Hello, fellow data nerds. My guest today is Michael Siebel. Michael is a group partner at Y Combinator, a managing director of YC Startup Accelerator. He also co-founded Twitch and currently sits on the boards of Reddit and Dropbox. Michael, welcome to World of DAS. I'm excited.
Thanks for having me.
Auren Hoffman (00:29.238)
Now, probably you've seen more data than anybody in what makes a company successful or unsuccessful? Like, can you give us a little sense of, a little taste of that?
One of the things we say to YC founders is that we see a lot more failure than success. So we really try to figure out how to help people not fail and then they figure out how to succeed. Some of the things that cause failure, let's see, I would say first, lack of a co-founder.
Startups are ridiculously stressful. I think people underestimate the amount of emotional stress there is, and just having another person there to share the load. We often joke that when single founders come through, I see, especially first-time founders, end up adopting their group partner as their kind of co-founder, which is not good for them and certainly not good for the group partner. So that's a big one. I think the other big one.
I don't think most YC founders who are successful don't have their unique insight pre-launch and they don't even have the unique insight during YC. And so one of the things we warn people about a lot is your job isn't to just build and scale your thing in your head. Most of the learning hasn't happened yet. And I think that that's really antithetical because, or at least counterintuitive because
You're pitching investors. And so you feel like you should know the right answer. Where when we go through all the YC examples of successful companies, almost none of them knew the right answer at the YC stage.
Auren Hoffman (02:11.326)
Is there a certain commonality in founders or something? But you know, probably all of them have like a decent idea. They're NYC to begin with. So they have a decent idea. They're smart people. They're probably all of them are smart people that you're involved with. Um, is there, uh, is there, is there some, but some of them are able to like find this great idea through, you know, tons of interactions over time. If you think of the Justin TV story where you were like, that was a, that was a huge evolvement that happened over time to Twitch.
One would hope.
Auren Hoffman (02:41.658)
Um, some, some of which maybe aren't able to find it. Maybe it's not their fault. Maybe they're just in a market that's tough, but have you, how, have, is there some sort of type personality type that you think is more likely to do it or some sort of type of founder where nowadays you're like, you have a higher, um, a rate of success.
So I would say this specific question is hard to tell when we're reading applications or interviewing. When we're working with a company, I would say there are two things. One is simple kind of logical thinking. I find that like the founders who can kind of...
communicate logically and not draw conclusions that can be easily disproven with basic logic really helps. Which you would think that's a skill that comes with being smart, but I would argue that it isn't really. It's a separate skill.
Auren Hoffman (03:41.962)
And what, because there's this, there's this other side of founders that people talk about this reality distortion field that you think would be intention with that.
Yeah, yes. And so I would argue, when I work with founders with the reality disversion field, it is much more likely to not go well. Yeah, much, much more likely. And then the second thing I would say is...
Auren Hoffman (03:58.421)
And this is maybe just another way of saying the same thing, intellectual honesty. Right? Like people who, when they say they're going to do something, they do it. Or like when they don't do it, they just say, we didn't do it. Um, I think that doing a startup is so stressful that, you know, some teams, people lie to themselves, right? You know, you don't want to admit that you didn't make those phone calls or you didn't send those emails or you made a
a spec list that was way too long, or you didn't write down your spec for the week, so you didn't really build the thing you wanted to. Like those things are embarrassing to admit. And founders who just kind of get over that and are just super honest with themselves, they learn faster. And so I would say that both of those, kind of the extreme honesty and the lack of the reality distortion field is a very good sign that we admitted the right person.
Auren Hoffman (04:57.134)
And those are probably traits that have been true forever. Are there traits that are, you think, more true today than were true a decade ago?
I think that there is a challenge that is more true today, which is that when we were doing YC in 2007, we felt lucky to be doing a startup at all. And so...
We didn't, I think young people nowadays feel like getting into the startup game is easy and so they only grade themselves on like how much progress they're making. And if that progress takes time, it's very discouraging. Whereas I would argue that for almost every YC company, it took longer than they thought. It was harder than they thought. The insights happened later than they thought. And so sometimes I see kind of young people today
a payoff that happens in the course of like a semester. You know, it's like, you know, like, like everything's related to a class. Oh, I started getting it three months in. And you know, we're like, it might be okay if you don't get it a year in. And a lot of young people are like, I, that doesn't, I can't do that. How would I stay motivated? And so I think that's a big challenge today versus back in the day. I would say that there's a big similarity.
Auren Hoffman (05:58.497)
Auren Hoffman (06:11.295)
Folks were excited about Rails and building web apps back, this sounds like the dark ages, right? But in the mid to late 2000s, yeah. The actual technology was exciting for people and AI reminds me of that today. We're like, there are a lot of people who are doing it because we're doing startups because it's their opportunity to play around with the most interesting technology. I'd argue we went through a phase, consumer fintech, crypto.
Auren Hoffman (06:28.194)
I remember that 2006, 2007. Yeah, yeah, yeah.
where a lot of the people who were getting into it were getting into make money, not to have fun, not to play around interesting technology. And that certainly was a disadvantage.
Auren Hoffman (07:00.834)
There's this class of people and I assume many of them or most of the people in YC are in this kind of world where they've been overachievers all their life. And many of them have not, because they're young, they may have not had an opportunity to experience much failure or much disappointment or many setbacks. Maybe some very small things like they didn't get the Rhodes Scholarship, but they got this other well-known scholarship or something, right?
It's a rough life.
Auren Hoffman (07:30.718)
Or, you know, they, they got into Princeton and not, you know, Stanford or something or whatever. Um, but, um, what, um, uh, um, w how do you deal with those types of people as they're kind of like running into their first few roadblocks? Um, and is there some sort of predictor? Some of them probably do extremely well with that. And some of them maybe get paralyzed by those situations.
One of the advantages that YC has is by funding people in a batch, you get to see all such a wide variation in outcomes during those months that I think it helps you handle the disappointment. The second thing that I think is really helpful with YC is the number of stories we tell of successful companies. I've always been shocked at how little history founders know.
