Results Ranking

views updated

RESULTS RANKING

Search engines are one of the most popular and widely used tools on the Web. When Internet surfers wish to locate something on the Web, they are more likely to use a search engine than any other method. After a user enters a search, most search engines utilize some formula for returning results that correspond with the user's search. The end result of this process is a results ranking, in which those Web sites that relate to the search are listed in a particular order determined by the search-engine. This issue has become tremendously important for Internet users and online businesses alike, because studies show that users are overwhelmingly more likely to visit those sites that appear early in the results ranking rather than those further down in the list.

Results ranking is the practice of listing search returns in order of their relevance to the search term or terms, with those most closely matching the search listed first. In other words, while a particular search term may result in thousands of site listings, those that most closely match the intentions of the user are featured prominently at the top. Of course, there are different methods of determining what the user may be looking for or find relevant, which amounts to the difference in the results rankings between search engines. According to Online, the practice of relevancy searching (also known as statistical or fuzzy searching) is one of the competitive grounds on which search engines work to distinguish themselves.

The criteria by which relevancy is determined are many. For instance, search engines may determine relevancy by

identifying the number of times the search term appears on a Web page
the location of the terms within the document (for instance, a site on which the term appears in the title would be tanked higher than one on which the term appears buried in a paragraph)
whether all or only some of the search terms are matched
whether an exact phrase or merely all the words in a search phrase are found
the number of times the terms appear relative to the length of the document, and a host of other measures.

Most sophisticated search engines determine relevancy not by any one of these criteria, but by some mixture of several of them. The particular weight given to each individual factor and the relationship between these factors constitute the ranking algorithm. Ranking algorithms, then, are the bread and butter of search-engine firms, and as a result are kept highly secret.

Search results are generally listed with only limited information about those specific documents—usually the Universal Resource Locator (URL), the document title, and a brief summary often pulled from the document text—so users rely heavily on the relevancy listing to determine how useful they might find the page to be. Sometimes results lists feature confidence rankings as well. A confidence ranking is similar to a relevance ranking, but adds a measure of confidence for the relevance of individual pages. For instance, again using a complex algorithm to determine relevance, a confidence ranking may list for each site a percentage of confidence, where a higher percentage indicates a more useful site.

E-commerce firms need to take results ranking into account when they design their Web sites. For instance, companies are more likely to be listed higher in the results ranking for searches for their products and services if the facts of who they are and what they do are clearly and prominently stated on their Web sites. Importantly, however, simply having a Web site isn't always enough to get one's site listed in a results ranking. Most search-engines are built on Web directories compiled by way of individuals visiting sites and deciding if those sites warrant inclusion in the database. Some search engines, such as Yahoo!, offer users a link through which they can apply to have their own site reviewed for possible database inclusion. While other search engines work not via human-compiled databases but via simple search-engine robots that will find and list most any site on the Web, a great deal of effort goes into getting one's site placed near the top of a results ranking.

There exists a tension, however, between the efforts of companies and others to take advantage of search-engine strategies to make themselves more prominent in results rankings on the one hand, and the need for search engines to maintain the integrity of their tools on the other. If companies employ savvy Web optimization firms to devise a working search-engine strategy, search engines need to avoid the appearance of favoritism, or at the least keep their search engines valuable to users as a whole. As a result, ranking algorithms tend to be under constant scrutiny and modification.