Google Continues Improving Search AI
In order to keep current with the SEO world, it is very important that you keep a close watch on Google’s plans. Last month John Giannandrea, who is the vice-president for search at Google, spoke about the increasing role that artificial intelligence was playing in search engine optimization. We have gained some very valuable insights from this interview as to what is on the horizon for the world of SEO.
In fewer than 20 years, Google went from processing approximately 10,000 searches on a daily basis to 40,000 per second (or about 3.4 billion per day). Both the Internet and Google have changed significantly over the past 18 years; there were about 2 million websites when Google first started, and there are more than 860 million today. Even more shocking than that, there are now more than 3 million Internet users – or nearly half of the world’s population that are online now. A majority of them use Google on a daily basis.
For Google, all of this has caused huge challenges for the search engine giant. It can not only rely on a search engineer team to adjust algorithms and monitor trends manually in order to provide search engine results that are the most relevant. The solution to this problem is a combination of machine learning and artificial intelligence.
Six years ago, before John Giannandrea joined Google, he was CEO for Metaweb Technologies Inc. That company was acquired by Google. He now leads Google’s Computer Science Research and Machine Intelligence that is at the heart of driving change for Google.
Google has made advances within many of the key aspects of AI already, like translation, natural language processing as well as image and voice recognition. However, it is now looking to use AI for improving how it ranks websites.
Giannandrea discussed what it means for both Google as well as its customers ultimately. He said the machine learning helps them improve the products they offer and makes things possible that only a couple of years ago were impossible. It is now possible to use your smartphone to talk to Google’s app. Deep learning has improved speech recognition, while natural language process (NLP) assists with understanding what someone means. Gamil is now able to block more than 99.9% of spam through the use of neutral networks, including some spam that has never been seen before.
Google being able to block email spam also is connected with being able to properly identify web spam. Already we have seen that algorithms like Penguin and Panda have been able to significantly reduce how much web (search) spam there is on the Internet. However, one thing that Google has been quite cagey about so far is in revealing how much of it is due to manual programming and how much machine learning governs.
Over the past few years, one of the largest changes at Google has been the Knowledge Graph’s increasing importance, which was the reason why Metaweb Technologies was purchased by Google.
Google Search will definitely continue focusing attention on either one or both of these two core aspects: improving the breadth and quality of instant results, or the Knowledge Graph, and the development of more reliable and faster machine learning algorithms for eliminating web spam as well as helping to rank the best websites and businesses at the very top of the search engine results pages (SERPS).
Google has definitely made impressive progress with its algorithms with the likes of RankBrain and Humming Bird. Giannandrea said that Google has gotten a bit smarter and doesn’t rely on only matching keywords any longer. Over just the past year alone we have seen plenty of evidence of this – many old SEO techniques such as SEO services that in the past used to help vault companies up to the top of the SERPS for their most important keywords are not working as well as in the past. This is where reputable SEO services in Houston can come into play.
How Google Has Been Changed By RankBrain
How Google Has Been Changed By RankBrainGiannandrea offered some critical insights on RankBrain as well, stating that it has helped Google better understand complex or long questions. And it is tied to real-world people, things and places by the Knowledge Graph.
Within the SEO world, it is a fairly gray area still. We just know that it exists and that it is massive, but what exactly is Google Rankbrain? No one really knows how it works or what it does! That poses one of the great challenges faced by SEO in many years and is no doubt something that Google is deliberately doing. Google doesn’t like SEO much – it wants to completely control its search index, and doesn’t like having any reliance on SEO professionals to assist businesses with ranking well.
Allo is the most recent update. It isn’t directly linked to search currently. However, since Giannandrea was discussing it, we can be confident that the data collected from the Allo app (which is a chat app) will be used for improving Google’s artificial intelligence even further. Giannandrea said that as they are developing their Google Assistant, that they will be using all of the experience they have from building Search, and that it will be extended to all types of other things that need to be done on a regular basis. And that they were still in the early phase of the journey.
So it appears that the future of search includes more machine learning, more artificial intelligence, and more user data being harvested to feed into the systems. SEO is becoming increasingly more challenging. However, ultimately, if best SEO practices are followed, then a website that is well-designed and has search optimized quality content, will be able to perform quite well.
Let our team at SEO Houston Pros keep you up-to-date on all algorithm updates from Google and better your online exposure within Google’s search engine. By calling us or contacting us today you can rest assured that you are provided with ahead of the curve SEO services in Houston. One of our architects is waiting to hear from you, don’t wait any longer!
If all other things are equal with 2 different pages, site-wide signals can benefit specific pages. For instance, if there are two articles on the same topic where one is in the Wall Street Journal while the other is on an unknown domain. Assuming that no other information is provided, the Wall Street Journal article looks better considering the information currently available. This is an instance where Google is propagating information from the domain level to the page level.