Karn Saroya on the Capital-Light Way to Start an Insurance Business

All right, and so today we have Karnes Roya, the CEO of Cover, which was in the Winter 2016 batch of YC. So, Karnes, what does Cover do for us?

"All, thanks for hosting me! I appreciate it. So, you can think of Cover as a multi-line national property insurance entity. Our customers download our apps; we ask a couple of simple underlying questions, and they take pictures and videos of things they want to insure. So, this could be cars; they could walk us around their homes, pets, jewelry, electronics. We basically make a market for just about anything you can take a picture of. And it's processed with computer vision, not human crowd. On the home side, it's computer. So, we use a TensorFlow-based camera to identify catalog your property so that when you need to make a claim, there isn't much of a fuss that's put up by an adjuster."

Okay, and now is it assessing more than what the object is, or is it just like, 'This is a bicycle'?

"No, no, that's actually— I mean, so the value of us being a visual app is twofold. One, you know, it acts as a sophisticated frontline underwriter. We're proving as a property existed in a given time, place, and condition. Okay? And that helps materially improve our loss runs. And then, for the customer, you know, what it means is that there can't be very much pushback in the instance of a claim. You have proof that your property existed; an adjuster can't come back and say that, 'Hey, that television that you're trying to claim is actually something that's like, you know, an inferior model,' or something."

Yeah, got you. And so are your models constantly adjusting per person? Rather, like, should I be photographing everything?

"Okay, so it again it's just a tool for us; it's not necessarily the central tenets of what we do. A big part of it is to simplify the onboarding of getting insurance, making it a bit more natural on native mobile. Okay? And then this is kind of an interesting divergence for you because before you had a style startup, yeah, before that you were in consulting, right? So maybe you should explain how you ended up here because I think it's— yeah, it's, yes."

"Yeah, sure. So I was a management consultant in a past life. So I was at Oliver Wyman, specifically in their financial services practice, and I got a CFA at some point. I went to MIT, studied finance. So, you know, I was in their finance and risk practice—a little bit of insurance work. It was great; it helped pick me— it helped me pick up a little bit of polish. The two years that I was there, I certainly could model things and, you know, build PowerPoint slide decks, and that, you know, that was certainly to my benefit. I, you know, I'm a person of the experience that I had there, but at the end of the day, you know, I kind of wasn't getting what I wanted out of the experience. The risk-adjusted returns to being a professional services like managing consulting or banking or private equity are great, but, you know, because they're great, I view them as like temporary. And so, you know, I'm looking at it, I'm like, the long run eventually I can, you know, use reversion to the mean, and I kind of want to make the jump to something where I'm building things, I'm taking risks, and you know, that just gelled better with my personality."

"So did you find that you were struggling to care while you were consulting or you just didn't feel there was sufficient upside?"

"Well, I was learning to code while consulting. Okay? And so I was like, well, I'd rather be a maker of things than an advisor, right? And so it's fine. One of the questions I kept like percolating in my head was like, you know, we're paid millions of dollars to advise banks to do things and, you know, to improve operational processes, and, you know, I kind of wondered why the partners that I worked with didn't build a bank, right? And I posed that question to a couple of the partners and the answer was, 'Hey man, like we make a lot of money doing this; we're not gonna end up taking these. We're single, quite frankly the scope of building a bank is way larger than the thing that we focus on whether it's like, you know, in an insurance context solvency-related in the banking context stress testing related or, you know, operational or something to that effect.' Yeah, I mean, I was like, well I'd be the guy that wants to build a bank, right? And so I made a jump."

"It was kind of— it was kind of ad hoc. You know, we— the very first thing that I built with an aunt, who's our CTO, was actually a body scanner. And so we use the, you know, three Kinect sensors because it's like a depth sensor webcam, three points of perspective, to get a body form, and we thought that we could use math to shave away clothes and, you know, use that body form to inform like sizing decisions on the Internet. Yep, right? So super creepy thing, you know, and I think if we had made a couple of logical checks before we had ventured off to do this thing, like why aren't there body scanners in every mall in America? I probably could have realized that if this was something that, you know, consumers would have done, and people would be there to be like a war for mall space, like mall space across America."

