Levels – Decoding the Bloodstream with Sam Corcos

In this episode of Just Raised, podcast host Joe Sweeny speaks with Levels CEO Sam Corcos. They chat about the science behind Levels, as well as the go-to-market and fundraising process. Sam describes why customer feedback is so important, how failing can make you stronger, and why you should dedicate your time and energy to relationships and causes that truly matter.

How Levels measures blood sugar

Thanks to CGM technology, Levels can teach people what elements of their diet and lifestyle and working and not working.

We use continuous glucose monitoring, which is a biosensor that measures your glucose levels in real-time. It’s really the first time you can measure a molecule in your body in real-time. Glucose is one of the most important molecules in your body. It is the primary molecule of energy. It affects a lot of things like when you feel tired in the middle of the day. It also affects things like long-term health. It is related to a lot of metabolic conditions if you have glucose that’s dysregulated, meaning not controlled effectively. So we use glucose monitoring to teach people how aspects of diet and lifestyle affect their long-term health and how they feel throughout the day. A lot of people end up discovering something. Like for me, my big discovery was that the breakfast that I’d been having for basically my whole life as the healthy option, which was steel-cut oats, turns out to spike my numbers to very unhealthy levels.

The benefit of closed-loop systems

With a dataset like Levels, users can close the loop on inputs and outputs related to their health and wellness.

Our thesis is that behavior change comes from closed-loop systems. So we optimize for as much of that as possible. It’s an interesting thing, we wrote about this in one of our blog posts, the Levels theory of behavior change. One of the things that we found particularly exciting is that the behavior change that we’re seeing is visceral. This was certainly my experience where if you don’t have a closed-loop system and you don’t know how your choices affect you, it’s very hard to associate cause and effect. I had a friend who started using Levels. He is a little bit overweight. He probably drank five cans of soda a day. He started using this mostly just the curiosity, and everyone kind of knows, yeah, sure, soda is bad for you. Like my mom tells me all the time, sure. But it’s a different thing to see it in data. We’re not being prescriptive, but you’re seeing numbers that you’re not supposed to be able to see as a healthy person. It helps reframe what that means for your body.

Don’t chase new features

At a former startup, Sam learned the importance of not getting bogged down in feature after feature.

One of the hard lessons that I learned from Sightline Maps was we had a lot of people who were fairly excited about what we were doing and they would just say, “If there was just this one more feature, we would totally pay for this.” And then we kept just chasing these features. It was very hard to get people to pay. What we should have done in retrospect is just said, “All right, if you’re not willing to pay now, like cut that, move on to the next thing. Just keep going.” So for us, it was very similar, which was, how do we position this? What is it that people want and how do we make something that solves that problem for them? The jobs to be done framework is a helpful way to thinking about it. Like what is the job that you are doing for this person? What problem are you solving from them?

Iterate until people like your product

It’s perhaps the most basic pursuit of product-market fit, but Levels has shipped various versions of their app to beta users to see what sticks.

People ask, what was your social strategy? The answer is somewhat lazy, but we learned as much as we could from people. We kept iterating on the product until people liked it. When people like your product, oftentimes they’ll post about it on social. That was pretty much our whole strategy was just to build a product that people really liked. We actually just brought on our first marketing hire maybe two months ago. This is part of the theme of closing loops. When somebody says I would really like it if the app solved for X, and then we ship a feature a week later that does that thing and we tell them about it, that’s something that gets people really excited.

Shining a spotlight on metabolic fitness

As interest and awareness in metabolic health grows, the Levels team hopes the price of their product will be driven down to an accessible range.

Our target is to get the price under $100 by the end of next year, which I think is largely achievable even with current technology. It doesn’t require any big improvements. Right now it’s $400…What gets it under $100 is still kind of up in the air. I’d like to get the marginal cost of this as close to zero as possible. Right now there are 90 million Americans that are pre-diabetic and 70% of those people will be diabetic within 10 years. The scope of this problem I think is really underappreciated. If we’re going to solve the metabolic health crisis, we need to get the cost down. And so one is going to be just on the hardware side, making it cheaper, and I think on the other side, it’s going to be around accessibility related to insurance coverage and other clinical capacities. So, we’re going to be pushing on pretty much all fronts on corporate wellness, anything that increases accessibility.

The importance of engineers

The vast majority of the Levels team is made up of talented engineers who are also deeply embedded in the product itself.

One of the things that we’re somewhat unusual is that our engineers are very deeply involved in the product process, especially early on. We have one product manager, one person in product in total at the company. By the end of this year, I think our ratio of product to engineering, and that’s to say all of the product org, which includes designers and UX researchers. I think typically the company has like two or three engineers for every person in the product org. And I think we’re going to be more like 10 to one. We put a lot more emphasis on engineering being involved in the process and that allows for much more rapid cycle times. It can be frustrating, but you’re always going to have constraints. My team is probably sick of hearing me say this, but you’re always going to have problems, you just get to choose which kind do you have. So I would rather experience the failure mode of having too much engineering time available and not enough specs ready rather than we have way too many specs and we’re being constrained on engineering.

Take time to step back

Sam has learned the value of big picture thinking, and has acted as the compass for Levels as it grows.

I found that it’s helpful to take a step back and think more holistically about where the company is going in the next five years. It’s a difficult thing where you both have to be extremely in the weeds and executing all the time for the thing that can kill you tomorrow because we’re default dead. And so you always have to be executing on the very short term all the time or you die. But somebody has to be thinking about the direction that you’re going. I’m fortunate in that for the last few weeks, I’ve had some time to step back and think about core changes to business model, team requirements, product direction, and other aspects of the business. So I would say for me, the thing that’s on my mind is thinking through our strategy, and we do a lot of long-form documentation on that to get everyone on the same page.

