Venture Capital: Kyle Harrison

Summary

The explainer episode, and the most useful thing in the corpus for someone who doesn’t already speak venture. Colby Howard’s show is built on the premise of finding out what people actually do all day, so Kyle spends 45 minutes on mechanics rather than theses — the funnel, the fee structure, the returns math, the decision processes, and what the job feels like from the inside. Recorded 8 November 2023 (Kyle dates it by the day OpenAI went down for two hours), published 5 December.

The job, described honestly. Kyle pushes back on the romantic version immediately: “I actually don’t spend a lot of time sitting in a chair smoking a pipe pondering the future state of things.” He resists even the word predicting — the founders are the ones inventing the future; his work is understanding what’s happening and judging which behaviours have staying power, which makes it “quite a bit of human psychology.” His days are organized around a handful of specific companies that are currently forcing him to think.

The three jobs: finding, picking, winning. The cleanest statement of the VC job description anywhere in the corpus. Finding is anything that puts a company on your radar; picking is moving it down the funnel (and choosing not to is equally a pick); winning is earning the right to invest, because the ones you chose may not choose you. Contrary’s difference is on finding: a ~500-person talent community that gives signal about where sharp people are going. SpaceX is his talent vortex example — and the heuristic is concrete: “if I find a company that has 20 people and six of them came from SpaceX, that’s interesting.” He contrasts this with genuinely thesis-driven firms (USV) and incubation shops (Lux Capital, Founders Fund) that “bust out the pipe and imagine the future” and then go find someone building toward it — a different strategy, not a worse one.

The prepared mind, with the luck caveat attached. Via Arthur Patterson at Accel (Kyle hedges the attribution on air): the discipline of having thought hard about a space beforehand, so you recognize value when it appears. Accel and Facebook is the canonical case — having worked out what actually drives network effects versus what didn’t work for Myspace and Friendster. But he immediately adds the honest part: “Facebook could have gone really badly just as well as it could have gone really well.”

Who the customer actually is. The clearest thing here. Founders are often described as VCs’ customers and in one sense are — you’re building a product for them, even if that product is only money. But “I am making money for my LPs”; they gave you capital and the product you built for them is returns. The carry is a margin like any other: “when Walmart sells you something they had to pay somebody to be able to get that thing… that’s my margin.”

The returns math, stated plainly. Below 1x you lost money and probably won’t raise again. Most people benchmark to the S&P 500 — and “the vast majority of venture funds don’t beat the S&P 500. You would be better off just putting your money in the stock market.” What sustains the industry is that everyone, “myself included,” believes they’re in the top decile. The structural cover is the fund cycle: deploy over two or three years, hold for ten, and raise fund two and three before fund one has produced an outcome — “you could be on your third or fourth fund before you’ve ever actually seen an outcome.” Distribution of a single fund: roughly 80% to zero, 10–15% middling one-to-two-x, and ~5% doing all the work at 10x-plus. Asked how many of those 5% companies exist across a decade, he says maybe hundreds — noting that Moderna and Apple were venture-backed too, so the category is broader than people assume.

Cold email, warm intros, and the homogeneity problem. Sub-10% of cold inbound gets a response, and Kyle doesn’t defend it so much as explain the trap: warm intros come from people you know, and “who do I know? Probably people that look like me.” He cites the statistic that something like 80%+ of VCs come from about three schools — “which is insane… it’s just these bubbles.” Colby’s line lands it: Harvard has filtered them already, and then they filter for themselves. Kyle’s response is not a defense: “I recognize that that is a problem.” (He notes BYU is not one of those three.)

How diligence actually runs. For a hypothetical AI recruiting company: talk to twenty recruiters, using expert networks at roughly $500 a call when he has no contacts in a space. Reference-check the founder on whether they can build it or hire someone who can. Ask genuinely technical questions — if you need LinkedIn data and LinkedIn doesn’t want you to have it, how does this work? Separate the enthusiastic end user from the person holding the budget: “the recruiter might say I’d love this, but the head of recruiting is like, we don’t have the budget.” And Warren Buffett’s too hard pile — not bad, just not something he’ll figure out. On being early: “there is such a thing as being right but early and therefore wrong,” because if the technology needs four more years, “what am I doing? I’m just paying your bills for four years.”

Why revenue is not the safety people assume. The counterintuitive middle of the episode. Past a threshold, companies mostly don’t die — they become lifestyle businesses, which is a bad outcome for the investor even though nothing failed. Debt is the thing that actually kills, which is part of why VCs have historically been allergic to it. Investors with enough ownership can force a sale for liquidity. His illustration is InVision versus Figma: when Figma was around $4M of revenue, InVision was at ~$100M and looked like the obvious winner — “why would we ever invest in Figma?” Six or seven years later the answer is clear, and InVision didn’t die so much as stop being able to become what its investors needed.

