How did people find anything on the web before Google?

Before Google, people found things on the internet in two ways: by browsing lists that humans had written by hand, or by using programs that collected names or words and let you look them up. The very first of these, Archie, appeared in 1990 and did not even search the web. It searched the names of files stored on FTP servers. Over the next five years the tools got steadily better, first with hand-made directories like Yahoo's, then with robots that read every word of every page, like AltaVista.

Archie came from Alan Emtage, a postgraduate student at McGill University in Montreal. It gathered the file listings of public FTP sites so you could search them in one place. It never looked inside the files, so you had to already know the title of what you wanted. Even so, at its peak it accounted for half of the Internet traffic in Montreal.

Then came the web, and at first it was indexed by hand. Until September 1993 there was a single list of webservers, edited by Tim Berners-Lee and hosted at CERN. As more and more servers came online, one list could no longer keep up.

Two answers appeared in 1994 and 1995. Jerry Yang and David Filo, two Stanford graduate students, kept a human-edited directory organized as a hierarchy: that was Yahoo, and by 1998 it was the most popular starting point on the web. Others built robots instead. WebCrawler, from 1994, let you search for any word in any page. AltaVista, launched in December 1995, was the first full-text searchable index of the web, and it drew more than 80 million hits a day by 1997.

Even that was not the end. In 2000, AltaVista was still used by more Internet users than Google. The rest of this path follows the machine route: how robots read the web, how an index finds a word in a blink, and how the results get sorted.