But then I think back to myself at 23 and I was like, I didn't know shit, I wasn't researching anything, right? Of course not, right? But when you tell people that Coinbase didn't allow people to buy Bitcoins when it launched, they're like, no way. They're like, yeah, it did, they didn't figure that out. Or Airbnb didn't collect payments when it launched. They're like, oh shit, that seems pretty obvious. And so I think a lot of the times, seeing people around you fail and recover.
Auren Hoffman (08:38.504)
Auren Hoffman (08:45.634)
is the easiest way to feel like, okay, this isn't our game yet. And to see smart people you respect around you, not doing well, and them stay motivated. Um, I'd say the other thing is that doing a startup's lonely. I, you know, I remember talking to one founder who didn't do YC and he's very successful. Um, but it became clear he didn't have a lot of startup friends because his relationship with his investors was, you know, a one-to-one relationship.
And it was fraught with, I don't want to say the wrong thing, right? Like these people can really screw me over. And I realized for YC founder, there's just a whole set of people they naturally get to hang out with who can't judge them. There's no, you know, I thought they're bored. They're not going to write checks and they can just become friends with them. And they can handle that stress as a group versus as an individual. And this is especially good when, you know, founders are having a dispute.
Auren Hoffman (09:48.151)
So like you can't go to your co-founder, but like y'all have friends within the batch. So I think that's kind of a really big help for YC, but even with that, your central point remains, right? If you get punched in the face and you can't get back up, Startup World's not for you.
Auren Hoffman (10:03.946)
Yep. Now, if you were gonna, if you were starting a YC for 13 year olds, where you can get a percentage of their lifelong earnings, um, and you know, you're competing with all the top investors in the world, all the top hedge funds, 0.72 is now in the market and all these other things. And so you can't just like invest with people with high SAT scores. How do you select people or, or is there something like people don't like that you would find like, Oh, that's maybe not so bad. Like, I don't know.
Auren Hoffman (10:33.454)
They got in trouble with the law or something.
Well, I'll say this, I wouldn't answer this question directly. I would never start a YC for 13 year olds. This is what I will say though. I remember a story from one of our university LPs where they said, I bet you think you monetize our students pretty well. And I said, oh yeah, like, you know, a lot of your students come through YC, I guess, you know, we do a good job. And they said,
You get 7% of the company that they do that does YC. We get 10% of everything that they make when they die.
Auren Hoffman (11:13.806)
And I was like, oh my God, like I'd never heard the university business model like summed up so succinctly. It's like, oh Jesus, if you can just wait, you know, 60 years. So I think the universities are probably doing the best job at this kind of program.
Auren Hoffman (11:23.69)
Yeah, that's a great business model.
Auren Hoffman (11:39.766)
Yeah, that's probably true. Um, now you, one, one thing is if you think of YCE as a, there's, you can think of YCE itself as a venture capital firm, but you also see YCE itself as like a business and where I feel like it's different in venture capital firms is venture capital firms are always advocating for their companies to create differentiated products. Um, but they don't necessarily take those ideas themselves, that advice themselves, whereas like
Auren Hoffman (12:06.242)
YC seems to really be the core exception. Like how do you think about that? The actual venture capital business and YC is placing it.
I used to think that venture capitalists were misguided for not innovating. And being in the innovation industry and not innovating, I used to blame them and say, what are you guys doing? That was a lack of understanding. YC's advantage was it was self-funded to start. And the founders didn't need to make their first buck.
they'd already sold the company. And by the time we were seriously pitching LPs to fund YC, we already had a successful model. Most funds are built in the exact opposite way. Before they have a successful track record, they have to pitch LPs. And that process kind of forces them into established ways of doing business that are the only ways that they'll get funded.
And so I realized like, oh, wow, the system of funding and the fact that like, if you can't sell fund your firm, you really have to do it like everyone else in order to get money creates this lack of innovation. And it's not surprising that some of the best firms like Sequoia or A16 get a little bit more leeway and innovating. But it's just a little. And I would say YC is...
extremely different. I would argue that YC is the combination of a university, an investment fund and a software business. That just doesn't seem to exist out there otherwise.
Auren Hoffman (13:48.386)
To push back a bit though, a particular venture capital firm, while they might not self-fund the fund, they usually own 100% of the GP and the LLC that kind of owns the fund. And so it's almost the opposite of a particular founder where a founder has to take all this dilution. Usually the fund is not taking dilution in the core fund.
in the business, in the actual business. They're only taking business in each individual fund.
Yeah, but to get any investment capital, they have to present a pitch that LPs are interested in. And if that pitch is too innovative, like, or if it conflicts with LP strategy, they won't get money.
Auren Hoffman (14:26.858)
Yeah, that's true. Yeah.
Auren Hoffman (14:35.854)
Now there are a lot of accelerators, but like when I think of like the accelerator university analogy, like if a student goes to Yale or Harvard, Princeton, Stanford, MIT, like, you know, they might choose any of those options and they're all great options. And if they get into one, it doesn't mean they're not going to choose the other. But for accelerators, like it seems like YC is by far the most prestigious. I assume it'd be very rare for someone who got into YC to choose a different accelerator. Like why, why is that the case where?
it's so dominant in accelerators, whereas like you can have many great universities or VCs, for example, there's many great, you could get benchmark and Sequoia and choose one versus the other for many different reasons.
So I'd argue the university analogy might hold, but you might want to look at the population of people who are interested. So there's a lot of people who are interested in going to university, so there's a lot of potentially good options. If you look at business school, most people would say there are way fewer good business schools than there are fewer, right? Because fewer people are interested. And I would argue that whereas in our world, startup seems huge, in the real world, it's the tiniest of tiny, tiny group of people who are interested.
Auren Hoffman (15:33.27)
That's true. Yep.
And so it just can't support as many institutions. And so I think that's, you know, the unfortunate reality is that you get a distribution based on how many people are actually interested. And it's pretty rare to wanna go through the pain of being a startup founder. And the long payoff cycle, right?
Auren Hoffman (16:05.206)
Now the scale of venture capital has changed like massively over the last decade. Um, is being able to scale up and expand a prerequisite for firm survival or, or how, or do you see these like more niche? How do you, how do you see that kind of walking out?
