Once again, web skeptics are leveling complaints that Google is doctoring search results to favor its own services.
A study by two Harvard professors claims that Google links to its own services three times as often as other search engines do. Google, like other search engines, uses its own algorithm to determine search results, incorporating user-given material like clickthrough rate.
As the country's dominant search engine, Google's longheld stance that their search results are objective and unbiased have come under scrutiny before. Especially given Google's continual expansion into all corners of the Internet, worries that the site is giving preference to its own services (such as Maps and Mail) at the expense of other sites proffering similar materials have only intensified.
According to the study, Google and Yahoo both favor their own services at a disproportionate rate, while Bing remains the most "objective." But some critics doubt the conclusions the professors drew from their data.
One concern: when the searches were performed--in August 2010, Yahoo was not yet powered by Microsoft search engine Bing. Moreover, search results change perpetually, making the use of any historical search result at least somewhat inaccurate to any time but its own.
The sample of search terms was limited--only 32 terms were used (of the near-unlimited number of queries that people are inputting each minute). Additionally, terms which revealed favoritism, like "email," "video" and "RSS reader," all point to Google services which are already the dominant platforms on the web--hardly surprising. Does anyone question that YouTube is the leading video site on the web?
Given the proliferating densification of the Internet, Google's supposed curation may point to a different, increasingly important issue--search engine "noise," a.k.a., too many irrelevant results that cloud the user's ability to find the information they need.
Though it can't be denied favoritism must be a concern when it comes to maintaining fair competition, Google runs on an algorithm that at its core is designed to favor certain sites over others--the relevant over the irrelevant.