Blog

GiftRocket combines the thoughtfulness of a gift card with the flexibility of cash.

YC Application Checklist

Posted 04 October 2011 by
Photo by Ciro Cattuto

Now that YC applications are open and people are scrambling to put together their apps, I’ve had a bunch of founders ask me for advice on getting in.

Some Math Behind YC Applications

At current growth rates, there’s a good shot YC will fund over 80 startups in W12 (the last batches were 36, 44, and 63). PG and Harj have said that acceptance rates have stayed constant between 2% and 3%. So there could be 4000 applications this round. Assuming a partner reads each application, that works out to at least 400 applications a partner. Applications are about 2000 words. So each partner reads at least 800k words of application content, or the equivalent of 10 novels. [1]

For a given company, the partners just need to identify that there is some slight promise that makes an interview worthwhile.

The most efficient way to get through this number of applications is to create a set of heuristics that gives a quick sense of whether an applicant is worth interviewing. They’re going to be looking for the presence of any disqualifying information that might let them move on to the next application, or qualifying information that is uncommon enough to help the application stand out. [2]

PG and Harj have written about some of these heuristics in various places. I’ve compiled what I think are the major ones below.

Avoid Red Flags on your YC Application

  • Uncommitted founders. If determination is the biggest predictor of success, then lack of commitment from the founders is a sure-fire flag. One applicant mentioned that she was planning to return to her MBA program after YC. No bueno. The best signal founders can give is that they’ve quit their jobs, moved in together, adopted a dog, and are working on the startup regardless of YC.

  • Single founder. PG has said the odds of being accepted are much greater (roughly 4x) with a cofounder. So it isn’t insurmountable, but it places considerably more importance on the first founder, and his / her ability to get others on board within a short time.

  • Skewed equity splits, or weak founder relationships If one founder holds 90% of the equity and the other two are splitting 5%, there’s probably a strange dynamic going on between the founders. The same is true if the founders haven’t known each other for a long time.

  • Founders working on software who can’t code. “Chuck Norris can get funded by YC without a cofounder who can code. Others we decide on a case by case basis, but the odds are a lot better if you have a hacker cofounder.” If you are working on a tech startup, one of you should know how to code, or be actively learning.

  • Outsourced development. Maybe 20% of YC startups change their idea at some point during YC. It’s hard to be nimble when you need to explain your pivot to your Yugoslavian dev team.

  • Video that fails the airplane test, or no video. I’ve seen a video that an applicant posted online that was 2 minutes of incoherent rambling. It isn’t hard to pass this one– just look at the camera and talk like a normal person for 30 seconds.

  • Working on something that no one wants. Your idea doesn’t have to be amazing. But if you can’t convincingly answer who would use your product and why, that isn’t good.

  • Incoherent writing. PG wrote an entire essay about applying where the primary conclusion was “be concise and matter of fact”. Poor communication is a red flag because it means that it is X times more difficult to read the application. But it also indicates that the founders may not be great at communicating in general, may not be focused on the startup, or haven’t thought through the idea well enough to be able to describe it clearly to someone else.

Notice that there are a number of things that won’t be huge strikes against you. For example, ideas are hard to judge off of just a few paragraphs. Most could be promising, but its too hard to tell. The focus is really on the founders at this point.

Raise green flags!

The absence of red flags is a good sign, and when accompanied by some of these green flags makes an application a great candidate to interview.

  • Prove you can get shit done. A prototype, good GitHub profile, and good answer to “tell me something impressive…” help make your point. Talk about the side project you launched on HN. An MIT degree won’t hurt, but don’t expect it to guarantee you an interview either.

  • Work on a real problem. The world probably has enough local business discovery startups. YC cares if you’re working on a problem. Bonus points if it is one you’ve experienced yourself.

  • Be qualified to work on your problem. MemSQL’s founders are two guys with deep technical back-end expertise; the GrubWithUs founders used to run a pastry shop; the Earbits founders are former bandmates from LA. Those teams just make sense.

  • Show an acute understanding of the problem. PG has said the line that sold him on Dropbox was what competitors didn’t understand.

  • Have someone in YC vouch for you. A glowing recommendation from a respected YC founder will make the team pay more attention to your application than they would have otherwise. That helped AirBnB get in after the application deadline.

  • Be a past YC founder, or a founder / first employee with a prior exit. It helps immensely for any sort of investor to know that you’ve been through the life cycle of a startup before.

  • Traction trumps. A crazy looking growth curve can excuse a lot of other issues.

The distribution

There’s a fair amount of correlation in the flags. People who have had past exits will probably also write well and have a better idea of what sort of things would make good startup ideas, and have an easier time finding talented cofounders. This ends up creating a bimodal distribution of the quality of applications.

There’s some evidence of this when PG says that chances of getting an interview are either extremely high or 0. So given that distribution, and these heuristics, it is probably very easy for the partners to make decisions 95% of the time, with maybe some debate about the remainder.

So with a week left, hopefully you have a good team in place. So just focus on writing clearly and intelligently. Our first drafts were unintelligible, as were most of the other apps I’ve reviewed. But they get better with review.


Notes

[1] There’s a LOT of content in the applications. The word count on Dropbox’s YC application came in at about 2000. 400 applications is 800k words. The average novel is between 80-100k.

[2] Not too different from reading resumes and cover letters for analyst positions at I-banks and consulting firms. At my former consulting firm, I’d be given a stack of 300 and need to have them scored in a week. I had the process down to about 20-30 seconds an application. I think I could have had a computer do the process entirely using SAT score, GPA adjusted for major, and number of legit-sounding leadership positions in activities. Even university admissions aren’t that different. The academic index described in A for Admission describes the formula in detail.