illustration for The Elegant System of Management - with Will Larson

episode 53

The Elegant System of Management - with Will Larson

Joel Hooks

Everything is a system, and every system is a box in another larger system. It's up to managers to think in systems to make choices and understand their consequences. The manager has an obligation to their team. The decisions of managers make a tremendous impact on folk's lives.

But managers are only human.

"When we go into school there's kind of this sense that authorities are these all-knowing kind of perfect figures that are responsible for everything." The reality is that managers are humans that are dealing with their circumstances. It's easy to put all of the responsibility on them, but they need to be seen as people.

People fail. Managers aren't the only people burdened with responsibility. There's the "10x Engineer" or "The Hero Programmer," that end up with severe burnout.

Managing is an ethical and moral profession. Managers have to make ethical decisions in regards to deciding who gets promoted, fired, given a raise. How do we get justice and results? That's what makes management an elegant puzzle.


Episode Quotes

"At companies that are struggling there is a lot of opportunity. And sometimes it's not easy opportunity, and it's not structured super well it's not packaged up, or you just get to have an easy entry into management."

"There's not someone teaching you there's not an easy path. You just have to go learn by trying, thinking, um going back to the principles and trying to figure out how to make it all come together."

"When we go into school there's kind of this sense that authorities are these all-knowing kind of perfect figures that are responsible for everything."

"Really any fast-growing company is the best place in the world to learn quickly about systems and how they work in reality."

"if you get good at the cold email of asking one short, crisp question that's a good question you can meet anyone."

Resources

Will Larson

Joel Hooks

Transcript

Joel Hooks: Hey Will.

Will Larson: Hey Joel, how are you?

Joel Hooks: I'm doing very well. Thank you, yourself?

Will Larson: I'm doing very well. Excited to get to have a conversation with the individual who gave me my first $500 I made working in technology.

Joel Hooks: Isn't that weird? I remember it because I was just starting too and you helped me with this Django project. That was my first application I ever wrote and I distinctly remember you asking me for $15 an hour and I insisted on paying you more.

Will Larson: I think I deliver some sort of MooTools contents which is the first time I've thought about MooTools and probably like 10 plus years and I don't think it worked very well. I think I delivered that full $15 of value, I don't know if I got up to 50 but I certainly learned a lot along the way.

Joel Hooks: I mean, effectively assigned you with rebuilding the Django admin panel as a UI front end. It was really weird, here's the funny thing that that software actually still lives. Let's sleep with that every night, the fact that that software still lives. It's good to finally talk to you, we worked together a long time ago, over a decade ago. We were both starting out and then we've both had our careers and come to where we are now. What I'm really excited to talk to you today is about your book An Elegant Puzzle, which I've read and quite enjoyed. Which is about systems of engineering management. Can you tell me what systems of engineering management means?

Will Larson: I can certainly try. First, getting to the title on the book took a lot of bad ideas. I think the first 10 to 15 were pretty bad and actually we didn't get to the title until I wrote the introduction to the book and it included the term an elegant puzzle and someone was like, "Your title is terrible, why don't you use this?" I was like, "Yeah, that's definitely better than what I have right now. Then we use that." That was where the title came from. Fortunately, part of writing a book is people who are smarter than you give you feedback. That definitely helped a lot in that case.

Will Larson: In terms of the idea about the subtitle Systems of Engineering Management, I've really found that a lot of management decisions in engineering in particular don't think about the longterm consequences and the longterm components that are interacting. Something that really changed my life for the better is when I read Donella Meadows, Thinking in Systems: A Primer. It gave me this vocabulary in the set of tools to think about longterm consequences and the stuff that I already had these intuitions about as able to actually talk about and codify in a really powerful way. That's where a lot of my approach has come from and that's where the title came from as well.

Joel Hooks: Going with that, what is a system?

Will Larson: That's a good question. I think there's probably like 15 different definitions that we could go with, but maybe the two I would go with that I think are useful. First an academic answer, which is I think systems are basically a reservoir of state and flows in and out of those reservoirs to other reservoirs. A good example on the technology side might be you have a bunch of servers and then they're getting requests and they come in at a certain rate and they get processed and completed at a certain rate. If they come in too much faster than they go out, then you are run out of capacity to service them, that could be like one simple system.

