#88 – The company is the product (Sam Corcos, Josh Clemente & Michael Mizrahi)

A strong culture shapes your company, but a weak culture can make you lose your values altogether. If you embed your principles and values at the root of the entire company, your culture will naturally spread as your company grows. In this episode, Levels Head of Operations Michael Mizrahi and co-founders Sam Corcos and Josh Clemente sat down to chat about the importance of focusing on culture, how culture evolves as the team scales, and why you should think of your company as a product.

04:50 – Don’t just focus on the product Josh said it’s easy to focus on developing the product and neglect the development of company culture, but both are equally important to cultivate a healthy, growing company.

When you’re in this early existential pre-chasm crossing, the pressure is all to focus on the “product product.” And so it feels like a distraction to focus on the “company product.” Everyone’s like, well, we’re going to die if we don’t fix this market fit problem. So why are we thinking about this at all? And I think there’s two types of losing. There’s not getting to product-market fit, which is just existential no matter what. And then there’s, I think, a degree of self-destruction that happens where there is product-market fit potentially, but the company can’t stand on its own legs because it’s now dealing with the chaos of just unintentional culture.

05:29 – Develop your core group Miz said the mark of good company culture is having a core group of people who can operate together under any circumstance.

There are things you can change about the product-market fit problem. There’s pivoting in general, which is an approach if the company needs to make adjustments. But using the argument that it’s too early to focus on the company or we have other priorities for now sets you up for failure if the plan A doesn’t work out. And so you need to have a group of people that can operate together, that can execute regardless of what flavor of mission you’re working on. And so it’s just a more resilient, robust organization if we can take this team that we’ve built and of course there’s specializations and different skill sets that we have. But generally speaking, we can apply that in different areas and be able to tackle different problems as this group of people that knows how to work together and is functional and efficient towards solving different problems. So we don’t know what those will be, but the upfront effort in building that team today will hopefully pay off as the challenges change over time.

09:42 – Size impacts agility Miz said the bigger your company gets, the harder it becomes to communicate and pivot quickly to new projects.

How this manifested on the personal level is that you do have a lot of employees who have really good intent, are fully mission-aligned, understand culture deeply, and are fully bought in and excited, but necessarily with size things slow down. You can’t get communication down more than a few layers. It’s hard to know what people are up to. And the kinds of projects that used to feel like you could just launch them and get 6 or 10 miles down the road within a day and be thrilled about that and be doing everything involved now requires, let’s say, 10 different people to move something an inch. I mean, that’s the feeling. And you get drained of your energy if you don’t know how to work in those larger environments. And so some people do really, really well and can adapt to those changes. And for others, it’s really hard to understand what changes come about and how to still get job satisfaction and feel energized and encouraged in an environment where you need to get satisfaction and celebrate away when you move things in inch, even though you used to be able to go much, much further. And so size plays a role in this.

19:55 – Culture fit is more important than skillset Josh said it’s better to hire a less-skilled person who is a great culture fit rather than someone who has all the necessary skills but isn’t in alignment with your culture.

In my experience, if you have someone who shares a similar perspective of the world and subscribes to the same ways of doing work, you can typically learn and, or most likely select for people who are very good at rapidly picking up the way things are done inside of that cultural environment. They might have difficulty in a completely different environment doing so, but the skills can be learned. It’s more so the principal disagreement or misalignment is not something that is going to be learned because that’s something that comes down to almost an emotional or personality level where that alignment is there and if that alignment is not there, rather, it’s not something necessarily that they want to fix. So I think that there’s… It’s a difficult situation when you’re in that hiring manager seat and you see just rockstar credentials, but the fit isn’t there. But we’ve had a few examples where it was so important that we stuck to our principles here and recognize that, the culture alignment is not here. And even though the CV looks great, this is not a good fit. And so I think we just have to continue to invest in that philosophy.

25:30 – The company is the product Josh said the goal with any company is to build systems and processes that result in building a good product, so the culture you build and the product you build actually mesh together.

People often draw distinction between the company is the product and the product is the product. And I was trying to think of why that is the case. I believe it’s because we think about it as emotional or abstract or philosophical reasons for elements that you’re discussing, but actually they are all practical. The goal is to build a machine that can build the machine. And so it is actually a very, very practical reason. We’re not just philosophizing here about what an organization should be. It’s actually, what is the most efficient version, most effective version of what we are trying to build. And effective means people want to stick around. They don’t get burned out and chewed up and spit out by a machine that is built on fear and reputational anxiety and things like that. We personally want a machine that allows for deep work and allows for transparency and allows for asynchronous lifestyle organization and so on. So I think that’s what we need to reframe is kind of get people thinking it is not distinct. The culture that we’re building and the product we’re building actually mesh perfectly and they are all one goal. It’s just that you have specs and requirements for the product and you have specs and requirements for the company. And we call those principles or values.

27:57 – Focus on how the company works Miz said the success of a company doesn’t depend on the raw materials, but on how the company and culture work as a whole.

The sum is greater than the parts. And it’s not about the assets, it’s not about the raw metal that goes into a Tesla. It’s about the organization from behind it and the nexus of contracts that creates the company between the employees and between the suppliers and between all the obligations. And so the company is the product is then we want to build really effective product that can change and adapt and have feature releases and be nimble and iterative and adjust for who the current users are. Just to play that out. And so I think you were onto something there, Josh, with not necessarily leaning into culture and values, but really just focusing on the company OS. Here’s how we work. Necessarily, we’re interpersonal and we have relationships. And so some of this is emotion on cultural. That’s how we reinforce it. But shifting our mindset from culture being the culture bucket, or the values to being how we operate as a team to get our work done, maybe brings it into the spotlight a little bit more and creates less of an afterthought, even though we put a lot of thought on this afterthought.

38:04 – Build a culture of experimentation Sam said it’s hard to build a culture of experimentation, but when you are open to change, you can test out theories to see what works.