Scar tissue takes you to the next level

When you’re a founder, failure once doesn’t mean failure forever.

I had several people DM me on Twitter and I had some long conversations with them about a couple of them. They were in the dying stages of a startup and one probably had three months of runway left and it was probably going to fold. It was interesting to see, because I remember experiencing something similar where he felt like everyone views him as a failure. He’s a YC founder and he’s like, yeah, I’m not going to be able to get another job. It’s like everyone’s going to think I’m a failure. I’m not going to be able to start another company or raise money again. I remember that and just thinking, it is amazing how our perception of ourselves is so different than others’ perceptions of us. I said, do you realize how incredibly hireable you are, a technical founder of a YC company? The scar tissue that you have from these failures is what gets you to the next level.

Make a running list of “asks”

If you don’t ask for what you need, you won’t get it. That’s why Sam is strategic about determining what resources and connections will help the company grow.

I will regularly solicit those from my team and say, “Hey, I’m going to send out a list of asks to our investors. Is there anything that you need?” And they’ll say, “Yeah, I’d love to talk to a partnerships person at a CPG company that has sold into these markets. Or, yeah, I’d really love to talk to somebody who has managed a big health data project who knows something about databases.” You don’t necessarily know who is going to be that person or who’s going to know the people who know those people, but the more disparate your networks are, meaning the more disconnected they are, the more valuable your network is. So for example, if you know 1,000 people and they’re all investment bankers that work at Goldman, they probably also already know each other. And so if you wanted to get in touch with somebody who’s in the fabric industry, you’re probably not going to be able to because you have one dense network that’s all concentrated.

Invest in meaningful relationships

The trajectory of life has taught Sam that you shouldn’t waste time of things and relationships that don’t add value and meaning.

I remember there was somebody from middle school that was the cool kid. I remember I would like wear skateboard shirts because he wore skateboard shirts and he was cool and I thought wearing skateboard shirts would make me cool. It’s interesting when you look in retrospect at just how ridiculous all this behavior was on trying to gain the affection of people who really dislike you or don’t care about you at the very least or completely indifferent to you. Realizing that you should lean into the relationships that matter to you and just emphasize and invest a lot more heavily in those, that’s been a tremendous source of value for me personally.

Sam Corcos: I’m a pretty aggressive minimalist as well. One of the joys of being a programmer is that you can be in Estonia or in Argentina or wherever you want to be. I spent a lot of time traveling just because I could, and I started out with a lot of stuff. I have like a checked bag and a carry on and a backpack. You just realize how little your actual material needs are on a lot of these trips. So I think I own a total of three t-shirts and one pair of jeans at this point. That was something that I’ve just sort of picked up and learned over time. Over time, you just start to adapt and you get a lot more comfortable being uncomfortable.

Joe Sweeny: Today on the show we’re talking to Sam Corcos, founder and CEO of Levels. Levels is building a fascinating new wearable that I think is more powerful and more promising than any wearable device that’s come before it, mainly because they are seamlessly and painlessly measuring blood glucose, the molecule most intimately tied to diabetes and colloquially known as blood sugar. We dive into the science that makes this possible, the hardware subscription business model, and the go-to market that has athletes, celebrities, and seemingly every investor who can get their hands on one raving about the company. Levels just raised $12 million from Andreessen Horowitz and is hiring engineers across the full stack, a general counsel and a head of clinical research.

Joe Sweeny: Sam, great to have you on the show. I would love to start by diving right into the science, and then we can go back and talk about your go-to market, the fundraise. I know you have some pretty interesting tactics focused on mapping networks and making connections. But first, let’s talk about how the Levels patch works. Can you give us a high level overview of the product?

Sam Corcos: Yeah. We use continuous glucose monitoring, which is a biosensor that measures your glucose levels in real time. It’s really the first time you can measure a molecule in your body in real time. Glucose is one of the most important molecules in your body. It is the primary molecule of energy. It affects a lot of things like when you feel tired in the middle of the day. It also affects things like longterm health. It is related to a lot of metabolic conditions if you have glucose that’s dysregulated, meaning not controlled effectively. So we use glucose monitoring to teach people how aspects of diet and lifestyle affect their long-term health and how they feel throughout the day.

Sam Corcos: A lot of people end up discovering something. Like for me, my big discovery was that the breakfast that I’d been having for basically my whole life as the healthy option, which was steel cut oats, turns out to spike my numbers to very unhealthy levels. I would crash into hypoglycemia and my hands would get shaky. I’d feel tired, I’d lose focus. In my head I’d always associated this with too much caffeine or not enough sleep or something other than diet. It really never even occurred to me that diet could have such a profound impact on the way I feel throughout the day. But once you can connect the dots, we talk a lot about closed loop systems at Levels, or when you can connect how you feel with a decision that you made using data, not just waving your hands, it really leads to some pretty powerful mechanisms for behavior change.

Joe Sweeny: Yeah. That is fascinating. I have some experience in this space. My undergrad was in biomedical engineering and then I covered wearables and digital health at Lux Research Consulting firm. When I covered wearables, noninvasive blood glucose monitoring was pretty much the holy grail. Google experimented about 10 years ago with a contact where you could theoretically measure glucose. But as far as my knowledge, non-invasive blood monitoring where there are no needles penetrating the skin does not exist yet. And yet your product seems to do almost exactly that. I know it’s a patch you stick on your arm that connects to an app, but I don’t know if you’re using micro needles that have been developed over the past five years or a really high end version of the continuous glucose monitors made for diabetics. So let’s dive into the product. How does the Levels patch work?