Will Larson: On the people side though, one way to think about it could be your hiring funnel is a system. Your onboarding funnel is a system, your conversion funnel. If you think about users purchasing coming to your site, considering purchasing and actually purchasing. These are all funnels where they have people in different states that are transitioning across them in certain ways. The powerful thing is when you start looking at across the system, it forces you to try to think about not just I want to double sales and have that target. It forces you to try to make a change upstream in the system and understand how that would get you to double its sales. That's a really powerful forcing function to actually understand the depth, the size of work you want versus just focusing on the outcome that you'd be really excited about.

Joel Hooks: I think it's interesting because you talk about hiring as a system and onboarding as a system and sales as a system. In fact, they're all factors in a larger system, right? If we make any changes to these, it affects the others.

Will Larson: Every system it turns out is just like a box in a larger system. I think one of the... to your point, one of the interesting problems is how do you model something useful when the reality is just extraordinarily complicated. To your point, our ability to hire in part depends on our availability to actually get people to buy our product because we have more money. We have a brand that people are excited about people are passionate about the products and so it does all connect. I think, if you look at the full complexity it gets pretty overwhelming. Trying to figure out how to simplify into something that is both useful but also possible to reason about I think is one the specialist skills of systems thinking that you have to develop over repeated failure over the years.

Joel Hooks: Systems thinking to me has been, I don't know, it's one of those things like when you get a little bit of it and you start to buy into this concept of systems thinking, you start to see it everywhere, right? Like everything we do, everything on this planet or even our universe, right? To expand it to the very far out there. It's part of the system and there're actors and things change and we can model and we can try to make predictions. Then, just kind of chaos every day, individual actors making decisions that there's no predicting that. It's really interesting, then at the same time to me it's practically the only way I want to think anymore.

Will Larson: I think prediction is hard, when I think about prediction I think about what is the Isaac Asimov Foundation Series where they develop a mathematical foundation for prediction about the future. I don't think systems thinking helps us predict super well, but I do you think it helps us make choices. For me, the place where I first learned about systems thinking very on time ago was really from Rachel Carson and Silent Spring, which is a look at how pollution in ecosystems surprisingly small amounts can poison the entire ecosystem as like larger fish eat smaller fish, et cetera. I think that that was a thing that was before she did all that research was very hard to understand. It seemed impossible to predict, but then after doing that it became like very easy to reason about. I do think there're cases where these tools make it possible to understand and understanding can lead to some predictive properties anywhere.

Joel Hooks: I wanted to rewind a little bit because we jumped right into your book. I want to talk about that more a bit, I'm curious because this book is about management and you are, and one could assume that he knows something about management. You have experience in this area. I was wondering, I know you started out as a front end developer and I'm wondering what's the path that you took to go from being an individual contributor into managing software and engineers?

Will Larson: To your point, my first full time technical work as an engineer was being a front end developer at Yahoo. Worked on this, actually a product, I have a lot of love for it still, but I was Yahoo builder on search service Yahoo BOSS. It was a pretty cool way to expose Yahoo search index indexes to anyone who wanted to build on top of that, we had an API you could use, et cetera. I did that for about a year and then as I got into my second year there're folks started asking you to work on lots of other pieces of the system. I ended up working in a bit of everything on my second year on that project. Then, Yahoo going through a weird set of times, a bunch of layoffs, trying to understand its place in the industry.

Will Larson: I went to Digg and I think the day before I started at Digg or the Friday before I started on a Monday, the CEO left very suddenly and my friend called me was like, "It's okay. It's going to be fine." The classic Silicon Valley like myth, which the worst thing that can possibly happen is your equity be worth like hundreds of thousands of dollars. That's just the downside risk. It could be way bigger than that and you're not very smart, you're younger, and you're like, "This seems like a great opportunity." You jump in any way. It ended up really, I think maybe three to four months later, I was managing a team, not through any intention to be managing a team, but just because there were a bunch of folks leaving the company.

Will Larson: Some of the advice I give folks, which I think no one has ever appreciated this advice but I keep trying anyway, is that I really think that companies that are struggling, there's a lot of opportunity. Sometimes it's not easy opportunity, and it's not structured super well, or it's not like packaged up, or you just get to have an easy entry into management. It takes a lot of privilege to have the opportunity to fail as badly as I think did in my first management role there. A year into my Digg experience, I was managing the entire engineering team, which was also much smaller than it was when I started, but still I was the only engineering manager at the company and that was how I started. It was hugely informative. It was really the foundation of a lot of what I've learned and my approach which was there's not someone teaching you, there's not an easy path. You just have to go learn by trying, thinking, going back to the principles and trying to figure out how to make it all come together.

