Google’s dominance of internet search has been uncontested for more than 12 years now. Before Google, search engines such as AltaVista indexed web pages and allowed for keyword search with an interface and functionality superficially similar to that provided by Google. However, these first-generation search engines provided relatively poor ordering of results. Because an internet search would return pages ranked by the number of times a term appeared on the website, unpopular or irrelevant sites would be just as likely to achieve top rank as popular sites.
PageRank Algorithm
Google’s PageRank algorithm used the number of links to a page to determine relevance. The more links a site had, the more relevant it was in results, and the more its links would determine the relevance of other sites. PageRank resulted in radically more useful search results, and Google search rapidly dominated AltaVista and other search engines.
Variations on the PageRank algorithm are still a significant part of Google’s approach. But over the years, PageRank has been supplemented by collective intelligence algorithms which take into account the links actually selected by web searchers, and the relative popularity of search results. Because of its massive search traffic, this “wisdom of the crowds” approach allows Google to outperform other search engines.
Competitors to Google—notably Bing and Yahoo!—have largely attempted to beat Google at its own game: refining the essential Google search approach rather than attempting to innovate radically new algorithms. And, since Google has access to a greater volume of search traffic for analysis, Google often can overcome these competitors using brute force data crunching.
Facebook's "Graph Search"
Facebook’s recent announcement of “Graph Search” may represent the first significant innovation in web search for which Google lacks a competitive response. Facebook’s graph search rates the relevance of searches using—among other factors—“like” ratings from your friends and the Facebook community at large. For some categories of web searches—most notably those seeking commercial goods and services—the results of such a sort could be superior to those from Google. After all, the number of “likes” accumulated by a restaurant probably speaks better to the quality of the food than the number of links from other pages.
Google has been trying to achieve similar results from the members of its Google+ social network. However, the participation in Google+ is only a fraction of the participation in Facebook, and is heavily weighted toward a highly technical community (aka “geeks”). If the Facebook system generates superior results and gains in popularity, Google may not have much defense.
Of course, the massive brand recognition of Google means that it will be hard for Facebook to replace Google as the default search engine. After all, “Google” is as much the verb that means “to search the internet,” as it is the name of a company. But, if we’ve learned anything from the history of technology, it is that disruptive change can take down even the mightiest incumbent.
Whatever happens with Google, it seems likely that as Facebook rolls out its new search (which was in limited beta at the time of this writing) companies will become strongly motivated to obtain “likes” for their Facebook properties. Not only that, but as “liking” a company or a brand becomes a more effective way of rewarding good service and quality, we may all feel more strongly motivated to hit that like button.
I’m drawn to the idea of allowing search results to draw on the feedback of my friends and the wider community. It certainly beats the current situation, where providing very poor service can actually improve your PageRank by increasing the number of negative reviews that point to your property. If Facebook Graph Search catches on, Google may have its first serious competition in a decade.