How to Interview at Y Combinator
by Jarrett Streebin
Back in 2013 we applied to Y Combinator for the winter class and got an interview. We interviewed with Paul, Jessica, Robert, and Trevor. We felt like the interview went really well and were very confident. We had already done a successful , had SV Angel invested, and gathered a lot of good recommendations. We didn't get in.
Six months later we applied for the Summer class and got an interview. This time our traction had drastically improved; we'd gone from signups to a working product and paying customers. We got in.
Here's what we learned between those two interviews and what helped us get in.
First, it's not an interview. It's a presentation. Go in with the mindset that you're going to present your case, piece by piece, and answer questions as they come.
Second, you are not a beautiful and unique snowflake (but possibly will feel like Jack's raging gall bladder on interview day). You're one of many in a very long day. Put yourself in their shoes: you've been pitched a few dozen times already today, you're hungry (or digesting). Worst of all, if you're an introvert you're drained from talking to people all day long.
Given that, you should go in with energy and excitement about what you're building. Don't expect the other side of the table to provide the spark to get you going. The partners are great, just understand interviews is a huge undertaking for them and it takes a lot.
You'll need more than just energy though, your presentation needs to be confident and convincing. What's the best way to prepare? We've heard of many preparing by compiling long lists of questions and answers. This is the worst possible strategy. This forces you to know umpteen questions and answers, and god forbid you didn't think of one, you're left without a prepared answer.
The best way to prepare is to focus on perfecting your pitch from five or six key angles:
These work well because nearly every question the partners will ask can be related to one of these.
Q: "Is the market big enough?" A: Market
Q: "Will you be big enough?" A: Strategy
Q: "How do you know?" A: Traction
Q: "Are you charging enough?" (Hint: this means they see the value in what you're doing.) A: Customer
Let's look at each key area for some tips on how to prepare.
This is what you're solving or hope to solve. You're fixing or creating X by building Y. (e.g. Carrier APIs are so dated that customers end up not integrating them and losing money on shipping, or paying for over-priced software.)
The best way to address all manner of "customer" related questions is to empirically learn everything you can about them. Don't guess, gather evidence. What you know is invaluable, what you guess is beyond useless, it's dangerous. You should prepare to talk both about what you know, and how you've learned it.
This is this most important item. Think of your company and all the others on a spectrum between high risk and low risk. We'll use how far along you are in the product to define the risk. Have a beta product with no actual paying customers or users? That's about as high risk as you get. Have paying customers and rapidly growing revenue? That's very low risk (provided it's not software for regional newspapers).
Where are you on that spectrum? This matters because the closer you are to low risk, the less all the other stuff matters. The further you are from low risk, the more all the other stuff matters.
You're just a week or less away from your interviews which means you have little time to try to change anything. However, if you don't have any customers, or even signups, you still have time to do a , This, more than those 5 or 10 friends who really love what you're doing, will actually mean something. This has launched many YC companies, including one of their .
The way to talk about your traction is in terms of numbers and growth. We have X customers and are growing Y percent month-over-month. We have X revenue and it's growing Y percent month-over-month. (e.g. We have 20 paying customers and we're growing 100% month-over-month.)
A startup with committed customers, prepaid customers, or signups is already better than the legions of startups with no customers or signups at all.
Find a reliable figure for the total addressable market you're going after. For example, shipping is a $300B industry in the U.S. and we can expect to gain 5% of that market. That means our TAM is $15bn. If your target market is a subset of the total addressable, make sure you've identified that beforehand. Be honest with yourself about these numbers because you won't be fooling anyone by pretending that you can address a gigantic market that you clearly cannot.
Market × reasonable percentage you could acquire = the opportunity.
Or, "how it gets big". After TRACTION, this is the most important thing to think about. For those of you that haven't read , they're only looking for companies that are going to get big. Their economics are such that they get a piece upfront and that's it. Given that, they need the pie to be huge (this is slightly different from the strategies of traditional VCs, although VCs often mistakenly apply the same metric. I digress).
Talk about what you're doing today, and what you plan to do in the future. How do you intend to transition from things that don't scale to things that do?
This is the least important area, although there will likely be questions about it. Best possible answer: we have known each other from working together for five years under Nobel Prize Winner X. Second best, we have known each other for five years. Worst: we just met. If you've just met, make sure you know who the CEO is and make sure there's no tension about it.
Also, unless your degrees are directly related to what you're doing, save the spiel for your dating life. They're not looking for degrees. Dedication and resilience is what they're looking for.
Practice these. If you nail them, you'll have the best interview possible. Plus, the great thing is even if you just get to four out of five or six, you'll still get a big chunk of your idea or business out there. During our second interview we only got through three or four of them, but it worked.