Y Combinator Startup Podcast - Silicon Valley's Cargo Culting Problem

Episode Date: July 13, 2023

When it comes to building a startup you’re never doing it entirely from scratch. Inspiration and ideas can come from a variety of places, including other successful startups. But there’s a thin line between borrowing smart ideas and copying them blindly - otherwise known as Cargo Culting. In this episode Dalton and Michael break down the problem with Cargo Culting and offer advice on the right way to draw inspiration from other successful companies. Apply to Y Combinator: https://yc.link/DandM-apply Work at a Startup: https://yc.link/DandM-jobs

The idea that superficially copying Uber and copying the things that they said in interviews, pretending that you're Travis, whatever, doesn't work. It's just like wearing a black turtleneck does not make you Steve Jobs. Yeah. That does not make you Apple. That is cargo culting. Yes. This is Michael Seibel with Dalton Caldwell.

And today we're going to talk about cargo culting. We encounter this way too much in our life. So there's classic variety and then there's new age, modern calgary culting. Let's start with classic. What's classic cargo culting? Well, let's start with what is cargo culting? What are we even, what are these guys talking about? All right.

So the history of the term, as I understand it, is when you superficially copy something without understanding why you're copying it. Yes. That is cargo colting. I think the history, if someone looks at us up in Wikipedia is like, folks would create a fake airfield so the planes came back. So it was something that looks like a landing strip because they wanted the airplanes to come back. Yes.

They didn't really understand why the planes were coming. So they're like, oh, if we make a fake landing strip, then the airplanes will come back. Okay. And so in this classic cargo colting mentality, it's just I'm going to copy something. I don't know why. I'm going to copy it. And then I will be as successful as a thing I'm copying, right? Anything to add there?

Well, I think in the classic variety, what's essential is the thing that you're copying is good. Yeah, we're not, no one would debate whether it's good or not, right? And so we're just debating how you copy it. Yeah. I think what's funny about this modern new age car car welding that we'll get to later is that like, we could even just debate whether the thing you're copying is worth copying at all. Fair? And so then it becomes this like kind of double-sided, you're screwing yourself.

But let's just go with classic. And the classic variety, what goes wrong in car car carolting is because you don't understand, really deeply understand the product that you're copying, you miss some like obvious things. Yeah. And I like your like, you know, airport example, right? Where it's kind of like, yeah, if you make a landing strip in the middle of the jungle, airplanes won't land there.

You're missing me. There's a part of the kind of. There's something larger going on. Yeah. The whole system. And I think founders did this. I think we have victims of this. And so let's talk about kind of the companies that were classically agriculture

in our era. Well, the first and most obvious one was Google. Yes. Everyone in the 2000s was obsessed with Google. Yes. It was the best startup. Yes. And so, you know, everyone you talked to was like, well, Google did this or are you the next Google?

Well, this is how Google did it. Yes. And so all these ideas that may or may not have been good ideas, we were kind of trained. We're obviously the right. We're things to copy. So the most obvious one is just like the office culture of having an open office and like bright colors and like free snacks and like all this like super. super Googling stuff. Yep.

That was considered, like, if you wanted to do a real startup, you had to be like Google, right, man? Not only that, that's what made Google great, was the lack of offices. Yeah, you're right. No cube farms. That's why it was considered their culture. Like, this is actually, there was arguments that copying down to the T, every aspect of the Google office culture was key to building a successful startup. Exactly.

Exactly. I think the other thing was this kind of flat org. Yeah, no managers, don't need management. another classic one that I love was just to hire as many smart engineers as possible. If you ever come across a smart engineer, no matter what, give them a job offer, you'll find a way to use them. I think it's funny because, like, this is what we came up. Yeah, like this was conventional wisdom. There was either examples more silly than that.

It was like the name of your startup, you need to be cute. Yes. Or like drop vowels, remember, flicking or started this where there was no E in it? And so it would be, you need a Google, like, logo. that looked like primary colors and had like a shiny, remember what was the turn? You would put like a lens flare on your logo. Do you remember this?

Everyone did this. Go look at logos back in like 2006. And that was because Google and these other startups had lens flare on their logos and they had cute names. Yes. So you had to do it. When I love it because it's like obvious

in the pursuit of building the world's best search engine, the logo styling was essential. I mean, really it was a vital step. So number two, of course, Facebook, even closer to kind of our generation. They're the ones that brought you the going viral. Oh, yeah. How is your thing going to go viral, Dalton? How's it going to go viral?

They're the ones who brought you the where's the share button? Yeah, you have to have a share button on everything. What else did? What else was the Uber? Well, the Facebook classics. I think just trying to not make any money directly from users. and build up as big of a user base you can. And then build an ads business.