Sam Corcos: We’re using off the shelf continuous glucose monitors. I actually have one on each arm right now. They’re each made by different manufacturers. They’ve been under patent since the ’90s, so they’re fairly widely used now. I think millions of people use these for diabetes management. We’re using the same hardware. The underlying mechanism, they are technically we’ll call them minimally invasive. There’s a filament that goes under the skin. You can see I have them on, I’m touching them right now. I don’t feel it. I’ve probably put 30 or 40 of these on myself. I’ve only even noticed it maybe once or twice. The underlying science is it’s electrochemistry. The filament is coated with glucose oxidase as the underlying enzyme. And when glucose oxidase binds with glucose, it changes the conductance of that filament, and that allows you to measure molecular concentration.

Sam Corcos: There have been many attempts at measuring glucose noninvasively or actually tracking on the order of 10 fairly distinct different technologies that are attempting to do this. Some are using spectroscopy, some are using a resonance, some are using micro-needles. There’s actually a really great paper. I shouldn’t call it a paper. It’s really a book. It’s like a 300 page paper called the Hunt for the Deceitful Turkey, and it’s about why non-invasive glucose monitoring is hard. So if you’re really interested in digging into it, that’s a good paper to read because it talks about some of the specific aspects of just physics on why this is a hard thing to do.

Joe Sweeny: All of your hardware is actually off the shelf then, and it’s also used by diabetes patients?

Sam Corcos: That’s right. That’s historically what it has been used for.

Joe Sweeny: Gotcha. Interesting. The core innovation here isn’t necessarily the hardware, it’s actually the business model innovation where you are bringing this CGM, continuous glucose monitor, to the athlete, the professional quantified health conscious consumer, and eventually the mass consumer market.

Sam Corcos: Data by itself is not particularly useful unless it’s contextualized. So if I was to tell you 160 milligrams per deciliter, that doesn’t mean anything unless you have some context for how that is affected by a behavior that you made. We are the contextualization and data science layer that makes that data meaningful and actionable. That’s the core value proposition, sort of like the difference between having access to GPS data, which tells you your coordinates, and having Google Maps or having Uber, which tell you how to get from these coordinates to those coordinates. It’s still necessary and useful to have your GPS coordinates, but the softer layer that’s built on top of it is really, I think, where most of the value is when it come from in this space.

Joe Sweeny: The software layer brings me to another question on the wearables market more broadly. It seems like one of the challenges of the quantified self space is doing exactly that, transforming biometric data into something valuable for the consumer. Most wearables companies have shut down at this point: Pebble, MisFit, Samsung Galaxy Gear, and the Nike FuelBand. Fitbit’s stock price tells a pretty negative story. Now, Levels feels different and is different. Like I said at the top, blood glucose monitoring, kind of the holy grail of wearables. You’re not an activity tracker counting steps, blood glucose is an inherently more powerful metric. But it still bears asking, how do you build that killer app, that Google Maps like UX that goes on top of this glucose data?

Sam Corcos: Yeah, I think one is that the data itself is very actionable. I think the biggest one is contextualization and education. We have event detection where things happen and we surface some insight as to what that means. We also do some textual analysis on whatever you log, what words you use in those logs, to recommend alternate pathways. If it’s something that contains alcohol or something that we know contains a lot of sugar, or if people do a really intense workout or they go on a walk after a meal, we can surface information and we can educate people as to how that affects their long-term health. This is one of the core features of the product and is something that there’s still a lot of work to be done on this front. There are some aspects of statistics and machine learning. The way that we’re using the data is much more around how do you make it actionable and how do you close the loop on a specific behavior?

Joe Sweeny: So for you, the UX is all about tightening that feedback loop?

Sam Corcos: Our thesis is that behavior change comes from closed loop systems. So we optimize for as much of that as possible. It’s an interesting thing, we wrote about this in one of our blog posts, the Levels theory of behavior change. One of the things that we found particularly exciting is that the behavior change that we’re seeing is visceral. This was certainly my experience where if you don’t have a closed loop system and you don’t know how your choices affect you, it’s very hard to associate cause and effect. I had a friend who started using Levels. He is a little bit overweight. He probably drank five cans of soda a day. He started using this mostly just the curiosity, and everyone kind of knows, yeah, sure, soda is bad for you. Like my mom tells me all the time, sure. But it’s a different thing to see it in data. We’re not being prescriptive, but you’re seeing numbers that you’re not supposed to be able to see as a healthy person. It helps reframe what that means for your body.

Sam Corcos: One of the things that he said that was most interesting is that when he started seeing numbers that were outside of normal bounds, he didn’t decide consciously that he was going to stop drinking soda. He said the experience was much more like getting food poisoning where when he was able to connect the dots for the first time that these energy crashes were caused by soda and that when he had the hypoglycemic crash afterwards and he was craving sugar, he realized that he was in this death spiral, he said the experience was more like his body decided that he wasn’t going to have soda anymore.

Sam Corcos: It’s like when you get food poisoning from clams and your body just decides you’re not eating clams anymore. The smell of clams is now repulsive to you. He said he had the same experience towards soda. He’d go to a soda fountain and the idea of having Sprite again was repulsive, even though he used to drink this every day for many, many years. It’s interesting to see that because I think for many people, we have these vague notions of like, yeah, Doritos, probably not good, soda, I’ve heard that’s bad, but is it really? How bad is it really? When you see it in the data, it leads to behavior change that I don’t think is possible in any other form.

Joe Sweeny: That’s fascinating. How did you start Levels? I know you come from a software engineering background and not academia, and you said your core competency is building consumer software. After your last company, how did you start ideating on Levels? What was that process like?

