How Does Google Work?

While no one outside of Google can claim to know exactly how the Google search engine arrives at the list of results you see when you type something into the search field, the general process (minus the crucial percentages and details) goes something like this…

First web content is uploaded to a server. Be it a large or small business redesigning their website, someone updating their blog or their Twitter, or a news site adding a story. When the content is there, Google’s bots discover it. How fast they get there will generally depend on how important they feel the website is, or how many links from valued sources are pointing at this new page. Google gets to new content by following links. If there are no backlinks pointing at a site, it will get crawled less frequently. If there is a robots.txt at a domain instructing the bots to not crawl it, it won’t get crawled at all. An XML Sitemap at a domain can let bots know about crawlable content that has no links pointing at it.

Googlebot crawls the new page and adds it to the Google database. It is this database that searches at Google are run on. While it is more up to date than ever before, the Google database is not a 100% currently accurate representation of the data it has crawled. It is not “live”.

Here in the database, Google makes some decisions about a page. Based on links – how much of an authority is its domain? How popular is the page itself? Is the page or domain breaking any of Google’s policies? Does it contain duplicate content? Has it been reported by users? Based on these decisions, a PageRank score is assigned and any penalties coming would be applied now.

Now the user types their search query into Google. Before they’re finished, Google may offer some popular search suggestions based on what has been typed so far. The user clicks “search” and Google gets to work.

First it looks for synonyms and spelling variations/corrections for the query to include, then creates its initial set of results. Now Google sorts this initial batch of results by PageRank, moving its choices for most “authoritative” higher up. If a website in the results set is linked to by many of the other sites within the set, it may be moved up higher in the results, regardless of its current PR. If localization is possible for the search, local results will be moved up. Google searches for duplicate content, aiming to remove it from the results set. If Google thinks other types of results apply to the search, it may add in video results, additional local results from Google Maps, product results from Google Shopping, or news results. If Google thinks the search relates to something that is currently “blowing up” or trending, as it’s detected a larger than usual volume of searches for the term very recently, it may opt to add in more “fresh” content from news sites and blogs. If the user is logged in, websites they have clicked through to from Google before may end up being featured in results.

All of this is searched, calculated and sorted, and then the organic search results are displayed.