Interview with Sam Lessin, Co-founder & Co-CEO of Fin

Teja: So, why don’t you tell us a little bit about how you got interested in, let’s say technology, entrepreneurship, like what’d you do before your first startup that you sold to Facebook?

Sam: Yeah, so I was always very interested in it. I mean, I honestly have to credit my father a lot.

You know, he was kind of one of these guys in the nineties. Who saw the internet and thought it was the most amazing thing you’ve ever seen and kind of indoctrinated me on that from a young age. And so for me, it was kind of obvious growing up that we were at one of the great moments of discontinuity in human history and that people, everyone kind of, knew the internet was amazing.

It was even more amazing than people realized and even more impactful than people realized. So I did do a brief detour in my life through Bain and company after college. Which was historic, I think I learned a lot. It was a cool thing. I wouldn’t do it again is what I would say. And then kind of quickly found my way back into the tech world.

Where again, I think there was then and continues to be now, incredibly exciting and important and kind of crazy things going on. Right? Talk about what sort of defines humanity for the next hundred years.

Teja: Totally! So you went to Harvard, did you study computer science?

Sam: The program I did at Harvard was social studies, which is a mixture of history and economics and things. I grew up, in high school, and even before with my dad, programming some, but I was never serious about it. So, you know, I’ve always been around computers and did a little programming, but it was never something I studied super seriously. You know, the way obviously a lot of people do.

Teja: Totally. It’s really fashionable these days, I think, to suggest that kids take a technical education or computer science education. I also studied liberal arts in school. I went to the Harvard of Boston, BU.

Sam: That’s great. To be clear, I think that learning to code and skills like that are actually awesome.

This is really funny. You might appreciate it. When I left Facebook and I was VP of product there, when there was only, I guess, two other VPs of products, you know, very, very senior level. I went to a coding boot camp because I hadn’t really coded. I mean, I never was a serious coder, but I hadn’t really coded at all in years. And this technology changed a bunch. So it was very funny because you go into one of these coding boot camps, and you’re working alongside a bunch of pretty young kids who, all they want is to get a job at like Facebook or whatever. They could not understand why I left a very senior job at Facebook and then was doing the coding boot camp.

So I think technical skills are actually really important and awesome. Both because it allows you to prototype on your own. It’s important stuff to understand. We live in a time where the internet and technology are changing everything, right? To the core of our society. We’re having this podcast and a hell of a week, and I think that it’s important. from a liberal arts perspective, and a philosophy, and an economics perspective to have some appreciation on the flip side of the context in which you’re working and what you’re building. To your point, I think you need to be able to go up and down the whole stack, not just the technical stack when you’re out building stuff.

Teja: Yeah, totally. And I mean, I think as maybe software development becomes more ubiquitous and maybe even more commoditized, the skill is actually in making it all work together rather than just building the software.

Sam: That’s true. But again, I think it’s funny because I am a liberal arts guy by background, but I appreciate hard skills.

The reality is, there’s so many smart people running around with great ideas and they don’t know how to execute them. The beauty of coding and the beauty of being an engineer is the freedom to not need anyone else and just start figuring stuff out and you learn by tinkering. So I thematically agree with you, but specifically, I’d say that being an amazing software engineer and spending time on that is highly worth it.

I do it for fun. Even, literally yesterday, I got frustrated with the contact solution I had.
So I’m scripting a different approach.

Teja: Yeah, it’s cool, and thinking about information in the context of computer science, I think is a powerful framework, for sure. The models that you get from computer science are really useful.

Sam: Absolutely. I mean, Claude Shannon, who’s a personal hero, did a lot of what you can think of as computer science today. You know, what he wrote is somewhere between math philosophy and engineering. I think having a grasp of those disciplines and how they intersect is pretty impressive.

Teja: Yeah, totally. That makes sense. So you studied social sciences but got interested in technology because of the impact that it was having on the world.

Sam: Yeah. This generation of technologists is going to define the next several centuries of human history for the better.

Teja: Yeah, totally. Even the government systems that we are probably operating in, Right? It’s going to be a function of information science.

So after Facebook, you became a professional investor? or what’d you do next?

Sam: Well, when I left Facebook, you know, Facebook’s an incredible company. I love them. I love it. I’m a huge fan of it, but it just got really big and I really liked the early days of things in a lot of ways.

