Google Passage Based Update.

[Barry]

I was wondering what thoughts were on this new “passage based” ranking update, being rolled out next month.

What I found concerning is that Google seems to be moving away from using exact keywords in title and heading tags, as seen with the Bert update, and even more so with this new passage based update.

Here are some points from the article that stood out:

“Google’s systems will consider the content and meaning of passages when determining what is most relevant versus previously we were largely looking at the page overall”

“often things like page title were very strong signals that helped us provide the best overall pages. Now Google can find that “needle in a haystack” and surface the most relevant result based on information within passages.”

The advantage we currently have over powerful big brand authority sites, is that we can optimise an entire page to increase relevancy, to outrank them.

Here’s my fear:

Let’s say you rank for “best laptop”.

Your entire page - title, headings, alts, body, etc, is optimised for that keyword, and you’re able to outrank strong authority sites that have massive link power, because of this page optimisation.

Google seems to be saying they are trying to bypass these strong signals and rank “passages” within content, which would then give the upper hand back to these big authority sites with heavy link power.

This update may also make tools like surfer SEO and pop redundant, since google won’t be looking or counting a hierarchy of keywords on a page - instead, they will rank according to page / domain authority and just a written mention answering the keyword query on the page.

A high DR site will only need to mention your keyword query in a “passage” to outrank your fully optimised page - or maybe I’m just being paranoid.

https://searchengineland.com/how-google-indexes-passages-of-a-page-and-what-it-means-for-seos-342215

 

[secretagentdad]

Bet you 20$ it makes the serps worse not better.

I've been working on how to do a low effort search index for a while now.
If you do anything to devalue page titles your relevancy goes to shit fast and results start to suck.

 

[bernard]

So this is basically like implementing Featured Snippets for all searches?

Could go either way. I could see this having some benefits for those who do proper contextual keyword analysis, while a negative for SEO optimized, but contextually weak content mills.

I would probably be worried if I was a question/answer, low hanging fruit type of site.

 

[Philip J. Fry]

@Barry Calm down. It only affects 7% of searches. It also is only for super long tail keywords where the information the searcher is looking for does not have a whole page devoted to it but only a paragraph within a larger page.

@secretagentdad I disagree. This is only for ultra long-tail queries. I'm guessing it's for queries where, at the current moment, none of the results directly answer the query since the query is so specific. It'll be queries that have too low volume for a keyword research tool to target. It's good IMO as more and more users are searching via mobile with natural language.

@bernard https://searchengineland.com/how-google-indexes-passages-of-a-page-and-what-it-means-for-seos-342215 says no. The featured snippet algorithm find the featured snippet from within the top 20 or so ranking pages. Then the one with the highest confidence score becomes the featured snippet. This algorithm considers passages within all pages in Google's index and if Google deems that a certain passage is relevant to a query, even if the page itself is not, it'll rank that page higher for that query.

So, let's say there's a query for "How many feathers does a penguin have at 10 days old?" There's probably no URL about that topic specifically anywhere on the Internet. However, there might be a page about Baby Penguin Development, which contains a section about the second week development, which contains a passage about feathers. That page might get a +20 point bonus in ranking for this query. Other pages with similar topics might get additional points too. Before, the page itself, Baby Penguin Development, wouldn't rank at all for the query. That'll prevent it from being ranked in the top 20 for the query, which would give it 0 chance of being the featured snippet. The two algorithms are totally different.

On another note, the reading comprehension of BuSo is at an all time low. Wow.

 

[alwicz]

@Philip J. Fry So you think that Google will still rank more relevant articles to these long-tail queries regardless of the website authority, etc.?

Let's say we have the keyword you mentioned: "How many feathers does a penguin have at 10 days old?"

You have a website with the article "Baby Penguin Development" with the passage about feathers, and this website has high authority.

Can you still outrank this website when you have an article about this topic, but your website has lower authority?

So your title, H1, URL slug, is "How many feathers does a penguin have at 10 days old?" and your content is way more related to this keyword.

It makes sense that you should be able to outrank that behemoth website, but I am a little bit worried about that.

 

[DanSkippy]

@Philip J. Fry is convinced that this algorithm considers passages within all pages in Google's index and if Google deems that a certain passage is relevant to a query, even if the page itself is not, it'll rank that page higher for that query.

That's my TL/DR.

 

[Philip J. Fry]

@Philip J. Fry is convinced that this algorithm considers passages within all pages in Google's index and if Google deems that a certain passage is relevant to a query, even if the page itself is not, it'll rank that page higher for that query.

That's my TL/DR.

I'm not convinced. I haven't seen this algorithm yet. It's what the article and Google's statements say.

 

[Ryuzaki]

In my opinion, this is entirely about their stupid Link to Text Fragment crap.

They want it so they can attach a something like #text=This%20Header when indexing your content. They'll break it up by header, most likely, and then link straight to those sections of the site instead of simply to the page.

