Y Combinator’s batches are getting bigger and bigger. Its first batch back in 2005 was made up of just a handful of companies — this latest batch came in at over 400. But that’s out of over 17,000 that applied!
What gets one team in where another is turned away? If anyone can answer that, it’s Dalton Caldwell, managing director and group partner for YC. He’s been with YC for its last 19 batches, with much of that time spent overseeing admissions. He puts his tally for the number of applications he’s reviewed in the “tens of thousands.”
Caldwell recently joined us at our TechCrunch Early Stage event, where he shared many of his insights on the application process — what works on an application, what doesn’t and the right ways to stand out.
Caldwell’s session was made up of two parts: a 20-minute presentation on how he thinks about YC applications, followed by an audience Q&A. Below are some of my notes on his presentation, along with some highlights from the Q&A.
The team
“The first thing I look at when I read an application is the team. What I’m looking for is technical excellence on the team,” Caldwell said.
What that means can vary a bit from company to company. For a SaaS startup, that might mean a programmer who’s proven they can build and ship; for a biotech company, perhaps it’s less about coding and more about field expertise. But most of the time, he wants the team to have someone who can actually build the thing.
Our teams that rely on trying to hire outsourced engineers or consultants or whatever to build their product tend to move much slower than folks with a technical founder. They tend to get ripped off. Dalton Caldwell
“I like to see some kind of founder/market fit for what the idea is,” he clarified. “If you’re building a developer infrastructure type of thing … if you’ve built related tools in the past, or your prior job is related to it, that’s great to see. Even if you’ve contributed to an open source project that’s related to the thing you’re working on, that’s great to see!”
“I always get questions about this — why do we care so much about having a technical founder on the team? Our teams that rely on trying to hire outsourced engineers or consultants or whatever to build their product tend to move much slower than folks with a technical founder. They tend to get ripped off. It’s not great, and I would really encourage you to try to have a technical co-founder on the team”
And generally, Caldwell suggested that technical founders shouldn’t be doing it alone.
“Ideally there is more than one co-founder. It’s much harder to be a single founder. This is something people always ask questions about, and I’m sure we’ll get questions today. But look: We have the odds. Even the folks that we do fund that are single founders, they have a very hard time. The fact is that startups are ultimately a battle against yourself. It’s an emotional battle and a battle of wills. Doing that by yourself is really hard.”
The application
“What else do I look at in an application? I look for something that is clear and well written. I don’t like buzzwords. Conciseness. Clarity. Bluntness. Those are all things we look for in a well-written application.”
It’s a message that held true throughout Caldwell’s talk: Put yourself in his shoes. He’s looking at thousands of applications; when you’re one among thousands, words are your currency.
“Think about it. I’m reading lots and lots of [applications]. Cut to the chase. Get to the point. Tell me what you’re trying to say. If it’s a wall of text trying to explain what you do, that’s not great. What we’re looking for is very, very direct, clear communication. We’re looking for evidence you understand what you are doing! Evidence you actually understand the problem.”
So less is generally more — but that doesn’t mean you have to obsess about character count. Or any other aspect of it, Caldwell said.
“You don’t need to spend dozens of hours on the application. You don’t need to contact 30 alumni; you don’t need to talk to me. There’s all these things I see people do because they’re anxious about it, or they think that it’ll help. I haven’t seen a great correlation between the folks that spend tons and tons of time on their application and actually getting in. No joke! A lot of the folks that get in … just apply. They put some thought into it, they do a good job — but they don’t obsess over getting into YC.”
But if you don’t feel like you’re getting your point across within the application, it generally doesn’t help to try to sneak around it.
“If someone cold emails us, or sends us a deck or something … I just archive it. We don’t read that; it’s just not how we operate.”
Likewise, Caldwell said, YC aimed to set it up so that in-network references aren’t necessary. As Caldwell later lays out in the audience Q&A, a particularly glowing recommendation might help, but obvious attempts to just rack up references might hurt. Why?
“I remember when I was a founder raising capital. It was very confusing and weird to try to network my way into raising money or thinking I needed to know people. We explicitly don’t want that to be the case, so I just want to make sure people hear it loud and clear from me: You just apply. You don’t need permission, you don’t need warm intros.”