Decision-making, and what Contrary does. Three models: dictatorial (one founding billionaire decides and everyone pitches them), egalitarian (every partner scores every deal 1–10, often with fives and sixes banned so nobody hides in the middle, and an average has to clear a bar — where a senior partner who made the firm billions can still get voted down), and Contrary’s, where each GP has independent check-writing authority. His caveat is the interesting part: that latitude doesn’t mean acting alone, because “I respect my partners, which is why I work with them.” The design intent — “we want everybody to be able to make contrarian investment decisions.”

The closing answer is the bluntest thing he says on any podcast. Asked what a 25-year-old entering venture misunderstands: that capital equals expertise. “Most VCs are pretty stupid — like everybody’s pretty stupid.” But VCs are often worse than average because they inhale their own hype: founders are deferential because you have capital, you conclude you must know something, and you start giving bad advice. “You’re actually probably wrong 80% of the time, 90% of the time — if all of your companies are going to zero, you must have been wrong.” And his demolition of pattern recognition: what gets called pattern-matching is usually “a very lazy heuristic” that isn’t finding patterns but looking for duplication — not someone with Zuckerberg’s characteristics but “literally a nerdy white dude from Harvard who learned to code when he was 12 and his parents are rich.” His example of where that leads: FTX.

Transcript

~45 minutes, hosted by Colby Howard on Colby Howard Wants Your Job. Recorded 8 November 2023 — Kyle dates it by the OpenAI outage — and published 5 December. ASR errors and names cleaned, text paragraphed; nothing reordered or summarized. Corrections applied: Accel, Arthur Patterson, Friendster, USV, Moderna, InVision, Figma, IRR, “too hard pile,” Colby.

Cold open

Colby Howard: Behind the scenes of venture capital. This week we speak with Kyle Harrison, general partner at Contrary. He goes through everything that happens in a venture capitalist’s day: how do these guys even find the companies they’re going to invest in, how do you invest over a 10-year time horizon, how often are they right — more importantly, how often are they wrong? He goes through everything that is a misconception about the industry from an outsider’s perspective. Kyle is fantastic, I learned a ton. As always, we’re learning together — let’s dive in.

Hello, hello, hello, and welcome to Colby Howard Wants Your Job, the show where we find out what the heck everyone else does with their day. Have a very special guest today — Kyle Harrison. How’s it going?

Kyle Harrison: Doing good, man. Thanks for having me.

Colby Howard: Absolutely. General partner at Contrary, super thoughtful about all things venture capital, which is why he’s the perfect person to go through exactly what a venture capitalist does.

We are filming this on November 8th — the day that OpenAI went down for two hours and shook the tech world.

Kyle Harrison: That’s right. No one could do their homework. No one could chatbot their chatbots.

Colby Howard: It’s a monumental day in the technology world. But to start out — quick detour, zagging right away — you went to BYU, correct? Undergrad?

Kyle Harrison: Undergrad, Brigham Young.

Colby Howard: I remember when I was interning at Goldman my sophomore summer, you had the Ivy kids — I was at Davidson, a 1,700-person school in the South — and you had Indiana business school undergrad and BYU kids, and the BYU kids crushed it. It feels like a super under-the-radar school. What’s going on there?

Kyle Harrison: Clean living, man. We’re not drinking, we’re not doing drugs, we’re just grinding on Wall Street.

Colby Howard: I love it, man.

Kyle Harrison: I think it’s just a very focused group of people who want to be very successful. They worked hard and were extremely smart.

Thinking on a ten-year horizon

Colby Howard: To start out very high level — most of us sit down at our desk and think about the next hour, the next day, maybe the next month. If you’re an investor in the stock market you’re maybe thinking about a six-month or year-long time horizon. Private equity, your hit rate needs to be pretty high and you’re thinking about maybe a five-year timeline. You’re a venture capitalist and your hit rate might be 10% over a 10-year period, and you are doing incredibly well. How do you think about that mindset being so much different from most other people’s?

Kyle Harrison: I think every investor has their own scorecard, so I don’t know that it’s a blanket “everyone’s doing well,” because there are certainly pockets of time. There are a lot of people who were doing super well in 2021 and now are feeling significant amounts of pain. So you don’t want to count your chickens before they hatch. Performance is relative.

But the job is unique, because you’re trying to predict very long time frames, and it’s very difficult to say what’s going to happen in ten years. I think it’s more a question of what do you see the early seeds of happening today, and what does that mean in the future — and what do you have to believe for a company to be big?

I spend a lot of time trying to predict what is likely to happen. I don’t know that we would ever describe ourselves as predicting the future, but you’re trying to predict specific behaviors and trends that have staying power. Is it likely that people will keep doing something they’ve been doing for a long time in the exact same way — is that likely to stick around? Or is it more likely that people are going to change their behaviors? So there’s actually quite a bit of human psychology wrapped up in a venture investor’s attempt to figure out what’s going to happen.

Colby Howard: When you walk into the office in the morning — Slack, email, meetings, personal thinking time — what does a normal day look like? What are you doing, who are you doing it with, and what’s the goal of each part?