I appreciate that we have people who understand our culture. It’s really hard to build a team of people who are willing to run these kinds of experiments. It really, really is. Most organizations I’ve been involved with, any minor change to how work gets done is an incredible headache. It’s so painful to do anything because everyone is just resistant to change. There’s just this… The default is no, as opposed to a yes, and I really appreciate how willing everyone is to try some wacky experiments. Sometimes the answer is, “Wow, that was a bad idea. Let’s make sure we don’t do that anymore. What did we learn from that?” But other times we learned things that changed the way that we operate, like the audio and video-only experiment that we did with a half a dozen people at the company. We learned some of the limits of what you can do in video and why sometimes doing writing is better. But we learned that and it’s now part of our onboarding. So figuring out all of these different things.

39:58 – Making good decisions is important Sam said it’s easy to suggest ideas, but what’s hard is thinking through those ideas to make sure you’re making a wise decision.

People are more than happy to throw out a half-baked idea and throw a couple engineers on it for a month, whereas you would’ve been much better off just taking a week more to think about it before you commit eight-plus weeks of engineering time, which is your primary constraint as a company on solving that problem. So I think this is something that is… It’s an obvious statement, but it’s generally underappreciated is that making good decisions is really important. And people, I think put too little time into thinking about what good decisions are, and instead just focus on what is achievable and they [inaudible 00:40:44] And solve the things they know they can win at, as opposed to trying to understand what is the really hard thing that needs to be done and really getting to the bottom of that. Making good decisions is really important and generally underappreciated.

47:03 – Optimize your delegation procedure Josh said delegation helps you systematize your projects so you can be more efficient and take on more responsibility.

Delegation is, it’s almost like making sure that you can systematize the project debt that you have and move it on so that you can take on more responsibility. That’s something that an effective company has to be good at. And I’ve never been at a company that optimizes for that. That really thinks how can we enable delegation in a procedure way. It’s really interesting, it feels like the elephant in the room. It almost feels as though there’s some, I can’t put my finger on what it is, but there’s some sort of inertial force that makes people… Maybe this is the giving away the Legos thing where people want to hold on to the things they’ve built. But there’s something else where culture, companies don’t necessarily want to create an environment where people are always churning, ratcheting upwards and clearing their plates and ready to take on more. Because it’s a little bit scary. It’s like, how will we keep them busy? What if everyone suddenly clears their plate? Now we need to think of all this new stuff for them to do. There’s something there where leadership should, across every company, want this outcome where everyone is able to delegate all their work into a more efficient version of itself and move up. We don’t see it and I can’t quite tell why. It almost feels like the risk appetite question where as companies get more successful, they get less risk-tolerant, which feels counterintuitive as well.

52:26 – The pros and cons of depending on your people Miz said that while it’s great to be able to depend on your people, it also means that when you’re a man short, things don’t work as well.

Leaning on people to solve this problem might be the best option today, but it’s also a double edge sword in that you, on the one hand, it’s great to have people. We are really good context machines. We can connect dots very effectively, but on the other side, it doesn’t pass our internal test of what I grimly call the hit by a bus test. If someone gets hit by a bus, do the Looms exist of their workflow and can we do their job in their absence without any input or guidance or direction? What’s really great about our documentation today, particularly in the op side I’ll speak for, is that it does cover that. Everything has a Loom, every workflow’s been recorded, so anyone can pick up and keep going. So we have to find the right mix of that, where we’re not relying on individuals too much or using them where they’re valuable and a value add, but not the core knowledge or the core connector.

Josh Clemente (00:06):

People often draw a distinction between the company is the product and the product is the product. And I was trying to think of why that is the case. I believe it’s because we think about it as emotional or abstract or philosophical reasons or elements that you’re discussing. And actually they are all practical. The goal is to build a machine that can build the machine. And so it is actually a very, very practical reason. We’re not just philosophizing here about what an organization should be. It’s actually what is the most efficient version?

Ben Grynol (00:45):

I’m Ben Grynol part of the early startup team here at Levels. We’re building tech that helps people to understand their metabolic health. And this is your front row seat to everything we do. This is A Whole New Level.

Ben Grynol (01:11):

When Levels first started, Sam Corcos, co-founder and CEO of Levels, he asked a number of other startup founders who had more mature startups, companies that had been around for a of years what’s the number one thing they would go back and do if they could. The answer was consistent across all companies. Every startup founder said culture. This might sound exhaustive. It might sound like something that we reiterate over and over again, but it’s one of those things that we can’t take for granted. People say, why are you focusing so much on culture? Why aren’t you just the building, experimenting? Well, essentially, any company, any startup is an experiment in itself. It’s not just the product that takes experimentation, it is the way of working that’s, for Levels, asynchronous and remote. How do we evolve as the team scales? How do we evolve as more people come on board?

Ben Grynol (02:05):

Well, it’s important that we focus on culture. Culture is something that scales as the team scales. If you embed culture at the root of the entire company, at the foundation that everybody understands what those values are, well, culture will spread. Mark Randolph in the episode that Sam and Mark did back in the summer of 2021, well, Mark said the first 30 people, they indicate the culture for the next 30 people. And those 60 people indicate the culture for the next 60. And that set of 120 people, it’s the next set to 240 and so on. Every time the team, you need to make sure that things are strong. The foundation is there. Fissures turn into cracks, cracks turn into broken foundations and broken foundations, well, you get the point. Things crumble pretty quickly. That’s why we focus so much on culture.

Ben Grynol (03:01):

So Michael Mizrahi, he Head of Operations, Sam Corcos and Josh Clemente, two of the co-founders at the Levels, they all sat down and they discussed this idea of why we focus so much on culture. It’s inherent to the way we work and inherent to our story. If we want to be successful as a team that’s connected and working to make change, well, culture is the foundation upon which everything is built on. Here’s where they kick things off.

Michael Mizrahi (03:34):

We’re here to talk about the company as the product. I think most recently, Josh, I saw you put this in to a job description, I think, for the technical operations manager role. I’m curious on your thought process on building it into that JD, because I hadn’t seen it so explicitly put out there before.

