Google works by having thousands of servers located around the globe. When a user attempts to connect to Google their servers determine which server to direct them to – based on geographical location and traffic load upon those servers. Once a user is connected their search is routed again to the least loaded web server before the search results are returned back. The size of such a network requires dedicated servers to achieve certain tasks. Examples of these servers are follows:

• Google DNS Servers €“ direct user traffic to the best placed web and data servers so the user experiences a fast response.
• Google Web Servers €“ process user search strings and return search results after contacting spelling and advert servers.
• Advert Servers €“ manage Google Adword advertisements.
• Spelling Servers €“ suggest spelling suggestions for user search strings.
• GoogleBot Servers €“ dedicated to crawling the web and indexing web pages.
• Document Servers €“ used to store copies of web pages when they are indexed. These servers store the cached versions of pages.
• Index Servers €“ links document server records to result pages.

With so many servers around the world it is a huge task to maintain consistent results. For this reason search results can vary by day to day and even hourly as Google updates its index continuously – this is commonly referred to as the ‘Google Dance’.

The search results returned by Google are based on pages they have indexed being passed through Google search algorithms. These algorithms are kept very secretive but the best known algorithm they employ is the patented PageRank (http://en.wikipedia.org/wiki/PageRank). The results having then been filtered and sorted are returned to the end user.

Periodically Google do major changes to their algorithms which can cause well ranking websites to suddenly drop in the search results but on the upside other sites can substantially gain. With ever changing Google algorithms and indexed pages it is difficult to view changes in search engine placement over time. Following known processes called Search Engine Optimisation over time should result in improved search ranking.

Search engines started to appear in 1993 to allow people to find web content matching specific criteria (typically those containing a given word or phrase). They provide results typically as a list of references of those sites that match the search criteria. From around 1994 search engines started to index full text from pages but it wasn’t until 1998 when Google was launched that search engines started to get more intelligent on how they indexed web page content and presented this in search results. Google’s search tried to rank sites and pages based on how many other sites linked to them. Combined with the minimalist user interface and relevant search results they have become the most popular web search in the world.

Google is the top search engine in Argentina, Australia, Belgium, Brazil, Canada, Denmark, France, Germany, India, Italy, Mexico, Spain, Sweden, Switzerland, U.K., U.S. – source Nielsen//Netratings 6/05, based on total number of unique visitors.

The Neilsen//Netratings August ‘US Search share’ rankings state Google has 50.2% of US searches:

Taken from http://www.nielsen-netratings.com/pr/pr_060919.pdf

Search engines work by using automated programs called robots (other names include spiders, crawlers and bots) to traverse the web and follow hyperlinks to new and undiscovered web pages. During this ‘crawl’ they index content from pages and this is used to create the search results. Over time they re-visit pages to see if they have changed and also to check that they are still available and can be accessed.