Kyle Harrison: Every activity you do is centered around a company. Because at the end of the day, I actually don’t spend a lot of time sitting in a chair smoking a pipe pondering the future state of things.

Colby Howard: It kind of sounded like that.

Kyle Harrison: That’s right — no pipe. Clean living.

It’s less about us predicting the future and more about us trying to understand what’s happening in the world. The people who are actually inventing the future are the founders, the people we work with. So most of the time I spend thinking about — right now I could list for you six companies that are forcing me to think about things in a specific way, and to understand what I believe this company is going to do, whether it’s going to work. Those companies center my thinking.

Finding, picking, winning

Colby Howard: To follow on that — how did you get to know that these six companies existed?

Kyle Harrison: What you’re talking about is deal flow, as people describe it — where does your deal flow come from?

My firm is a little bit unique. Most firms either have a strong enough brand that people come to them, or they’re doing a lot of cold outbound and reaching out to random people. Contrary manages a 500-person talent community. There are hundreds of people we’ve identified from top companies, great schools, whatever, and we stay very close to them, and they often give us signal into what’s interesting — what companies they’re starting, what companies they want to go work at, or if one of their sharp friends goes to work at a company they tell us about it.

So most of the companies I find, it’s because I’ve paid attention to some talent signal. Some smart person went to work somewhere, or there’s a really good company that produces good talent — we call them talent vortexes. SpaceX is a talent vortex; there are tons of really smart people who work at SpaceX. I spend a lot of time paying close attention to those talent vortexes, and then: okay, where do people go when they leave? If I find a company that has 20 people and six of them came from SpaceX, that’s interesting. I want to understand why that company has attracted so many good people.

Colby Howard: Sometimes you hear about having a thesis about the future — thesis-driven investing: I think the future is going to be this, let me see what companies fit into that. It sounds like what you’re saying, and it’s not binary, is: let me talk with as many smart people as I can, they’re seeing something — so much so that they’re starting a company around it or joining a company around it — and that’s a signal for me to be interested.

Kyle Harrison: They basically talk about three jobs that a VC has. It’s finding interesting companies, then picking which companies to back — because if I find six, I’m not necessarily going to invest in all six. And then even if I pick two of the six, I have to win the right to invest in that company. I might pick those two, but they might not pick me, because they have other options.

So finding, picking and winning are largely the jobs that every investor has to do. We just happen, on the finding side, to do it a different way, which is very talent-focused — we’re very focused on people, and where they’re going, and what we think of the people associated with the company.

Firms like USV, Union Square Ventures have a very specific thesis-driven approach. They’ll literally sit down and bust out the pipe and imagine the future and say, I think the future looks like this, and now I’m going to go find the companies that are doing that thing. Or you’ll see incubations from firms like Lux or Founders Fund where they do the same thing — they sit down and predict the future and then try to find somebody to build in the direction of the future they’ve seen. It’s just a different strategy.

The prepared mind

Colby Howard: Part of this is an investor mindset versus an operator mindset — 90% of people go in to their job to do a job, you execute that day and go home. Theoretically you could spend your entire day — you just said you don’t — smoking a pipe, thinking about the future, reading a lot of publications, and theoretically be a very good venture capitalist.

Kyle Harrison: I think it’s Arthur Patterson, who was an investor at Accel — don’t quote me on that.

Colby Howard: Don’t put it online. Video recording or anything.

Kyle Harrison: That’s right.

Somebody like that talked about this idea of having a prepared mind — we have spent a lot of time thinking about that thing, doing that, reading about it, so that when a company appears and we have a prepared mind, we recognize that there’s value in it.

Accel, if I remember correctly — one of the examples of this was Facebook. Accel invested pretty early in Facebook, and they had a very prepared mind when it came to social media. They had thought about what the network effects are that really drive a social network, as opposed to things that had been tried before, like Myspace and Friendster, that maybe didn’t work. They had spent the time disciplining themselves to understand social networks, so that when they saw Facebook they recognized: this is something.

Some people would argue that there’s also a significant element of luck in that — that Facebook could have gone really badly just as well as it could have gone really well. So a prepared mind can work to an extent, but then there’s a question of how you measure success.

The mandate, and who the customer is

Colby Howard: There’s a lot of money out there that needs to be put somewhere — pensions, high-net-worth individuals. One place to put it is venture capital. Everyone in venture competes to say, give me that money, I can invest it really well. How would you describe the mandate?

Kyle Harrison: You can think about venture as a value chain. A lot of people make the argument that founders are VCs’ customers, and to some extent that’s true, because as a VC I’m building a product that’s valuable for that founder. Whether that’s just capital — the customer interaction is me giving you money and that’s it — or maybe I’ve built some kind of talent or support or business development offering that’s valuable to you.

But I am making money for my LPs. The pensions and endowments give me money, and in many ways they are my customer: they’ve given me money and I have created a product for them.