Josh Clemente (03:50):

I started to think more about this as it relates to the individuals, because the only way to make the company is an idea manifested by the people that make up the company. And so all those people have to have the same idea, otherwise it doesn’t happen. So it’s become more clear that should be intrinsically in focus for anyone, especially in a leadership position. That you see that paradigm as important. There are many ways to build a successful company. Mark and Jason joined our forum last week and he talked about this. At the end of the day culture is winning but there are lots of ways to get there. And there are company cultures of all different shapes and sizes, but you kind of have to design the end goal, have some specs and requirements for what the company should be like and then make sure everyone’s moving in that direction.

Sam Corcos (04:37):

And to that point, if you can have this theoretically optimal culture, but if your company is losing, then it’s the wrong culture. Even if you can intellectualize that it’s the right one.

Josh Clemente (04:47):

I think that’s kind of the tricky part is that when you’re in this early existential pre chasm crossing, the pressure is all to focus on the product, product. And so it feels like a distraction to focus on the company product. Everyone’s like, well, we’re going to die if we don’t fix this market fit problem. So why are we thinking about this at all? And I think there’s two types of losing. There’s not getting to product market fit, which is just existential no matter what. And then there’s, I think, a degree of self destruction that happens where there is product market fit potentially, but the company can’t stand on its own legs because it’s now dealing with the chaos of just unintentional culture.

Michael Mizrahi (05:28):

Put another way, there are things you can change about the product market fit problem, right? There’s pivoting in general, which is an approach if the company needs to make adjustments. But using the argument that it’s too early to focus on the company or we have other priorities for now kind of sets you up for failure if the plan A doesn’t work out, right? And so you need to have a group of people that can operate together, that can execute regardless of what flavor of mission you’re working on. And so it’s just a more resilient, robust organization if we can take this team that we’ve built and of course there’s specializations and different skill sets that we have. But generally speaking, we can apply that in different areas and be able to tackle different problems as this group of people that knows how to work together and is functional and efficient towards solving different problems. So we don’t know what those will be, but the upfront effort in building that team today will hopefully pay off as the challenges change over time.

Sam Corcos (06:21):

I mean, it’s not a coincidence that pretty much every founder that I talk to and I think Josh probably has a similar experience when we ask what are the things you wish you had spent more time on. Almost all of them talk about culture as the number one thing. And there’s something about having… The analogy people often use as rowing in the same direction. Building a culture where everyone that is in the organization knows what to work on and understands how to interact with other people. There’s so much friction in most organizations. That’s certainly been my experience where the ability to just get anything done is just so incredibly tedious. It’s like… I haven’t worked in a very large company, but this is one of the things that I hear from them all the time. It’s like you just can’t do anything.

Sam Corcos (07:18):

There’s so much inertia. I was talking to… I don’t remember if I told you this story, but I was talking to a friend of mine who ran operations at a COVID testing company that had to go. They had to hire a lot of people very quickly. They went from approximately zero people to, I think, 7,000 employees in less than a year, which is just not a great idea if you’re trying to build a healthy, long culture. But they had to, they had no choice. And so they made a lot of compromises. And he said that they were just going onto Craigslist of people looking for jobs. It didn’t matter what job they were looking for. They would just hire them. They just need humans. And the challenge they ran into is he said that information could never make it more than one layer down. So if I was telling you… It’s like the game of telephone except just much lossier. And he was saying how the only way to get information transmitted was through a North Korea style, top level loud speaker that just everyone heard.

Sam Corcos (08:21):

It was the top level of the company just shouting to everyone at the same time, the same a message, which if you don’t have… It’s probably not a coincidence that it is North Korea that does this. In 1984, they have the same thing. The speaker that you can’t turn off in your house. It’s probably not a coincidence that if you have a low trust culture that’s not super high performing, that’s pretty much what you have to resort to. So ensuring that you have really good information pathways and that people are set up for success, I think, it’s an incredibly important part. It’s also one of the reasons why it matters a lot to people how they experience working at a company. It is a huge percentage of their life. And most people, I would assume, certainly myself included, don’t want to spend call it 10 or 15% of my entire life working in an environment that’s toxic and detrimental to my personal and mental health.

Michael Mizrahi (09:24):

From personal experience being at a company that started in startup phase, was relatively small, I’m talking about Uber, being 100 people and then fast forwarding to being an organization of 17,000 people five years later, right? Insane growth and there’s a larger story there, but how this manifested on the personal level is that you do have a lot of employees who have really good intent, are fully mission aligned, understand culture deeply and fully bought in and excited, but necessarily with size things slow down. You can’t get communication down more than a few layers. It’s hard to know what people are up to. And the kinds of projects that used to feel like you could just launch them and get six or 10 miles down the road within a day and be thrilled about that and be doing everything involved now requires, let’s say, 10 different people to move something an inch. I mean, that’s the feeling. And you get drained of your energy if you don’t know how to work in those larger environments. And so some people do really, really well and can adapt to those changes.

Michael Mizrahi (10:27):

And for others, it’s really hard to understand what changes come about and how to still get job satisfaction and feel energized and encouraged in an environment where you need to get satisfaction and celebrate away when you move things in inch, even though you used to be able to go much, much further. And so size plays a role in this. But bringing it back to why we’re focusing on it now and how much we’re focusing on it, I think the question that I’ve asked Sam and Josh, I think we’ve all spoken about this is how much is too much. Is there a point where we’re focusing on building the company more than we are the product and how do we identify that line? I think we’ve looked at this holistically by looking at what percentage of our entire team’s time is working on products that impact our members and what efforts are going towards the underlying company building foundation. But it’s still hazy and tough to find that point. So in your opinions, where are we on that spectrum? How will we ever know if it’s too much?

Josh Clemente (11:23):

I think it’s a lot easier to know if it’s too little because things start breaking. And at that point, it’s difficult to recalibrate. I think, it’s really, really hard for a set of mission driven people, this is not going to be quantitative, but it’s very difficult for a set of mission driven people to under invest in, sorry, to over invest in the company culture side of things, because they’re here to solve the main thing and it’s going to be hard for them to tear their attention away and just all of a sudden, they’re just naval gazing all the time and thinking about the company and have completely abandoned the main thing. I think that’s very unlikely. So the entropy direction, we like to talk about things in terms of entropy, the entropy direction is less and less focused and more focused on the product, product.