Joel Hooks: You have an appendix in your book that talks about all of the books that you've read. I assume that probably started when you realized that managing humans was going to be on your list of activities.

Will Larson: Yeah, I had no idea I was going to become a manager when I joined Digg. Interestingly, I joined Digg so that was my third year in the industry and I wasn't even a senior engineer. I was hired as just like a mid level engineer. I've actually never had the title of senior engineer, I made this unintentional lateral transition at that point and I didn't have the idea I was going to be one and I had no intention to be one. Once you are, I think something that I feel really strongly as this obligation to the team and you are really impacting folks lives in a tremendous way. To me that's a pretty heavy responsibility and I wanted to try to educate myself. Although, it took, it really feels only after six years of management, I started to have the basics down. I think the first five or six years were fairly rocky for everyone involved and I appreciate the folks that have taught me so much, particularly the folks that taught me while letting me try to manage them along the way.

Joel Hooks: Yeah, being patient with you. They're being managed, but they're being patient with you as a manager and helping you to succeed like a mutual respect and an ebb and flow last name.

Will Larson: I think one of the really interesting things is that when we go into school, there's this sense that authorities are these all knowing perfect figures that are responsible for everything. I think in a school that's sort of true, that's a very controlled environment, but in work your managers are often less experienced than you. They're often, dealing with some really complicated stuff in their personal lives. They're often, dealing with all these circumstances, like maybe their peers are struggling or maybe their manager is having a rough time for various reasons. It's easy to put all of this responsibility onto them, but I think looking at them as people who are going through their own life and at their own speed with their own circumstances has been for me really important. Early on I had this impossibly high bar for the managers that I worked with. I think that that was pretty bad for both of us until I started to understand they were people too.

Joel Hooks: You mentioned senior engineers and the title of senior engineer. I think about that that term a lot, because mainly I think about the inverse of that where people are described as junior engineers and sometimes it feels maybe a little condescending, but what is in your mind, how do you define a senior engineer?

Will Larson: First in terms of junior, I've been trying to move to early career as a replacement and that that feels a little bit better. Not perfect, but I think more accurate.

Joel Hooks: Language is hard.

Will Larson: It's a hard but also mobile or it keeps moving and as you learn the right ways there that are wrong again and you have to continue to stay humble enough to keep changing how you speak even though it's not the way you grew up or the way you relearn five years ago or even the way you relearned last year. It's constantly moving target, which is hard but I think important to come to terms with. I think what does a senior engineer and what does a staff engineer and what does a principal engineer and what is a distinguished engineer? These are really, really challenging questions, right?

Will Larson: For me what a senior engineer is wanting a career level. This might be for most companies the majority of folks will advance to this level but might not advance beyond it. Something that you could develop mastery at for decades in your career is one aspect of senior. I think some companies use it in the sense and some companies don't, but to me a senior engineer ought to be able to take an arbitrarily confusing and poorly framed problem, frame it into something that's solvable and useful and lead the execution against it of, like a modest size team, maybe a team of like eight.

Will Larson: To me, if you can do all of those things, then I think you're absolutely a senior engineer. Then how many different types of problems should a senior engineer be able to solve to fit that criteria? I don't actually have a great rule of thumb for that. I think we want our senior engineers, certainly at Stripe that'd be quite broad, but still we have folks who are extraordinarily strong front end engineers and we wouldn't ask them to work on a distributed infrastructure problem, but they're still easily senior engineers in terms of their technical depth, their experience and their ability to take these poorly framed problems and turns them into something useful.

Joel Hooks: I think another aspect to me of a senior engineer, and it can relate directly to experience sure, but then there's also this idea that it's somebody that... it's a leadership role and you're able to mentor, and you're able to teach and help other people. It seems like that, that's probably an important part of at least some job descriptions that include senior in the title.

Will Larson: Mentorship, teaching, being a multiplier, making the team better with you than without. There was obviously the Twitter out outrage about this 10x engineer.

Joel Hooks: Again, it's a recycled outrage, but yeah recently.