Sam Corcos: Levels started somewhat slowly. I track my time pretty religiously, so I can actually see how much time I dedicated to Levels at different points in time. From the time when I started working on it as a concept to the time when I actually cut all of the other projects and meaningfully committed to it, it was probably three months. The process was much more around market discovery. So I didn’t really start building a product until six months later. It was learning from people what got them excited. I think the biggest one is getting people to pay money.

Joe Sweeny: It’s massive when that happens, it changes everything.

Sam Corcos: It does. This is one of the hard lessons that I learned from Sightline Maps was we had a lot of people who were fairly excited about what we were doing and they would just say, “If there was just this one more feature, we would totally pay for this.” And then we kept just chasing these features. It was very hard to get people to pay. What we should have done in retrospect is just said, “All right, if you’re not willing to pay now, like cut that, move on to the next thing. Just keep going.” So for us, it was very similar, which was, how do we position this? What is it that people want and how do we make something that solves that problem for them? The jobs to be done framework is a helpful way to thinking about it. Like what is the job that you are doing for this person? What problem are you solving from them?

Sam Corcos: I think one of the best examples that I’ve seen of this is the founders at Truepill where they knew that the pharmacy industry is big and has a lot of problems. The founders just like grilled pharmacists for months, like what are all the problems? And they just like wrote them all down. They learned as much as they could. They started seeing gaps where they thought they could fill it with software. And then they built a very successful company out of that. I think a lot of people, and I’ve been fully guilty of this myself many times, you start with a product because building products is super fun and getting rejected constantly by people is super not fun.

Sam Corcos: So starting with the market and learning from people and being willing to take some relationship risk. I was hustling my friends to try to get them to buy Levels. A lot of them said no, other ones said yes, and they loved it and that was great and we got good feedback from it. You have to be willing to go out on a limb and learn as much as you can from the market and just be ready for rejection because it’s going to happen.

Joe Sweeny: I would love to dive into your go-to market because it’s pretty drastically different than you might expect for a digital health company. You’ve seen some massive growth with wild numbers of people on the waitlist, over 50,000 people, at a very high premium price because you’re still in beta. How have you been able to generate this kind of excitement?

Sam Corcos: I think the answer is going to be somewhat disappointing. It’s largely just straight from the Y Combinator motto of make something people want. I think that the market discovery is a lot harder than building product. You have to spend a lot more time thinking about that. I took hundreds of customer calls myself from people on the wait list to learn who they are, what job our product is solving for them. That’s how we developed all these personas docs, how we learned about the value propositions. We just kept iterating on the product and shipping things very, very quickly. It wasn’t until probably May or June that we really started to hit a stride where I think we had the feature set that people really liked.

Sam Corcos: This is another one of those things where we’ve been told that our social strategy has been really good. People ask, what was your social strategy? The answer is somewhat lazy, but we learned as much as we could from people. We kept iterating on the product until people liked it. When people like your product, oftentimes they’ll post about it on social. That was pretty much our whole strategy was just to build a product that people really liked. We actually just brought on our first marketing hire maybe two months ago. This is part of the theme of closing loops. When somebody says I would really like it if the app solved for X, and then we ship a feature a week later that does that thing and we tell them about it, that’s something that gets people really excited. So I think for a lot of people who are not in software or who are not in startups, things feel static.

Sam Corcos: They’re used to using a system every day that is the same, like Microsoft Word. It’s the same system that has been since probably they were in high school. But when they say, man, it would be really cool if I could search through all of my logs and find things and filter, and then we ship that feature a couple of days later or a couple of weeks later, it feels magical when these things can actually change while you’re experiencing it in beta. So I think that’s probably another part of it is we’re really engaged with our members. We interview a lot of them either before they start, when they’re in the midst of the program, after they leave the program, if they subscribe after they’ve been subscribing for some amount of time. We generate a lot of data on how people are using it. A lot of qualitative data as well.

Joe Sweeny: I had Phil Libin on the show a while back. He was the founder of Evernote, and now uh-huh, and he says the exact same thing about product. His product model was make something that people love for a very small target market. Make sure it really moves the needle for them, and then amplify their voices when they tweet about it and talk about it and tell you how much they love it. And then partnerships with big companies will come. Venture dollars will come. I’d love to talk about the price point a little bit and where you want it to go because I know your mission is to make this a mass consumer product and you’re aiming to bring down the price. But I’d love it if you could sequence out from where we are today to where it gets to be a mass consumer product like the iPhone, what has to happen between now and then?

Sam Corcos: Our target is to get the price under $100 by the end of next year, which I think is largely achievable even with current technology. It doesn’t require any big improvements. Right now it’s $400. If we can start to increase the volume that we’re doing, I think it’s very achievable that we can get this under $100 by the end of next year. Dexcom and Abbott are the two major manufacturers. Their existing set of technologies is just linearly heading in that direction. If you look at the price curve, it’s been dropping pretty consistently every year. Unlike a lot of aspects of healthcare which are prices have been rising pretty consistently, this type of technology, because it is ultimately a technological solution, has been coming down pretty substantially every year. Plus, the volume that we’ll be doing probably by the end of next year, it gives us a lot of leverage in terms of price negotiation.

Sam Corcos: What gets it under $100 is still kind of up in the air. I’d like to get the marginal cost of this as close to zero as possible. Right now there are 90 million Americans that are pre-diabetic and 70% of those people will be diabetic within 10 years. The scope of this problem I think is really under appreciated. If we’re going to solve the metabolic health crisis, we need to get the cost down. And so one is going to be just on the hardware side, making it cheaper, and I think on the other side, it’s going to be around accessibility related to insurance coverage and other clinical capacities. So, we’re going to be pushing on pretty much all fronts on corporate wellness, anything that increases accessibility.

