The Request for Startups
Paul Graham
16 August 2009

Every funding cycle we try to do something new. For the upcoming winter 2010 cycle we're introducing the RFS (Request for Startups).

There are a lot of startup ideas we've been waiting for people to apply with, sometimes for years. Recently we tried hinting at some of them, and that has yielded results. So now we're going to try being more direct: we're going to issue numbered RFSes with more details about the idea, and suggestions about the sort of group that would be a good match for it. Groups applying to YC will be able to specify on the application form if they're responding to a particular RFS.

RFSes won't describe exactly what we're looking for, partly because we want to see if applicants can figure out the other half themselves (ideally better than we had), and partly because describing the idea precisely might create competition for the YC-funded startup we hope will be working on it.

We don't expect responses to RFSes will ever be more than a fraction of the applications we accept. We wouldn't want them to be. Most good ideas should be ones that surprise us, not ones we're waiting for.

We expect RFSes to be relevant in two cases. If a group is already heading in the direction described by an RFS, we hope they'll be encouraged to apply by hearing that this is something we've been waiting for. And if a group hasn't decided what to work on, perhaps an RFS will spark their interest.

Responding to an RFS will never be the deciding factor in who we fund. As we've said many times, the people matter more than the idea. Plus you can't do a good job on something unless you're genuinely interested in it. So don't apply with an idea described in an RFS if there's some other idea you'd rather work on. You're more likely to get funded to work on the idea you actually like.

We might ask for a little more equity from startups responding to an RFS, because we'd expect to contribute more to them. But at most a percent or two, and often nothing. Ideas count for something, but execution matters far more.

Apply | What We Do | Blog | People | Jobs | Quotes | FAQ | Contact | Lib | Legal