Obviously, everyone was trying to do this. But Facebook really, we were constantly told by everyone we knew. And again, we believed it that following the Facebook playbook was what was right for every business, no matter what type of business. No matter what. Charging your users was against the rules. Yeah, because you wanted to get the most user data possible. You wanted to get the biggest database of users, all that of this stuff.

And then the nobody cares about privacy. There was a suck quote where it was like privacy. Doesn't matter. I forget. They were like, anyway, everyone that was blindly following what they did was trained to ignore all of those matters. Then maybe the last one, even more modern. Uber.

Yeah. This was the fun one. I think it started with spend as much money as possible. Which I love because Uber was a successful company. It's great. And the idea that we could boil down a key strategy, spend as much money as possible. But that was the takeaway.

They wrote a book about it. It's called blitzscaling. Spend all the money. As fast as it comes in. What's funny about maybe these Uber ones that are slightly different from the Facebook and Google ones is like, I don't think Uber even did these things. Yeah. Like Cargoculting kind of took on its mind of its own.

Exactly. And again, to be really clear on the point, Uber did a great job of what they did. They nailed it. It's the folks that were like, well, Uber did X, so we should do this without really understanding why Uber did the things they did, they would superficially scale the way they scaled, burn the way they burned. Yes. Ignore laws the way they ignore laws. Yes. Ignore union economics. Go to as many cities as possible as possible. Yeah. And so it was, it was the superficial wannabe Ubers. Yes. That did the silliest stuff versus Uber itself.

Yes. That worked pretty well, right? Now, let's be clear, all these businesses worked. Yeah. So let's talk about it, right? Like for Google, this was great advice. Yeah. For Google. For Google. hiring as many really smart people was essential because they were trying to do something really hard. Yeah, I think that the conventional wisdom pre-Google was that search was a commodity. Yes. And that Yahoo was good enough. And that incremental engineering breakthroughs in search wouldn't matter. And Google disproved that where it was actually a hard technical problem.

And they were able to put those engineers to work. Also, historically, this was right after the dot-com bubble crashed. So they were able to hire some great talent that was available because the dot-com bubble just crashed. And so if you really look at the decisions they made in context of history, it made a ton of sense. Made a perfect sense. But how many people did we know that blindly copied the Google Playbook and it did not, they did not replicate Google's success, did they? No, no.

They weren't even trying to do anything near as hard. I think in Facebook's case, another perfect example of, like, not charging users was really smart, right? They understood they could build an amazing ad business. They understood that scale was the only path to it. and like Facebook is an amazing ads business, right? Like that was really smart for them. Not clear that every business should be an ads business. Yeah.

And when you think about it, how many Facebooks are there and how many people that were blindly copying the Facebook cargo, you know, how many people that were cargo-culting Facebook succeeded? Yes. Not many. Again, maybe it's more than zero, but wow. That was more of like a one-shot epic company than something that people could just blindly copy

and be just as successful. Facebook. The second Facebook style profile you had was delivering infinitely less value. Yeah. Than the first one you had. I would also argue that people, you know, Facebook and this whole going viral, it's a lot of easy to try to make your product go viral if users spend two hours a day using it, right? Yes. That, day one Facebook, people were spending two hours a day using it. If people use your product five minutes a month, it's a lot harder to go viral. Yes, you're, you're, You need some different tactics. You can't just copy Facebook blindly and hope it works.

So it's like going by all was a great strategy for Facebook. Is it a great strategy for you? And the last with Uber, I think the hilarious thing about Uber is that most of the common beliefs about Uber just weren't true. People who were investing in Uber were looking at markets where Uber was doing very well in unit economics and purposely funding them to expand quickly. It wasn't people looking at Uber serving no markets well. burning money left and right and saying, oh, let's just throw more of our money on the fire. And it was weird to watch founders just be like, well, you know, we should expand our business to 17

cities when it's not working in one. Yep. When like, you remember the first time you took Uber? Like, I do. It was a huge moment. It literally was so valuable. Like, it was immediately successful in San Francisco. And so this idea that like, oh, Uber just like, nobody liked it.

And they just burned a bunch of money and somehow got successful. It's like, this doesn't make any sense. The idea that superficially copying Uber and copying the things that they said in interviews, pretending that you're Travis, whatever, doesn't work. It's just like wearing a black turtleneck does not make you Steve Jobs. Yeah. It does not make you Apple. That is carbo-culting.

Whereas like if you built a thing that people would use multiple times a week to get to the most important things in their life, like work, are their loved ones? That's pretty good. That's really valuable. So hard to replicate this stuff, right, man? It is. It is. It is hard. You have to be really smart to copy well.