Will Larson: You got to keep up, it's like fashion every seven years it's going to be back for you.

Joel Hooks: Yeah, exactly. The bell bottoms of Twitter outrage.

Will Larson: Yeah, that's the bell bottoms of Twitter outrage. Love it. I think there is that engineer is a real person. That engineer who comes in sucks up the oxygen or folks around them become less effective, but they become extremely effective. We've probably all worked with someone, not even just engineers like you see this person in any career. I think there is that person who is effective in part by making others less effective. I think we have to be careful of that person and not think of them as pure individual, "excellence" in that way as senior. Instead, to your point, making the overall some of the system, the team larger.

Joel Hooks: I think there's a section title on your book that is kill the hero programmer, which is not literal. I think that speaks to this point, I was wondering what do you mean when you say, we need to kill the hero programmer?

Will Larson: This is the oldest piece in my book, I first wrote that 10 years ago. I think if I was writing that today, I probably would not have used the term kill. I think I tried to find a less stark way to frame the point. The idea is that I think you find these really well meaning folks who have been captured by their company or their organization or their process in a way where they have to keep working harder and harder to sustain basic function and they don't have enough time to think and to solve the fundamental issue which is causing them to have to work harder and harder. Which is creating them as a single point of execution, a single point of failure. They get put in this really terrible spot where their lives are basically consumed by their work. They also get this outpouring of credit and love from the company or at least often the leadership which views them as this essential individual.

Will Larson: I've been in a lot of situations and I think, I've personally been in that situation myself as the hero. Where you feel like you're carrying it or your assistant office or you're maybe atlas is a better phrase, the world on your shoulders and you think you're doing a great thing and then you're actually breaking the system around you. The idea behind the article is how do we help folks who are doing this amazing work who are trying so hard to help the company or the team or the environment, but are actually undermining the system and themselves realize that and how do we avoid putting them in that situation and keeping them there.

Joel Hooks: It relates to the idea of the 10x engineer and that we've taken this hero and we put them on a pedestal, but then we were actually requiring them to carry this weight. When in fact and even if they are contributing at that level, it's not healthy and it's definitely not healthy over longterm and it might not even be where they want to be in terms of their career, just the lifestyle that comes along with it.

Will Larson: I look at the differences in the 10x engineer and the hero is that the 10x engineer, I think, sometimes they have with this identity as the person who is going to do it all or I think the hero tends to be more of an accident where they pulled into the role. I think you're right that the overlap here is pretty strong.

Joel Hooks: Yeah. Well, the hero is somebody that is maybe identified as a 10x engineer and given the responsibility as such, but didn't really ask for it.

Will Larson: The hero and the 10x engineer as well are this precursor to burnout culture, which is the obvious consequence. I think one of the downsides of folks changing jobs every couple of years is that you can often hide the signs of burnout for two years and then folks don't understand the consequences longterm when people disappear or they go, they're just switching jobs. In reality what you've done is like basically crushed this person's spirit for the moment and they're off like recuperating somewhere in private.

Joel Hooks: Yeah. Squeeze the joy out of the whole thing. Because it's great, and this job can be amazing and solving puzzles constantly is a really wonderful privilege that we get to experience. Then you can take that too far and burnout is real.

Will Larson: To your point, I think what you just described there about the job being amazing is one of the interesting challenges here where you know by and large folks working in software and particularly folks working in Silicon Valley and software or at a successful company like yours are truly some of the most privileged people in America, one of the most privileged countries in the world. Yet they still do have real challenges here created by burnout and figuring out how to have that conversation in a way where we can acknowledge that while also acknowledging the privilege that we're standing upon is I think always a difficult line to it to navigate.

Joel Hooks: I think this really leads to a question I had about the description that you put forth of management as an ethical and or moral profession. I was wondering if you could expand on that and maybe tell me why you consider management to be an ethical moral profession.

Will Larson: To me, what it means for a profession to be ethical is that your decisions matter and they matter in a very real way. For example, when you decide who to promote, I think there's a couple of different ways you could think about that. One way which is very pragmatic is that you promote people based on who you think will leave if they don't get promoted. That is optimizing for retention is I guess one way you might describe that, but then the consequences that you have, the folks who are doing the work you want, the folks who are not demanding are effectively getting punished by not getting promoted. You get into the spot where the people who are doing what you ask them to do are disproportionately disadvantaged. You've created a system where the suckers lose and the suckers here are people like me or people like you or people who have this strong sense of duty to their work.

