How 'content farms' beat Google, and what search engines should do about it
Saturday, January 29, 2011; 6:27 PM
Google can give you free long-distance calling and provide driving, walking, transit or bicycling directions to almost anywhere in the world. But can it find information on the Web when you ask?
The Mountain View, Calif., company has made an unusual confession: It's having some trouble with its original and primary task.
As my colleague Michael Rosenwald writes, Google's reputation for uncanny accuracy has been dulled by "content farm" sites that game its search system to boost the visibility of pages many readers say they don't want.
Google search engineer Matt Cutts's Jan. 21 blog post acknowledged their complaints: "we hear the feedback from the web loud and clear: people are asking for even stronger action on content farms and sites that consist primarily of spammy or low-quality content."
It's easiest to see this problem if you search for a review of a product or instructions on how to do something and find yourself looking at dozens of irrelevant results that don't answer your question or that rip off another site's work.
There's not much point in getting too mad at these sites. They're simply following a prime directive of the commercial Web: Get people looking at your site, then use advertising - often placed through Google's services - to transmute that traffic into money.
Meanwhile, plenty of writers, photographers, videographers and editors are willing to accept minimal per-product payments to crank out a large volume of posts that match up with common Google searches.
As one result of this dynamic, Yahoo paid $100 million for a content mill named Associated Content in May. The best-known company in this category, Demand Media, staged its initial public offering Wednesday and closed the day with a higher market value than the New York Times Co.
Both sites would dispute the content-farm categorization, and I've seen each publish useful information. But I've also read plenty of dreck at these sites, and there's no disputing their business model of mass-producing content to fit, Lego-like, with search terms.
In essence, Google has unintentionally been teaching to the test - and some of its students have learned all too well.
As Rosenwald's article suggests, this opens up an opportunity for social-networking sites that connect people with trusted, knowledgable friends to beat Google at its own game. But by requiring your identity to work, they incorporate privacy and security risks.