"Now, was this something that was intended to be a startup or did you just think it was a cool side project? It was intended to be a startup!"

"So we're going to see left at that point; yeah, we had left. We were pitching the gap and like a fair number of, yeah, you know, large, large retailers and brands and so we didn't get traction on that. The technology was particularly cool, and so we stepped back and thought about, hey, like what are we actually trying to do here? We're effectively trying to predict, yeah, you know, the sizing and stylistic preferences of our customers. And so, you know, over the course of like a weekend, we built a super simple app where we scraped a whole bunch of lifestyle bloggers and fashion bloggers' content and threw up outfits in a single— like, threw up outfits and allowed people to scroll and double tap on individual parts of an outfit, right?"

"Yep, so all we wanted to see that was that, you know, somebody would indicate a preference for like a blouse or like a handbag or like a hat or a pair of shoes. Lo and behold, like the activity was insane! Yeah, right? And so we were picking up users. So we event— we even eventually stepped back, we pulled it back, and then we built that something that would register the unique structural elements of individual items that people would interact with. So like if you double tap them like a handbag, we know the price point, the colors."

"Huh, the unique structure— always computer vision again?"

"Yeah, so actually it was even more rudimentary than that; we were just tagging everything. It was super manual. Okay? And so that was working, and eventually, we start— we start against a point where we were predicting what people would want to see next. All right? And so with StyleKick, actually we applied to YC like four or five times leaner. We interviewed three times, actually. Yeah, and by the last time I was interviewed we had 400,000 active users and, sort, in the business was based, sort— was it sort of like supply? So it was like a—it was a marketplace. So eventually what we had was designers, you know, influencers uploading looks, you know, and we were earning affiliate income doing that."

"Okay, and at that point, so you had four hundred thousand active users? Yeah! How much money were you guys making out of those?"

"Not nearly enough. Just barely to cover our hosting costs, and the reason for that is a lot of the stuff that was being sold was super high-end. So, you know, back then on a foreign screen or smaller, yeah, you know, selling a pair of 300 jeans wasn't an impulse price purchase. You know, it was something we ended up acting much more likely to a desktop purchasing experience. You know, and in retrospect, you know, what we probably would have done is step back and looked at was actually working within that. But you know, the marketplace native mobile e-commerce system—the things that were working like Wish, right? So impulse price purchases, 5 bucks, 20 bucks, where you didn't need to think twice about what you were buying. Yeah, it was also working. We know we relied heavily on influencer marketing to grow to eventually a million actors—this was 2014. It was like 2013."

"Yeah, and so if we step back, we would have recognized that there were a lot of these, you know, individual brands that were starting to stand up on top of Instagram. Yep, and we probably could have gone the route of actually building a fashion brand or, you know, selling products of our own record. So the Lossy is a great example of that. Those are the types at work either you went vertical or you went impulse price purchase. Mm-hmm, because something in the middle which is never gonna happen. Did you consider, you know, a full desktop interface where people could save and then purchase?"

"We did; I think, you know, so I can— Etta Port and Farfetch and a bunch of guys went that route. Yeah, I think we just ran out of steam. Yeah, I mean, okay, yeah. And you know, in and around the time we were running out of steam, you know, we got some other folks at Shopify, particularly Craig Miller, who's the current CEO—"

"So T shoes a VP, yeah, like a GM there. And you know, we had a long-standing relationship and ultimately we decided to join Shopify and run a mobile product team. So, you know, we were there for a little under a year. We’re building experimental marketplace apps for them. We still had the itch, right? Like we knew we could build, you know, really beautiful mobile products. StyleKick itself was, you know, featured just about every market, you know, on the App Store on the page repeatedly. We were at number one for style in France. It's crazy. You know, we’re driven millions of people through the app. So we knew how to drive distribution of your native mobile, and so we started thinking about, you know, what is the next thing that we want to do?"

"We knew we wanted to be mobile because that was one of our core strengths. We knew we could drive customers. We wanted to make sure at this point that the basic economic model worked, right? And so we did; we built things on weekends, you know? We drafted like Northern Ontario and Muskoka and just have a couple of beers and think about what we wanted to build. So we tried, you know, things in healthcare. We considered entertainment against Natalie, work for Russell Simmons, and it's kind of the impetus of us moving forward StyleKick."