I don't know. It's very, very hard for me to think about how a typical VC fund should be constructed because I've never really worked in one. And it's so different than YC. I feel like I'd be giving someone advice without having enough context. I do think that one thing that I'll comment on is that in the zero interest rate environment,
I think that everyone expanding their growth funds and writing huge checks into early stage companies didn't help the companies. I think that we were at fault just the same as everyone else with our growth fund. I think that in hindsight, it's somewhat embarrassing what was done. If you were just to say facts out loud, giving a pre-product market fit company $25 to $50
Auren Hoffman (16:59.03)
They basically got too much money.
ain't going to fucking help them. Giving a founder enough secondary that they can live the rest of their lives and never have to work is going to have an impact on their execution, especially if the company is going to struggle in the future. I think that whole thing was just an embarrassing example of investors, ourselves included, thinking that
Auren Hoffman (17:17.42)
You could model a business in Excel spreadsheet and cash is a primary input. I've never seen cash being a primary input in a successful company. It's definitely an input, but it's not the number one most important.
Auren Hoffman (17:55.434)
Yep. There's certain and there were certain businesses that maybe require a bunch of like, if you didn't raise the cash, you will die. Like a, some of these ride sharing businesses or food delivery business or something like that where they're extremely, yeah, rockets obviously, or, you know, or something like that. But, but, but probably for many others, like, um, you know, the winner was not, not the one who raised the most cash.
Yeah, or rockets.
Well, and I would argue even in those businesses where a certain amount of money was a required catalyst, I still don't think that money was the primary mover. For every one of those industries, there were a ton of companies funded and only a couple of winners, and a lot of companies got money. So I think that unfortunately, it was crazy to watch it happen because all founders know this, right? Like all founders know,
Auren Hoffman (18:35.959)
If you raise your series B six months after your series A, you haven't done shit in the interim, right? Like, we all know that, right? But, you know, on the other hand, if you're advising a founder and like money is available at three X evaluations, 60 days after your series A, you're gonna tell them not to take it, right? It's tough, but like I might argue that for, on the order of 80% of founders, that probably was the right decision in the end.
Auren Hoffman (18:53.314)
Auren Hoffman (19:07.53)
Yeah, yeah, it's very tough.
There are so many founders now with a ton of cash and no idea what to build.
Auren Hoffman (19:24.994)
Well, the other thing is, at least right now, is because it's harder to raise cash, you will have less competition, which also means you have less worry about not raising cash.
Well, and I think that's interesting, because I definitely know a lot of companies who have successfully cut their burn. So they have tons of runway. To your point, they're not gonna have a lot of competition. But I would still argue that by far the hardest thing is to figure out what customers want and to build it for them, right? And it's like, you can get all those other things right. You can have all the cash and the good hiring and da da. Yes. So sometimes I think it's like, I always think about this like.
Auren Hoffman (19:51.278)
Auren Hoffman (19:56.254)
I've struggled with that of everything I've ever done. Yeah.
I say to founders, what item on your to-do list makes your stomach hurt when you look at it? You should probably be working on that one. And I think a lot of founders kind of go to, and myself included, it's easy to go to item two, three, four, five, six, seven, eight on the list. And ignore number one.
Auren Hoffman (20:11.213)
Auren Hoffman (20:19.79)
Well, because we want to check it off. We're all in that kind of like we got to check off the list. Inbox zero type of thing, right?
Yes, exactly. But like if item one is like by far the most important item and you never work on it, we're about to see a lot of that.
Auren Hoffman (20:36.234)
Now, if you were, if you were like going to create a competitor to YC, like, is it like some sort of like really niche thing? Like we're going to do something like an accelerator and like this crazy niche area or like, or do you think it's even possible to do it? Like, how would you, you know, what are you worried about as someone who's in YC?
I would say startup accelerators are generally not good businesses. I would say that YC started in 2005, but if anyone bothers to go out there and do the math for a second, a lot of the big IPOs literally happened 15 years later. And so I'm always confused when people want to compete in the accelerator model, especially if they're not already rich.
Auren Hoffman (21:18.454)
Because it's a pretty hard way to make money, I would say. So I think that if I were interested in doing something else in the startup world, it would not be starting Accelerator. No, no.
Auren Hoffman (21:33.483)
Auren Hoffman (21:41.847)
What do you think about these incubators? We had Jack Abraham from Atomic on World of Deaths. How do you think about these incubators? There's not that many of them that have been successful, but what do you think about the incubator model?
I think that, um...
I will say this, I think that startup investing is very different depending on what scale you're doing it at. I've certainly seen people do it at very small scale with their own money and work with a very small number of companies and see some amount of success doing it. I think the tricky thing is it still takes forever to pay out and that model breaks at scale.
Auren Hoffman (22:24.45)
And usually the people who are rich want to make something at scale. So it's this constant kind of battle. Um, but I think that.
Auren Hoffman (22:27.764)
Auren Hoffman (22:31.222)
Like if you did, you know, one every two years, it's one thing. Or, but if you start doing one every two weeks, it's a whole other thing. Yeah.
It's a different thing. Yeah. And so I definitely see people, you know, I think the people who are doing incubators as kind of like their lifestyle fun thing, you know, I think that's probably the right idea. I just don't think that model scales, unfortunately.
Auren Hoffman (22:52.958)
Yeah. Okay. Interesting. And one of the things I admire about YC is that, you know, cause the hard thing in all of these financial firms is, is the CEO handoff. And, you know, YC still relatively, I'm coming out, I at least have had like four CEO handoffs, I think, right at this point, which is four more than most venture capital firms have ever had. Um, like, is that just like baked into like the founding? Like how did that happen?
That's a really good question. I think YC has a hidden advantage that people don't realize, which is the people who are partners at YC, for the most part, were YC founders. And so there is a huge part about working at YC that's about giving back to the institution that kind of gave to you.
