I started off by reversing the links, and then I wanted to find basically, say, who links to the Stanford home page, and there’s 10,000 people who link to Stanford. Then the question is, which ones do you show? So you can only show 10, and we ended up with this way of ranking links, based on the links. Then we were like, “Wow, this is really good. It ranks things in the order you would expect to see them.” Stanford would be first. You can take universities and just rank them, and they come out in the order you’d expect. So we thought, “This is really interesting. This thing really works. We should use it for search.” So I started building a search engine. Sergey also came on very early, probably in late ’95 or early ’96, and was really interested in the data mining part. Basically, we thought, “Oh, we should be able to make a better search engine.”