Google is the biggest, fastest and most advanced search engine in the world. It also has the most complex algorithm, enabling it to return search results that usually match closely with the user’s query in a very short space of time.
Bearing in mind that there are literally billions of websites on the internet, all interlinked in one way or another, it is a very complicated process to determine which sites will be the most relevant to the searcher and the order in which they should be displayed. Google has its own term for this – PageRank.
How It Works
Google’s overall methodology for ranking involves three separate processes: examining web pages, indexing them through SEO and eventually putting them into its SERPs (Search Engine Result Pages) in order of relevance.
The first part of the process is where individual web pages are ‘crawled’ by the ‘Googlebot’, which is assigned to ‘fetch’ information from web pages that have been updated or to find completely new ones to add to its index.
Google’s algorithm determines how often the process is carried out, on which pages and how many pages in total should be fetched for indexing. An important part of Googlebot’s work is noting new hyperlinks, as well as those which are dead or point towards bad territory, such as known spamming sites.
This part of the process is the one which takes the longest, as there is a huge amount of data for Google to evaluate based on Googlebot’s findings. In very simple terms, the index is compiled using all of the textual content Googlebot has identified and its location on an individual web page.
Textual content also applies to the HTML source code, hence the importance of accurate title tags, meta descriptions and ALT attributes. It is also worth noting at this point that Googlebot is unable to ‘read’ web pages built using frames, tables or those which are created in Flash.
Google’s primary concern for its search results is relevance. Although overall relevance is determined by a minimum of 200 separate algorithmic factors, PageRank is a big part of the process.
PageRank was named after its developer, Larry Page and it is a link-analysis algorithm. In short, the algorithm is designed to assign a weighting number to each part of a set and place them in order of importance, with part of the algorithm being dedicated to identifying spam links (from known link farms) and discounting them.
In the context of search results, the weighting factor (PR1 to PR10) is relative to incoming hyperlinks. But it is not just the quantity of inbound links that has value, it is the quality of them as well. For example, one link from a relevant website with a high PageRank is worth more to the recipient than several links from a lower-ranked website. Results are eventually determined by several iterations of an extremely complex mathematical equation.
It is the sheer complexity of Google’s algorithmic capabilities which allow it to outshine some other search engines in the efficacy of its search results. Nobody knows for sure just how much importance is placed on any particular function of the algorithm, but most experts agree that PageRank and the ‘value’ of links definitely plays an important part.