Auren Hoffman (23:39.852)
as opposed to kind of extracting from it. And I think kind of in a normal corporate environment, it would be nice if people felt like they were trying to give more than they would take, but I think the reality is people are trying to take more than they give. And so that makes a lot of these issues a lot harder to deal with. But if you're a group partner at YC, you've already sold a company, you're doing it to give back, I just think that's like a way more
Auren Hoffman (24:05.855)
that team has way more foundational strength. And I think that's really, really important. I think the other thing that's important is...
we think about YC way more long-term. The reason why I was interested in working at YC was because I think it can thrive for a hundred plus years. And I've already seen, I've already been shocked to see that the lifespan of the companies that all my friends started and the ones that we're funding now is fairly short in the end of it. And so the opportunity to work for an institution that can be...
thriving 100 years from now is somewhat unique, and so you wanna treat it with respect.
Auren Hoffman (25:07.126)
It is interesting that like, I think most people, most people who are familiar with Y Combinator cannot name the CEO of Y Combinator. And I think that's a feature, not a bug.
Well, and I think this is a way where we're a little bit more like a university. Like, I don't know the current, you know, I went to Yale. I don't know who the guy, maybe I know who the current president is, yeah, like, but it's not, it's not even something you think about as the primary thing, right?
Auren Hoffman (25:27.822)
Provost or president or whatever Yes
Auren Hoffman (25:35.566)
Correct. Yep. Okay. Yeah. Good boy. I have no idea who the pet of my university is currently. Yeah. Okay. That makes a lot of sense. Now, you're a big proponent of the MVP concept that founders should focus on getting out some sort of minimal viable product as soon as possible and then kind of iterating. Where do you draw the line between like bad rust product and MVP or how do you kind of think about those?
I think this whole debate is missing the point. I think that.
When you think about bringing in a new employee in a company, oftentimes you want to deliver some type of emotional win to them pretty quickly. So you put them on the bug fixing team or you give them a small task, right? Yeah, exactly. And the reason why is because you want to get them into a pattern of launching and then learning. And because you assume that they're true kind of
Auren Hoffman (26:30.998)
customer success or something, yeah, service.
superpower moment in the company is not going to be in the first three weeks or even three months. I would argue that when we take a company through YC, it's very similar. We know that major insight they're going to have is probably not going to happen for a year. And it'll happen slower if they spend that entire year trying to build the perfect thing in their head. Don't talk to users, don't talk to customers, don't launch. It'll happen faster if they...
get stuff out there and see some of the assumptions they have in their head be proven wrong and start learning. Now, I think that most successful founders hear me say this and they think of their story when they look backwards as linear. Right? Like, oh, it was just A, B, C, D, E, F, G, right? And so obviously doing A poorly would screw up B, right?
I just think that you have to change your perspective. When you're sitting at A and you don't actually know what B is, it's just a different, it's not a linear progression. And so I just think this debate's kind of, it's silly because on the other hand, I would say it really also depends on the founder and what they're building. When Parker is building Rippling, okay, right, well, a third time founder who's rebuilding a business that they've already built once.
Auren Hoffman (28:13.023)
should probably approach this differently than a 24 year old who's building a business for the first time. So, I think that I run into this problem a lot at YC where a founder will want every piece of advice to apply to them and they'll want that piece of advice to apply at every stage of their company. And I'm like, wouldn't it be nice if advice were like laws of physics? But then I'm saying physics doesn't even work that way.
Auren Hoffman (28:43.257)
Yeah, like that's not how the world works. Like so sorry that like everything we say doesn't apply to your situation. And the thing we try to do this year won't work three years from now. Like, sorry.
Auren Hoffman (28:55.214)
It's also a, it may also be like, if you're a company that has zero product, well, then like maybe launching quickly and getting users and learning, you know, but let's say you're already like even somewhat established, we had Amjad Massad, who's the CEO of Replit on this podcast, you know, they already have like a great product. They already have tons of users. And so, you know, maybe the next product that they launched, maybe does have to have, has to be a little bit shinier. The X can't be as half, it has to be.
has to be more put together because the users have a certain expectation of what a replica product would look like today. Whereas like when they lost, it maybe didn't have to have that.
I'm going to go to bed.
Certainly, I would argue the rules for Amjad are going to be different than the rules for a founder who's three months in. Got it. Of course. Yeah. Of course. Yeah. So I think that, I don't know, it's interesting. I actually think there's something else going on with this debate that's a little bit deeper. And I think the deeper thing isn't like, oh, your MVP has to be better.
Auren Hoffman (29:40.906)
Yeah, yeah, yeah. And then, of course, like Microsoft may have even the next or whatever another bar on there.
I think it's like market forces coming at work. Pre this AI game, we hadn't really had a major platform shift since mobile. And so I think what a lot of young founders don't realize is that many ideas have been tried. And so I think you get this kind of thing that happens, where it's like they don't realize that this has been tried three. Yelp 2.0 has been tried 80 times.
Auren Hoffman (30:28.804)
they release Yelp 2.0, no one uses it, and people are like, well, your MVP sucks. And it's kind of, is it your MVP sucks, or is it that the market doesn't want Yelp 2.0? You know, and like, you're just running the 81st like confirmation experiment on that. Or with the current tools that are available, you can't build a Yelp 2.0 that's sufficiently better. And you're the 81st company in the experiment. And so I think that's actually what's really going on, is that behind the scenes,
Auren Hoffman (30:35.562)
It's not like you can just build an Uber disruptor every like two years, or a DoorDash disruptor every two years. Like that's really not how the game works. You need some environment or technology change to open the door to disrupt an incumbent. And if one doesn't exist, I don't care how good your MVP is, ain't gonna matter.
Auren Hoffman (31:19.39)
Is, is there some sort of sense of like, you know, maybe some sort of a bastardization of the classic Peter teal question that like, where the founder sees the future that most smart people doesn't see, um, everyone sees like, Oh yeah, it'd be great if like we could build a better salesforce.com. Like the UI is terrible and you know, it'd be awesome if we could, you know, but is it something where they, they can see something and then they, and still most smart people completely disagree with them.
But somehow they can still see it and they can then, or is that not as important? It's just kind of executing on something that like most people can see.
I think I don't have a good answer for you because I feel like I have success cases for both. And I think that...
