Google's Lawsuit Yields Deep Search Insights

Photograph of a judge's gavel with the scales of justice


by Paolo Vidali
Search Engine Land has been releasing a series of articles covering the Google antitrust trial over the last few months, gleaning information that Google has been forced to enter into the public record via depositions and cross examinations. What they have found by pouring over transcripts has produced some fascinating information about how organic search actually works.

Among these, some highlights include:
  • There are hundreds of separate algorithms that run to build every Google search result, some sequential and others simultaneous
  • After those have run and parsed the data, there are around 100 ranking factors which include the page text, quality, topicality, reliability, locale and click history
  • The user's device (mobile or desktop) affects the actual list of results they may receive, with separate "slices" of page sets based on device
  • The concepts of "Information Satisfaction" and freshness come into play repeatedly, being parsed across dozens of different algorithms in different ways and weights
  • Google records and includes click, swipe, hover and scroll data into determining what's sticky for users

While it's never possible to have a true open book with a closed system, reports like this help us understand how Google looks at page content and gives us context on how to shape our own strategies to maximize potential returns.

Read the full article.