"Okay? There were a couple of things we tried, you know, and eventually we just settled on insurance and the reason was, you know, we had been working with insurance brokers in the past to help us pick up policies for StyleKick and for ourselves. But this is not necessarily obvious, right? Because you were in Toronto, yeah, and now Cover is in the U.S—yet with a more competitive market! Right? So, I mean, what really got— what convinced you that this could work?"

"Yeah, so I think what we wanted to prove was that the pie was big enough to drive a significantly sized business, right? I mean, at the end of the day, it was a back of the envelope math that showed us that this could work. Right? A single auto policy that we sell in California, the average premium is something like $1,600 just as a distributor of insurance, never mind underwriting or moving down the margin stack. You know, we earn anywhere between 200 and 300 dollars of that policy in perpetuity, right? And so if you think about what churn looks like in insurance, like a very, very good broker is churning 10% per year, right? A good broker is turning 15 percent. So the LTV on a single auto insurance policy is thousands of dollars, right? Which is why you see Geico and State Farm spend billions of dollars a year trying to acquire a disgusting."

"Yeah, there, because they're just massive! But you also, in the beginning, knew that you couldn't afford to underwrite this stuff? Yeah, correct."

"Yeah, correct. I mean, so we were thinking about what's the capital-light way of venturing this business. And so the very first version of Cover was super simple, right? It was—we didn't have any licenses because we weren't selling any insurance; we didn't know anything really about insurance distribution. It was just a handful of views leading to a camera view with the preamble that said, 'Take a picture of something you want to insure.' That's it! And we launched it, and we used the methods that we used to grow StyleKick to grow Cover, you know? We ended up being a number one ranked insurance app in Canada the second day, and during YC, you were actually featured for like two weeks? Until launch, we were featured for like 'under the best new apps' in the United States!"

"Right. Now again, look, we didn't have insurance licenses like we didn't know anything about insurance. All we would see is that every 60 seconds, we're getting an insurance request."

"Right? So how do those conversations with investors go? Because I think you could— yeah, just as a distributor, I think a common pushback would be like, 'Oh wow, isn't this just like running to zero, Mark?'"

"Yeah, yeah. So I think like any purely Gen business is just relying on a marketing arbitrage, right, to continue to exist. And there are lots of examples of that in insurance. I don't think that you can build a durable business just as a lead-gen service for a couple of reasons: one, those MO arbitrage opportunities eventually dissipate; the second is, you're not building any brand, equity or value, and you're certainly not, you know, doing your customers a service by vacuuming up their contact information and selling it on ten times the average revenue per user, right? Especially as a startup venture back, yeah."

"And so, and so we were like, well we knew exactly what we needed to do. We're like, okay, well we need to start owning the actual customer service, like the selling experience, and eventually more, yeah. And so during YC we were like, okay, well we're committed to getting licensed; we're committed to getting licensed across the country; we're committed to getting as many insurance markets as we can on the Cover platform so that when a customer comes in the door, irrespective of what it is they're looking to cover, we're a competitive place. And so we stood up a national insurance brokerage in under 12 months, right? So we were in 49 states with 30 carriers in under 12 months, which is I think enough to show that, you know, we could de-risk ourselves from an execution perspective because that's a non-trivial thing to do."

"And so I know, how much did you have to raise to accomplish that?"

"So that was YC; so if you're in Winter 2016, yeah, so by Winter 2017, you were in that position? Yeah, right, and so we raised—I mean we raised 3.2 million bucks, and I—so slightly some of it before demo day, a lot of it after demo day, huh? We actually had most of our seed round still when we raised hooray, and so a big part of it was we knew that there would be—there was going to be a ramp. So we didn't make our first hire for probably the first year and a half, right? We knew that it would take a little bit of time to get everything before we could start to scale. But yeah, we raised our A as soon as we were in a position to be able to be selling an insurance business."

"Okay, and so when you raised your A, did you feel that you had product market fit?"

