The rise and advances of new automation and artificial intelligence software algorithms, including those that learn predictive models from historical data, are making increasingly consequential predictive decisions. Such decisions are having an impact on the lives of people in domains as diverse as digital media, advertising,finance, credit, employment, education and criminal sentencing.
Marketers and companies need to understand that we are entering into a new generation of optimizing websites for search engines. While many principles and tactics stay the same in search engine optimization, we cannot deny the current path of technological breakthroughs, such as the introduction of complex machine learning algorithm systems like Google’s RankBrain.
To the non-search engineer, CTO or data scientist, the concept of RankBrain may seem technical and intimidating, but it is one that chief marketing officers for brands — not just technically savvy search marketers — will have to understand to be competitive in 2017 and beyond.
An intro to Google’s RankBrain
RankBrain is a machine learning artificial narrow intelligence system that was introduced to the public in 2015. Bloomberg was the first news publication in mainstream media to break the news about Google’s newest search ranking factor in an interview with Greg Corrado, a senior research scientist at Google.
RankBrain is designed to better understand the meaning behind the words a person uses and types into his or her search engine because 15 percent of queries per day had never been seen by Google before.This concept is explained in this article detailing their algorithm’s ability to better understand word relationships. You will definitely want to learn more about this technology!
The rise of mobile: A primary driver of RankBrain’s existence?
Voice search has been on the rise for a while now. Mary Meeker, a partner at the VC firm, Kleiner Perkins Caufield Byers, states in this presentation that voiced search is up 7 fold since 2010. RankBrain deals well with the conversational long-tail queries that are common to voice search today. Mobile search behavior has been a game changer when it comes to queries with AI assistants: Siri, Google Now, Cortana as well as those coming from voice assistants such as Alexa and Google Home.
A rundown of what RankBrain actually does
RankBrain was designed to better analyze the language of websites in Google’s index in order to apply that analysis to a particular search query. Because there are a number of core algorithms that exist to parse queries in Google’s search pages, it is RankBrain’s job to learn what mixture of these core algorithms can best be applied to each type of search result. By better understanding the search query, Google’s search has a more precise ability to match users with websites and pages. The purpose is to better understand the meaning of content and the intent behind a search query. Once it understands the intent, it can presumably apply appropriate Google algorithm signals that deserve the most weight to determine what results to show based on the search query.
A search result will bring up websites that include a mixture of hundreds of algorithm ranking factors used to determine the relevance of the page in relation to the search query. Along with the ability to better understand concepts on a web page, RankBrain also allows for a better understanding of the association between multiple queries, such as:
- “Where is the Eiffel Tower?”
- Followed by:
- “How tall is it?”
How does RankBrain actually learn?
During an SMX West Expo conference, speakers Marcus Tober, founder of Searchmetrics, and Eric Enge, CEO of Stone Temple consulting, respectively, shared examples of RankBrain working in action. One of the studies showed how RankBrain better interpreted the relationships between words.
This can include the use of stop words in a search query (“the,” “and,” without,” etc) — words that were historically ignored previously by Google but are sometimes of a major importance to fully understanding the meaning or intent behind a person’s search query. It’s also able to parse patterns between searches that are seemingly unconnected, to understand how those searches are similar to each other.
Let’s say we use the example of the comedy TV series “The Office”. This is an example of a search that would be taken out of context without reading and incorporating the all-important “the” word, if it was indeed typed into to the search box.
The Next Web
Comments are closed.