And be wary of anyone suggesting their intro is the key:
“Watch out for fake gatekeepers that say they can help you get into YC. I see these pop up sometimes; people claim they have some inside angle, they’re alumni or something. … I don’t know, I don’t think that’s a real thing, so don’t fall for that.”
And if you don’t get in? Try again. Most of the companies that get in didn’t make it their first time around.
“Most — more than half — of the companies in the most recent batch had applied before, at least once, and not gotten in. So it’s totally cool to apply and not get in! No sweat. That happens to most people. Sometimes the people overthinking it are like ‘I only get one shot at this!’ … Literally not true.”
The idea
It’s less about just coming in with a great idea, Caldwell said, and more about doing something that works because of you.
“I read dozens, if not hundreds of applications that are all the same idea … every batch,” he said.
“Your startup idea is probably less original than you think. That is not a bad thing! [But] since these ideas are not necessarily unique, what matters more is finding an idea that fits you like a glove — that you have great founder-market fit for.”
Too often, the “great ideas” that people come in with are those that have been tried time and time again — ideas that Caldwell refers to as “tar pit” ideas. Who has tried this before? Why would you and your team succeed where so many have gotten stuck before?
“The lack of a good solution does not always mean no one has started a startup there,” mentioning things like music discovery — a topic Caldwell knows well, having co-founded the music-centric social network imeem in 2003 — as one idea that pops up a lot.
“A lot of folks think that music discovery is an unsolved problem, and they want to fix it — and I apologize if this is too close to home for anyone here — but just because new music discovery is not well-solved does not mean there has not been … probably tens of thousands of founders that have tried to solve music discovery.”
“Be aware of these tar pit ideas,” he said. “See how many people have tried it before you, and try to understand why they failed and — at least in your application — talk about what you’re going to do different.”
“It’s hard for you to know what the most common ideas are, but I recommend doing the research. And if you are in a crowded space, try to show that you know more or have done more work than other folks that are in this space.”
Why some companies don’t make it to the next step
Caldwell said that applicants regularly contact him for feedback as to why they didn’t make it past the initial application stage, and his answer usually boils down to one of a handful of things:
- “It was a generic, low-effort application — usually it’s whatever the most popular idea is at that time. You’ll see people apply every batch and the idea will be different — it’s just whatever is raising money on TechCrunch.”
- “The team doesn’t seem committed or serious about doing a startup. … There’s no evidence [they] actually want to do a startup, [they’re] just applying for the heck of it. Evidence that you’re serious about it is something we’re looking for.”
- “The team doesn’t have the tech skills to build the idea they described, or they would have to entirely rely on outsourced engineering.”
- “It’s a common ‘tar pit’ idea with no evidence that you’ve done any research or have a better take than other people who have tried the idea before,” citing restaurant discovery or social apps meant to help you find people to hang out with as other “tar pit” ideas. “Everyone applies on that idea,” he said. “And they have for the past 15 years. … If you’re going to do that, you need a higher burden of proof that you will actually succeed where so many have failed, right? I’m an optimist; everything is possible. But if you’re trying one of these ideas, there’s a higher bar for you.”
- “Finally, the fact is … there’s thousands of other companies applying to YC. So a lot of the time the reason someone doesn’t make it to interviews is just that there’s other stuff we’re more excited about. And there’s actually no actionable feedback! And I know this is depressing; I know [people] want something to hold on to for feedback here. But with just the sheer number of folks that apply … the folks that do all the things I mentioned earlier to stand out rise to the top. The folks that are just sort of in the middle? They don’t make it to interviews, but there’s actually no actionable feedback.”
Audience Q&A
Following Caldwell’s talk, I moderated a Q&A session with most questions coming straight from the audience. Here are some of the highlights:
About 30% of the companies in the latest batch came in at the idea stage. What got those teams in?
I think for the idea stage, the most important thing is the technical skills of the team so that there’s no risk they can build something great fast. Say they built something like an open source project before, or they’ve built cool products at a prior company before … that is a great thing to look for at the idea stage.