The product varies for every firm. For the average venture firm, that product is taking their money, identifying companies at various stages and sectors, investing that money, generating a return, and giving back the return. As a reward for that, I get to keep a certain percentage — which is the same thing as a margin. When Walmart sells you something, they had to pay somebody to get that thing; what’s left over is their margin. My margin is, in most firms, about 20% — whatever we make in excess of the money you gave me, 20% of that is what I keep.

There’s no one-size-fits-all venture firm. There are a lot of firms trying to be everything to everyone — investing across any stage and any sector, from crypto tokens to biotech to application software to the OpenAIs of the world; from very late-stage companies a year or two from going public, all the way down to a person with an idea, or even before they have an idea. You have to dictate your strategy — where is our focus area — and then you hope that strategy is effective.

Just recently somebody launched a $30 million venture fund whose explicit focus is fighting misinformation online. That’s their whole strategy — very, very niche.

Colby Howard: So the only companies they invest in are directly or tangentially related to fighting misinformation.

Kyle Harrison: That’s right. Or you’ll see the same thing with funds explicitly focused on longevity — increasing the average lifespan of a human being. A lot of pharma stuff. That’s their only focus, and those are usually smaller funds, because there are only so many longevity companies. And then you have dramatically broad firms managing tens of billions of dollars, investing across stages and sectors.

Colby Howard: If I’m starting a company, I’m making a bet that my idea is good and I can execute. You’re making the bet, when you launch the longevity fund, that longevity will be valuable, that companies will emerge, that you’ll be able to invest in them and they’ll do well. But the reason they’ll accept your money is because you’re the longevity expert?

Kyle Harrison: Yeah, if that’s your strategy. It might be that you have really specific relationships with big pharma companies — the companies you invest in need to go through distribution with big pharma, and you were an exec at one for 30 years and know everyone in the industry, so they really want to take your money because you’re really good at that. Or maybe somebody worked in the defense department for 30 years and then raises a fund explicitly focused on defense firms. Or you are Contrary and you’ve built a multi-hundred-person talent network, and as a startup you want to focus on hiring the very best people and we have a track record of placing exceptional people in great companies.

So everybody has their elevator pitch.

What counts as a good return

Colby Howard: Let’s say there are 100 venture firms and 10 of them generate most of the returns. What’s an okay return, what’s a bad return? If firm number 100 raised $100 million and invested it, five or ten years later they’ve done what?

Kyle Harrison: The worst is that you lost money — anything below 1x. You’ve lost, and you may never raise a fund again.

The only exception is that you invest a venture fund over the course of two or three years — that’s when you deploy it — but you keep the fund for ten years. You invest up front and then you wait.

Colby Howard: Which is a wild concept, by the way. That’s crazy.

Kyle Harrison: And what else you can do is invest in a bunch of companies that look like they’re working, or that are just so early it’s hard to tell. So you raise a fund, you invest fund one for two years, and then you wait. And while you’re waiting, you raise fund two. You invest that for two years, and then you wait. You could be on your third or fourth fund before you’ve ever actually seen an outcome — a company getting acquired or going public. By that point you’ve had many shots on goal, so even if your first fund is a zero, maybe your second or third fund has some good companies, and that can keep you going.

But for the most part, if you raise a fund and return less than the money you invested, you probably won’t stay in the business very long. So 1x is baseline — you at least want to return the money. And then most people benchmark to the S&P 500: I could have just put my money in stocks. What would that return have looked like over ten years?

The vast majority of venture funds don’t beat the S&P 500. You would be better off just putting your money in the stock market.

The difference is the hubris of most VCs — myself included — that you can be one of those top firms. And maybe there’s an element of it not being the same ten in everything; it shuffles depending on strategy. But what you believe is that you can drive outsized returns, better than the average person. The data indicates that the vast majority of people fail to do that.

Cold emails and warm intros

Colby Howard: Let’s go from high level to very nitty-gritty. Say I’m starting a company — an AI-backed recruiting company — and I’m emailing a whole bunch of people in venture capital saying, I’ve got a super interesting idea that’s going to change the world, and in parentheses, but I’m not putting it, also make you a lot of money and me a lot of money. And you might get that email.

Kyle Harrison: I’ve gotten lots of those emails.

Colby Howard: How are you reacting to that email?

Kyle Harrison: In that exact example, the vast majority are probably not getting responded to.

The thing that gets criticized about VCs is that they often look for warm introductions — they don’t really want people to cold-email them. I don’t know the data, but I would imagine sub-10% of cold emails to an investor get responded to. And some of them are super weird — borderline illegal. It can get really weird.

There are a lot of firms that pride themselves on saying we don’t accept warm intros. Because if you think about it: okay, who can I get a warm intro from? People that I know. And who do I know? Probably people that look like me.

I think something crazy like 80-plus percent of VCs have gone to one of three schools —

Colby Howard: Jesus Christ.

Kyle Harrison: It’s like Harvard, Stanford and Penn, or something like that. Which is insane. It’s just these bubbles. (BYU, I should say, is not one of those schools.)