Will Larson: When I think of management as an ethical profession, I think about it in terms of how do we take accountability for the systems and the behaviors that we put out into the world and make sure that we're creating the justice and the outcomes of our work and not just results. I think results come in line with justice in a lot of companies but I think there're ways to get results without justice in ways to get justice without results. The difficult part of management is finding the way to get both. That to me is where the elegant puzzle component of management really comes into play is trying to find a way to get both justice and results on your teams and in your company.

Joel Hooks: As a manager or even as an independent contributor like incentives play a part in all of this, right? We have incentives to keep people in particular positions or retention. There're incentives to take somebody that cares and leverage that caring to expand their position to meet our business goals and that sort of thing. I'm wondering, how do we manage that? How do we manage incentives and put in the right incentives as managers or people that are contributing to a project?

Will Larson: Part of incentives is that there is real scarcity and so you can't just pay everyone as much as they want without a certain consequence here. Budget, usually you don't have enough roles for everyone that wants to have large roles. This is most true for management roles where there are certain properties that are a card to take. For example, I want to get experience managing managers, there's only so many manager roles that manage managers. I'm working talking about impact in different ways to create this ongoing larger and larger role as larger and larger impact. At certain points there're just constraints that are actually limited. How do you assign those? Is to me, like a big part of that.

Will Larson: I think part of this is you have to let people know honestly and openly about the opportunities that there are and the opportunities have to be aligned with what the company needs. Sometimes I think that means folks won't be able to stay with you at your company and that's okay. Where this gets tricky though is the financial incentives. I think a lot of companies with these stock options and the short exercise periods, you can get into like pretty perverse scenarios where folks can't get the opportunity they want, but they're still doing work of value to the company and they personally feel like they can't leave because the financial loss often just perceived financial loss is too high.

Will Larson: I think that's in particular where a lot of companies, I think optimize their financial compensation around not losing folks, like how do we retain folks? They don't necessarily optimize it for folks being impactful and engaged. I think Netflix is a good example of the ladder where they have not gone towards the option model but instead focus on short term cash comp and rewarding folks while they're there, but also in a way where it's cheap for folks to leave if it's not working out for them.

Joel Hooks: Yeah, that cut ties quick. It goes either way like you can leave, we'll cut you, but we'll also, while you're here you're going to get paid and treated very well.

Will Larson: Well, although something really interesting about the Netflix model that I didn't really understand is part of the reason it works and part of the reason that their engineers are very productive is they really don't spend much time on mentorship, on training by virtue of only hiring later career folks into the company. There are certain elements of that I think work because of the can tradeoffs they've made, but they're pretty extreme tradeoffs I think not most folks are interested in making.

Joel Hooks: Something you discussed and this is related to company culture speaking of that is the idea of opportunity and membership. This is I think the idea of building an inclusive organization where everybody gets opportunity and membership. I'm wondering what are your favorite ways to build a more inclusive workplace with both opportunity to membership for people that work there.

Will Larson: When you start thinking about inclusion and diversity, I think we're a lot of folks start is this idea of how do we hire more folks from underrepresented minorities and more women into the company. Then you often get stuck on this, "Ah, if we could just increase top of funnel." I think there's a lot of ways to increase top of funnel home. Cold sourcing is the most effective one that I've personally found, but events, lots of ways to do it. I think this is taking the easy way out because you're looking at these short term measureables and you do need to do that. This is critical work, but it's only one piece of it and ultimately the success in terms of thinking about an inclusive organization is that folks from any background can come and can be continued to get promoted and rewarded for the work and to feel like they belong to the environment at the same time.

Will Larson: I started thinking about inclusion in terms of, as you described, access to opportunity and then access to membership. For access to opportunity this has been the one that I spent the most time on over the last two years. I've learned I think a fair amount about one anti-pattern you'll see at some companies is that every time there's a critical initiative, like the same two or three people get pulled into it. It's always Sara. It's always like Jeremy or something. It's always the same person getting pulled in. People were like, "I want to do that, but I can't get access to it." Then this is the hero pattern exposing as an inequality across the company for other folks, because most people aren't the hero. How do you create the possibility for other folks to participate but in a way that's also safe where you don't have, you do have to get the work done. It's not okay for just anyone who wants to do it to get these critical projects. You need someone who have confidence can do it.