In addition … the ideas, and I know this is hard for people to discern, what is common and not common? When the ideas are uncommon, or really interesting, or very relevant for their backgrounds, it’s easier to get in at the idea stage versus when it’s one of the ideas I mentioned earlier, one of those ideas that everyone applies with — that’s harder. Those ideas that are sort of the tar pits … much much much harder.
Particularly in the face-to-face or video interview stage of the process, how important is personality, confidence or the ability to present?
Sometimes folks get this wrong, where I think they just watched too much “Shark Tank” and so they approach it like it’s frickin’ “Shark Tank.” [sigh]
Honestly, what we’re looking for is folks who we can have a conversation with, and that we feel like we are excited to spend a lot of time with. Part of this job is that we spend a lot of time with founders that we fund, right? We don’t just fund ’em and disappear.
These are folks that become literally part of our lives! Literally, to this day, folks that I fund are part of my life. I talk to them all the time.
Confidence is important but show that you can have a conversation and that you can tell me your idea and be excited about it. That if you get a hard question, you don’t just spout baloney — you actually think about it. You’re like “OK, that’s a hard question — I don’t know.” That’s good! What we’re looking for is that it feels like an authentic conversation, that it feels like someone you’d want to spend time with, and where you think they’ll be able to tackle the problems of a startup. None of these things are like a “Shark Tank” pitch, right?
We have a few questions here about “in-network” referrals. How important are they; do people need an “in”?
You definitely don’t need one. The vast, vast majority of companies that get into YC don’t know anyone in YC. They don’t know a soul, and that’s cool.
Sometimes you see folks that know alumni, or [see] the alumni be like “Oh, I can give you a recommendation,” or something like that. … I think that stuff is fine. Sometimes if they know you really well and they actually vouch for you, that can be something that I notice more on an application … but on the other hand, if it’s someone that doesn’t know you well, that can backfire.
I sometimes see people try to play games by trying to get tons of recommendations or solicit phony recommendations. I … would not recommend that. It’s like anything else in life; if you try to game the system too much, it doesn’t work.
I swear you don’t need that stuff. Most folks that have ’em don’t get in; most folks that get in don’t know anyone. They don’t know any alumni; they didn’t do any of that stuff.
If someone applies a second or third time and gets in when they didn’t before, what usually changed?
That’s an excellent question!
Sometimes the idea gets better. Sometimes they added a co-founder that they really needed — they applied as a single founder the first time then applied with a co-founder.
Sometimes it’s the same idea with the same team, but they’ve really gotten a ton of traction and validated that they are serious — especially if it’s a common idea.
I can look at the delta between applications and what you want to see is, again, people that are for real and serious and have heart to work on their startup. That’s different from when you just see people apply and it’s [essentially] the same stuff; they may change the idea, but they’ve made no forward progress.
How does YC feel about hardware companies that have to go through the FDA approval process?
We do fund those. It’s a hard road. One of our more recent group partners, Surbhi [Sarna] … she is an expert on medical devices and really understands how to evaluate those. She built one of those companies herself and was super successful. So how would we evaluate that? Surbhi would look at it, is my answer
And, look, it’s really hard to not run out of money getting through FDA approvals. … All of the universe wants you to run out of money before you get to market with an FDA-approved device.
Another audience question here: I’ve heard YC is mostly worth it for B2B startups that can sell into the other batch companies. What are your thoughts on that?
Yeah, I mean, I completely disagree. If you look at our portfolio, what got really big … most of them are not B2B startups that sold to the other companies. Look at the recent IPOs: Ginkgo Bioworks? Not that. GitLab? Not that. DoorDash? Not that. Right? These are really, really big companies! OpenSea? Not that.
I think that’s a meme that’s out there, but I don’t know where it came from. It’s not evidenced in the data. Right? It helped Brex! But if you look at our top companies list, if you look at our recent batches, for the vast majority that is not what their go-to-market was.
I don’t know if you’d have a specific number here, but someone asks: What percentage of recently successful applicants have been web3 or metaverse focused?