Josh Clemente (12:09):

And so it almost has to be a degree of uncomfortable investment in it, I think, to be right, because the failure mode is that you get too far along and you realize, we’re in a bad place. And a lot of the conversations Sam referenced between myself and other founders and Sam and other founders are kind of this backwards looking, you know what? We thought we were right. We thought we had this thing nailed in the first 25, 30 employees and so we kind of took our hands off the wheel. And it turns out that’s when everything’s fine. It’s the next doubling and the doubling after that, if you were off by one degree, suddenly you’re off by 100 miles and you didn’t even see it coming.

Sam Corcos (12:47):

I think that’s a great analogy. What is the one in 60 rule? What is it Josh?

Josh Clemente (12:54):

It’s one in 60. It’s been [inaudible 00:12:58]

Sam Corcos (13:00):

The question of how do you know if you’re investing too much in it. I think as with almost all of these things, people index it on what they’re used to or on what other companies are doing. And just being objective about how much time is being spent on things is one really useful way. We spend low single digit company resources on thinking about culture, which is probably not nearly enough. This is one of the big surprises for me, because we spent so… I would say from an external perspective, most people say that we have a lot of documentation and that we put a lot of time and thought into long form writing and strategy. And I’ve… For the first couple years, I wrote most of the strategy docs and it was still only 5% of my actual time was on writing strategy, which means most companies are probably spending less than 1% of their time on company strategy.

Sam Corcos (14:09):

And so it felt like a lot. And there were concerns early on that we’re spending too much time philosophizing and thinking about things and not enough action is being done. I think empirically there’s a lot of action being done. Our track record so far speaks for itself. We’ve made a lot of progress. We’ve created a new category, which is a hard thing to do. And comparing it to others, given that the base case here is failure, almost all startups fail. So comparing ourselves to most other startups which don’t succeed is probably not a great way to go about it. I think this comes down to reasoning from first principles. It’s like, what are the important things, do those. And it’s okay if you get criticized. You have to be able to have an opinion, stick with it. This is one of the things that I remember leading into our seed round when we were talking to Andreessen Horowitz and a bunch of other funds, I did a lot of research on all of the partners, Marc Andreessen included.

Sam Corcos (15:15):

And one of the things that he mentioned, which it really worked to our advantage because this was already the case. But he says he has a lot more respect for founders and teams that have a very strong position and stick to it to the end. They have an opinion, they have a very strong reason for that opinion, and they’re not going to change their mind just because Marc Andreessen said, “Well, I think remote is stupid.” It’s like, “Well, I think you’re wrong. And I guess we’re going to find out.” And having that strong opinion and sticking to it. In our case, we think we have a strong opinion that culture is important. This is informed by lots of other conversations. We’re also probably not spending nearly enough on culture. So we need to figure out what is a maximum that we’d be willing to invest in culture. I’d say probably something like 10 to 20% of company resources if culture includes people ops, feedback processes all of that stuff.

Michael Mizrahi (16:20):

Bringing this down a little bit to get practical and thinking about the levers and the controls that we have for actually building the company product. There’s the explicit culture documentation and memos, which are defining the norms with thinking and discussion and putting that in writing into our memos database and sharing that. So it’s the written culture. There’s the reinforcement of positive behaviors and the kind of feedback on detracting behaviors between people, interpersonally. So what they’re actually doing. There’s the onboarding components, which is the enculturation, teaching people how we work and making sure they understand the details there. We’ve spoken about re-onboarding potentially a year in and going through a refresh training program to make sure that you’re still up to speed on the details and the high level concepts. But all of that is great. I think the biggest lever we have is really around hiring.

Michael Mizrahi (17:12):

Increasingly I think that is the best control we have for getting the right people in early on. And what’s difficult as the company is that you’re hiring profiles start to change. Early on the company’s small and we can hire these multi tools that can do a lot of different things, have a ton of mission alignment, are fully bought into what it means to be at a young startup and on a small team and they can do a ton of different things. As the team gets larger, you start opening up job descriptions to for really specialized roles. And this is where you start to make trade offs between we really need this specific skillset and this is going to sit on a specific team and they can learn the culture. They can adjust to what the company needs them to do on the kind of the larger end.

Michael Mizrahi (17:55):

And that’s where some of those sacrifices start and those decisions start to come up. When you have two candidates, one who might be specialized and perfect for the role and one who might be probably a better culture fit, but has room to grow on that specialization, which do you hire? And that’s a tough situation that I’ve seen a bunch of our hiring managers in lately and members of the team as we’ve grown the team, making these tough decisions. And I think that’s the best control we have for maintaining culture, kind of the front gate. We can do things inside, but that’s a strong one.

Sam Corcos (18:25):

Hiring is everything. And one of the things that I think we’ve done… I’d say the thing that’s been most encouraging is how I feel like every hire we make is a better culture fit than the person before. And because we put so much information out into the wild, people opt in to the way that we do things. I think a company that does this is Buffer where they… If you’re the kind of person that is willing to have public compensation data, you’re probably the kind of person that’s going to be okay with a lot of the other stuff that they do that most people would not be okay with. So having those cultural lightning rods to attract the kind of people that we want, I think, it’s a really good thing we keep doing.

Josh Clemente (19:15):

I think that there, like Sam said, there is an increasing attention on our attention to these sorts of things, which is helping us. But Sam’s got this 2D plot vector space for talent versus cultural alignment, which I think is really helpful here. Where there’s the always hire quadrant where they’re super high talent and super high cultural alignment. And then there’s the more ambiguous and, or the never hire, which is obviously the low talent, low values alignment. And it implies that in the direction of hire is values alignment more so than talent alignment if you look at what those four quadrants turn out to. And I think that’s right. In my experience, if you have someone who shares like a similar perspective of the world and subscribes to the same ways of doing work, you can typically learn and, or most likely select four people who are very good at rapidly picking up the way things are done be inside of that cultural environment, right?

