Here debuts the Google Passage Ranking: what it is and how it works
It was one of the most amazing ads of the last edition of Google Search On and it immediately aroused the interest of the community for its potential scope: after about four months, officially debuted the Google Passage Ranking, the new algorithmic system thanks to which the search engine becomes able to locate and classify in SERP specific passages within the text that are significant for the search intent of the user.
What is the Google Passage Ranking
Historically, Google Search used to evaluate “all the content of a web page to determine whether it is relevant to a query”; however, in order to offer more and more accurate answers to users, It has become necessary to study a solution to find valid information that could “be buried deep in a web page”.
In practice, thanks to the developments of machine learning and artificial intelligence systems, Google algorithms have now become able to better understand every text published on the web and extrapolate even the parts significant to the query of the user, which are therefore independently ranked.
The Google Passage Ranking (“official” name given to the system instead of the previous Google Passage Indexing, because in fact it concerns the classification and not the indexing of pages) It is useful to identify content elements with a particularly informative value for the user within large text pieces (such as long form classics) and covering multiple topics.
A new opportunity for long form content
Previously, getting out of focus in a content could decrease the possibility of a page being displayed in the Search, even when it contained the correct answer for a specific query.
Now, instead, with the Google Passage Ranking the search engine algorithmically sifts the long pieces of content and is able to understand that a specific step or a series of steps concerns the X query, while another specific step or series of steps in the same document may concern the Y query.
Therefore, “Google is able to classify the same content but several steps within that content separately and for different queries”, notes Barry Schwartz: it is not “a separate indexing, but a separate classification”.
The effects of Google Passage Ranking
Starting from February 10th, Google announced the release of the Passage Ranking for US queries in English, anticipating that “it will be available for more countries in English in the near future and later also for other countries and languages“.
According to the first news, Google predicts that this update will have a significant impact, improving about 7% of search queries – which is why it is one of the most anticipated and feared Google updates.
Basically, it will not be possible to “see” the change in the SERP, because the positioning of the passages does not produce results with a different aesthetic look, but a classification in all similar to that of the other snippet of the search results.
What changes with this update
Roger Montti repeated some Martin Splitt‘s (Google’s Developer Advocate) considerations in presenting the Passage Ranking, beginning to clarify which types of content will benefit from the novelty.
That is, long forms that deal with several topics within the same text, which thanks to the new method of positioning will be able to appear in the SERP also with a small section of the article, or sites that have no particular care in SEO optimization and do not use headings or a structuring organized in sections.
In the example provided by Google – purely indicative, at least from a design point of view – the query “how can I determine whether the windows in my house are UV glass” now provides an immediate response rather than a more general result.
How to manage content: what to do for passage ranking?
Google’s intent is to “help those who are not necessarily familiar with SEO or know how to structure their content or content strategy” to have better results; it has been noted that “many people end up creating wordy pages that have difficulty classifying themselves for anything, because everything is diluted in this long content”.
Google then decided to improve more granular understanding of the content of a page and developed a method to “score different parts of a page independently“: therefore, the Passage Ranking is a purely internal change that does not require any intervention to publishers or creators of content, because you do not have to “make changes to your site, to any of the pages, to any article or markup”.
At the same time, you can’t “ask” to exclude a page only from the placement of steps because, ultimately, it’s just “another blue link” like the classic Google ones.
Ultimately, the Passage Ranking means that Google gives more chance to pages with long content and that they deal with various topics to compete for more queries: that is, for any type of content and enough “a little focus on semantics and a structure in the content, so that it is easier for automated systems to understand the structure and parts of the content” and determine that “this part of the page is relevant for this query, while the other part of the page is not so relevant for this query”.
An example of how Passage Ranking works
Martin Splitt also tries to simplify the way Google systems analyze content with the new algorithmic update.
Let’s take the case of a huge page – for some reason or because the author didn’t really know what he was doing – that talks about lots of vegetables, covering five different topics of which one is a type of tomato, while the rest of the page talks about cucumbers and gardening in general.
Historically, Google would have identified “vegetables” as the main topic of the page, but would probably not consider it for more specific queries, privileging pages that dealt more in detail with the individual vegetable, for example, instead of this that “is not very focused on what the user asks”.
However, overcoming the previous limitation of language comprehension, Google is now able to understand that the single passage concerning tomatoes on that all-inclusive page is relevant as a response to the user who asks about this vegetable, and then classifies in that SERP the page “about all vegetables”.
This is what the feature means to “classify different things in a more granular way“: it is not a different way of indexing pages nor is it necessary to do something specific about it, but only create quality content that “are relevant and suitable for the queries for which you want to be classified”, even if that portion represents only 10% of the entire content.
The difference between passage ranking and featured snippet
Another important information provided by the Googler concerns the difference between passages and featured snippets, which initially (also thanks to the image shown above) seemed to have common traits.
Actually, Splitt explained that these are two completely separate systems: the featured snippets take an “autonomous” response that is located on a page and then inserts it in a context of “instant answer” type, or a search query that does not need an in-depth answer but that can be answered with a few sentences or less. Generally, the user does not need to click on the link in the featured snippet to access the site, because it “content” with the information obtained directly in the search results.
In contrast, the passage ranking feature identifies new pages with “worthy” content to enter the SERP as classic blue links, thus increasing the competition for certain types of queries.