Joe Sweeny: It’s interesting to think about market size here. In some ways it feels like you have to be almost a power user to understand these metrics. For instance, right now, I know your Level users track the glucose graph over time. They also plug in all of the meals that they eat, which has been a chronic challenge for wellness companies to get people to do. As you think about expanding to the broader market, are there things you plan to do to make the product more accessible to the average consumer?

Sam Corcos: It’s a challenging question to answer because that is the core of the product. The question is, as you expand to different market segments, how do you adjust the product to fit the needs of those people? I think the answer is you have to do it iteratively. I would describe it in terms of process rather than in terms of specific outcomes is you have to continue to learn from your customers, understand what they want. An interesting case study in this is around macro tracking. This is something we had in the app at one point where it had a much more robust analysis of all of the meals that you logged. But the thing is we found that most people really just cared about friction. They didn’t want to spend five minutes inputting their food every time they ate something, they just wanted to add something as like a placeholder as quickly as they could and get on with their lives.

Sam Corcos: And so we did something of an anti pilot. We removed that feature and NPS went up. We just made it really, really simple to add stuff. Some people still request it’d be great to do macro tracking, but it’s a nice to have. It’s not something that meaningfully alters the product experience. When we started working on this, we thought that it would just give people access to the data, give people CGMs, and they’ll automatically know what it means. It turns out that is an incorrect assumption. I think of the first maybe 10 people that we gave these to, eight of them were like, okay, so I have some food and then my numbers go up and then my numbers go down. And now what? What am I supposed to do with this information?

Sam Corcos: We realized that as a team, we were really suffering from the curse of knowledge where we just said, well, surely you know the long-term metabolic consequences of glycemic variability. Like I have no idea what you’re talking about. This is one of the challenges that we get from doctors all the time where they just say like, yeah, all you need is just access to the data. Everyone implicitly knows the long-term consequences of glycemic variability. The reality is they just don’t. So, one of the big challenges we have is building a program that helps people contextualize that information so they understand how those choices affect them and to close that loop. The data is only useful insofar as you can contextualize it and make it useful for people.

Joe Sweeny: Talking about closed loop systems, I think that’s actually a really good segue into your fundraise. You just raised $12 million led by Andreessen Horowitz on a pretty massive valuation. You have some very interesting, specific and tactical strategies about how you go about fundraising that I’d love to touch on. When we first connected on the first email that you sent over to me, at the bottom you had this amazing list of asks of something like 50 different things that you need covering topics like people you wanted to connect with, management problems, specific hires you are trying to make, a whole ton of different things that I found I have never seen a list of asks that was so useful. And you send that out to, I think, every person you connect with. So, if you could just talk a little bit about your strategy for connecting with people and fundraising, I think that’s a fascinating aspect of how you operate Levels.

Sam Corcos: I think the biggest one, and this is something I think Mark Suster talks about. He wrote a post on people should invest in lines and not dots. For me, it was about building relationships early, looping them in with our investor updates, and having them see progress every month. I remember one of the investors that we were talking to when we finally started the process, she had been following us for six months. She said, “It feels like you’ve made five years of progress in the last six months because I can just see every month you’re shipping, you’re consistent and you deliver on the things that you say you’re going to deliver on.” It really builds that confidence.

Sam Corcos: I think that especially when it comes to raising capital, there’s a huge component of trust and confidence that you have to instill in order to pull the trigger on it. Sometimes these things happen very quickly. I think for the best relationships with a partner or for the best funding rounds, like ours I would say was pretty solid, it comes from just execution consistently but also being transparent and showing progress every month. I think for us, some of it we lucked out a little bit. We were an all remote team focused on digital health and I think pretty much every fund came out of COVID in September looking to invest in all remote teams in digital health. That certainly didn’t work against us, but we’re probably going to raise a healthy round anyway.

Sam Corcos: I think the biggest thing is about being honest with people. Integrity is incredibly important, especially when you’re talking about tens of millions of dollars. So being transparent, showing not just the good but also the bad. If you have challenges in a given month, be honest about it and say how you’re addressing those problems. And then hopefully over the course of the next couple of months, you can show that you did address those challenges and that you did it effectively. That builds confidence. There’s a saying about you shouldn’t build a fire department when there’s a fire. It’s very similar for fundraising where you shouldn’t start building relationships when you need to raise money. You should already have those relationships and those people should already know who you are, that you’re a good executer and that they can trust you, before you need to raise capital.

Joe Sweeny: You’ve talked about moving fast in building software, as in someone sends you an idea that they’d like to have in the app, and then it’s shipped the next week. You also talked about how in your investor updates, you show consistent progress. Do you have any tips on pure execution?

Sam Corcos: I mean, that’s the million dollar question. Execution is hard. I think one is culturally. I think that most companies have too many meetings. We have almost no meetings and that’s I think a big aspect of why we’re successful in execution. Another is hire really good people and trust them to do things on the spectrum of micromanagement to not paying attention at all. We’re probably too far on the realm of like full trust. We’ll check in if you want us to check in. We’ve had some of our engineers have asked us to do performance reviews, so say, “It’s great that you trust us that we’re doing good work, but it would be good to get some feedback eventually.”

Sam Corcos: The other thing I would say is you have to really focus on your customer. There are a lot of failure modes that I have experienced myself in past companies as well as failure modes that I’ve seen in companies run by friends of mine. We are a product and engineering company. This is my first non-technical role. Our whole company is really built around maximizing development velocity. What that means is shipping features that improve our members experience. So if you’re working on scaling issues that is important and needed, but if we’re not shipping things that positively affect the experience of our members, we are losing. If we’re spending time on internal tools, there’s some amount that is necessary but that should not take up the majority of our resourcing.