Not only do I have success cases for both, I actually think that the solution part isn't the part that people have trouble with. The more I work with founders, the more I realize that it is really hard to understand what problem a customer has and what their interest is in solving it. And time and time again, I'll meet a company who says like, oh, well we can save this enterprise a million dollars a year.
and they haven't really come to the realization that the enterprise might not give a shit about saving a million dollars a year. Or they say, oh, we can build this tool that's gonna replace 50% of the people on your data team. And they might not come to the realization that the head of data doesn't want to fire 50% of the people on their data team. So I often feel as though the misunderstanding happens in the problem space, everyone we fund can build software. And so like,
Auren Hoffman (33:00.171)
Building good solutions is way more, that skill is way more distributed within the YC population. Understanding customer problems, that's actually really hard. Most people don't even want to engage with that intellectually. In the story of Twitch, it took us five years to understand that a subset of our users, gamers, wanted to make money. Nowadays, that sounds like...
the dumbest thing in the world because every social platform supports influencers and pays them out. Back then influencers didn't exist. Like it wasn't a thing. And so I remember, you know, Emmett tells a story all the time where he was talking to streamers and he's like, you want us to give you a cut of your ad revenue? Like that's going to be like $6 a month. And they're like, yes, I want that. I want $6 a month. And you know, eventually we realized, oh shit, these people want to.
Auren Hoffman (33:38.781)
quit their jobs and do this full time. And like, once we knew that, we could build the thing, like building the things wasn't hard. It was figuring.
Auren Hoffman (34:07.135)
Auren Hoffman (34:16.65)
It was just being in enough repeated interactions to even come up with that. And being open to it when you hear it, right? Being in the right state.
Yes. Yes, and being open to think like, oh, maybe we can actually do that. Yeah, no. And I think what's so tricky is that for most of the history of Justin TV, we didn't talk to our users and we didn't.
We didn't understand that knowing their motivations and what they cared about, like really caring for them was an essential part of our job. You know, we thought building features was the most important part of our job. And, oh, you know, well, that requires people, well, that requires money, da da, like, you know, that whole, you know, building a bridge to nowhere kind of thing was what we were doing for the longest time.
Auren Hoffman (35:04.91)
There seems to be like, if you think of like the companies that pivot, there seems to be kind of three types of companies. There's the companies that are like doing well, and then they don't need to pivot. They should just like focus more on what's going well. And then there's the companies who are like doing, like, it's just not going well. Like don't have private market fit at all. And then, okay, well, that's clear that we should like do something radical because we don't have product market fit. But you have all these companies that are kind of in the middle. They've got like something like, okay.
I'm gonna go.
Auren Hoffman (35:33.486)
It's not growing super fast. It's, it's, it's there. It's a business. Um, if you, if you start going on this new thing, that means you, you have the potential to kill this okay business. Cause you're not, you're not focused on it. Like, I imagine you have a lot of companies in that tweener stage. Like, how do you think about that?
Well, this reminds me of what we tell founders when they're fundraising. They always say, well, we have these people who said yes, these people who said no, and these are maybes. And we're like, the maybes are nos. Move forward. I would argue the tweener businesses are more often than not nos. And especially coming out of the last environment where there was so much cash around, I've started to see this phenomenon where, I would argue that like most smart,
Auren Hoffman (36:05.236)
Auren Hoffman (36:16.503)
founders that apply themselves can get a business to a million dollars in ARR. I'd argue if you give those people 25 to $50 million, they can probably get a not very good business to 10 million in ARR. That doesn't mean it's a good business. That doesn't mean it's profitable. That doesn't mean it could ever become a business generates a hundred million or a billion in revenue. It just means that like, if you spend a dollar enough, you can figure out how to extract 20 cents from it. And
I would argue that those businesses are not going to get product market fit and they should not be furthered. They're in the category of businesses that should be at the minimum milked. We could milk Justin TV because it was profitable. We could milk the revenue from Justin TV to build Twitch. But if that business that's not working is also burning capital, then we're going to
Auren Hoffman (37:05.484)
Auren Hoffman (37:12.726)
That is a problem.
It's a big problem, yeah. And people don't say that enough. I think people are afraid. You know, I've had founders who are like, well, I know this thing isn't gonna work, but if I don't grow 10% month over month, we're screwed. I'm like, you've just given me two scenarios where you're screwed. Like, you probably need to stop growing this thing 10% month a month. If you're absolutely convinced it's never gonna work.
Auren Hoffman (37:29.425)
Auren Hoffman (37:39.102)
And in the Twitch example, you had something like Justin TV, which was like, okay, it was there, but you knew it wasn't going to change a massive outcome and stuff like that. And so how did you do? Did you just be like, okay, we're going to put two guys in maintenance mode over here and the rest of us are going to go like, how did you think about it?
Yeah, well, I'll tell you how we did it. I think that we couldn't actually think clearly about Justin TV until we got a profitable. And we didn't get a profitable until I pitched 50 investors and they all said, fuck you. So, and until we had two months of runway. So it was like, pitch a bunch of investors, they say, screw you, two months of runway. All right, let's make this thing profitable. Yeah, let's make this actually, like we don't wanna die.
Auren Hoffman (38:19.052)
Auren Hoffman (38:22.91)
Let's get some ads on there or something.
If we could just bottle up the motivation that founders have right when they're close to death and just give it to them all the time, game over. And I think that once we got into that clear air of profitability, so that would have been the end of, oh God, my dates are all screwed up now. Maybe end of 2010? Is that right? I don't know. Anyways, once we got there, then this great guy by the name of Gideon Yu came by our office. He was a partner at Coastal at the time.
had no interest in investing in us. Zero, he thought our business sucked. It was absolutely correct. But for some reason, dropped this dime on us where he basically said, look, if you keep on doing what you're doing within the next three years, your revenue is gonna slip, you're gonna go unprofitable, no one's gonna fund you and you're gonna die. And no one's gonna remember what you've done.
That's it. That was all the advice. And because, yes. Well, and I'd argue that because we were profitable, we...
Auren Hoffman (39:23.782)
You're like, shit, better do something now.