When I left, I basically started doing two and a half things. One thing is, I’ve always been an investor. I had put in little tiny baby angel checks when I was too young to be doing such things.

Things like Venmo, for one, where I was kind of the first investor. But also, Do you remember back in the day, maker bot? Printing and all that sort of stuff. So there was an interesting budding New York ecosystem where I kind of seeded or helped seed some companies, even if you remember, Birch box, which is the first subscription box. We started in my job PO office and I gave them their first check. So I have always been involved, but it was more casual.

When I left Facebook, a few friends and I started scaling up a fund called Slow Ventures, which had kind of been like a friendly thing, just a few friends. It was almost like a little club and we started turning it into a real thing. So now it’s about a half a billion-dollar AUM early-stage fund and we’re working on that.

The second thing I did, well, I teamed up with my buddy, Andrew Kortina who started Venmo. He and Ingram are the real founders of Venmo though, for sure. And we started messing around with what I’ll call “human in a loop” systems.

I think that machine learning is really interesting, and you can do really powerful things with it, obviously. But AI is kind of all shitty. The reality is, especially back then, when we started working on this, everyone was like, “AI is gonna solve all the world’s problems.” Not really, there’s a pretty big gap there.

So we got really interested in those types of systems and we built an assistant service called Fin. We were basically saying, look, we want to use a human in the loop approach to do really sophisticated, high-end, thoughtful work; Book me the right flight, get dinner at my place, remind me of the stuff you’d want a super Siri to do. In the process of that, we actually figured out, what has become Fin analytics. Which is a company that’s growing really quickly. We’re very excited about it.

The third thing is, I do a bunch of writing. I love writing. My wife runs a publication called The Information. So, I get to write columns there and that’s kind of my half job if that makes sense. But the real thing I spend time on has been Fin, obviously, and investing.

Teja: Yeah, I saw that you consider yourself an intern at that company, your wife’s company. It’s funny because you’re a GP, you’re a co-founder of a scaling company, and then you also have time to write on the side. I guess they all kind of work together.

Sam: I think the reality is, the dirty reality, is at a certain point, all CEOs just want to be interns.

Teja: Oh my gosh, one hundred percent. I actually thought about going to apply to Blizzard and just figuring out how they write game software because I think that’s cool. I play a lot of computer games, so I would love to, but I don’t know if I want to see how the sausage is made. I feel like maybe that would ruin my love for the past time.

How did Fin evolve into Fin Analytics? Is it still the same company?

Sam: When Kortina and I started fin, we did a thing, which is kind of a luxury to be able to do. We were both second-time entrepreneurs and we had done some notable stuff. We basically took that, and we told our investors, we’re probably not going to make you any money, and we’re going to go explore and figure out what the biggest opportunities are in the human in the loop world.

We spent some time figuring that out and what we figured out in the process, I think is a very, I mean, I’m very excited about it. I think it’s one of the bigger ideas in the world, if I do say so myself. But here’s the basic theme, which is, when you start putting humans into systems that can be doing back-office, customer service, really anything, sales marketing, engineering lawyers, doctors, you name it, you lose measurements. So, if you’re from a computer science background, or you like working a big social app, you’re used to having incredible data. I want to know everything you’ll be used to clicking on, I want to know how all the millions of machines in our fleet are performing, and where to perform them. It’s a very data-driven activity.

The second you put humans in the loop, you’re like, well, then the human did something. I had no idea they did, right? And then, you have no idea how to optimize it. If you’re in the technical world, you have things like new Relic, et cetera. We got super frustrated from our backend being like, look, we want to build better and better systems and keep optimizing, but we don’t actually know exactly what the people are doing, so we can’t optimize it.

So, that’s what fin analytics was. This directed us to solve our own problems. We basically started saying, okay, how do we instrument? Literally down into the clicking, the scroll, the how people are completing work, and what are the inputs to doing that? There’s a bunch of cool stuff going on in the world that makes this possible in a way that was never possible before.

One is the browser. The reality is, it’s a super hard instrument at a super low level without some standard system that everything’s being piped through. And in case it turns out the Dom and what’s going on in the browser is super big, you need a ton of data that’s super expensive, or it used to be. There are all these things that are making it possible as work moves into the browser and becomes digitized in part.