Sam Corcos: So, a lot of the things that we use like Retool is an excellent product that allows us ops to build their own dashboards so that engineering doesn’t have to get involved. Marketing and growth do a lot of things manually and it would be better, it’d be more efficient if engineering was involved, but you have to recognize that as an early stage software company, engineering is the fundamental constraint to progress of your company.

Sam Corcos: So you should think of every engineer’s time costs something like $10,000 an hour. So when marketing says, “Hey, I’d really like a new automation here.” It’s like, “Cool, is that worth $50,000 of engineering time?” And the answer is like, probably not. So, how about we just do it on a spreadsheet? Great, let’s do that. Sometimes the answer is yes, like this is actually worth that much money. And so we bring in engineering. So you have to be ruthless with making sure that engineering time is being spent on the things that are improving the experience of your members.

Joe Sweeny: How do you see structure your team? This seems relevant to a company like yours because your core competency is, again, consumer software, but you’re also measuring blood glucose. I know you have scientists on the team and are very scientifically minded even in your approach to fundraising, so how is the team structured?

Sam Corcos: About half the team is engineering. I think over the course of the next year, it’ll be probably two thirds of the team will be engineering if not more. We’re adding maybe 10 engineers this year and maybe only two or three other positions in total. We’re very much focused on adding more engineers. We’re also really focused on bringing on product-minded engineers. That’s something that I think is really important is that I have seen the failure mode of a company that has too many really good product people and they have lots of really good ideas, but their cycle times are too slow because of that where let’s say you’re a product person and you have this great new idea for something that can really improve the member experience, and you say, “Cool, how soon can we get this on the roadmap?” And the engineering team says, “Wow, that’s a really great idea. How about two years?”

Sam Corcos: It’s like, well, if you’re a startup, that might as well be infinity. So you have to be at a point where your cycle times, which are the time from learning something about your members where they say it’s really frustrating how long it takes me to log these things, to the product person puts together a spec and says, all right, here’s an incremental improvement that we think will reduce the friction to logging by three seconds, to engineering implementing that, shipping that to the customer, that needs to be on the order of days, maybe weeks. If it’s more than that, you’re in a world of hurt.

Sam Corcos: So, I think one of the things that we’re somewhat unusual is that our engineers are very deeply involved in the product process, especially early on. We have one product manager, one person in product in total at the company. By the end of this year, I think our ratio of product to engineering, and that’s to say all of the product org, which includes designers and UX researchers. I think typically the company has like two or three engineers for every person in the product org. And I think we’re going to be more like 10 to one. We put a lot more emphasis on engineering being involved in the process and that allows for much more rapid cycle times.

Sam Corcos: It can be frustrating, but you’re always going to have constraints. My team is probably sick of hearing me say this, but you’re always going to have problems, you just get to choose which kind do you have. So I would rather experience the failure mode of having too much engineering time available and not enough specs ready rather than we have way too many specs and we’re being constrained on engineering. So, that’s the way that we think about it.

Joe Sweeny: To call it out, it sounds like you’re almost entirely hiring engineers, but any key roles that you’re looking for or specific types of engineers, people that you’re looking for?

Sam Corcos: Yeah. I mean, it’s really engineering. We’re looking to bring on really excellent engineers. We have a job description at levels.link/eng, and that gives a pretty good description of what it is we’re looking for. I think product minded engineers. People who are former technical co-founders are really excellent in that because they tend to have the exposure to many different stakeholders and they understand and empathize with customers. So those have been really high value ones for us. I think the biggest gap we have right now is maybe product minded front end developers who understand mostly mobile but also web interfacing. So interface developers is probably the biggest one, but we’re hiring 10 people, whether that’s just with a front end, back end full stack, or kind of across the board.

Joe Sweeny: So you’re scaling this engineering team. I know you just hired Ben Grynol, who is a friend of the show and introduced to us as your head of growth. He’s fantastic. With that in mind, what’s top of mind for you over the next few quarters?

Sam Corcos: Personally, I’m in a fortunate position, really since Ben started. Now that Ben is fully onboarded, I largely have no more responsibilities at the company. If you saw our latest Friday forum, you noticed that I didn’t really speak at all. I don’t really have a speaking part. I can think a lot more strategically about plausible future business models. I found that it’s helpful to take a step back and think more holistically about where the company is going in the next five years. It’s a difficult thing where you both have to be extremely in the weeds and executing all the time for the thing that can kill you tomorrow because we’re default dead. And so you always have to be executing on the very short term all the time or you die.

Sam Corcos: But somebody has to be thinking about the direction that you’re going. I’m fortunate in that for the last few weeks, I’ve had some time to step back and think about core changes to business model, team requirements, product direction, and other aspects of the business. So I would say for me, the thing that’s on my mind is thinking through our strategy, and we do a lot of long form documentation on that to get everyone on the same page. For the company, it feels like a lot of it is just execution risk right now. We have David who’s my co-founder and head of product has a really great vision for the product. It’s really just about execution against that plan. And so making sure that we’re continuing to learn from our customers as we iterate and just shipping features that customers love and keeping that going.

Joe Sweeny: Plausible future business models might be my favorite topic. Right now you’re a hardware subscription, the HaaS model, hardware as a service. What other possible future business models are you considering?