Josh Clemente (20:18):

They might have difficulty in a completely different environment doing so, but it’s like the skills can be learned. It’s more so the principal disagreement or misalignment is not something that is going to be learned because that’s something that comes down to almost an emotional or personality level where that alignment is there and if that alignment is not there, rather, it’s not something necessarily that they want to fix. So I think that there’s… It’s a difficult situation when you’re in that hiring manager seat and you see just rockstar credentials, but the fit isn’t there. But we’ve had a few examples where it was so important that we stuck to our principles here and recognize that, the culture alignment is not here. And even though the CV looks great, this is not a good fit. And so I think we just have to continue to invest in that philosophy.

Sam Corcos (21:03):

I can think of a specific example. I won’t name the role, but I do recall there was a particular role we were down to just a couple of candidates that we really liked and we picked the one that was a better culture fit. And Josh and I reflected on it, I think, three or six months later and we realized how much better it was that we made that decision and how there could have been just so much more friction in decision making. And you almost never regret hiring for culture fit when people are humble and they understand how to add value. Those people can usually grow into something much bigger or at the very least find ways to add value. The people who are very technically capable but are not a culture fit are… and they’re also just so incredibly hard to justify getting rid of once they’re in the organization because they’re adding value.

Sam Corcos (22:02):

You don’t fire your best salesperson, that’s ridiculous. But it does kind of remind me of in the movie Moneyball when they got rid of Jeremy Giambi, I think was their best player at the time because he was just creating a lot of chaos within the organization. And although on paper he was their best player, there were a lot of other externalities that he brought with him that they couldn’t manage. Is also one of the things that I’ve been thinking about is this ties in a lot to what Josh mentioned around taking your hand off the wheel at a certain point, which often seems to happen. Where approximately these doublings, if we use that as a rule of thumb, which seems to roughly coincide with our experience so far. We’re pushing up against really what I think is our third phase change of complexity right now.

Sam Corcos (23:01):

And we’ve been experiencing some of the friction there for the last three-ish months, where prior to that, it was pretty easy to just pay attention to every message that was sent to the entire company. And we didn’t have a context problem. If you wanted to just block off a few hours a day, you could keep track of what everyone is doing at all times. And now we’re getting to a point of context collapse where we can’t really keep track of everything that’s going on, which also starts to degrade trust in execution for other people on the team. Because what is the research team even doing? What is our market team doing? What is partnerships doing? What is editorial doing? And you used to have full visibility into everything that was happening every step of the way and now we’re becoming much more specialized and isolated.

Sam Corcos (23:50):

And so I’ve been trying to think about, we tried adding OKRs about a year ago. And there was a lot of overhead, a lot of thought that went into them and it just ended up not being very useful because as the team was sufficiently small, that we all kind of knew what these things were already and we already… If you just were paying attention, you would just know what everything was. And so updating it on a spreadsheet didn’t add any value. We might now be at that level of complexity with about 50 people where we should reconsider adding a system like that for adding a level of abstraction that reduces friction, that builds trust, that enables whatever that next level of scalability is. So I think that’s a big part of it is that it’s… You really do have to treat it like it’s a different company and the types of problems that you solve are completely different. And you have to always revisit and make sure that the processes you have in place that still serving your purposes.

Josh Clemente (24:49):

I think that’s true. That the implementation of the principle evolves, but I think there are distinct, you can call them the first principles of the culture that you want to retain across all phases. And it’s just a different implementation. It’s important that everyone has some degree of insight and that you don’t have silos of information. And so what we’ve done is we opted to transparency and OKRs was one means of providing additional insight, but it was redundant to what we already had. But maintaining that transparency looks like a different implementation later on in company life. I actually think, I’ve been reflecting during this conversation, people often draw distinction between the company is the product and the product is the product. And I was trying to think of why that is the case. I believe it’s because we think about it as emotional or abstract or philosophical reasons for elements that you’re discussing, but actually they are all practical. The goal is to build a machine that can build the machine.

Josh Clemente (25:56):

And so it is actually a very, very practical reason. We’re not just philosophizing here about what an organization should be. It’s actually, what is the most efficient version, most effective version of what we are trying to build. And effective means people want to stick around. They don’t get burned out and chewed up and spit out by a machine that is built on fear and reputational anxiety and things like that. We personally want a machine that allows for deep work and allows for transparency and allows for asynchronous lifestyle organization and so on. So I think that’s what we need to reframe is kind of get people thinking it is not distinct. The culture that we’re building and the product we’re building actually mash perfectly and they are all one goal. It’s just that you have specs and requirements for the product and you have specs and requirements for the company. And we call those principles or values, which I think is maybe a little too emotional. Maybe we should just call them all specs and requirements or something. You have product specs and you have company specs or something. I don’t know.

Sam Corcos (26:51):

It’s funny. I’m reminded of some… One of the things that Elon Musk talks a lot about, especially as it relates to Tesla, is how building the machine that makes the machine is 100 to 1000 X harder than building a prototype. And I think there’s a lot of truth to that and a lot of wisdom there because building a company is, I think, a lot harder than building most software products. The reality is that most software products are just crud apps. I’m currently working on a prototype of a new communications tool and it’s low double digit hours to get something working. It’s pretty easy. Most software is built on top of tools that make these things very, very easy to do. But building the messaging layer, scaling things, building the team that can continue to iterate sort of like the… If you want the first or second derivative to be positive in terms of how your product is going to function, you need to have a team. And that’s much, much harder than just building and executing as a simple product.

Michael Mizrahi (27:57):

The sum is greater than the parts. And it’s not about the assets, it’s not about the raw metal that goes into a Tesla. It’s about the organization from behind it and the nexus of contracts that creates the company between the employees and between the suppliers and between all the obligations. And so the company is the product is then we want to build really effective product that can change and adapt and have feature releases and be nimble and iterative and adjust for who the current users are, right? Just to play that out.

Michael Mizrahi (28:32):

And so I think you were onto something there, Josh, with not necessarily leaning into culture and values, but really just focusing on the company OS. Here’s how we work. Necessarily, we’re interpersonal and we have relationships. And so some of this is emotion on cultural. That’s how we reinforce it. But shifting our mindset from culture being the culture bucket, or the values to being how we operate as a team to get our work done, maybe brings it into the spotlight a little bit more and creates less of an afterthought, even though we put a lot of thought on this afterthought.