We could hear that and do something with it. And what we did was we split the company into three groups. One group was maintenance for Justin TV, keeping the revenue flowing. One group started working on Justin TV gaming, which became Twitch. And one group started working on Social Cam. We set a goal where we wanted to have this much traction in one of the two businesses within six months of launching it, or we would kill the one that wasn't having traction.
The story we like to tell is that both were doing well. And so we spun off one. The reality is that Twitch was doing way better than Social Cam, but investors really liked Social Cam. And we really liked the idea of Social Cam, right? Instagram for video just seems like a home run idea. And so why give it up? And so we actually ended up pursuing both. And I would argue that it was somewhat of a home run idea. Snapchat was made right at that time.
Auren Hoffman (40:17.022)
And had we had better insights, we could have competed with them more effectively, but we did not.
Auren Hoffman (40:35.602)
Interesting. Yeah, in 2010, a similar scenario where this company, LARAP, wasn't doing that great, got it to profitability, but we had $16 million in the bank. So we did this pivot to LiveRamp. And I guess the pivot was easier because we had a relatively functioning base thing. We had $16 million in the bank. We had a good team of 30 people, mostly engineers. So it just made it easier to do a pivot than kind of a classic pivot would be to do.
Yeah. But we still got lucky. I mean, that shit was, didn't have to go that way. Ha ha ha.
Auren Hoffman (41:06.646)
Um, yeah, exactly. Now, one thing, one thing I think interesting about YC is that, is that YC generally advocates equal splits for founders in, in the batch companies. So if you have three founders, they would get like the same shares each. Right. Why, why do you do that?
So I would say this goes back to the teams that we are historically seeing. So we're historically seeing a three-founder team, all first-time founders, all kind of mid-20s to late-20s. And probably last but not least, none of them true experts in the field they're going after, right? Really smart, technically talented, but not deep industry expertise.
Auren Hoffman (41:54.754)
So they're all like roughly at the same level. One might be called CEO, but they're all roughly at the same level.
Exactly. And I think that because those are the scenarios that we see the most, we kind of give advice for those scenarios. I don't think equal equity splits work in all situations. One, if it's a two founder set up, I often recommend some slight unequal so there's some kind of like, one of the two founders on the board or one person has 1% more just to deal with if there's co-founder dispute. And then certainly there's different
you know, founders of various difference levels of experience and da da. But I would say that we like that advice for the teams that we mostly are seeing.
Auren Hoffman (42:35.806)
Yeah. Now let's say Michael Siebel starts a company that you decide, I'm quitting YCM, starting a company and you start it with an up and comer, 20, 24 year old, the next Michael Siebel who you meet and you're like, oh, obviously there would be some sort of equity imbalance because you have all this experience. You have, you've, you've started these super successful companies and stuff like that. Have you seen anything like that?
So I don't spend a lot of time thinking about those scenarios. I think that if I were to do that, the equity balance would come from me investing in the company versus...
Auren Hoffman (43:11.826)
Okay, got it. So it's like, okay, we have even equity and then I'll put money in at a five pre or whatever. And then okay, yeah, okay.
Exactly, yeah. But I will say this, right? Being a non-technical founder and working with two amazing technical founders with Emmett and Kyle.
Justin TV used to break every Saturday and Sunday morning. And we always used to we all would get SMS is we'd all get the page duty SMS is like, you know, video systems down websites down. And like, I can't fix it. I can't like the only thing I can do is make sure that Emmett and Kyle are awake, getting the same. Yeah. And so I saw that so often where like,
Auren Hoffman (43:42.892)
Auren Hoffman (43:47.366)
Nothing you could do. Yeah.
Auren Hoffman (43:52.958)
to get their coffee and their donuts, yeah.
I'm the CEO, sure, but if my partners are not incredibly motivated, then this shit ain't going to work. And giving people equity, I think everyone, I think a lot of the people who do a lot of unequal equity splits are business people like me, business guy. And they're like, oh, well, they only negotiated for this much. Or like, oh, well, it was my idea or some shit like that.
Auren Hoffman (44:26.475)
And I'm always just like, who is waking up in the middle of the night to fix this? The equity is gonna wake them up or you're gonna wake them up. Like the equity does a way better job of waking them up. And it's a mistake that people don't realize early. You know, I always say, you know, a level one CEO is thinking about, you know, the next year. Oh, well, they're gonna be perfectly happy with this equity split over the next year.
Auren Hoffman (44:42.041)
It's a good boy.
A level 2 CEO is thinking, well, this might take 5 to 10 years. How are they going to feel 5 years in? I should be making sure they're going to feel good about 5 years in, even if they don't think that way right now. It's my job to think that far out. So that's why I always like giving people more equity.
Auren Hoffman (45:17.622)
Now, a lot of what we're talking about are software related companies, companies with bits, which have really been what, so the Valley's been very good at funding. I've heard you say that it's not so good at funding things that are more atoms-based or bio-based or et cetera. Like what's holding us back and how do we change that?
Of course, only software. Yeah.
I don't know that something's holding us back. I think that they're harder. I think.
Not only are they harder when you look at the examples of the people who have won. Let's talk about three different types of businesses. Let's talk about lending businesses. Let's talk about hard tech businesses. Let's talk about bio businesses. I'd argue that in lending businesses, it's interesting that a firm started by someone who's incredibly wealthy who can...
go and raise debt based on their reputation and their personal balance sheet would have a massive advantage, right? Very interesting, right? If you think about hard tech, it's interesting that Elon could go raise a bunch of money to build rockets, right? Like based on previous success. Very helpful. When you look at bio, the entire bio VC ecosystem for the most part, like invests a ton of money in early stage companies.
Auren Hoffman (46:24.022)
buys way more ownership than the software world, often replaces the management team. Yeah, right? It's like, so it's like, I think that, you know, this software model is different, right? You know, there aren't as many Mark Zuckerbergs in those other three worlds. And so, you know,
Auren Hoffman (46:48.63)
Yep. Kicks out the founders. Yeah, yeah.
Auren Hoffman (47:04.833)
Auren Hoffman (47:08.47)
Now there are these like mini YC, if you think of Bob Nelson, he's kind of a famous biotech investor. He's kind of like a mini YC in a way, or mini atomic, or kind of like starting a bunch of these things. And of course, we have accelerators there, like Moderna came out of accelerator or an incubator. So there are these like other structural types, but the capital needs are very different for those companies.