Will Larson: We've worked on building this selection process where we send out publicly letting people apply, we give them time to apply so they can talk with their manager, talk to their peers and that's outside of the company to get a sense about whether it's the right time for them. We explain the level of commitment and then we ask them to privately apply to see me or whoever's running the process. Then we have a selection criteria to pick the folks, and we give feedback to each person who applies and can double check, double confirm with the person who we choose that they're willing to take on the project, say like an architecture group or something. Leading architecture group something like that.

Will Larson: This has been really powerful because all of a sudden people are aware of these projects. I think they're aware of the downsides of the projects where I think when you only see the glamor, you don't see the risk, the level of effort, the amount of time on the tight timelines or whatnot. Also, it allows us to instrument and measure the cardinality of folks getting these projects and actually understand is this a problem or not? When we started one of our senior engineer leadership groups, one of their goals, one of the metrics they measure themselves against was increasing the number of distinct people leading important projects. That goal and measuring it was only possible because of this selection process we put together.

Will Larson: I think that's been really powerful, it didn't work super well initially we've had to iterate on it a couple of times. For example, I thought people would just apply what I've learned you really have to go out and privately nudge everyone who wants to apply. If you don't personally invite someone to apply they probably won't. It has been different people have different risk tolerances and you have to be thoughtful about that and make sure you encourage folks who might have a higher risk for a certain type of project to take it on.

Joel Hooks: I would assume if you don't do the nudging then you get the usual suspects so to speak.

Will Larson: It is not that effective if you don't do the nudging and it's also time consuming. If you don't do it, if you don't do the nudging it's both slower and it doesn't actually work. It's pretty easy to decide it's a terrible process at all.

Joel Hooks: It's just broken.

Will Larson: Yeah, throw it away.

Joel Hooks: Hiring us is hard in general, I see the technical interview that comes up so much in terms of people just hating it. I think mostly they're railing on the idea of the whiteboard interview or are these high pressure quiz like game show style interview processes, which still exist and are hopefully moving away from that. Is it similar to what you're talking about in terms of opportunity is it like internally for projects, but then externally when we're hiring. Because hiring is hard on both sides, it's hard for candidates, it's hard for firms and organizations to do it. How are you working on making that process better?

Will Larson: I agree with you, I think hiring is hard on all dimensions. I was talking to someone recently, they were, how do I find the company I want to join? Part of the conversation we had that I thought was quite interesting is that if you look at how often women are underrepresented minority is that company they're applying to versus how white men often that company is. On one hand you see a huge amount of back channel exploration trying to understand and then the other you often see effectively know research and attempting to understand. I think it's interesting how the security and privilege of one of those situations makes it possible to not that the company that you're applying too much because you're confident that'll probably work out and if it doesn't you can just switch to another. On seeing how different those approaches are has been really informative to me as well as I've gotten further in time management career and understanding how hiring impacts different communities so, so, so differently.

Will Larson: In terms of how we work on hiring, to me the most important thing for any endeavor is having feedback loops to make sure you're getting better. For example, very early on like five, six years ago, our interviews didn't have rubrics. As we debriefed on trying to understand whether the interviews had gone well or not getting aligned was extremely hard. From that process we realized we need rubrics for every single interview, they should be numerical. They should have four or five points with clear criteria that are objective, and we can use those to start having a shared vocabulary about the performance when choose interviews. Then we iterated on each of the rubrics over time to narrow in and to get more and more consistent.

Will Larson: Then another one is you can still have people who are inconsistent across interviewers. Then, we rolled out this candidate review system where we have folks who review all the feedback, and look for inconsistencies. Then we can go coach folks who are out of alignment with a rubric. We can also see places where the rubrics just aren't working well for us. Layering in these feedback mechanisms where you have a learning process that can keep getting better has been the most important system in terms of the access to opportunity component of it. I think rubrics are quite important, I think having mechanisms where we can see what the rubrics are not getting applied consistently is quite important. Then also just whenever folks decline, we try to do a decline, a post decline call with them and learn what they decided turned them to pick a different company to go with. Although, that's obviously the power dynamics in that situation or are pretty weird where folks often just don't give you the full perspective on why they might not choose to work at a given company.