Yeah, I don’t have a hard number there. I think the other reason that’s confusing is because people change ideas so much, and so many people are pivoting into the space … so it’s a moving target.
What something looks like in applications is different than what it looks like at Demo Day! … But definitely double-digit percentage of the Winter 22 batch, and I would expect it to be the same for the summer batch.
Are you any more skeptical of web3 or blockchain-focused companies when they apply? Do you feel like there’s more snake oil there right now?
I think what happens is founders like to pivot to things where there’s lots of investment activity, whatever that thing happens to be. So I do think there’s people pivoting into it that are like … running out of money or are in trouble and this is a last-ditch effort to raise money. We’re definitely seeing that.
But I’m totally an optimist. One thing I worry about in that space — and this has nothing to do with the idea itself — whenever ideas get really hot and everyone starts copying them and pivoting to it … the failure rate tends to be pretty high.
Do you remember Uber for X? Or chatbots? You end up with these memes where everyone, every startup pivots into this one idea space. The idea space could actually be excellent! And Uber for X was a good idea space; chatbots were good … or machine learning for x, two or three years ago. The spaces are inherently good, and a lot of the companies work out, but because everyone pivots into them, the odds for success feels a little bit lower for any incremental startup that enters the space.
It’s just noisy — versus if you’re in a space where there’s no competition, the odds of success are higher. Like when OpenSea was actually doing the work back in the Winter 18 batch, it was not cool to work on this stuff. So they got two or three years to work in obscurity when it was no hype, and there weren’t really that many competitors, which I think helped them a lot — now that it’s caught everyone’s imagination, OpenSea is well positioned.
Here’s a fun one: How many people turn down YC?
You know, it’s very rare to go through all of the application process and interview process and decide you don’t want to do it. There’s always a few like, maybe right before interviews they got an investment offer … but for the most part, people that don’t want to do YC don’t apply to YC. Why would you go through all of that if you just want to raise an angel round?
What is the most unique application you’ve seen? And, I should add, should applications be unique?
Yeah, that’s tricky. I think when you try too hard to be unique, it always is … unique in a bad way. Like if you’re trying too hard to be weird or something.
I think truly unique things are where your domain expertise is on something very rare, or interesting, and you’re not trying to be unique — you just have a unique interest.
Rigetti Quantum Computing … I funded them, they went public. Not many people, at the time, applied to YC with quantum computing ideas. The guy worked at IBM Research doing quantum computing and his Ph.D. was in quantum computing. So that was a very unique application — I hadn’t seen many like it, it was very successful … but he wasn’t trying to be weird or different. He just happened to be an expert in a thing that was very unique. That’s the good kind of unique, you just have a unique interest or expertise. I love that stuff.
Virtual versus in-person — what does the program look like in the future?
I think there’s announcements coming for how we’re doing this summer batch, but I don’t wanna leak that.
You sure?
[laughs] Yeah. But it seems good that you don’t mandatorily have to move to Mountain View for YC if you’re an international company, and there’s no plans to make it mandatory.
How important is it to have revenue before applying?
I think it depends on the idea space. With lots of ideas, it’s good to have a first customer.
B2B SaaS, where it’s easy to get first customers, if you don’t have a customer, that’s a little weird, so I’d expect to see first customers. Versus like, consumer internet, a new mobile app, I don’t expect to see revenue.
It’s very contextually based on the vertical you’re in and how difficult it would be to get that first customer and how much it shows that your idea is good.
We’re running out of time, but I want to get a few more of these audience questions in. In the application process, should [applicants] be telling YC what [they] want to get out of YC, or why [they] want to get in?
There’s a section on the application that’s “Why apply to YC?” And yeah, I think it’s good to be honest there. … You don’t need to write us a novel on why you’re applying to YC, but I think it’s nice when people put something thoughtful in that section of the application.
One last one! What are the most common things that can kill an application or detract from it?
I think … if you lie.
Yeah. Sometimes people … You wouldn’t believe the stuff that people put in applications that like, doesn’t make [them] look good, or they think it’s good to put in that they did bad things. Saying that you lied or ripped people off. … You’d be surprised at what people put on an application, but hopefully no one here does that.