Colby Howard: Harvard has filtered them already, and then they’re filtering for themselves, and everyone just gets to stay together.

Kyle Harrison: That’s right. It’s not great. So I recognize that that is a problem. But the reality is that most introductions come from I respect so-and-so, and so-and-so introduced me to a company.

I even actively try to curate relationships with people so that they trust me and feel like they want to send me things. It’s the same reason I write a lot and put out my writing — so that people get to know me, and then the people who read my writing and like me send stuff to me.

Moving a company through the funnel

Colby Howard: Let’s say it does get to you and you go, interesting — this is worth the conversation. What’s the bar?

Kyle Harrison: It goes back to finding, picking and winning. Any time I come across a company I’ve technically found it — whether I read about it in an article, or somebody sends it to me.

Picking is how far you move it down the funnel. I get an email, I look at it, I make a judgment call — that is actively me picking. And I might be wrong. I should have taken that intro and I didn’t, and it goes on to be a great company. Or I spent too much time on a company that in the end I said no to and shouldn’t have. Or I lose — I spent a ton of time with that company and they decide to go a different direction.

There’s a huge funnel of opportunities for that company to fall off my radar. I don’t know the numbers, and different firms are different, but I’d imagine that over the course of a year — if you count anything you find, meaning anything I read about, anything I think about that’s interesting, anything I see on Twitter and save to look at later — thousands of companies every year.

Colby Howard: And there are thousands of you.

Kyle Harrison: Everybody’s looking at different things, and there’s a whole network of people looking and making judgment calls. Some people like something, some people don’t. And there will obviously be overlap.

Colby Howard: There’s an upside to venture capital that’s more philosophical: you are betting and giving money to an extremely high-risk prospect. The odds of my company not being around in two or three years are pretty damn high.

Kyle Harrison: Like 90-plus percent.

Colby Howard: And you’re looking for companies with those odds.

Kyle Harrison: That’s the cost of doing business. You just acknowledge that that’s true.

Colby Howard: In gambling, in public market investing, if you have a 51% hit rate you can do extremely well. And yet here it’s: yep, I know a lot of these aren’t going to work, but there’s going to be one bet or two bets, and I’m constantly trying to find ten of which one could be that thing.

Kyle Harrison: Roughly — 80% of the venture investments you make probably go to zero. Maybe 10% or 15% of them are middling outcomes, one or two-x, where we didn’t lose money but my LPs would have been better off putting their money in the stock market. But then there’s that 5% of companies that cover everything else — they’re 10x-plus investments.

Colby Howard: And what’s interesting is the general public does end up hearing about that 5%.

Kyle Harrison: Unless they’re in enterprise software or something. But there are a ton of companies that started with a 90% chance of failing that you’ll never hear about — and then all of a sudden this one did well.

Colby Howard: How many 5% companies have there been in the past ten years, do you think?

Kyle Harrison: That’s a good question. You can also broaden it — Moderna raised venture money back in the day. Apple raised venture money back in the day. There’s a broad swath of companies you might not even think of as venture-backed startups that technically fall into that pocket.

So it’s tough to say, but maybe hundreds of companies. And there have been — with that math — thousands and thousands of investments over the course of years. So it is a tiny fraction of companies that end up becoming the massive producers of outcomes.

Prioritizing, and diligence in practice

Colby Howard: I emailed a thousand people, five got back to me, and you’ve had a first conversation with me. How am I in your mind space now?

Kyle Harrison: It very much depends on the timeliness of the situation. Sometimes the first time we meet you’re actively raising and need to finish within the next month, or else you have issues — because you want to hire this person, or whatever. If it’s very timely, that shoves everything forward, and I’m trying to do a ton of things in the course of two or three weeks.

If it’s something where we’re just going to get to know each other and you’re not raising right now, it might be more casual — but it also might be more in depth. I might spend more time going deep with you on a specific thing. Or if you’re trying to hire somebody, I might help you hire.

That’s interesting, because it not only helps you — which increases my odds of you picking me when you have five people to choose from — but it also helps me get to know you. I’ll help you hire, so I can see who you hire. And if I don’t think very highly of the way you hire, or the people you hire, or your heuristics for making decisions, that informs my decision.

So the more time you have, the more you’re able to build a relationship, and that informs your judgment on the company. Sometimes there are companies you know for years before they’re in your strike zone. You have to radically optimize your day for: who am I going to spend time with that progresses a specific relationship, because I think what they’re doing is really interesting?

Colby Howard: How do you sit down and prioritize that day, given how many people you know and how many companies you could be thinking about?

Kyle Harrison: It’s one of the reasons venture often gets compared to a sales job. If you have a sales pipeline, you have tools to keep track of stuff — when did I last talk to this person, what did we do, what did they say. Most investors have the same thing. Here are the companies I think are going to raise soon, I need to prioritize those. Here are the companies I think most highly of, but they’re not raising soon — I still need to prioritize them, but maybe not as much. Then I have companies I kind of think are interesting.

