Finders Keepers

How Internet search engines control our access to information


| May-June 1999



Yahoo!, Infoseek, AltaVista, HotBot, GoTo.com. If you’ve ever done research on the Internet, chances are you’ve used a search engine. The process is easy and often quite satisfying: Just type in one or more keywords, hit return, and then watch as a slew of hyperlinked results scrolls down the screen. It’s like having your very own card catalog wired into your computer. Or is it? As the lucrative Internet search business evolves, the information you seek may be tainted by the almighty buck.

First, a few basics about how search engines make money. According to Lisa Allen, an analyst with Cambridge, Massachusetts-based Forrester Research, search engines generate revenue in several ways. Besides selling banner ads that run across the top of their web pages or striking distribution deals with content providers, search engines also sell what are referred to in the industry as “targeted keyword buys.” What this means is that Ford can buy the right to have a Ford Explorer banner ad pop up anytime someone types the word Ford into a search engine. It also means that if Ford declines to buy its brand name, General Motors can snatch it up.

While this extremely targeted sell job is enough to make any marketing-leery Web user nervous, it stops just short of skewing the hierarchy of your search results. GoTo.com has taken it a step further. The Pasadena-based search engine has designed an ad-free site that makes money by taking bids for “priority search-result placement,” as Dow Jones Interactive (Feb. 18, 1999) puts it. In other words, companies pay a premium to be listed at the top of the search-results page—whether or not the link is relevant to the search.

The impact of this practice cannot be underestimated, says Andrew Shapiro, director of the Aspen Institute Internet Policy Project and author of The Control Revolution (PublicAffairs, June 1999). “Studies have demonstrated the existence of a phenomenon called ‘screen bias’ where users—not surprisingly—are most likely to choose the information options that are presented to them first. Given the amount of data smog we’re exposed to, it would be strange if we didn’t choose among the first few options that are presented to us.”

There is nothing inherently wrong in paying for premium placement, says Lisa Allen, noting that it’s a routine practice in both supermarkets and bookstores. Still, selling keyword searches is more insidious, because the majority of users don’t understand that they are receiving information that has commercial goals. “When you open a magazine,” she says, “information that is provided by advertisers is clearly marked ‘advertorial,’ which gives readers a heads-up on how to evaluate the content. When users rely on a piece of information and think it’s a critical judgment when it’s a piece of puffery, they get burned.”

The real problem, says Shapiro, is that the Internet industry has yet to develop guidelines on how to distinguish advertising from editorial content. “Norms haven’t developed online the way they have in other media, where the church-state divide between ads and editorial is established enough that we notice when it is breached.” And until such guidelines are accepted, the Internet will continue to slip “dangerously close to an environment that resembles the oligopoly of traditional electronic media. The dream of a media world in which ‘everyone is a publisher’ may well go unfulfilled unless there is a way to preserve some space for the voices of small commercial outlets, nonprofits, and individuals.”