"I mean, if you take a look at our reviews— reviews! The reviews? Okay, like for Cover, I think we do—I mean, we do everything we say we're gonna do, right? You know, a customer comes in the door, we make the underwriting application process super simple. You know, we do not spam our customers with phone calls or emails. Every single one of our customers gets a dedicated text line where they can text us back and forth to ask questions. It's a—it's truly like, I think—I think millennial-first experience, which is a cliché thing to say, but it's like the thing that we do. So, yeah, I mean, like we do what we say, and you know, we’re doing well because of it."

"Mm-hmm, and so when it comes to growth, you guys just raised a B—"

"Yeah, on 18, correct. It's still compared to Geico, you know, nothing! Yeah, a tiny amount of cash, yeah, no, we have meet their other two, sure. Text editor of raises hundreds of millions of dollars now, right? Yeah. Slow drop in the bucket. Well, in the insurance market there are five hundred billion dollars in insurance premiums that are written just in the property market in the United States every year, right? And so, look like that—that's not of concern to us, because one, we don't take balance sheet risk yet. I mean eventually we may be because we can be pretty opportunistic about the types of requests that are coming in and whether we want to stand up products for those."

"So in practical terms for people who aren't— it's what does that mean? That means that like if we see— if we see a lot of customers who are coming in for a specific thing, right? Like maybe it's home or maybe it's auto in a given geography and we're writing it profitably, we should be—we can actually just step in and underwrite the business ourselves, meaning that we take the risk or we price the risk, take part of the risk, and sell the rest of it on, okay? And if we take any risk at all, we need to hold the capital in reserve to be able to take that risk, right?"

"Right now, we—we did not, right? And so, which is why we can be fairly capital-light in focus, mostly on, you know, the distribution aspect of the business and act as sophisticated frontline underwriters, right? So we're being actually pretty prudent about the types of risks that we do send to our carrier partners. Mmm, okay."

"Yeah, I mean you can think of insurance as three discrete businesses, right? Like distribution, which is a business center of itself, underwriting, claims servicing, and, you know, regulatory overhead. Folks who are attacking, you know, every aspect of this at once; we're probably not gonna get any of them right and burn a lot of money right out the gate, and those are the folks that have tended to raise hundreds of millions of dollars to get to that point right."

"Well, I think our goal right now is to make sure that we're continuously iterating on the best possible front-end experience, making sure that we're proving out that the product itself is a good way, a mechanism to filter out bad risks, yeah, and then move into underwriting and maybe the rest of the stock. Okay, yeah, gotcha."

"And so how does your lead-gen work now?"

"Yeah, we acquire exclusively through our apps. So, the App Store, Google Play, really just highly rated! So it's all search, yeah? So, I mean, yes! So, we were the beneficiaries of like pretty decent ASO very early on, yeah? And we run a lot—we run a lot of paid, but we take it generally speaking like a portfolio approach to growth, right?"

"So, manok, like you, we would acquire via all the channels you expect; they say consider and read it, like what have you. We're good at ASO. We're starting to build up Cover.com; that'll probably be a fair material entity for us, yeah, and then we build things that are like what we consider to be pre-insurance, right? And so these are tools that we think are useful to our customers irrespective of whether they buy insurance from us right away or not. Yep."

"A good example of that it's like in chin in the Cover app price drop alerts. You don't have to buy an insurance from us. What we can do is just programmatically market you em every time we see like a violation fall off of your record, you know, or a claim fall off of your record. Oh, you're coming up for renewal, that's a service you don't need to pay for, but it's pretty valuable because you don't have to think about it. Yeah, that's pretty—another example is, you know, for certain markets, we're giving away access to a defensive driving school, right?"

"So if you go driving over calm, you know, if you— a very stylized example in Texas, you know, if you go to defensive driving school, you can get a violation knocked off of your record or get access to an insurance discount, right? So it can be material! If you're like 10% of the overall premium, so they're talking about hundreds of dollars, right? Usually, those things, like there are entities that charge anywhere from like 25 to 150 bucks to go through that course. It takes like a couple hours; we’re a technology company; we just build it and give it away for free, yeah?"