Joel Hooks: Can you give me an example of a rubric? For myself, I'm just curious what would be a simple rubric that you would use in this process?

Will Larson: We redid our manager of manager interview loop maybe six or seven, maybe nine months ago. Who knows, time flies I'm not good with dates. We were thinking about do we think about in particular folks who have dealt with rapidly growing organizations. A rubric here might be a one hand have never dealt with a team that shrunk or grew, have only dealt with small static size teams and that might be like a one. Then a four might be, has dealt with a team that has grown rapidly at least 50% year over year and they can speak to the metrics showing how they did that in a healthy, sustainable way without burning out the team and that could be the top of the line. They'd have gradients in between with specific accomplishments that they've made or approaches that they demonstrated. For example, anyone with money can grow a team really quickly.

Will Larson: That's not like a four, it's just like hired a great abandon with no foresight is not the goal but instead of being able to talk to the counter pressures were they part of our rubrics. In fact, like an example of not hiring us fast as someone wanted them to and instead being able to explain how they paced the hiring in a thoughtful, effective way.

Joel Hooks: I mean the mass hiring can be, that just gives me that runs cold shivers down my spine actually, just even thinking about it. I need low scale, like small teams that's where I'm best at. I've never had to experience the hyper growth of a successful startup or I guess a potentially successful startup in some cases.

Will Larson: Well, I mean it seems like you're doing something right. It seems, I guess maybe you don't think of yourself as a startup which I think is a particular term that means a lot of things for a lot of people, but clearly you've been able to scale the business really well over time without hiring a ton of folks to do it, which to me seems like even more effective.

Joel Hooks: We're totally not growth oriented we're oblique in a lot of our approaches, just by nature kind of a slow simmer is how we're approaching and it's been interesting. Like you said when you were thrown into management, I had the same experience, right? Like I don't know how to run a business or manage humans and had to learn on the job and it's ups and downs, but overall it's working out pretty well. I want to know about instrumentation of a hiring funnel because this is something that I hadn't, this is the first time I'd heard of this idea. I've heard of instrumentation and software where we're sending data so we can analyze performance. I assume it's similar, but how do you instrument your hiring funnel so that you get valuable metrics back that you can actually use to improve that process?

Will Larson: There's a category of software called Application Tracking Systems like ATS and Greenhouse is one, Lever is another. Those are the two really big ones and I think one of the upsides of these is they support the high end of easy automation. The other upside, which I think is even larger is they actually set up workloads you can instrument and what's happening along the way. How many people are sourced, how many people are referrals, how many people pass the phone screen, fill the phone screen and you have the whole conversion funnel all the way through. Then I think that that gets you 80% of the way there. I think these ATS is trying to keep the data in, which can be a little bit tricky to actually get like good data out of it. You can definitely things like GDPR for example have complicated some of the things you're able to do with your data.

Will Larson: In general, that's 80 or 90% but it's not the entirety of it and that's where you need a system on the other side. Maybe it's your learning management system or your LMS or maybe it's your workday setup or maybe it's a spreadsheet, maybe it's anything but some system where you don't just end the metrics at higher. I think it's really important to first there's like, looking at candidate success, I find this like a little bit hard to get enough data from those companies. They can actually say like, "This person is a strong performer and there's enough of them." You can reach significance to say that this interview loop is or is not effective or something. You can eventually do that at a certain scale, but most companies aren't aren't that size. You can instead start instrumenting like how long did it take them to get to their third pull requests with more than 15 lines of code change or something like some sort of metric around being spun up.

Will Larson: I think that's really powerful because it helps you understand your hiring rate in large part depends on your ability to ramp people up effectively. If you're able to ramp people up quickly, then those people can quickly start training other people and if not, you can't. I really think the hiring funnel that every company should at minimum track through being ramped up to be effective.

Joel Hooks: As a systems and data modeling nerd Stripe must be just an amazing like test bed for all of this information and data and process and just learning and growing in that way, I would think.

Will Larson: Yeah. Any fast growing company, but particularly it's a lot easier when you're fast growing on the way up. I think when I joined Digg was fast growing on the way down and that was less instructive in this regard. Although, also interesting to see how systems like fail to scale down, but really any fast growing company I think is the best place in the world to learn quickly about systems and how they work in reality and not just in principle.