Josh Clemente (29:05):

I mean, there is a category of most companies call them process controls, or there’s a bunch of different ways, waterfall versus agile development. It’s actually not cultural. For some reason, there’s a distinction drawn there because it directly relates to how you build product. And then anything outside of that, as it relates to transparency and boldness and so on, we call those a different type of thing. That’s an abstract cultural value. That doesn’t actually in anything. It’s a lot of corporate lingo. People dismiss it. I don’t think that’s true. And actually it is more like a framework for developments, framework for developing the team that can develop the thing.

Josh Clemente (29:40):

And so in the companies like SpaceX or Tesla, where Elon recognizes that the machine has to be built in order to build the machines, well, there’s another layer up. It’s machines all the way down, but there’s a layer above that which is that Elon has to build an organization that can build the machine that can end up building the ultimate machine. And so I think reframing, it is something that’s really important and a lot of companies would benefit it from thinking about it in that way. Stepping up one layer and recognizing that this has to happen. And it’s either going to happen sort of as a byproduct of your decisions without you focusing on it intentionally, or it can happen intentionally. But in the end, it is contributing directly to your goals.

Michael Mizrahi (30:15):

Interesting analogy, we’re currently going through the process and the steps to get compliant with GDPR as we look at some international markets. And so the steps to do this are implementing a list of maybe 15 policies about how we operate. And these are just template boiler plate copies about what happens around access control or how we handle data management, incident response plan. If we have a power outage, what happens in the office? And so it’s this whole list of policies that this vendor is dropping on us to implement, which is funny because our way of doing this typically, we have process stocks, we have memos around culture, and this is so far removed for a company of our size of how we would actually operate in these kinds of situations. We would absolutely do the right thing and follow these policies.

Michael Mizrahi (31:02):

But the way in which it’s broken out is very foreign to our culture. But this is a requirement and this is… If you have all these policies in place and all these Google docs exist, check, you are now compliant. And so there’s something about the same list of culture or company processes that if these exist, this company is focused on culture or has a process by which they can operate. And so those might be, there’s a way to make a decision and people know how that’s done. Here’s the process for giving feedback, which may be assumed to be self-explanatory, but we have a different way of doing it or set guardrails around how we approach it and plenty more. But I think that’s what we’re talking about very practically is just the list of agreements on how the company operates and having those explicitly defined and trained and reinforced, which is where you can go awry if you just assume that’s in place.

Sam Corcos (31:54):

The… Maybe enforcement in my mind is not the right word, but there’s something about how they are lived rather than… There’s how things are written and then there are how they are applied. Almost all countries have laws against corruption, right? But when you go to a lot of countries outside of the US, corruption is just rampant and people just get paid under the table for things. And in a lot of countries, the black market is actually bigger than the legal market in the entire our country, because going through the normal pathways is so cumbersome. And so this is one of the reasons why we put so much time into making sure that we are all living the values of the company. Because when people start to notice that some people are maybe… As they would say in 19 or in animal farm, some pigs are more equal than others, but maybe the rules don’t quite apply to everyone.

Sam Corcos (32:54):

It starts to undermine trust in the entire system and then you start to realize that none of these rules actually mean anything. That the people who are writing these roles don’t even believe them. They’re supposed to undermine trust in the entire system. And then you can’t make assumptions about anyone else. There’s a lot of indication that country GDP is pretty tightly coupled with the amount of trust that people in that society have towards the people around them. If you can make the assumption that you can leave some, I don’t know, expensive piece of machinery outside and people aren’t just going to immediately steal it, you start to have to build in these guardrails if you can’t trust everyone. It’s like you can’t have expensive infrastructure to provide lighting on streets because people will just immediately tear it down and steal it and they will no enforcement of any of that.

Sam Corcos (33:42):

And so having basic rule of law. There are places that I’ve lived that they have a lot of things written down that are ostensibly the law, but everyone who lives there knows that if you actually call a police officer to enforce one of these rules, they will probably laugh at you and there’s nothing you can do about it. And that’s an incredibly… It’s a sinking feeling when you realize that all of these… You imagine society is based on all of these rules that people are following. And you realize that the people enforcing it. In fact, in… I was reading an interesting book on Russia and modern Russia, there was… I’m trying to remember the exact context, but he brought a crime to the police. And in the course of doing so recognized that the people who were leading these criminal rings were actually the police. The police were the criminals. And so bringing these crimes to the police was incredibly counterproductive.

Sam Corcos (34:47):

And so having trust in the system, there’s a reason why all of these entrepreneurial endeavors happen in the US. For all of the flaws of our legal system, it is still probably the best in the world. Almost everybody has confidence that our judges are not on the take. There are news stories when this happens and it’s incredibly infrequent, and it’s a huge deal because it’s so rare. You can imagine the same news story in Venezuela. It’s like everyone would go, well, of course they are because they’re judges. The judges work for the president. So why wouldn’t they take orders from him? They don’t have an independent judiciary. So anyway, that’s one of the reasons why focusing on culture is so important and continuing to emphasize it. Because if you don’t continue to fight that degradation, you eventually end up in a really bad situation.

Josh Clemente (35:40):

Interesting anecdote about the GDP versus trust kind of correlation there. I’d be curious to see… We’ve had a bit of a plateau in GDP here in the US lately. And I wonder it also feels like there have been trust issues that have developed, or at least they seem to be manifesting. And so I’d be curious how that maps. Wouldn’t surprise me if those are scaling at the same time. But it kind of goes to exactly what we were talking about, which is this is very pragmatic and practical concept. If you want to increase the GDP of the US, should you be focusing on community trust and how to get people to relate to one another more effectively, as opposed to thinking about the machines that are producing the widgets?

Michael Mizrahi (36:19):

It’s all about the people. Always comes back to that.

Sam Corcos (36:23):

One of the things that I really appreciate about our team is that we… Who was I talking to? I’m trying to remember. I was talking to a founder the other day about how there are companies who philosophize and they talk about all these things that they wish they could do, that they think are really important, but then they’re like, Bob, I can never do it here because of X, Y, and Z. And we have a habit of just doing those things and not being afraid of it, which I think is really a positive sign. We’re willing to take the risks, do the experiment, figuring out what… We’re currently struggling with this context collapse problem. And so most companies would just say, well, everyone uses Slack, so we’re just going to do that. But I actually am about 95% of the way through a world without email, which I think both Miz and Tim Kimball recommended to me.

