My thoughts on passage ranking
Google's passage ranking could be a new paradigm shift in organic search - or not. In this post, I elaborate on what it could be and the hints Google gave us about it.
Google's passage ranking could be a new paradigm shift in organic search - or not. In this post, I elaborate on what it could be and the hints Google gave us about it.
What?
Two weeks ago, Google announced "passage ranking" in Search On 2020 (I reported).
At first, people thought it was indexing (because it was written as "indexing"). Then, Barry Schwartz clarified that it's about ranking, not indexing.
And this is as much information as we got from Google. Unlike Amit Singhal, who wrote about building the search engine of the future, or Ben Gomes, who wrote about the next 20 years of search, Raghavan doesn't provide a lot of information about Google search features as head of Search. His Search On 2020 presentation focused much more on reinstating trust in Google.
Why?
Google wants to help people find specific information: "By better understanding the relevancy of specific passages, not just the overall page, we can find that needle-in-a-haystack information you’re looking for."
That seems like a shallow reason to me. I suspect that Google either looks for more ways to display ads or to keep people in search results longer to display more ads.
How?
How this will turn out or look like is still very unclear. I see two options.
Option one. Passage Ranking is basically "Featured Snippets on steroids" and the SERPS carry more Featured Snippets. The top spot gets weaker and the click-curve flatter. The "breakthrough in ranking" Raghavan mentioned is that they applied the Featured Snippet algorithm to sub-topics and all three types of user intent (dominant, common, minor; you heard me ramble about this before).
Option two. Snippets will be redefined and redesigned. Instead of titles, Google displays passage headings. Instead of meta-descriptions, snippets come with a larger passage above the title. Google has been rewriting a lot of descriptions for a while, anyway.
So what's it gonna be? We don't know (yet). But Bill Slawski, kind of the Indiana Jones of patents, looked at a patent that might be used for Featured Snippets. It describes how Google might be able to evaluate passages rather than full web pages and whether they're a good fit for a question.
Two points stick out to me. The first one is the importance of and hierarchy of headings - they provide context to passages. The second one is how Google might use query-independent signals to rank passages. In my mind, that would be links or page speed, but it could also be the content relevance of the overall page, instead of just the relevance of the article.
In other words, similar to Featured Snippets, passage ranking might take more than just the content into account. It also matters how well the site ranks, how good its backlink profile is, and the overall content quality.
In my mind, Google follows a very similar strategy with passage ranking as with "key highlights in video". It's often difficult to find key pieces of information in long-form content and one Featured Snippet might not be enough.
What started with Featured Snippet highlighting (the yellow passages you see when clicking on a Featured Snippet result) will be extended beyond a single result.
It's interesting how Google highlights passages, anyway. As you can see in the screenshot above, Google also highlights a sentence of the next passage. That is natural language understanding! Google "understands" that the next sentence might also be relevant for the query (it was "post mortem"). I think that's the same technology that will apply to passage ranking.
This highlights the importance of semantics in the future of SEO: clear and logical header structure, concisely written passages, high overall content quality, and coverage of related entities.