So we talked about this classic cargherculting, and at least with the classic cargrafting, you're copying successful companies. What concerns me more and more nowadays with the explosion evaluations is that we see a lot of founders copying companies that haven't succeeded yet. And this is what's really scary to me is that like some founder will be like, well, we have to do it this way because so-and-so coming to this. And I'm like, oh, why are they good? It's like, well, they're a billion-dollar company.

They just raised a series B. They just did this. They heard they read some fundraising announcement. And then I asked them, well, how much revenue were they making? And the most common answer is I have no idea. And it's like, so you're basing a huge chunk of your company's strategy.

Yeah, your strategy. You're the CEO of the company. Let me get this straight. You're the CEO of the company. Yes. You're telling me the strategy you're betting the farm on. Yes. Is based on you read a fundraising announcement.

And you have no other context, but like a podcast interview with a founder. And it has a lot of GitHub stars. It's one thing to try to copy something that's working. Like, you might get lucky in when I'm like, right? Like, if you copy something that's not working, I don't even know what are we doing here. And I think that, unfortunately, these valuations have gotten people all messed up in the head. Like, they think that if investor thinks is worth a billion dollars, it's working. Well, again, just to be super concrete, how many people do we see that were influenced by WeWork and building WeWork for X?

And their entire strategy was based on WeWork. Yeah. We saw the same thing, Zoom Pizza. Yeah. Like robotic pizza making. And so you'd see these things where it eventually turns into like, you know, sadly a flame out for the founders. But there were a lot of people that were copying these companies blindly. Because they raised a bunch of money.

Yeah. And I think this is part of this kind of certainly weirdly bigger trend of like people doing the modern cargo culting on. amongst startups where they're like, oh, if my startup looks is going to be real, it has to look like other startups. Yeah, it's like the superficial. They're really focused on how people will perceive the startup superficially. Yes. Like the logos thing we were talking about. Yes. Yes. Like what are some examples of the superficial things that you want to do to pretend to be a startup? Yeah, well, I have to raise money, right? I have to have good advisors. Maybe someone, God forbid, told me I should have a patent, right? I have to have an amazing picture.

I got to be able to pitch my company anywhere. Maybe a couple conferences I get invited to, some press. And these are things that founders do to impress themselves to impress other founders. Yep. The user is not essential. Has nothing with the product or something positive users. It's all the superficial facade of this looks like a startup that looks like other startups that raise money. So let's think about because we're talking about Tiger Holding on three levels, right?

We're like, okay, garden variety cargo culting that's still really bad as you cargo code Google. Then we're like, okay, the modern thing is to cargo cult, the modern unicorn that has very little revenue. And now the super, super modern is just to cargo cult the other struggling startup. I love when someone references a startup that we know and we're like, no. We're like, no. And I think that like maybe the base conclusion here is that you have to do something. copying. You cannot innovate everything. But I think that like I like when founders start with what does the user want. Starts with who are the companies the user's paying for, paying today

to provide this service, and what can we learn from those companies? And then like, what can we learn from those companies writing the service who are successful? Okay, now let's think through it. And I think once you think through it through the user's eyes, you can tell which things are essential. Do people care about the Google logo or do they care that they get search results? They're good. Hey, how about that? Fast. Do they care about the Facebook, I don't know, or do they care that their friends are on there and they get updates quickly? Like, do they care about, you know, what city Uber just expanded to or do they care that like Uber shows up in five minutes in their city? So I think when founders start with the user, they won't go into

this carvolting. I think when they start with the, like you said, fundraising announcement. It's almost impossible. Look, the metaphor I use for this is think about the difference between plagiarizing a paper that you read. Imagine you're like, hey, I got to turn in a report. And what you do is you go and Google for things and copy and paste blindly all the stuff. You copy and paste three different things. Just put in your paper and submit it. Yes.

That is not evidence of original thinking. No, no. Versus you go read a bunch of places. Yes. You digest the information. You synthesize it into your own original work. You maybe cite the sources of what you're learning from. But you actually took the time to digest and process these things.

Same thing with the music industry, where if you plagiarize a song, that's a problem. Like if you just rip off someone's music, you will get sued. Right? But if you listen to a lot of music and you get influenced by some of the sounds you hear and you integrate these new sounds into the new music that you want to put out, totally cool. That's literally... That's the game.

That's music. That's the music business, is that everyone's being influenced by everyone else. And so again, the really important thing to tease apart is copying completely without thought or any original thinking or original work versus being influenced and borrowing good ideas to integrate into your own thing. Right? Doesn't the thing that's talk about startups is that you do have to think about a lot of things. There isn't just like... You can't just copy and paste. There isn't just a checklist.

You do have to think carefully. But I think that's why really smart people are. attracted to it. Yeah. Because like you have to think really carefully. Yeah, it's an amazing test of your skills. All right. Great chat. Great. Thanks.