So I first prioritize the companies, and then for those companies I prioritize what am I going to do for them that is going to move the needle — and also things independent of helping the company that I need to do to deepen my understanding of it.

Colby Howard: For my AI-enabled recruiting company, what are you doing besides company analysis? Thematic or sector analysis — am I riding a wave, what wave, and is this company able to ride it?

Kyle Harrison: The first thing you do is go talk to 20 recruiters — whoever you would sell your thing to. I’m going to say, would this help you? If it did this thing, would that actually be helpful?

Colby Howard: Let’s say it’s a rare earth minerals mining software company and you don’t know anyone in the space. Are you just hitting them up on LinkedIn?

Kyle Harrison: Sometimes there are expert networks — companies that exist that I pay money to, and I’ll say, go find me a rare minerals expert so I can talk to them, and I’ll pay 500 bucks and you take a cut.

Colby Howard: So you’re getting up to speed on the space. And there’s a chance you just say no because you don’t want to get up to speed on that space?

Kyle Harrison: Warren Buffett has this quote about a too hard pile. It doesn’t mean it’s bad — it just means it’s too hard for me. I’m not going to figure that out, I’m not going to touch that particular business.

For a long time venture was 100% focused on software and mobile apps. If you told me you have to buy anything physical, I’m out — it’s too expensive and too scary, maybe you have to raise debt and debt is scary. More VCs are comfortable with that now; we’ve invested in a number of companies that have hardware or manufacturing operations or physical capex. But to each their own — every VC decides what goes in their too hard pile.

What you’re actually underwriting

Colby Howard: So you’ve talked to a lot of recruiters and they say, if Colby could build this, that would be incredible. What risks are you looking at in terms of the company?

Kyle Harrison: Mostly in the early days I’m vetting your ability to build that thing. I would reference-check you, talk to people you’ve worked with before, and ask: is Colby capable of building something like this? If you’re an engineer, it’s suss out your technical capabilities. If you’re not, it’s suss out your ability to go hire somebody technical — because eventually rubber hits the road and somebody has to build the thing.

Colby Howard: It’s a great PowerPoint.

Kyle Harrison: That’s right. So eventually I’m sussing out: can you actually build this, either yourself or by hiring people who can — and also, can this be built? If you say you’re going to use GPT or OpenAI, then okay, that’s going to require fine-tuning for specific recruiting tools. Or: recruiting requires a lot of LinkedIn data, and LinkedIn does not want you to have their data. So can you actually build a model that helps your recruiting tool if you can’t get access to LinkedIn data? How are you going to do that? I’m going to ask you all these technical questions about how you plan to build it.

Colby Howard: You might get a lot of dreamers. I’m going to reshape the homebuilding business. Do you know that wood is involved and that you have to source it?

Kyle Harrison: Right.

Colby Howard: Would you ever take a bet that they’ll figure it out because they’re ahead of a trend in technology that will be ready in three or four years?

Kyle Harrison: Probably not at that level. Because there is such a thing as being right but early, and therefore wrong. You can’t be so far ahead of the curve. Not everybody has the patience of James Cameron to wait ten years to make a sequel because the technology has to catch up. If I’m going to fund you, I’m going to be paying your bills — and if it takes four years before that’s possible, what am I doing? I’m just paying your bills for four years while you wait for technology to catch up.

So I need to be confident you can do this thing. But sometimes there’s an element of: I think we can figure this out. Maybe you don’t have every aspect figured out, but I’m trying to get a sense of whether you’re sharp enough to figure it out and solve the problems.

And then you move from I think they could build this to: people want it, I think they can build it — who else is building it? Who would they have to compete with? That’s both status quo — are there recruiting agencies they have to compete with — and other software providers selling recruiting software that maybe isn’t AI-driven but works really well even though humans are doing it.

So I’m going to look at all those things and ask: is their capability going to be so much better than what’s out there that people would be willing to switch? Because that person can say I would love that — but you also want to talk to the people making the buying decision, and that’s not necessarily the end user. The recruiter might say I’d love this, but the head of recruiting says we don’t have the budget for that, I don’t care how much you’d like it. So you have to understand whether there’s market demand.

Why revenue isn’t the safety people assume

Colby Howard: When you say pre-seed, seed, Series A, Series B — most people imagine a company with no revenue that’s just starting is infinitely more risky and infinitely more likely to be a zero than a company with $10 million of revenue. But companies with $10 million, $100 million, and zero — all of those fail all the time. How would you describe that?

Kyle Harrison: There’s a sliding scale of risk to reward. High risk, your reward goes higher. The odds of you failing when you’re zero revenue and just getting started are super high — but if I invest early… Lance Armstrong made something like a $150,000 investment in Uber and I think it ended up being worth like $40 million. Huge risk, because he invested super early, but a massive outcome. The multiple on that investment is enormous.

Then as you move up a company’s life and it has $10 million of revenue, that company is less likely to go to zero. It could — there are lots of businesses. But actually, once you cross a certain threshold, companies don’t just die easily unless they have fundamentally bad business models.