Third, you need things like CRMs. I need to know people. It’s fine to know someone’s spent this much time in Slack or whatever, but that doesn’t mean anything. Like, I need to know what they were working on. The reality is that systems of record. Anything from a Zendesk to a Salesforce, to Asana becomes more standard. You can start doing really cool stuff with that. So there’s this confluence of things going on, where all of a sudden, human knowledge work starts to get measurable. When you start measuring human knowledge, where do you find the huge opportunities to improve people?

Processing tools is the way they look at it. So, people being like, I want better coaching. Imagine going to a job where you got no feedback ever. It’s the worst, right? Where we are right now, most people, maybe, get quarterly feedback. But, what you really want is actionable, real-time feedback on what you’re doing and how to improve, based on real data and how your coworkers are doing things. The end of the process. To do this thing, you should do these steps. Well, is that the best practice? Then someone else figured out something better. How do we make that, kind of, more shareable?

Then, tools. You just want to always improve your tools. So you have to understand how people are actually using them in order to figure that out. I really believe that. If you think about where we are in history and what’s going to happen next, you know, think about having manufacturing. It’s like all productivity comes from the cycle of measuring things and optimizing them, and rinsing and repeating. Then, you look at knowledge work and there’s nothing. All the stuff we do every day, no one has that. I think building that instrumentation layer effectively, and starting to help optimize systems like that at that level is incredibly exciting.

I think it’ll be incredibly important for the future of how we work. So, that’s been a really big mission of mine. It’s going really well. We have some really, really great customers that are big. A lot of unicorns whose names you’d know. We keep growing as people figure this out and realize how big of a deal it is to have.

Teja: That’s awesome. What would be a couple of examples of applications? Customer service?

Sam: Yeah. Customer service. That’s the easiest way to think about a lot of them. In customer service or something, requests come in. Inevitably, it’s going to go to one of 10,000 customer service agents around the globe, right? Depending on the company, they’re going to pick up that task. They have something like an SOP, a standard operating procedure, where they’re supposed to go. They might not do it. The tools might be broken because they frequently are. There are a thousand variances you’re getting, you know, your door dash order is cold. You need to know how to do that. Possibly using the data’s first approach where you say, okay, let’s measure exactly how it was doing all these things, build out, and understand the highest customer satisfaction scores. Where do they come from? And then how do we replicate it? Which tools are broken?

You know, on customer service, you’re typing and there’s a pause, and you’re like, what’s going on half the time? The systems are super slow. Like, some resources are slow to load. That adds up. That creates bad customer experiences, et cetera. You can go deeper on this, but that’s a really simple way to think about it.

If you think about it, anything’s like this. Lawyers are the same thing, right? Like, I need this contract done. How did you review it? You know, what systems are you using? Who’s the best in the world that’s reviewing NDAs? How did they do it? Right. And then, how do we learn from that and apply it out to everyone else?

Teja: Right. You can even think about applications, maybe in sales, where you can be held to a quota, but you’re not sure how long you need to spend on an email with a specific opportunity, how the best performers, even in the organization, are doing that, which should be super instructive.

Sam: I think that’s another great thing, to call out. Which is, the worst thing in the world I’ve always found, is someone measures you purely on your outputs, but those outputs are out of your control, right? So someone’s like, you have to do this, you know, the customer will have to be happy or this has to happen. But the inputs to that aren’t actually happening, or there’s a randomization factor, or their input, you need tools that don’t exist. It’s the worst feeling. The more complete data where you understand exactly what’s happening, you’re not just looking, at this task like, it took too long. well, did it? Why?

That creates a much healthier, more open culture, rather than what normally happens in customer service. Engineers all blame the operations people and say, Oh, you know, these people are terrible or whatever. The operations people are all super frustrated. The engineers are like, these guys don’t know that this is broken, and the communication breaks down.

So the more we can be like, look, here’s the data. Let’s just knock these problems off, the better.

Teja: So do you think that this approach is antithetical or let’s say in opposition to the Netflix or Spotify approach of, hey, we’re not going to tell you what to work on or how to do it? The only thing that matters is you do it well, and they don’t really define what well is. That seems to be a popular management methodology, right? Work on what you want, and we’ll make sure to hire good people and they’ll tell us what to do.

Sam: A few thoughts on that.

One is, that only works at the richest companies in the world that can afford a lot.