"And so that can be funneled directly into our products. So, you know, in Texas there are like four hundred eighty thousand people over the course of the last eighteen months that went through defensive driving school. Wow! Just in Texas, right? And and so it's like if you think about, like, hey, where can I be like where are my customers not and how could I build product to, you know, facilitate transition of those customers from those sources to the actual the apps, you can build fairly durable distribution."

"And were you thinking about these kinds of programs when you were in YC, or is this just by talking to your customers, figuring it out?"

"I think it was observing the types of things that, like, you know, folks were—we're asking: like, do you have a defensive driving—really? Yeah? Like, will literally ask us, do you have military discount? Do you have this? Do you have that? And we're like, oh, well, maybe we haven't considered this, and I should go build something."

"Okay, to support this. The other thing I do is like, I really— like I read a lot of insurance rate filings, and so it's a weird thing to do, but I get a fuller appreciation of the way that, one, the filings are structured, but also discounts that are generally speaking available to our customers, geyser, and how we can build product to make it easier to access those discounts, right? And so, yeah, when someone might be shopping around, they're gonna start comparing these things, and this fringe benefit might matter a lot to them, yeah?"

"And yeah, me, and Geico, for example, has this weird, you know, affiliate discount where if you own a share of Berkshire Class B stock, you get a percent discount on Geico, right? Really? Just quite easy!"

"Yeah, well, if there you go, now you know I should take advantage! So, you mentioned Cover.com, and I forgot to put this in my notes, but you guys just bought this domain for recently, right?"

"Yeah, so this is—we've never talked about this on the podcast, but you spent a lot of money! Yeah, and I got all that crap for it too on news. Yeah, really? Yeah, I mean, look like the crux of it is like what's the ROI on buying, you know, something like that? Yeah, look, buying a very expensive domain is not gonna make or break your business! Like, because we bought Cover.com does not mean we're gonna be successful; it also doesn't mean that we're gonna fail."

"Right! We're gonna—we would fail if we did something fundamentally wrong or, you know, continued to build on something that didn't have product market fit. We know we have product market fit; people buy our products, people spend thousands of dollars on those. So from my perspective, it was, hey, look, right now we're super surgical about how we acquire, right? Like we have to be because they're capital constrained at some point or another. We know that like an equilibrium, we are competing on general awareness, right? We're in the trust business, and any minor change, right, in conversion, given super high LTVs is gonna make a huge impact!"

"Right! If you have a $2,000 LTV and conversion moves up a couple bits, yeah, up or down a couple bits when you have tens of thousands of people coming in through every single day, the domain pays for itself, yeah, very, very quickly, yeah?"

"And so, you know, outside of that, if we're competing on general awareness, yeah, we care about discoverability, right? That matters! We care that we're discoverable, you know? And we own—it's like one of the more common words in the English language. And so, in other words, you would say if you have the cash, do it?"

"I think it's on a case-by-case basis. I think if it makes sense for your business and you have to have like a very good sense of what the underlying economic model is, right? Like we can do the sensitivity analysis; it's like minor changes in conversion and how that eventually affects the bottom line for us. Yeah, and for us, it's like if we're writing a million policies a year, it’s gonna actually take like a couple thousand policies like less than a couple thousand policies to pay for this, cause it was a nine hundred, so we paid 750 for it."

"Okay, a broker peace? Yeah, okay, gotcha! If you have any pro tips, it's just—I mean look, like the guy that we worked with if you're actually considering, you know, buying a domain who’s great, you know? Super as a professional. I was introduced to us by Arjun, set the tribe, and it's like on our board, just our lead investor. But a lot of the folks in this bit are super shady, right? You should be pretty careful; I would recommend using a broker irrespective—like the general trading aspect, you know, the initial bids that we were getting were in like the million and a half to two million dollar range, right?"

"And I can guarantee you that if we had tried to buy it after our B or C, we'd be looking at like five, six million dollars, right? Okay, so did they know it was you? No, okay, yeah, we proctored it, we had a broker who proxy through another broker. Really? So that—yes, I've heard about people who owned many of like these big deals, they will only work with the customer! Yeah, yeah."

"I mean, I—you know, I think we were lucky enough to have network access. Yeah, it's not necessarily afforded to everybody, but it saved us a lot of money! Yeah, that's great. So, to go off on a tangent a little bit, you are engaged to one of your co-founders! Yeah! Yeah, right?"

