We just passed the number of applications from last cycle, so it looks like this will be another record, considering we usually get around a third of the applications on the last day.
Hard to say. The quality of the top group of applicants (say, the ones we invite to interviews) has definitely gone up, but it's hard to say whether it's because the total pool is larger, or because more smart people want to start startups, or because as YC seems more legit, better founders apply.
Unless my clock is wrong there's still half an hour left. But we never quote the number anymore anyway, because now that we have competitors we don't want to tempt them to get into a number war.
Absolutely! I doubt that if we don't get accepted it'll be because we didn't tweak our application enough. At the same time, it doesn't hurt, and on occasion I do find things that really could be done better. (As it turns out, I added a good deal to one question that I'd breezed over the first time.)
Yeah, I was just musing that us engineers love to focus on things which don't really matter. It's an interesting phenomenon, and I do it all the time too. I wonder if the solution is as simple as continually asking ourselves, "Am I spending my time on one of the most critical aspects of the project?" Anyway, I didn't mean anything by it. :)
[I was going to say it's not closed but it says PST not PDT, so nevermind--but that's why this comment is attached here in case you're confused]
PG - it would really rock if they were confirmation emails. The button to submit being gone and replaced with tiny text doesn't inspire enough confidence in me that it worked.
It worked. The "Thanks, you did" isn't just something that gets generated automatically as a result of clicking submit. The code that prints this doesn't know where you came from; it decides whether or not to print "Thanks you did" by checking whether your application has been modified since the last time it was submitted.
Given that picking the right startups is hard how will this increase in applications be handled over time? Increasing the reviewers will not necessarily mean that the quality is still the same. Will the number of days for review be increased in future?
It's always been just the 4 partners reviewing the applications. We've added a few improvements in the software that manages the process, but mostly it just takes us longer. We are probably getting a little faster with experience, though.
Rtm and Trevor and I each read them online independently and vote on them. Then they're ranked in order of their total score (normalized because Rtm always gives low grades). Then we meet to argue about borderline cases. At the end of that meeting we have a clearly defined set of 50-60 groups we want to interview.
I went back and looked at the applications for the current cycle, and roughly half the groups we invited to interviews tied for the highest grade from Rtm. Which was a B+.
Do you think you could make this data public, maybe starting off with startups that didn't make it, including YC's comments when they were evaluated during the interview?
I am sure there is a lot of great info in there that would help other startups. It would be like using etherpad for writing essays :-), we could peek into YC's thought process.
I imagine picking the people to interview is an exciting part of running something like YC. And although the act of reading paper applications is probably more prosaic in reality, the videos should make it more interesting.
Well they are planning on interviewing and accepting more... So that might help a little.
Paul also mused about getting the YC alumni more involved --but I read this in execution rather than selection (not that there was any indication either way).