Sam Corcos (37:21):

And there’s a lot of really good stuff in there. I actually think that the core premise I think is correct. I think one of things that he pushes in the book is actually wrong, but not for… I think it’s wrong because he had not considered an alternative. He pushes for a lot more synchronous meetings like daily huddles that you have in scrum. But I think what he didn’t consider is that it’s less about the synchronicity and it’s much more about the medium, which is writing versus audio and video. And I think doing asynchronous audio or video could solve most of those problems in a way that written communication couldn’t. So I appreciate that we have people who understand our culture. It’s really hard to build a team of people who are willing to run these kinds of experiments. It really, really is. Most organizations I’ve been involved with any minor change to how work gets done is an in incredible headache.

Sam Corcos (38:28):

It’s so painful to do anything because everyone is just resistant to change. There’s just this… The default is no, as opposed to a yes, and. I really appreciate how willing everyone is to try some wacky experiments. Sometimes the answer is, “Wow, that was a bad idea. Let’s make sure we don’t do that anymore. What did we learn from that?” But other times we learned things that changed the way that we operate, like the audio and video only experiment that we did with a half a dozen people at the company. We learned some of the limits of what you can do in video and why sometimes doing writing is better. But we learned that and it’s now part of our onboarding. So figuring out all of these different things. One of the things that I know Miz, you and I have talked about is this concept of re-onboarding, which might be an annual thing for people to get re-acclimated to how we do things and take a week off or two weeks off even to just re-onboard. Go through the checklist again. Make sure that all the things.

Sam Corcos (39:34):

And just one of the things that I wrote this in a memo on memos, on why long form memos are important is I think most people underestimate how much work is done on things that should not be worked on at all. And you do the math, especially when it comes to things like engineering time. People are more than happy to throw out a half baked idea and throw a couple engineers on it for a month, whereas you would’ve been much better off just taking a week more to think about it before you commit eight plus weeks of engineering time, which is your primary constraint as a company on solving that problem. So I think this is something that is… It’s an obvious statement, but it’s generally underappreciated is that making good decisions is really important. And people, I think put too little time into thinking about what good decisions are, and instead just focus on what is achievable and they [inaudible 00:40:44]

Sam Corcos (40:43):

And solve the things they know they can win at, as opposed to trying to understand what is the really hard thing that needs to be done and really getting to the bottom of that. Making good decisions is really important and generally underappreciated. And Lauren’s clinical and regulatory strategy document that she just put together was a really, really good illustration of this. And it’s a long document. She spent several weeks on it. Somebody could read that and say, why did we spend all of that time on this particular subject? And the answer is that if we make the wrong decision on there, she spent whole roundup. We’ll say she spent two months writing that document, which she spent less than that. But we’ll say she spent two months writing it. That’s one person. She’s not an engineer. So she’s not the primary constraint of the company.

Sam Corcos (41:36):

If we had picked the wrong path, we would’ve probably committed 200 engineering weeks to the wrong thing. And that’s from our bottleneck. That’s way more expensive than somebody spending the time to really understand what a good decision is. And that’s a really, really strong document. And so making sure that we think through these things and make good decisions is important. And something that Jeff Bezos talks a lot about on Amazon, it’s leaders make good decisions often.

Michael Mizrahi (42:07):

Leaders are right often is the principle. Are right often. I think that ties the line or it draws the line very clearly to a good company, a company that is built well is one that has processes in place to make good decisions. They can make the right bets and then be right about them because they’ve considered it. And if you pull that back a little bit, there’s the writing culture, there’s the memo culture. Those are just building blocks to building a company that can operate and make good decisions. And that is ultimately what we have to do. To your earlier comment about why big companies can’t do this or why this feels so difficult to do in larger environments or in different environments, I think upfront some of these experiments can seem kind of hokey and, is it some eye rolls?

Michael Mizrahi (42:54):

Sam’s saying, I’m only going to communicate via video for an entire week. It’s like, “Okay. I guess go do that. Is that worth our time right now? We have big priorities to focus on. We’ve got this roadmap, we’ve got other things on our mind to execute against, why is this important right now?” And it’s impossible to have seen how that would’ve played out, but now it’s such a core piece of how the company communicates and it’s greased the wheels for all the people joining to be able to communicate more effectively when they need to. We would’ve never figured that out.

Michael Mizrahi (43:24):

And so a lot of times people might hesitate to run these experiments or feel like… We recently put out this medium article about how to intentionally structure and scale company comms, half the reactions are, this is so awesome I can’t wait to try this, and the other half is, this is so awesome I could never do this at my company. And it’s because there’s not a willingness to give things a try and to experiment when it’s hard to see the connection to a bottom line or a return. So there’s a bit of curiosity that has to be built in order to make progress here.

Josh Clemente (43:56):

There’s two things I wanted to hit there. The first was exactly your first point where if good decisions are important, then we have to structure the organization to enable that or to support that. And that’s the deep work philosophy of making sure people have the time to really sink their teeth into a problem and make a good decision as opposed to an ad hoc in between their synchronous meetings. But it kind of reminds me hearing you talk over the experimentation that goes into some of these emergent company culture decisions or values. And it’s very similar to finding product market fit. It’s almost like company mission fit where we need to get to a certain point. And so we’re almost inflicting user testing on ourselves in order to figure out what are the jobs we’re trying to accomplish and what are the most efficient ways to do that.

Josh Clemente (44:42):

And this gets inherently harder once we get to a certain scale. Just like if you hit growth mode and you’re trying experiment with main product features in the midst of an exponential increase in user volume, that doesn’t work. Right now we’re optimizing for upfront feature experimentation to get our product to a point where we feel very confident that we should scale it. We need to do the same thing with the company. We shouldn’t be hitting the gas on growing something that we don’t feel is ready for growth because this is all going to get infinitely more challenging. And for the reasons all those people say on the media article, I could never do this at my org. I think there’s some truth to what you’re saying that, well, that’s because they aren’t willing to try, but also it’s very, very difficult at that scale. It gets exponentially more challenging once you’re across a certain size.