You already see this in Youtube SERPs where they single out a 10% length of the video instead of sending you to the full video.

The impact of this is that they'll ruin the impact of copywriting, above-the-fold optimization, drop RPMs of display advertising, etc.

If you think they're going to stop at 7% of searches, you've not been in the game long enough. It always starts with "some tiny amount of searches you don't have to worry about" before it becomes 50% or more and a fundamental part of the algorithm / index.

I hope that all the data they collect with this 7% test shows them that people hate it and reject it wholesale.

Because once they index every little "passage" and link to it with "text fragments" they're one step away from only showing featured snippet boxes, not only for every query, but for every result in each query. Imagine the 10 organic results all being featured snippet boxes and you'll see where this is going without government intervention (not that I think that's the solution).

Google doesn't want to be a search engine any more. It wants to be an answer engine, and they want us to provide the answers while receiving none of the traffic.

 

[DanSkippy]

Google doesn't want to be a search engine any more. It wants to be an answer engine, and they want us to provide the answers while receiving none of the traffic.

Guess by then BuSo will need a separte section on Data lakes and Data warehouses...

What I also found interesting was Prabhakar Raghavan et al. yapping on about privacy. Most of Alphabet's users don't even realise that this is a winner-take-all market, nothing nice about it.

 

[Philip J. Fry]

In my opinion, this is entirely about their stupid Link to Text Fragment crap.

They want it so they can attach a something like #text=This%20Header when indexing your content. They'll break it up by header, most likely, and then link straight to those sections of the site instead of simply to the page.

You already see this in Youtube SERPs where they single out a 10% length of the video instead of sending you to the full video.

The impact of this is that they'll ruin the impact of copywriting, above-the-fold optimization, drop RPMs of display advertising, etc.

If you think they're going to stop at 7% of searches, you've not been in the game long enough. It always starts with "some tiny amount of searches you don't have to worry about" before it becomes 50% or more and a fundamental part of the algorithm / index.

I hope that all the data they collect with this 7% test shows them that people hate it and reject it wholesale.

Because once they index every little "passage" and link to it with "text fragments" they're one step away from only showing featured snippet boxes, not only for every query, but for every result in each query. Imagine the 10 organic results all being featured snippet boxes and you'll see where this is going without government intervention (not that I think that's the solution).

Google doesn't want to be a search engine any more. It wants to be an answer engine, and they want us to provide the answers while receiving none of the traffic.

To me, it appears that this is only for queries where there's not pages devoted to a topic. It'll make more sense, if there's in-depth pages devoted to a topic, that the pages would be displayed in the SERP instead of pages that only contain a paragraph or two about the topic.

This'll be reserved for the long tail IMO.

As for the general cynicism against Google, I'm not worried. I think that's more of an issue with the grey hat mentality the members here have. This was a form born from WF after all.

Most of Alphabet's users don't even realise that this is a winner-take-all market, nothing nice about it.

Tech is a winner's take all market but, just to play devil's advocate, it's also open to disruption. It's the nature of startup culture. MySpace was a dominant social network and quickly died within a few years. A series of bad decisions could spell the end of a tech giant just as well. Hence why they want to hire the best and get the most capital as possible.

IMO it's better for the consumer as it makes Information Retrieval much easier. It's way easier than going to a library and asking a librarian where some books are.

 

[DanSkippy]

Tech is a winner's take all market but, just to play devil's advocate, it's also open to disruption. It's the nature of startup culture.

Agree. Big leaps are my specialty.

But in all honesty; I guess my biggest flaw is a love for a person like Paul Otlet, in large part reintroduced to the American public via the great work of W. Boyd Rayward. This is also an automatic disqualification when it comes down to possessing a certain level of ruthlessness needed.

Having said that, it saddens me that you do not hear (more) from a person like Luciano Floridi. To be more precise; At least not in public and at a level that is understandable for the average consumer. It is my believe that sometimes the King has to speak to the peasants at their level. Here again I am letting my personal convictions get the better of me.

 

[Barry]

7% sounds small, but not so small when you see that Google processes billions of searches per day.

Remember the Penguin update affected 3%, and that caused carnage. And as Ryuzaki states, these things grow over time.

In the Google video, they discuss “specific searches” – long tails.

But she then gives examples of “broad searches”.

The search query example is “home exercise equipment” and the search results returned pages like:

• Best gym equipment
• Home exercise equipment
• Home gym equipment 2020
• Home gym guide
• Best workout gear
• Best exercise equipment

Now, she says that these results are helpful, but she wants to see a “wider range of results”.

The new results after applying the update are:

• Home exercise equipment
• Small space equipment
• Premium workout gear
• Gym guide for small space
• Affordable exercise equipment
• Small space equipment 2020
• Exercise equipment – premium

What she seems to be saying, is that she wants to see a variety of sub-topics displayed in the search results around her broad search, without having to search for each of them individually.