Well, and the skill sets of the founders are often, you know, way more specialized. Like I would just argue it's harder if you're doing a bio business, a hard tech business or lending business, it's harder.
Auren Hoffman (47:39.681)
Auren Hoffman (47:48.522)
Okay, interesting. Um, now how do you think of this like con insider outsiders? Like I think a lot of the well-known founders have kind of historically been outsiders, um, and, um, but the, but why and why, but, and it may be that in the early days of YC, anyone who would join YC was probably somewhat of an outsider. Um, but today, like getting into YC is kind of like getting into McKinsey or Goldman Sachs or Google or something. It's like, it's a.
Auren Hoffman (48:18.262)
There there's a, well, there's a badge on your resume that everyone is familiar with. Right. So if you say, Hey, even if your company didn't go well, if you say, Hey, I got funded by YC, that's a, that's a, that's kind of like, you know, it's like going to Stanford business school or something or whatever, it's kind of thing that maybe an insider would want to do as well. How do you differentiate or do you differentiate between these like outsider and insiders?
Sure. There's a credential.
I would say this, I would say YC is often perceived as a better deal for outsiders. And I would say that YC was built for outsiders. And I think the core thesis of YC is that there are more outsiders and so there is a bigger talent pool of outsiders than there are insiders, and that there is less competition there. All the VCs in the world are trying to go after former Facebook execs.
Auren Hoffman (48:58.551)
you know, top of the class Stanford CS kids, etc, etc. Much smaller pool, many more competitors. I would also say that outsiders more often than not have a chip on their shoulder. And that is very useful as a startup founder to feel like you've been underestimated and to feel like you need to prove to and this isn't healthy, but if you're looking to prove to the world or yourself or your parents or someone.
that you're better than they think you are is extremely useful. And I would argue that insiders sometimes have the same problems that kids of rich parents have, which is like, they've always been skating on the easy lane. And so I think that what we have seen as YC has gotten bigger, and the advantages of YC have gotten more clear, is we see some insiders saying,
is better than it looks, you know? And I think you start seeing this a lot with YC alums who come back, right? I did YC twice, Justin did it four times, Parker came back and did it, Kyle came back and did it. You start seeing this like, oh wait, I'm an insider, but the thing that Brian Chesky told me when I was considering doing YC again was, YC is a toolbox full of tools and you know how to use the tools better.
than you did when you were 23. And you can get way more out of them. And I was like, oh shit, you're right. Whereas I think a lot of people wanna see YC as like a status thing. It's like, oh, it's like undergrad. I'm already past that. It's for young people. It's for first timers. It's like, no, in this way, it's not like a university. In this way, it's very different than a university. In this way, it's much more like a software company or like dev tools where like a bad engineer can take some dev tools and build something.
Auren Hoffman (50:37.672)
And a great engineer can take some dev tools and build something much more amazing. And so I think that what's been interesting is to see the insiders kind of flip a little bit and say like, oh shit, this deals better than it looks. But it was still designed and it's still primarily for outsiders.
Auren Hoffman (51:24.202)
One of the things is like what as YC itself evolves and becomes more of an institution, um, in some ways it's a hoarder for somebody with an outsider personality to manage. I have this theory. The reason why Paul Graham left YC is that he wanted to remain a true outsider. And it would be hard to do it as like running, you know, cause you can't necessarily say whatever you want to say if you're running like an institution. Um.
So, um, and it gives you, you know, you have a little bit more freedom if you're, uh, if you're slightly on the outside or, or something like that. Like at some point, like these things become, you know, if it's going to be a hundred year institution, um, it's a, um, uh, you, you have to watch a little bit, what you say, you have to be a little more polite, uh, right. Or, or no, you don't agree. Yeah.
But that's not, I don't think being for outsiders means that we have to say radical things. I think it has to do with our process for funding companies. Typical VC funds celebrate the idea that you need to have a warm referral to talk to them.
What about everyone who's not one degree separated from someone who knows Roloff at Sequoia, right? It's easy to argue, oh great, they should be excluded because they don't have the skills to do that networking. YC would argue, please continue excluding them. Like we would love to talk to them. But I would say that the thing that would make us not for outsiders is if somehow how we solicited applications, read applications and interviews changed. And.
Auren Hoffman (52:40.01)
Auren Hoffman (52:52.842)
Yep, yeah, yeah.
That's something that stayed fairly consistent, from PG's first batch to today. You don't have to know us, you don't have to, you just fill out an application, you don't have to build a deck. Like it's an experience that's a lot more like applying to a college. And so it is familiar and common to almost every smart person in the world.
Auren Hoffman (53:28.819)
What is a, last couple of questions, what is a conspiracy theory that you believe?
What's a conspiracy theory that I believe? I'll tell you one, it's not really a conspiracy theory. I think Doug at Sequoia let a big secret slip in a podcast once that we literally, I recorded and put onto YC's user manual because it was such an important secret. And he basically said, Sequoia can't help you if your pre-product market fits.
Like it's your job to find product market fit. We can help you with all the shit you have to do after that. But the product market fits the black magic part. I think that my conspiracy theory is just, you know, some version of investors are way less helpful than they appear or present. And founders deserve almost all of the credit because finding product market fit is so fricking hard. And like no one who's like,
Auren Hoffman (54:08.428)
Auren Hoffman (54:31.756)
doing board meetings once a quarter is really contributing to that effort at all. This idea of founding investors, it's like, sorry, no. In many ways, especially in the early stages of companies, your investors are more like your lawyers. No one would give your lawyers credit for figuring out the company.
Auren Hoffman (54:37.63)
Yeah, yeah, totally.
Auren Hoffman (54:52.226)
Well, maybe in the scenario where you mentioned the Gideon use scenario where they could like scare you into doing finding something else or something.
Maybe, but I would even argue the thing Gideon said to us was obvious. We were running a company that had mostly copyrighted content. We were being sued. He wasn't giving us a secret. He was just holding a mirror to our faces and said, this is what you are. So anyways, I think that maybe my conspiracy theory is that pre-product market fit, the value-add investor doesn't.