"So, I got engaged to Natalie who heads up all the product at Cover when we were at Shopify because we think we were thinking that we would, you know, be there for a while, okay? And this would be a little bit more stable than, you know, the ups and downs of—for a wedding budget, for a wedding, yeah. Give him a house! Shopify stock is done! Yeah, yeah."

"So, we were at Shopify like, okay, this is gonna happen, like we'll plan a sale! We have some time now! Yeah, then we applied to YC, right? And so we were like, oh, so many times like what's another time? This—we applied in like, you know, we got in! All right!"

"And it was like a basic prototype! We didn't have any users really, which is, I guess, the way the world goes! And then, like, two weeks later, we flew from Toronto to California, right? And you— Zeus to the last minute to get some housing! Yeah, and since then it's been a rollercoaster, right? Because like we've, you know, we did YC, raised our seed, started scaling, built up a business, and it's been a pretty aggressive run. We're probably gonna get married this summer."

"Okay, yeah! Very exciting! And so how do you manage that relationship, you know? Because there are obviously, you know, relationship tension, co-founder tension, and then ya deal with, you know, employees at the same time, yeah?"

"Yeah, I think like— look, commingling professional and personal risk, especially if you're starting a start-up—yeah, I mean, it's not necessarily a recipe for success, right? To be quite frank, I think we're the beneficiaries of having worked on two businesses together and being—having been acquired in Shopify and working on the same product team, right, and there's like a clear division, you know, the scope of responsibility, a role. She’s an exceptional designer, she's an exceptional product person; she got, you know, StyleKick and Cover featured on the front page of the U.S. App Store simultaneously, which I think very few people like a vet can have claim to. And I trust her completely to take care of that function, right?"

"I work on the rest of the business, so I work on the insurance aspects of the business. I'm thinking about what do we need to do to continue to deliver an exceptional product experience, what products should we be standing up, who should we be working with? Yeah, they're very different roles, right? So that helps a lot. But I think it's been a net benefit, right? Because the cohesiveness of our co-founder team I think is like not as, you know, it's pretty atypical, you know, and I went to high school with Natalie; I am engaged to Ben, who was our first hire into StyleKick, and so we've been working with him for six years, right? That helps a lot!"

"Yeah, and it's like one—the strength of that core team because they're all, you know, like awesome in what they do, it's helped us hire very well. And when it comes down to the actual tactics that you use to make sure the relationship, you know, the conflicts will obviously happen but it doesn't implode— like, what are the tactics? Say?"

"Yeah, yeah, I mean like eventually you feel out, you know, with your partner what you need to do to de-escalate, you know? You will know who exactly to put buttons you can push and you shouldn't be pushing, and you step back and you realize that you're on the same team."

"Mm-hmm, and that like helps! Certainly, you have to put any rules in place in terms of, you know, like living together, working together. I don't think there are no hard and fast rules. Actually, our co-founders still live too, you know? Yeah, so I mean we're as tight as you can get. I think, yeah."

"So now that you guys have raised your series B and, yeah, are quite large, you have two offices! So you have two, Toronto and your San Francisco office. Now, as the CEO, how do you manage that, and how do you make sure everyone's come?"

"Yes, so Anna is full-time in Toronto. Natalie flies between the two offices. Okay, I mean most of products it's in Toronto. We have some—we have seen some senior engineering in San Francisco; sales lives in San Francisco; insurance operations lives in San Francisco. Mhm, so I mean the folks are running their respective functions or where their teams are, right? And then I make an effort to be wherever we can be, and we run these off-sites— like, we're heading to Vegas next week, okay? The entire—what, the entirety of the team will be together!"

"Okay, yeah! So again, division of labor! We invest— we've invested in firms in like a very high quality, you know, AV equipment to make sure that, you know, at least communication is a solved problem. Okay? And in terms of culture, how do you keep people on the same page?"