Joel Hooks: My outside offers duration is that it wouldn't be just like any, and that Stripe is special in several different ways. I'm looking at your book, and I'm looking at what this company is doing and I'm so totally like floored because this doesn't make sense, right? Like the Stripe press doesn't, I think it's like an oblique strategy, right? It doesn't really make a lot of sense but it's amazing. Your book is amazing and Tyler Cohen's book is amazing. I just look at this company from the outside and I've read the hiring is done in their cohorts and they start every couple of weeks and they're significant, right? Like non-trivial amounts of people being hired to like feed this global engine of this company that changed my life for one, but I don't know. It's interesting, my Stripe fanboy spiel a little bit.

Will Larson: There's probably 50 people joining every two weeks or something that might even be low. It's humbling to see just how many folks are joining us at this point. It's really shocking considering where we were when I joined not that long ago.

Joel Hooks: Yeah, it's cool. I'm curious about this idea of building a community of learning both inside of the workplace and outside of the workplace, and what that means and how we as participants can help facilitate that sort of thing.

Will Larson: First thinking about the inside versus outside of workplace trade off. I think one of the interesting things about becoming more senior in a large company is that you're not really allowed to express the negative feelings you have. You really have to find a way to contain those. It's not very inspiring for you to be complaining in front of your team a lot. You have to find the most positive version of the truth that is still the truth and really try to stick to that.

Will Larson: I think that to me why as you get more senior having folks outside of the company is so incredibly helpful. Where you are able to be, "Hey, we're doing this. This doesn't seem right, what do you think?" People are like, "No, that's how it works everywhere. You'll just got to get over it." Or like, "No, that that one actually is pretty weird. That you should probably push back on more than you are." Just this grounding sounding board where you can go and get feedback from folks that you're not getting the day to day feedback from. Just like a different era with different perspective for you. I found that incredibly helpful. This is something for myself I've worked pretty hard to cultivate or the last couple of years I realized at one point I didn't have much of a support network in this regard of peers who are just doing similar work.

Will Larson: I've been reaching out, pretty intentionally to one or two a quarter over the last couple of years just trying to find and develop these relationships so that I have them when I need advice, need perspective. If there is one thing, if there was only one piece of advice I can give myself like 10 years ago would have been to start doing this then not trying to pick it up now. It is easier when you've been in the industry longer you know more folks to introduce you to people. I really think if you get good at cold email of asking one really short crisp question, that's a good question you can meet anyone. It's a superpower to write an email that people actually respond to, but it's one you can learn pretty easily if you put a lot of effort into it.

Will Larson: That's the external component, internally I think so many companies are so focused on getting done what they today, this week, this quarter that they don't invest into this community and this idea of longterm learning. To me that's what really makes a company longterm sustainable for you, is that you are doing and learning new things every 18, every 24 month. Where you don't have to switch companies or even necessarily switch roles because what you're able to learn continues to grow with you and around you. To me these manager meetings where we meet every two weeks to discuss an interesting topic, and we can have off sites. We're picking up... we have a couple of reading a book reading groups already and we're picking up another one.

Will Larson: There's all these different ways like paper reading groups where you talk about interesting architecture papers or really anything and it doesn't have to be technical. It can be management oriented. It can be about the industry that your company's in. Just getting folks to break out of the day to day mastery of what they're good at and get back into this learning mindset of engaging with something they aren't super familiar with.

Joel Hooks: Yeah, I agree with that completely. It's important and I think just important as a human to keep learning and growing and pushing the boundaries as well.

Will Larson: Something I've been thinking about a lot for the last year is that I think in Silicon Valley in particular, but in tech in general, part of, we talked about senior engineers before, part of the getting the senior title, often four or five years into your career. I think a lot of folks think they're done after four years, but really it's just the beginning. There's so much more to learn about so many things that I think are really gratifying to keep spending time engaging with and not calcifying at that point.

Joel Hooks: Yeah, and it's amazing. I think, I'm always amazed by the level of information and just cool stuff there is to learn and discover.

Will Larson: It's endless.

Joel Hooks: Yeah. Well, I really appreciate you taking the time out of your day to chat with me. It was great, thank you very much.

Will Larson: Likewise, it's a real pleasure. I hope to do it again sometime soon.

Joel Hooks: Cheers.

More Podcasts