That’s one of the reasons VCs are scared of debt — debt is one of the things that will kill you. You have to pay interest, and if your revenue is fluctuating and you can’t make interest payments, you’re screwed. But if you’re a company that generates a bit of revenue, and you’re not crushing it, you’re not exploding, but it’s just software — you don’t have that many costs, and you could fire most of the people and just keep living. There are a lot of technology companies that become lifestyle businesses.

Colby Howard: Given you invested in me, five years in I’ve got a million in revenue — you’re not happy.

Kyle Harrison: If you suddenly flipped the switch and said, this thing prints money for me — I’m not happy about it, because I was shooting for a big outcome. But that happens a lot.

Colby Howard: If I have control, then I’m allowed to do it.

Kyle Harrison: If your investors own more than a majority share in your business, they could force you to sell the company so they can get their liquidity.

Colby Howard: I for some reason thought that was just not allowed.

Kyle Harrison: Sometimes people will do dividends over time. If they’re profitable they’ll pay their investors back each year a little bit out of profit to make that investor whole — but it probably takes several years to pay them off, and so their IRR, which measures return over time, goes down. But hopefully they’re getting a multiple that’s good enough.

A maybe good example — and I’ve never invested in or worked with either of these companies — is InVision and Figma. These are companies that build software for prototyping. If I’m a product designer and want to build a product, I can prototype it in these tools and play around — make sure the screen looks like we wanted, the buttons do what they should.

Figma started after InVision. They basically do very similar things — they took a slightly different approach but operate in the same space. InVision got started earlier and grew way faster. By the time Figma was maybe $4 million of revenue, InVision was like $100 million of revenue. And for a lot of people the thought was, well, InVision is the clear winner — why would we ever invest in Figma?

Fast forward six or seven years. InVision — I don’t know where they’re at, it’s still around as a company. But Figma is a dramatically better business. Just dramatically. InVision is smaller than they were from an employee count. There’s a lot that went wrong, to the point where InVision is no longer likely to become the outcome its investors wanted it to be. But they’re not just going to shut down a $50 million revenue business. You do something with it, you figure it out.

Colby Howard: And you would have been thrilled. You would have told people I invested in InVision, and people would respond, oh, that’s going to be super good for you guys. And then someone else comes.

How the decision actually gets made

Colby Howard: If 90% of the population got that email, a lot of people would think, oh my god, this is my opportunity, I could be a billionaire. Is your initial reaction, after being in the seat as long as you have, skepticism or promise?

Kyle Harrison: You would get really depressed in this job if you were immediately skeptical. I’d describe it as analytical optimism. I’m optimistic, I’m curious to hear how it goes — but I’m going to be analytical, I’m going to ask what about this, what about that.

If you’re pitching me on a company that sells software to typewriter repair shops, I’m immediately going to be struggling to see it. So there are instances where it’s difficult to be optimistic. But in general I try to take a very curious approach.

Colby Howard: You’re part of a firm with multiple general partners, multiple salespeople all with their own funnel — but you don’t get to close it on your own, because it’s the firm’s capital. What happens at the very end of the funnel?

Kyle Harrison: Every firm has different decision-making processes.

I have been at firms where it is pretty dictatorial — one person making a call, and everybody else is pitching that person, and that person decides yes or no. A lot of times it’s a billionaire founding partner. That’s that guy’s firm, and we know who we work for, and we’re pitching that guy.

I’ve also worked at firms that are super egalitarian, where every single partner votes on every deal. You can vote one through ten — ten being you love it, you want to drop everything and invest; one being this is garbage and I hate you for bringing it in. And maybe you can’t be a five or a six — they don’t want anybody in the middle. Either you like it or you don’t. Then maybe the average score has to be above a six. You could have a very senior partner who’s made billions of dollars for the firm and its LPs, and that person gets shot down because they couldn’t get the votes.

There’s a broad spectrum. Contrary is independent — each individual general partner has check-writing ability. If I want to do something, I can do it; I don’t need everybody else’s buy-in.

Now, that doesn’t engender me a lot of goodwill with my partners if I’m just doing whatever I want. I respect my partners, which is why I work with them, so basically every time I’m going to take their perspective into account when making a decision. But at our firm, we want everybody to be able to make contrarian investment decisions.

What a 25-year-old gets wrong

Colby Howard: As we wrap up — if I’m 25 years old and I go into venture capital, what am I not thinking about? What are the misconceptions?

Kyle Harrison: I think most people think that venture investors know more than they actually do. That’s the biggest misconception — that capital equals expertise, or knowledge, or power.

Most VCs are pretty stupid. Like, everybody’s pretty stupid. In the grand scheme of things we all make really bad decisions, and no one knows the future. Our psychology is super messed up — we make decisions that go against our own well-being. There’s a myriad of reasons we’re all pretty stupid. VCs are also pretty stupid.