Auren Hoffman (55:18.03)
I don't want to say the value I've done investor doesn't exist. 90 plus percent of the lift is giving product market fit, and the investor can't help you do it.
Auren Hoffman (55:37.642)
Yep. All right. Last question we ask all of our guests. What conventional wisdom or advice do you think is generally bad advice?
I used to think that everyone who had the skills to do a startup should do a startup. And I used to attempt to browbeat people into doing startups, who obviously didn't want to. And I would do it even more for underrepresented populations, like, oh, man, like you could be doing this, like you should be doing this. And I think it took me a long time to realize that this isn't skills are important.
but they aren't the actual real game here. Everyone in the game has got skills. Anyone who's serious in the game has skills. There's a little bit of, you have to be broken. Like wanting pain and stress, like seeing life as more exciting when there's pain and stress is not rational. And so I think I've just come to this conclusion that there are...
you know, 1% or whatever of the population has this like weird thing that's broken about them. And they're inclined towards startups, and they probably should go acquire the skills so they can go do them. And everyone else isn't. And they probably shouldn't do them. And I bet one day, you know, we'll figure out what that thing is. And we'll see it as akin to, hey, you know, most people who are six, eight to seven foot should think about basketball, right? And most people who are five, two.
Auren Hoffman (56:58.335)
Auren Hoffman (57:14.882)
Should have. Why? And I think that, you know, there is this idea that everyone can be an entrepreneur. And like, sure, everyone should have the right or the freedom to be an entrepreneur. But I don't think everyone can process this amount of pain and disappointment and not only remain motivated, but like can use it to motivate themselves. And it's funny, cause on the team at YC, you know, the people I work with, they always joke. They're like, you're waiting for bad things to happen.
you want bad things to happen, you like when bad things happen. And I'm like, that's not wrong, you know, and they're always they always complain when it's like, well, you got good feedback on this and you're dismissing it. I'm like, yeah, I want the bad. Like, that's what good feedback that's useless. And I think that
Auren Hoffman (57:45.902)
Auren Hoffman (58:02.166)
probably slight amendment. You're really talking about, you're not just talking about like being an entrepreneur because probably lots of people will start like a little thing. It's more like being a high growth startup, like kind of person, right? It's a very, which is a very, very different animal than, okay, I'm going to start a firm with five people and we're gonna put my kids through college with it, which is great.
Yes, yes, yes. Yes.
I completely agree with you. I'm not talking about starting a small business. I'm saying trying to be an entrepreneur that builds one of the largest businesses in a country or in the world. And I think that that's probably the piece of conventional wisdom that I got wrong, that I should be, instead of talking about how amazing this is and trying to convince people, I should be talking about how bad it is and have people being like, no, that sounds like fun. Yeah. They'll be like, oh, great.
Auren Hoffman (58:45.654)
Select out. Yeah, yeah, yeah.
pain and suffering sounds like fun, we're of the same mind.
Auren Hoffman (58:54.402)
Well, there's also a certain class. There's a lot of people who have a desire to be successful, but there's a certain class of people who just would not have an extremely tough time being successful in the confines of an already existing institution. Um, and they just don't have the personality to do that, um, in the right way. And, um, and so those people almost forced to do a startup whereas somebody else who, um, maybe has quite, quite a good personality to be successful in those.
in the confines, maybe it doesn't actually ever make sense for that person to start a company.
You know, when I talk to college kids, I make this exact point you made. I basically am like, it is rational to like a job at Google. You don't have to work very hard. You get laundry, you get food, you get like, it is not bad to want that job. But if you think that sounds horrible, maybe, maybe this world's for you.
Auren Hoffman (59:48.478)
Yeah, exactly. Yeah. Uh, this has been amazing. Thank you, Michael Siebel for joining us. The world of DAS. I follow you at MW Siebel, uh, on Twitter. I definitely encourage our listeners to engage with you there. This has been a ton of fun.
Thanks for watching!
Thanks for the invite again.
Auren Hoffman (01:00:05.694)
All right, amazing. Thank you. This was really great. And we'll give you a full overview of when we launch.
Michael Seibel is a group partner at Y Combinator and the managing director of YC's startup accelerator. He evaluates thousands of startups a year and sees hundreds go through YC, which means he has better data on startup success than just about anyone in tech. He’s also the cofounder of Twitch.
In this episode, Michael and Auren dive into the startup landscape and the inner workings of Y Combinator. As a partner at YC, Michael sheds light on the top qualities he seeks in applicants, drawing from his experience of sifting through a staggering 20,000 yearly applications to their accelerator program.
They also discuss Michael’s journey as co-founder of Twitch (formerly known as Justin.tv), which now boasts an impressive 140 million monthly active users and was acquired by Amazon for roughly $1B. Michael delves into the mistakes Twitch made along the way and the surprising advice that saved the company from collapse.
Amjad Masad is the founder and CEO of Replit, an online coding environment.
Amjad and Auren discuss what the rise of AI means for coding, developers, and tech companies. Amjad shares his thoughts on consumer vs creator computer culture, why development hasn’t moved to the cloud faster, and why everyone should be able to call themselves a programmer.
They also discuss Amjad’s experience coming from Jordan to start a company in the US. Amjad shares whether he still thinks Silicon Valley is the best place for startups. They also discuss unique considerations for CEOs that aren’t talked about often: free speech as an executive, assessing expert opinion, and how investing makes you a better CEO.
Avlok Kohli is the CEO of AngelList and the founder of FastBite and Fairy.
Auren and Avlok discuss the underestimated impact of AI and how it could be creating a tech mega cycle. Avlok explains the importance of unique data as a moat for companies building on top of large language models, and how this technology can disrupt incumbents.
They also discuss the rise of Dual Threat CEOs and their outsized success in early stage venture capital. Avlok makes the case for why more current operators should start funds, and explains why they’ll thrive in the changing VC landscape.
You can find Auren Hoffman on Twitter at @auren and Avlok Kohli on Twitter at @avlok.