"Yeah, so that's interesting! I mean, actually I think the cultures are—the individual cultures are pretty different, right? Like one is very product-oriented, and we have a sales office in San Francisco that is like super rambunctious, right? Like I actually don't have a problem with that at this point. I think what my job to it primarily is to make sure that we're all aligned in what the ultimate goals are for our organization, right? And we're trying to build it a fair and sustainable insurance entity at the end of the day, right? We know what we need to do; we need to make getting insurance easier; we need to make it fairly priced for individuals for whom it should be fairly priced; we need to be in and around when claims are being made and help facilitate those claims so that our customers have positive experiences around that, because that's the ultimate function of an insurance company, right? And yeah, effectively act in that advisory capacity and scale that across just about any line of business."

"Okay? An insurance that we can—that make sense! And so as an international founder, we have a million questions coming in from Twitter all the time; every—it's like, oh, how do you deal with this as an international founder applying away? So you're doing my CEO or post-YC, essentially? I think the most common place of anxiety where people are like, I think I can figure this out in a three-month term—longer term—what is your advice to those people? Like how do they make it work?"

"Yeah, so look like a—we were all in San Francisco and decided to open a Toronto office, right? Because we know the market, we—yeah, we knew where to look. We have a pretty good—and it's a super high-quality team, right? That decision we made concertedly because we had intention in doing that. Okay?"

"I mean if you're talking specifically about, hey, like if I'm an international founder, should I stay in the Bay Area or should I be somewhere else? You know, my objective opinion on that is this is the beating heart, right? Like this is a beating heart of everything technology, and you should have a presence here, you know? Absolutely, if you can get your visa issues resolved."

"Yeah, I mean even, even we had issues—like Anna was stopped at the border like three times, turned around. I had to get a one, which is kind of ridiculous, but that stuff happens, you just have to work through it. Yeah, it happens to everyone, and it's like you make it through! Yeah."

"But obviously, neither of us are lawyers, so yeah, consult. So the new batch is starting; yeah, you guys haven't been out for that long, Winter 2016, so it's in recent memory still. What advice would you give to the people just entering the batch?"

"Yeah, so, so actually I ran a Start School batch, right? So there are a couple of the companies that were in my batch are now in YC; and certainly, like I've been doing this for a little while, like which, yeah, pre-coaching folks who are writing YC. One of the folks—the first folks I did it before was Austin Atlantis. Cool, okay, cool! And I have been like one of the first investors, giving me the opportunity to be one of the first investors in Land of School too."

"Look, I think I— the core advice that I give is actually not that different from what YC partners give, right? It's like, hey, focus! Pick a KPI, make sure you're growing that KPI. You know, everyone implicitly knows that a couple inches deep into your startup, everything is held together by duct tape and glue. The bet that folks are making on you is that eventually you're gonna figure it out, right? So, so go into it with that in mind; make sure you hit your KPI, and you talk about it, right? Because it's important to actually talk about your successes."

"And outside of that, the typical advice, you know, is don't hire anybody because you shouldn't be hiring; don't go to conferences or parties that are not directly related to the KPI. So if it's gonna make you revenue, if we're gonna build revenue doing it, fine! But otherwise, don't be doing it, yeah? And otherwise, like, don't get distracted! This is like a three-month sprint for you to truly accelerate what you're doing, right? Yeah, and either get to, you know, YC is like a forcing function, right? It's like if you're not—you don't have product market fit, it's gonna help—you’re gonna help you determine you don't have product market fit pretty quickly, and if you do have product market fit, it's gonna be like, okay fine, it's an accelerant."

"Yeah, and what about post-YC? I know there are many companies who kind of go through a lull and they need to decompress for, you know, a week or something?"

"Yeah, but—yeah, you guys find your rhythm again? Yeah, I think, I mean we're the beneficiaries of having done this a second time; the ups and downs are not as aggressive, right? Like, the highs are not as high, the lows are not as well, and so we kind of just kept our heads down. All the lessons we learned from the first time around were compost. Mm-hmm, and so we could—we have a pretty good cadence of just making sure we're hitting our goals and communicating that to investors just why we raised in a pretty quick way, right?"

"But yeah, you just—you keep doing what you were turning during YC, yeah? You should be okay. Okay, cool man, well this has been great! Thank you so much!"

"Yeah, thanks for having me! I appreciate it."