations and some of the paved path work in the next quarter. And we just sort of at an engineering leadership level, those are some of the things we might be able to point out and say, you know, do you want to balance your time slightly differently?

17:47That's really interesting. And I think it brings up this concept of the constantly shifting demands of what the product needs and resources. And it sounds like what you're doing is maintaining a balance kind of organically based on the prioritization that you're providing to engineers and their own judgment and knowledge of the systems between introducing new features, supporting existing systems and other needs. Would that be how you would frame it or would you think about a little differently? No, I think that's a reasonably accurate framing of what we're doing, which is it's meant to be organic.

18:21We don't have a tool that says put in what percentage of time you spend on what. We don't have that kind of structure. And honestly, I hope we never have it. Because all we have is a dock at the beginning of the quarter. It used to be the beginning, and I think it'll continue to be so. Engineers commit to what they're going to achieve in that quarter. We have a midpoint review. And then some teams actually like sprint. So they do sprint planning every two to three weeks. but that flexibility is left to the team to do in the way that makes sense for most of them. So even in my teams, we have three different ways of doing roadmaps and we were just talking about, maybe we should have a little bit more structure that talks about what we're committed to and what we're stretching.

18:59But beyond that, we also want the engineers to feel that power, that freedom and that empowerment to say, I'm going to commit to this and this is my commit, this is my stretch. And then we can hold them accountable for what they were able to do and learn from that and do better next time. And I think this speaks to Netflix's broader strategy, as well as the specific tactics you're talking about. So the strategy of hiring high performers and enabling them to do their best work. And as you mentioned, building up trust among them. And this feels like a crucial way to do that, where people are enabled to really make an impact.

19:29And they know that what they are doing, the decisions they're making are impactful on, you know, whether it's a new product feature or supporting existing systems. Yeah. And I mean, there's so much research now that shows that high performing teams want that enablement. And then it really, that enablement and the understanding that they're making impact, the trust that they get out of it is what drives high performance. Is really what drives that performance. And as engineering leaders, it doesn't mean that we are not technical or we step away from the day-to-day. In fact, we are very involved with the details, but it's really in much more of a supporting role than it is with the I'm going to tell you what to do role.

20:03In fact, even escalations, we have a way of talking about, you know, if you and I have conflicting priorities, how do we resolve that? And one of the ways is you go talk to the leaders, but you don't talk to the leaders as saying escalating. Oh, I need you. I need your resources. It's more in terms of get more context and make the decision. But you are really responsible for making the right decision for what you need to do. And I think this context piece, too, that you're mentioning it is incredibly important because you need a high degree of context to enable this kind of trust. Because without it, maybe decisions will happen that are not ideal.

20:38And then you have to jump in and say, oh, well, we have to we have to adjust. But if you can give your teams the context and the understanding of the priorities, they can then go make the decisions. And self-organizing teams can be so powerful, particularly in engineering like this. That's absolutely right. So where are problems or challenges that you see? I wouldn't say this is a pain point, but I would say that some of the things that we're adapting to is that Netflix has recently hired our less tenured colleagues. and our senior engineers who are only focused on building product or building initiatives now suddenly have to also learn how to mentor, how to scale, how to scale their knowledge.

21:12It used to be that, especially for backend systems, you know, one system was owned by one engineer and then they traded that over time. But now suddenly you have a team of these junior engineers who are available to help you with that. So how do you scale your knowledge? How do you scale your context? How do you mentor? So that's like another, I would say, paradigm shift. for Netflix and in how we onboard these engineers and make them productive in the best way possible. It sounds like you're taking the same approach of saying, look again, we're hiring high performers. We trust them to figure out what's the right thing for them.

21:44And we know they're going to deliver because that's the Netflix culture. Yes, yeah, yeah, absolutely. And at the same time, our engineers, even, and we're seeing this, right, they have interned at companies that may not have the same culture. And one engineer that recently joined, had joined another company where the work was so much more structured. They were told what to do and how to do it and all of that stuff. And then they come to Netflix and they're like, I'm not sure what I'm supposed to do here. And we're like, so we there is this transition where it's although you may be a high performer, you really don't you really need some time to understand the culture, understand.

22:20And so we actually talk with them with the culture document like we have. For example, I have breakfast meetings with all my new hires once a month. And the first thing I asked is, what is Netflix's mission, which is transform and entertain the world? And I said, how does your work tie to Netflix's mission? And it was an interesting question. I could see the gears rolling and, you know, turning in the minds of these engineers. But that's really the idea is like, can you see how do you build highly trusting and adapting teams is to build inspired teams? And I'm giving away my talk for tomorrow.

22:52Thank you for the preview. I'm really excited. Yeah. which is, you know, building inspired teams and inspired teams have a strong sense of purpose. Yeah. And so this is about, really is about connecting their purpose to the organization's purpose, which ties to the company's purpose. And then really having a leadership that is inspiring around them. And that of course can manifest itself in different ways. But I think that you cannot treat all engineering tenures, I know that you cannot in my experience. and you if you really want to invest and if we really want to invest in the culture of any company whatever that might be and whether it's Netflix's culture you have to start in my you can't just like go into a conference room with 300 engineers and say this is the Netflix culture yeah you have to do it through these among you know regular reinforcements both through how you perform as stunning colleagues and you set an example but also I believe through these mechanisms where I meet for monthly breakfast meetings.