How Google Panda works is somewhat of a mystery, at least to the typical user. In the most basic terms, Panda is another way to reorder the search results based the website’s content quality. This leaves users with fresh, relevant search results based on their search query.
Back in 2011, Google announced that the Panda algorithm would not just reorder the search results, but also reward sites with high-quality content. With this approach, Google used a type of ranking system to give each eligible page a value. This value then determined where the page would fall within a particular query.
Fast forward to today–and over two dozen Panda updates later–and a patent has been created with the name Navneet Panda. The name was inspired by a Google engineer who basically wrote the book on the Panda update. From the patent, we’ve learned that there appears to be two main parts:
Generation of an initial score or value
Depending on the case, group based modification factors are applied
The first step is creating a quality score for every URL listing that relates to a specific query. When the initial score has been created, it is adjusted based on factors such as the quality of the group that the URL has been assigned to.
In the second step, an initial score has been assigned for each URL, and the scores are modified based on the quality of the group that they are associated with. For all URLs in a particular group, quality score is dependent on a count of reference queries and independent links.
Another aspect is the flow of Panda, which includes navigational and reference queries. It’s believed that navigational queries may have to do with mentioning the brand name, while reference queries are based on time.
Although the details can be interpreted differently, one thing is for certain: The Panda Patent has helped shed light on some of the things that webmasters can do see improved rankings. For instance, Panda URL ranking factors include relevance to the query and quality, while the Panda Group Level ranking factors include count of reference queries and count of independent links.
The Panda Patent is an interesting read, so check it out for more information.