See how the first results all return pages optimised for the search query “home exercise equipment”. All of the words in the titles are either exact match to the searched keyword, or synonyms of the keyword - gear and equipment, exercise and workout.

Post update - the results include titles with the words small, affordable, premium – none of which are in the search query.

Post update – they are returning pages that are not specifically optimised for the target query, but are subtopics that the searcher did not ask for.

Like I said, it feels like they are trying to move away from ranking 10 organic results, all optimised for a target keyword. They want “variety” and as Ryuzaki mentions, they want to rank and link to “text fragments” on a page.

And Ryuzaki is spot on with his observation that Google is becoming an Answer Engine, instead of a Search Engine. Google started as a way for us "search the web" – to find websites - by simply displaying a list of links. But now, their search results are thick, content rich pages at our expense.

It all stared back when they increased the size of meta descriptions, then they started re-writing meta descriptions and titles, then rich and featured snippets – gradually displaying more and more of our content in their results. It’s all about reducing the click on organic – give the answer on the results pages, and get more clicks on the ads - but that's business.

Anyway, we can only speculate about the pending update – but I don’t like the look of where things seem to be heading.

 

[Philip J. Fry]

7% sounds small, but not so small when you see that Google processes billions of searches per day.

Remember the Penguin update affected 3%, and that caused carnage. And as Ryuzaki states, these things grow over time.

In the Google video, they discuss “specific searches” – long tails.

But she then gives examples of “broad searches”.

The search query example is “home exercise equipment” and the search results returned pages like:

• Best gym equipment
• Home exercise equipment
• Home gym equipment 2020
• Home gym guide
• Best workout gear
• Best exercise equipment

Now, she says that these results are helpful, but she wants to see a “wider range of results”.

The new results after applying the update are:

• Home exercise equipment
• Small space equipment
• Premium workout gear
• Gym guide for small space
• Affordable exercise equipment
• Small space equipment 2020
• Exercise equipment – premium

What she seems to be saying, is that she wants to see a variety of sub-topics displayed in the search results around her broad search, without having to search for each of them individually.

See how the first results all return pages optimised for the search query “home exercise equipment”. All of the words in the titles are either exact match to the searched keyword, or synonyms of the keyword - gear and equipment, exercise and workout.

Post update - the results include titles with the words small, affordable, premium – none of which are in the search query.

Post update – they are returning pages that are not specifically optimised for the target query, but are subtopics that the searcher did not ask for.

Like I said, it feels like they are trying to move away from ranking 10 organic results, all optimised for a target keyword. They want “variety” and as Ryuzaki mentions, they want to rank and link to “text fragments” on a page.

And Ryuzaki is spot on with his observation that Google is becoming an Answer Engine, instead of a Search Engine. Google started as a way for us "search the web" – to find websites - by simply displaying a list of links. But now, their search results are thick, content rich pages at our expense.

It all stared back when they increased the size of meta descriptions, then they started re-writing meta descriptions and titles, then rich and featured snippets – gradually displaying more and more of our content in their results. It’s all about reducing the click on organic – give the answer on the results pages, and get more clicks on the ads - but that's business.

Anyway, we can only speculate about the pending update – but I don’t like the look of where things seem to be heading.

The video covers multiple new AI Algorithms:
16:44 - new spelling update
17:18 - specific search - the passage update
19:20 - broad searches

You're talking about a different algorithm, the one that broadens results for head terms, in a thread that's about the passage update

 

[Ryuzaki]

You're talking about a different algorithm, the one that broadens results for head terms, in a thread that's about the passage update

Perhaps he is, but at the same time the "Passages" one will feed into that.

Many don't realize, but Google has been "reserving" slots in the SERPs for "sub-intents" already.

An example could be if you searched "Internet Marketing". 3 spots might be reserved for informational posts about "what is internet marketing?". 3 spots might be reserved for big agencies that offer affiliate marketing as a service. 1 spot might be for "the history of internet marketing". Then you've got your additional sections like a spot for a video carousel. These spots are mixed and matched and shuffled constantly, but the sub-intents stay the same.

The shorter the tail, the more likely this is to happen because there's a lot of vague sub-intents built into those queries. People will refine their queries as they learn more about it, but at the start, everyone wants something a little different and Google can measure that by CTR and then serve all the needs (by segmenting the SERPs).

This type of internal SERP diversification is already happening, and these Passages will make it easier for them to bust up entire pages into different sub-intents. This could be a good thing for those of us who create monster posts. Could be bad.

 

[secretagentdad]

Really like the theory of this update is about linking to the relevant subsections of content.
Hope that happens actually. I never really used meta descriptions in the first place so what evs.....
Gonna have to get more serious about using some schema and shit.