What Google’s Patents and Acquisitions Can Teach Us About SEO
Posted on Thursday, May 8th, 2014 at 8:46 pm
Matt Cutts, the head of search spam at Google, recently released a video in which he discussed a future where links weigh less into the ranking of a website. To save you three minutes, Google knows that as long as they’re dependent on links for the majority of their algorithm, it will always be easily manipulated. So it would make sense for them to try everything they can to get away from links as the primary factor in ranking.
The First Step Away from Links as a Backbone
Back in 2005, Google filed a patent called Agent Rank for a technique that would allow them to rank a piece of content based on the person who authored it. In theory, when content is added to the web, there would be a digital signature connecting the real life author to the database’s digital author profile. This signature could be unique, and attached to every piece of content that author puts on the web, creating a portfolio of sorts for each author.
An authority score given to different authors, or “Author rank,” could then be used in the future to give weight to new articles and content authored by the author. For instance, the author could launch a brand new website, and because their author rank is high across 20 other websites, Google would associate this author with other great content and potentially give more authority to that website, even though it is new.
Some websites, though, don’t include authorship info, and Google doesn’t have any good way to assign value to these pages outside of links. But that could change with Google’s acquisition of Deep Mind, home to the world’s leading researchers involved with artificial intelligence and deep learning. The term “deep learning” has come around since the mid-2000’s to describe a programming architecture in which it could make connections between different sets of data. It would make sense then, that it deep learning is most effective when it has large quantities of data to sort through and analyse.
How Google Will Come To Know Us Better Than Ourselves
Well, as of Jan 2014, Chrome has a dominating market share at 55% of all internet browsing, Google Analytics is on over 15 million websites, the Ad Network reached over 2 million websites, and Gmail is the leader in web mail, so it’s no secret that the amount of data that Google is able to collect is simply unfathomable. Up until now, sorting through so much data and drawing informed conclusions has been troublesome for computers. From TechCrunch,
World-renowned artificial intelligence expert and Google’s new Director of Engineering, Ray Kurzweil, wants to build a search engine so sophisticated that knows users better than they know themselves. “I envision in some years that the majority of search queries will be answered without you actually asking.”
But now, Deep Mind’s AI program will play a role in all of Google’s infrastructure, including search, advertising, and social. With the end goal to document and draw smart connections between the real world’s people, places, events, and things, we must assume that Google is going to be using their mass repositories of data to create individual user profiles for each of us – including authors and readers.
Big Data + Deep Learning = Personalized Results
Users will see an increased level of relevancy in searches. For instance, if someone has emails in their inbox discussing the purchase of a new Honda Civic, and then they like Honda on Google+, and finally post pictures of their new Honda Civic to G+ with hashtags, then when that user goes to search for “change spark plugs”, Google will tailor the search results to include videos and tutorials specific to the user’s history, which is changing spark plugs on a Honda Civic.
Conversely, authors will see an increase in engagement from their users. Let’s say I’m a mechanic and operate a blog detailing simple maintenance on Hondas and Toyotas (Japanese cars). Google sees that I frequently discuss related topics to car maintenance, and that I mention Japanese name brands, not American. Hopefully the users who are searching for “how to change my car’s oil” and own Fords won’t see my blog, and by the same token, the time users spend on my site will likely increase because the content is more relevant to their lives.
In the future of the internet, where digital and real life become more integrated, it will be important that we associate ourselves and our businesses with others that are considered to be industry leaders, in hopes of being given credit and benefit of the doubt based on association. As such, we should all start building a digital reputation for ourselves now, so that we aren’t behind when the time comes.