Sam Corcos: Yeah. I mean, there are an endless number of business models. I don’t have an answer to that yet, but I would say that the way that I have been thinking about it is more from first principles of how do you align the incentives of your business with the incentives of your members? There are some great examples of this done poorly. Car repair is a classic example where the incentives are directly misaligned, where you bring your car in for a repair and the person that you’re talking to, your service advisor, they’re a salesman and their role is to try to trick you into spending more money than you should. There’s this incredible information asymmetry where they say, “Yeah, I mean, you definitely need to do your brake fluid. If you don’t, you’re going to die.” You’re like, “Whoa, that sounds pretty serious. I guess I need to do that.” Like, sure, I don’t know.

Sam Corcos: But most of the time that’s not actually real and they’re actually compensated, something I learned having worked in the automotive industry at some point, is that they’re actually compensated extra for things that are high margin like fluids. So it’s very much a sales role. Those are all upsells, most of them you don’t need. Another example of an industry that has this perverse incentive structure is gym memberships where they are strongly incentivized to get you to sign up for a long-term contract and then to try to get you to never show up because if you show up, you’re a cost center for them. So your incentives are directly misaligned with your members. So, just thinking through ways that we can build a business model that it might constrain us in some way, but I think for the longterm health of a company, it’s okay to have these constraints if it fundamentally aligns you with the happiness and health of your members. So, that’s the way that I’ve been thinking about it.

Joe Sweeny: Nearly any ad model where the incentives would come to keep eyeballs on whatever platform that-

Sam Corcos: Great. You become an attention merchant.

Joe Sweeny: That’s interesting. With a company like yours, when I think of a non morally challenging business model, it’s often direct revenue in that you make a product, someone buys it because they really like it and that’s very clear. But for a company like yours, that almost implies something like people paying less based on how healthy they are. Insurance is one part of the market that we haven’t touched on at all. Is that a direction you’re considering?

Sam Corcos: That’s what Virta is doing now, I think. Virta is doing outcomes based pricing for diabetes reversal. Virta is one of the most interesting companies in the space. A lot of the research that they’re doing as well has really changed the way people think about type 2 diabetes. There are some really great companies working in the diabetes space. I think there is a case to be made for insurance coverage of health and wellness products. It’s uncommon but not unheard of. When you work with insurance companies, you basically have two levers. Ultimately it is a financial decision for them. And so the two levers are numerator and denominator, which is to say you can either make it so cheap, that insurance coverage, it almost doesn’t even matter what the outcomes are. If it’s slightly positive, it’s worth it.

Sam Corcos: The other aspect is if you can show tremendous health outcomes, even if it’s really expensive. If the health outcomes save them $50,000 a year and your product costs $25,000, then sure, it’s worth it. So you have to figure out which aspects you can pull in terms of those levers. For us, it’s about getting the costs down as quickly as possible. And one of the reasons why we’re hiring for clinical is to start to figure out the degree of efficacy and how that might translate into insurance coverage.

Joe Sweeny: One of the closing things I’d like to touch on is you’ve spoken pretty powerfully on the challenges of being a founder. Can you touch on that?

Sam Corcos: Yeah. I mean, I actually solicited some folks on the podcast I do with Jason Yang talking about the emotional difficulty of being a founder. I had several people DM me on Twitter and I had some long conversations with them about a couple of them. They were in the dying stages of a startup and one probably had three months of runway left and it was probably going to fold. It was interesting to see, because I remember experiencing something similar where he felt like everyone views him as a failure. He’s a YC founder and he’s like, yeah, I’m not going to be able to get another job. It’s like everyone’s going to think I’m a failure. I’m not going to be able to start another company or raise money again.

Sam Corcos: I remember that and just thinking, it is amazing how our perception of ourselves is so different than others’ perceptions of us. I said, do you realize how incredibly hireable you are, a technical founder of a YC company? The scar tissue that you have from these failures is what gets you to the next level. The scar tissue that I have from my last experiences are what… There are times working in Levels where the path is just so abundantly clear to me even though it’s not to maybe the other co-founders because I’ve been down this path so many times and it’s like, oh, this is that failure mode that I experienced at CarDash, or this is that failure mode that I experienced at Sightline Maps. And you just, you can see it coming from a mile away.

Sam Corcos: A perfect example of this is something that I learned the hard way. I was the technical co-founder at Sightline Maps. We did topographical mapping, mostly for the US military. One of the things that we learned the hard way was around fundraising. We had decent connections, but we had timed our fundraising around when we would run out of money, which makes sense. It’s like, what’s the point of raising money when you already have money? That turned out to be a huge mistake. It’s something that you learn when you get terms that are like 25% what you expected because they know they have leverage, and nobody wants to put money in if you could just run out of money.

Sam Corcos: So it turns out cash is an asset and the more money you have, the more money people want to give you on better terms. And so when it came this time around, we raised more money than we probably needed to. It’s another one of these funny survivorship bias problems where when I talk to friends of mine who have started very successful companies, everyone says, “Yeah, I know. I wish we took on less capital.”

Sam Corcos: And so it feels like everyone wishes they took on less capital. But the people who wish they took on more, their companies died. So those people don’t exist in the conversation because they’re gone. So I think everyone always regrets the decision that they made because you could optimize it a little bit more, but really it’s just about making sure that your company survives getting to the next level. I would say for other founders, it’s important to have somebody to talk to who has experienced a lot of these things before. There were certainly times at this company and in other companies when there were founder dynamics that were really confusing or frustrating.