Michael Mizrahi (45:28):

100%.

Sam Corcos (45:29):

It’s also something that I think figuring out how to leverage people on the team is really important. The idea of investing in things like support staff, which we use Athena for outsourced EAs, which have been a really great way of leveraging team time. It’s something that comes pretty naturally to software developers because you’re already using lots of hot keys and aliases and macros. You sort of have to do this in order to be effective as a programmer.

Sam Corcos (46:07):

This is a controversial programming statement, but a lot of people who use VIM as their program environment are sort of, they’re often the pathological extreme of this where they spend a tremendous amount of their time optimizing workflows when they could just use a modern tool that solves a lot of these problems like VS Code, which is what most programmers use. And so there’s a natural tendency from software engineering to know that there is some balance between typing things out repeatedly over and over again, versus writing a macro to save you that the next time you do it. Building a template, when you realize that, wow, I’ve been, I’ve built this thing 10 times. It’s time for me to build an abstraction layer to solve this.

Josh Clemente (46:59):

Delegation, it’s one of the things that we’re working to optimize inside of the company. And delegation is, it’s almost like that. Making sure that you can systematize the project debt that you have and move it on so that you can take on more responsibility. That’s something that an effective company has to be good at. And I’ve never been at a company that optimizes for that. That really thinks has how can we enable delegation in a procedure way. It’s really interesting, it feels like the elephant in the room. It almost feels as though there’s some, I can’t pull my finger out what it is, but there’s some sort of inertial force that makes people… Maybe this is the giving away the Legos thing where people want to hold on to the things they’ve built.

Josh Clemente (47:42):

But there’s something else where culture, companies don’t necessarily want to create an environment where people are always churning, ratcheting upwards and clearing their plates and ready to take on more. Because it’s a little bit scary. It’s like, how will we keep them busy? What if everyone suddenly clears their plate? Now we need to think of all this new stuff for them to do. There’s something there where leadership should, across every company, want this outcome where everyone is able to delegate all their work into a more efficient version of itself and move up. We don’t see it and I can’t quite tell why. It almost feels like the risk appetite question where as companies get more successful, they get less risk tolerant, which feels counterintuitive as well.

Michael Mizrahi (48:21):

I heard an interesting story recently for a friend and I’ll try to recount it accurately. But they work in biotech and they work at a company that acquired an asset from a company that had shut down five or 10 years ago. And so they acquired this molecule and it had already passed a bunch of FDA testing and was approved and good to go to market. And now this company’s job was to bring that therapy to market. When they got into the labs, they’re dealing with pretty technical science here and biology and chemistry, they couldn’t figure out the process to reproduce the drug that they said they could. The institutional knowledge had been lost somewhere along the way. And so they owned the asset and they had acquired it, but the processing and manufacturing capabilities were lost. And it can’t be reconstructed.

Michael Mizrahi (49:09):

You can’t… The FDA would like to see them produce more of this and show them the manufacturing to prove that they can produce it to the same standard. And they simply can’t figure out how to do it, which was a fascinating story to me because they acquired this asset and they had it ready to go but there’s something about the company process that mattered. And I don’t know the details. It could be environmental, it could be really specific to where they were producing it. So the labs they were working in. But my understanding was that a lot of it came down to process that had to be done in order to create this thing and that was lost. And so part of [crosstalk 00:49:42] the process is gone.

Sam Corcos (49:44):

I remember Josh and I were talking about things like why you always have a tech transfer clause when you’re doing IP licensing. Is that just having the IP often doesn’t mean a lot. And it’s also why we think so much about documentation. If we’re building meaningful IP in our software, locking that up in institutional knowledge is not good. And we’re actually… Coincidentally, this ties into the context collapse problem. I’ve been talking with Dave and Ian on how we can do a better job of solving this context problem in engineering, because the engineering org is now 15-ish people. And we are getting to a point where there are so many different parts of the code base that it’s not reasonable for every engineer to know what’s happening in the code base anymore. And so there are… This is sort of a cliche, but Google has something like a dozen implementations of Gchat because there were so many times where in a chat application was the thing that somebody needed and they didn’t know that it was already built somewhere else.

Sam Corcos (50:53):

So they just built it themselves and they have all these redundant systems that end up getting really messy. So figuring out how to solve that context problem, where, what is the right layer? Andrew’s actually writing a document right now. I think the answer is actually right now we’re just going to solve it with humans, which is the eng manager is the context machine. Where the eng manager is going to be the stakeholder who has to be paying attention and they’re going to have fewer deliverables now at our current stage. In the past, the eng manager was half individual contributor, half manager. And I think now we’re switching more towards manager. Their job is going to be understanding context, reviewing things, being able to give input. Because we had this issue on one of our recent projects where several engineers started working in one direction and that one engineer was looking at the project and said, “Hey, well, we actually already built all this out. It’s over here,” which caused some internal churn within engineer.

Sam Corcos (51:50):

So making sure that we solve those context problems is super important. And just recognizing that the system that we’re building now within engineering, which is the hinge manager as context keeper, that’s going to break maybe when we get to 100, maybe 200, maybe 500. I don’t know. But it’s going to break again. And so we’d need to constantly be vigilant in understanding the problems that we’re facing. And we don’t need to solve the problem of 10,000 people today, we need to solve the problem of 50 people today. And the solution is probably going to be different than it is at 10,000 people.

Michael Mizrahi (52:26):

Leaning on people to solve this problem might be the best option today, but it’s also a double edge sword in that you, on the one hand, it’s great to have people. We are really good context machines. We can connect dots very effectively, but on the other side, it doesn’t pass our internal test of what I kind of grimly call the hit by a bus test. If someone gets hit by a bus, do the looms exist of their workflow and can we do their job in their absence without any input or guidance or direction? What’s really great about our documentation today, particularly in the op side I’ll speak for, is that it does cover that. Everything has a loom, every workflow’s been recorded, so anyone can pick up and keep going. So we have to find the right mix of that, where we’re not relying on individuals too much or using them where they’re valuable and a value add, but not the core knowledge or the core connector.