If you work in a company where you’re so rich and it’s making so much money, that you actually don’t even care if 99% of your employees are completely unproductive, as long as 1% of them deliver great outcomes. I can understand that that’s a rational model, but that’s not how business really works. That might be good for some people, but I think for a lot of people in the world, if you’re just like welcome, here’s a job, you get no definition of what to work on or how to create value. I think a lot of people would be very unhappy in those cultures. You want definition and goals and milestones.

There will always be experiments in management and there’ll be more. My sense is that kind of mentality, while it’s really great for recruiting a certain type of person, an engineer in that certain environment is going to be always a tiny, tiny, tiny percentage of the world. Frankly, I’m not even sure how long that stuff even lasts. At some point, you say, wait a minute. If 80% of my employees are less productive than they should be, and managers can’t organize big projects because everyone just does what they want? I’m obviously paraphrasing these things, but I’m not sold.

Teja: It’s a fair point because how many of those businesses that have these types of management models, just have stumbled into a moat that wouldn’t exist in other industries or some product that just doesn’t have an alternative? Like anybody else, besides Valve, make steam as a lot of their games, so they can just ride that.

Sam: Well, it’s that and it’s being in a position, where again, you can hire because you’re so rich. 10,000 people, and you actually don’t care if 9,900 of them are not productive, so long as someone produces a hit. That’s a really important insight and that does happen, but it’s rare.

Also, I feel for those 9,900 people. You really want to spend your time in life working where you’ve been given so little direction, you do whatever and everyone secretly knows that what you’re doing is worthless? That’s terrible. Do you know what I mean? I don’t think that’s as positive of culture as some people might think.

Teja: That’s a great point.

How does your work at Fin inform your decision-making, as an investor and vice versa? How do the two and a half roles play with these?

Sam: You get the sense, I’m a generalist and I’m a cross-pollinator. I like that mentality. When you’re a generalist, there are strengths and weaknesses. The weakness of being a generalist is you’re never the best in the world at anything. There’s always someone who’s better, who’s a specialist. Eventually, the benefit is, you hopefully can see patterns that other people can’t or start making connections because you’re just being broader. You’re seeing different stuff.

So when you say, how does being an operator at Fin influence decisions or investment decisions? One thing is, I think it’s really easy to say that operating is hard, right? Or understand what success is in that. The reality is, I think you have to experience it yourself on a continuous basis to really get how insanely difficult it is and where the problems are. There are all sorts of specific ways in which I think having an operator or someone who’s actively operating as an investor is helpful.

Someone the other day was like, I’m having a problem doing this recruiting. I’m like, I’m literally having exactly the same problem, let me tell you where I am and trying to solve it for myself. There is some cross-pollination to that, but I think it’s also just an appreciation for how incredibly hard this stuff is and hopefully, some sense of who runs through the walls and who doesn’t.

I think, on the flip side, if you’re an operator and you’re just building a company, it is so easy to miss the forest for the trees. Right? You get so obsessed with what you’re doing and whatever the metric is you set up, that you might even achieve it, but by the time you achieve it, it’s the wrong goal.

I think being an investor forces you to be a little bit broad and see what else is happening in the world.
Everyone will always justify their own life decisions, but from my perspective, I think that being a real operator, having built companies and ideally being in it yourself, just makes you a better investor and more empathetic with the teams you’re working with, in a whole bunch of ways.

Vice versa, I think the scariest thing for most early-stage founders, is you just got sucked into your vortex and you completely missed the game. It’s not that you miss X or you just missed the game. Having some toehold in investing helps you with that to not make those mistakes.

Teja: That’s powerfully put.

You mentioned before we started recording that, Fin Analytics was hiring engineers. Do you have a public job space or anything like that?

Sam: We do, at fin.com. I think there’s a link there to our job stuff.

At a high level, things are going great. It’s a really big problem. I think this is one of those things that’s sneaky big and sneaky important. Because you’re like, what are you doing? You’re putting a plugin, on customer service agents and measuring their clickstream, that on the surface sounds pretty boring, right? For most people.

The deeper thing we’re working on is how do you instrument knowledge work and help people work better, and have better jobs, and be more efficient? I think it’s a massive platform business that will happen. I’d like to win it. I’m pretty sure it’s going to happen in the next bunch of years. We’re very excited about that opportunity.

Teja: All right, Sam. Thank you so much for your time, I appreciate it.

Sam: Cheers, dude.