But one of the reasons they’re often dumber than the average person is because they are inhaling their own hype. Founders look at me and say, well, you have capital, I want capital, I don’t have capital — you must know something that I don’t know, because you have capital. So I’m going to be deferential to you. And VCs think: yeah, I do have capital, you’re right, I do know what I’m talking about. And they’ll give really dumb advice, or be very wrong about things.

They’re way more wrong than they are right. You’re actually probably wrong 80% of the time, 90% of the time — if all of your companies are going to zero, you must have been wrong.

There’s this idea that venture investors’ whole job is pattern recognition — the longer you’re in the game, arguably the more data you have to draw more patterns. Oh, I remember when I saw Zuckerberg in his pajamas, and I’m looking for that again and again. The more people you’ve seen, the more you can recognize what good companies look like, how they grow, how they start.

And usually what that pattern recognition actually is, is a very lazy heuristic — shorthand for the same thing. It’s not patterns, it’s looking for duplication. You don’t want somebody who has the characteristics Mark Zuckerberg had at that age — you’re literally looking for a nerdy white dude from Harvard who learned to code when he was 12 and whose parents are rich. You’re literally looking for identically the same thing.

And that gets you into problems like FTX, where you think you’re recognizing patterns but you’re actually fitting molds.

Most people think VCs must know something they don’t know. Most of the time that’s not true. They’re just shooting from the hip and trying to figure stuff out along with you.

Colby Howard: Kyle, this has been awesome. Thank you for stopping by.

Kyle Harrison: Thanks for having me. It’s super fun to talk about.

Colby Howard: As I think of all the things in venture capital — if you’re in the bubble, everything seems obvious, all the nomenclature is at the tip of your tongue, everyone talks VC, everyone talks tech. Outside that bubble there’s this world of, wait, people do what with what money, and fail how much, and still sometimes do well after a decade? You broke it down incredibly well, so I appreciate you doing that.

This has been Kyle Harrison at Contrary walking us through his day-to-day life as a venture capitalist. This has been Colby Howard Wants Your Job — appreciate you watching, we’ll see you next week.

Connections

The show

Colby Howard Wants Your Job · Colby Howard — a show built on finding out what people actually do all day, which is why this is the mechanics episode rather than a thesis episode.

The job, mechanically

Finding, picking, winning — the clearest three-part statement of the VC job description in the corpus. Note that declining is itself an act of picking.
Talent Vortex — the SpaceX example, with the concrete heuristic: six of twenty employees from one place is a signal worth chasing.
Contrary — the ~500-person talent community as the finding mechanism, versus brand-driven inbound or cold outbound.
USV · Lux Capital · Founders Fund — thesis-driven and incubation firms, described as a different strategy rather than a worse one.
Prepared Mind · Arthur Patterson · Accel — the Facebook case (network effects understood against Myspace and Friendster), with the luck caveat attached rather than omitted.
Expert Networks — ~$500 a call to reach an expert in a space where he has no contacts.
Warren Buffett — the too hard pile: not bad, just not mine to figure out.

Fund economics

The LP is the customer, and carry is a margin — the Walmart analogy is the most legible explanation of the 2-and-20 structure anywhere in the corpus.
The returns distribution: ~80% to zero, 10–15% at one-to-two-x, ~5% doing all the work. See Power Law.
Most venture funds don’t beat the S&P 500, and everyone believes they’re the exception — Kyle explicitly includes himself.
The fund cycle as structural cover: deploy over 2–3 years, hold 10, raise funds two and three before fund one has resolved.
Niche strategies as illustrations — a $30M misinformation fund, longevity funds — and why they’re necessarily small.
Moderna · Apple — venture-backed companies people don’t file as venture-backed, which widens the denominator on “how many 5% companies exist.”

Judgement and its failure modes

Pattern Recognition — the sharpest critique in the corpus: what’s called pattern-matching is usually duplication, not pattern. Leads to FTX.
Capital ≠ expertise, and the deference loop that makes VCs worse than average at knowing what they don’t know.
The homogeneity problem: warm intros come from people who look like you, ~80% of VCs from about three schools, and Kyle conceding the point rather than defending it.
Right but early is wrong — the four-years-of-paying-your-bills version of the argument.
Buyer versus end user — enthusiasm from the person without budget authority isn’t demand.

Companies and cases

InVision vs Figma — $100M versus $4M of revenue, and why the obvious winner wasn’t. The best worked example in the corpus of revenue not being the safety it looks like.
Lifestyle businesses, debt, forced sales and dividend paydowns — the ways a company that doesn’t fail can still be a bad outcome.
Uber — the Lance Armstrong angel investment as the extreme case of early-stage multiples.

Elsewhere in the corpus

Investing 101 — he explains here that he writes partly so people get to know him and send him things, which is a more instrumental account than the write-to-think version he gives on The Idea Exchange four months later.
Capital Inferno Heaven — the April 2024 Down Round episode picks up the LP-incentive thread from here and pushes it much further.
Venture Capital — the concept page this episode is effectively a primer for.