Taking a Look at Google's SEO Search Algorithms
Barry Schwartz over at Search Engine Land wrote a great article about Google’s RankBrain, and how it uses machine learning to improve Google search results. Incredibly important to any SEO Toronto campaign, understanding what is at the core of Google’s search algorithms will ensure that websites rank for particular search phrases. Gary Illyes of Google better explains the tools Google uses, and the interview with him has been transcribed in Search Engine Land’s article. But before we discuss the interview, we must first understand what exactly RankBrain is.
RankBrain itself is an artificial intelligence that has been designed to better understand the human search experience and interpret human input. It receives inputs and information from the millions of searches done daily, along with the websites and navigation behavior done by all Google’s users. Over time RankBrain collects, analyzes and acts upon this information by changing the websites displayed when searches are done, interpreting what the searcher is really looking for, and then refining and providing more relevant, higher quality search results. RankBrain has alluded SEO professionals as it is something Google has never made firm statements on. We understand that it is a learning program, but to what extent it applies what it learns, and to what extent it interprets information it is given has always been up for debate. We have some ideas, but no conclusive authority on the program.
With that explanation out of the way, let’s take a look at the interview...
First off, the most interesting piece of information is that Google is using over 200 signals for their ranking processes. Along with that, there are ranking systems that are machine learning such as RankBrain, and those that are not, such as the Penguin algorithm. For SEO in Toronto, this means that things like location, age, and search history all play a part. What isn’t a learning program would be the algorithm updates that Google has been using, like Penguin, where it has pre-determined parameters that affect search results. I would infer that these updates take into account the findings from RankBrain, along with the other many many signals Google has received, to create the parameters that affect Google’s search results. People and working with the information gathered, and even though machine learning is very effective, it still has to be filtered through human analysis.
Overall, the article offered some insight into the cohesion of Google’s practices, programs and algorithms. At Vestra Inet we’re constantly keeping up to date with the latest SEO techniques and news to provide our SEO Toronto clients with the best SEO services in the GTA.