Sam Corcos: I do some advising for companies that my friends have started and they talk to me about founder dynamics. They’re all the same. It’s really just about having a third party perspective and reading a lot of books. Five Dysfunctions of a Team is a classic book on this. The Phoenix Project is a classic book on this for software projects. I’d say Nonviolent Communication is another really great book on this. Just learning how to communicate and set expectations effectively. A big part is realizing that you’re not the only person who has gone through these problems and that these are all solvable and that a lot of the most successful companies have had to go through these things. So, definitely that you’re not alone going through these struggles.

Joe Sweeny: Yeah. Communication is so critically important to almost everything we do, but it feels like it’s critically understudied or at least under prioritized maybe. With that in mind, I know you’ve thought deeply about networks when it comes to raising this round. You’ve talked about this a little bit in the past, but can you recap that for us, how you think about social networks, networks for raising capital, peer-to-peer founder networks?

Sam Corcos: Yeah. It’s a question of when we structured our early cap table. I spent several months deeply studying network theory when I took a year off. One of the big conclusions is that networks are really important in that there were many different types of centrality. Degree centrality is the one that most people think of as like somebody who is popular, they know the most people. But the one that’s probably the most important is eigenvector centrality, which is to say, how many people do the people that you know know? And so it’s like it’s one degree removed from just how many people do you know, and also using that to get access to disparate networks.

Sam Corcos: It’s very hard to predict how networks operate. One of the things is we’ve executed on this fairly well. I solicit requests from my team. Like you saw the long list of asks that I have for our investors. I will regularly solicit those from my team and say, “Hey, I’m going to send out a list of asks to our investors. Is there anything that you need?” And they’ll say, “Yeah, I’d love to talk to a partnerships person at a CPG company that has sold into these markets. Or, yeah, I’d really love to talk to somebody who has managed a big health data project who knows something about databases.”

Sam Corcos: You don’t necessarily know who is going to be that person or who’s going to know the people who know those people, but the more disparate your networks are, meaning the more disconnected they are, the more valuable your network is. So for example, if you know 1,000 people and they’re all investment bankers that work at Goldman, they probably also already know each other. And so if you wanted to get in touch with somebody who’s in the fabric industry, you’re probably not going to be able to because you have one dense network that’s all concentrated.

Sam Corcos: If you have a person who works in film, if you have a person who works in retail, if you have a person who works in tech, who works in healthcare, you have people who give you access to all these disparate networks. It has been endlessly surprising to me how many times the least likely people were the ones who ended up making the most important connections for us. I can’t think of any off the top of my head, but examples where we ask for something very specific and somebody who’s in a completely unrelated industry would say, “Oh yeah, I went to high school with a guy who knew him. Let me see what I can do.” It’s been really remarkable how networks operate in that capacity and how that converts into outcomes. I think we’ve converted on something like 100% of all of the things that we’ve asked for from our investors. So I think it’s pretty unusual.

Joe Sweeny: And it also feels like a fantastic way to engage the team. The fact that you’re sending their ask to your investors at Andreessen Horowitz and the board feels powerful. Before we close out, can you tell us a little bit about tactically how you connect with people, who’s important, how you make inroads into these different networks? I know Ben mentioned that you aim to connect with 1,000 people a quarter. That feels like a valuable skill and from the people I know who have raved about you and Levels, feels kind of like you’re a superpower.

Sam Corcos: Yeah. So the 1,000 people a quarter is my target for… I probably communicate with well over 1,000 people a quarter, but 1,000 people a quarter is my goal for a number of people in my life who I care about that I need to stay in touch with, friends from college who I would otherwise fall out of touch with them or people who I like. It’s amazing how we can often go a very long period of time without reaching out to somebody that we care about. I’m bad at remembering things and computers are very good at remembering things. So I use computers to help me remember like, hey, you haven’t reached out to Jeff Wagner in two months. It’s like, oh, has it already been two months? I feel like I just talked to him, but I guess not. Let’s see what he’s up to. Jeff was my college roommate.

Sam Corcos: So using those tools to help me think through that. It’s much less about work connections although ironically, these things end up overlapping quite a lot. For example, Josh and David, who are two of my co-founders at Levels, are people that I’ve been in touch with for many, many years. Josh I’d probably been in touch with regularly for three or four years before we started Levels. Now, David probably six years before Levels. Like we’d stayed in Buenos Adiós together and like lived on a polo farm for a while. I’ve known David for a long time and he’s somebody that I’ve stayed in touch with for a very long time. Casey is actually another perfect example, kind of the exception that proves the rule. I didn’t know her before we started the company, but her brother is somebody that I’ve been in touch with for a long time.

Sam Corcos: When he found out what we were up to, he said, “You have to talk to my sister. This is what she’s been talking about for year.” And she was just a perfect fit. So staying in touch with people. There aren’t really professional outcomes that are intended, but I’ve just found in my life that being around people that I care about that I think are interesting tends to lead to these sorts of outcomes sort of unintentionally. One of the things that I emphasize for myself personally is I emphasize people who also want to stay in touch with me and who care about me. I realized at some point in my life I was spending way too much of my attention chasing the attention of people who didn’t like me. It’s like a weird thing when you realize that.

Sam Corcos: I remember there was somebody from middle school that was the cool kid. I remember I would like wear skateboard shirts because he wore skateboard shirts and he was cool and I thought wearing skateboard shirts would make me cool. It’s interesting when you look in retrospect at just how ridiculous all this behavior was on trying to gain the affection of people who really dislike you or don’t care about you at the very least or completely indifferent to you. Realizing that you should lean into the relationships that matter to you and just emphasize and invest a lot more heavily in those, that’s been a tremendous source of value for me personally.

Joe Sweeny: Well, I think that’s a fantastic place to end. Sam, thank you for coming on the show. This has been a fascinating conversation and I’m excited to get my hands on a Levels patch. Thanks for joining us.