We review all Google search ranking systems: validity and characteristics


Google is focusing a lot on being more transparent regarding how SEO ranking works within its search engine, which is why it has presented through its blog for SEO a «Guide on Google Search Ranking Systems», with which he hopes that users can understand the systems he uses to obtain search positions and rankings in SERP (Search Engine Results Page).

Although it is true that the information in the guide does not bring much new to those who follow the Google algorithms, it is the first time that all the information is obtained. information in a unified, orderly and very well explained way.

Google Ranking System Updates

First of all, in the Google guide it explains that it uses automated sorting systems They are responsible for analyzing various factors of web pages, using them to present “the most relevant and useful results, all in a fraction of a second.” With this we already have the fundamental point when doing SEO: define what will be the most relevant and useful content, both for clients and readers, as well as their search intentions.

Similarly, Google claims that they are doing tests and adjustments continuously in all its ranking systems, sending notifications in case they could be useful for content creators or other users, that is, they really affect the search results.

These are the active Google search results ranking systems

To begin with, we will discuss the different search ranking systems that are active currently.

BERT (Bidirectional Encoder Representations from Transformers)

Bert, is managed through an AI system that helps Google better understand how the combination of words can express different meanings and search intentions.

Thus, Google is able to contextualize the content, relating them to the search intent of users and ensuring that the content with the best ranking is more relevant to the queries.

If you do a test and change a word, nuance or even a single letter in the search, you will have different results. Google claims that this system is in the process of improvement and it still does not cover all searches.

crisis information systems

Google has developed a system capable of provide useful and timely information in times of crisis, whether personal, natural disasters or generalized crisis situations. We have already seen its use in recent cases such as the Covid health emergency or the war in Ukraine.

Personal crises: Google’s system is able to identify when a person is looking for information about personal situations such as suicide, sexual assault, poison intake, gender violence or drug addiction. In these cases, the search engine offers the user direct telephone lines authorized by the relevant authorities or content from organizations to make inquiries. SOS Alert: in times of natural disasters or general crisis situations (floods, forest fires, earthquakes, hurricanes, etc.) the SOS Alert system displays updates from local, national or international authorities. Updates point to phone numbers, emergency websites, maps, translations of helpful phrases, and donation drives. Deduplication systems

Because Google searches can bring together thousands or millions of web pages with similar content, the search ranking system is responsible for displaying only the most relevant results, avoiding unnecessary duplicate content.

An important novelty is that the deduplication also applies to the featured snippets,“If a web page listing rises to become a featured snippet, we don’t repeat the listing later on on the first page of results.” Thanks to this system, original and useful content for the audience gains a lot of value.

Exact Match Domain

This Google classification system takes into account the domain name words, to assess whether the content is useful for a search. Although this does not mean that if a domain exactly matches the keyword to be ranked, it will appear in the top positions or that it matches the queries.

Google explains that its system can identify websites that have been created with an exact domino and keyword and the weighs with other filter systems.

updated content systems

Google has systems based on “questions that deserve updating”. These are designed to offer more up-to-date content on certain searches. As the blog explains: “If someone is looking for a movie that just came out, they probably want recent reviews rather than older articles from when production started.”

useful content systems

With this useful content system, Google places an emphasis on showing results whose content has been made by people, for people. Instead of offering content that has been over-optimized for SEO and whose sole purpose is to get traffic from search engines.

Link analysis system and PageRank

Google has several systems that can analyze how pages link to each other in order to identify what they are about and determine which are the most useful for a search. PageRank is an example of this type of system, being one of the main ones.

local news systems

These systems identify and display local news sources, as long as they are relevant. It does this through functions like “Featured News” and “Local News”. This applies to virtually all content on the SERPs.

MUM (Multitask Unified Model)

This AI system is capable of understanding and generating language from information from text and images. Although it is currently not used for general ranking in search, it is used in some specific applications such as “improving searches for information about the COVID-19 vaccine and to improve the featured text snippets we display”.

Other active ranking systems

Neural Matching: An AI system that helps Google understand the representations of concepts in queries and pages, to relate them to each other. Original content systems: this guarantees that Google shows first the original content in searches, ahead of those who quote them. This does not mean that information cannot be cited, but the most important thing is to add value to the content by including context, opinions and examples. Duplicate content can be between multiple sites or within a single site, so Google recommends using canonical tags to tell which page is the parent. Removal-based demotion systems: Google has policies that allow certain types of content such as plagiarism, piracy, or offensive content to be removed. Two assumptions are considered. On the one hand legal removals, when they receive a large number of valid takedown requests that affect the entire site and not just the content being removed. And on the other, the deletion of personal content, when they receive requests for the use of personal information without the consent of the users, also linked to virtual harassment content and explicit images. Page experience: this system evaluates aspects that guarantee a better experience for users, such as speed of loading, compatibility with devices, if the pages are free of intrusive interstitials and if they publish safely. This isn’t a ranking factor on its own, but it does give preference to pages with a better experience. Passage classification: identifies one page passages to understand how relevant a page is to a result. It is recommended to use subheadings, paragraphs and a clear structure. Product Review: This system identifies and rewards pages with a review of high quality products. It is better to provide useful information, rather than promotional content. RankBrain: an AI system that allows you to understand how They connect words with concepts. That is, the relevant content is displayed, even if it does not have all the exact words that have been used in the search, but that are related to the concepts that are defined. Reliable information: Google defines it as “multiple systems work in various ways to display the most reliable information possible, how to help display more authoritative pages, and demote low-quality content.” Thus, Google rewards what it calls quality journalism and curbs sensationalist media. For topics with little reliable information, Google displays a content advisory that warns about topics with rapid change or overall low quality of available results. Diversity of the site: with this system, Google prevents display more than two results from the same website in the top positions, so that no one site can dominate the top results. Spam Detection: The internet is filled with large amounts of spam, which also affects search results. They have several systems to detect and eliminate spam, such as Spam Brain, which are constantly updated.

retired systems

There are other search ranking systems that have been retired. Although Google emphasizes that «have been incorporated into successor systems or have become part of our main classification systems.”

Hummingbird: made in 2013, He was the precursor of relating and understanding complex concepts. Mobile-friendly rating systems: was in charge of prioritizing results that looked better on mobile devices. It is now part of the page experience system.
page speed: it was announced in 2018 as “Speed ​​Update”, prioritizing websites with optimal loading speed. It is now part of the page experience system.
Panda: it was the system that began to prioritize original and high-quality content, being a thorn in the side of many sites since 2011. It became a key piece of ranking systems in 2015.
Penguin: designed to go after link spam. It was announced in 2012 and since 2016 it is part of the main classification systems.
Secure systems: was released in 2014, and when everything was equal, sites protected with HTTPS did better in ranking systems. Currently, it is part of the page experience system.

Image: Depositphotos

Stay informed of the most relevant news on our Telegram channel