Google Announces AI Search Updates – Analysis – Search Engine Journal

Google announced a series of changes to search that will change how sites are ranked. Some changes are live now while others are coming by the end of the year.

Google BERT Now is Nearly 100%

One of the biggest reveals is that Google is employing BERT in virtually every search query.

BERT is a technique for natural language processing pre-training that helps Google understand words within the context of the surrounding words. Google has said that BERT helps Google Search better understand the intent of a search query.

When BERT was announced it was said to be used in 10% of search queries, particularly on longer types of search queries.

Moving forward, BERT will impact nearly ever English language search query.

According to Google:

“Today we’re excited to share that BERT is now used in almost every query in English, helping you get higher quality results for your questions.”

New Spelling Algorithm

Google also announced a spelling algorithm that helps Google to better understand misspelled words. Google said that it’s the biggest improvement to spelling in five years.

What makes the spelling algorithm of interest is that it helps Google understand the context of misspelled words.

Passages Indexed

This is a very important change to Google Search. It may make the most visible change to Google Search Results Pages (SERPs). Google is now indexing passages in a web page, not just the web page itself.

So essentially, passages from a web page can be treated as web pages themselves when returning a search query. Google says that this will impact 7% of search queries.

Screenshot Showing Google Ranking a Passage from a Web Page

Screenshot showing how Google ranks passagesScreenshot of a before and after search results showing how Google ranks passages from a web page instead of a web page.

This update enables Google to surface pages where the answer to a query is deep within the content.

Passages Example Live on Mobile

The example that Google showed was not live on desktop search. But the example is live on mobile Google search and it does show a different search result.

Screenshot of a Passages Algorithm Web Page Result with Highlighted Text

When diagnosing traffic drops, it may be useful to check the differences between mobile and desktop to see if there’s a change in search behavior related to this algorithm.

According to Google:

“We’ve recently made a breakthrough in ranking and are now able to not just index web pages, but individual passages from the pages. 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.

This technology will improve 7 percent of search queries across all languages as we roll it out globally.”

Google says that rather than rank a broader web page about the topic, Google can now rank a specific passage for a search query. That seems to be a huge shift in how web pages are ranked.

This is another area that may dramatically impact how search queries for broad and general search phrases are ranked. Broad search queries, like “home exercise equipment” can mean lots of different things to different people.

It’s like trying to understand what people mean when they search for jaguar: is it the car, the animal the football team?

This update, coming by the end of 2020, will show a greater diversity of web pages for broad search queries.

This is one of those trade-off type changes, where it’s shuffling who the winners and losers are.

For example, Subtopic Ranking is going to make it harder to rank for high traffic broad keyword phrases.

But it will benefit businesses that optimize for specific subtopics and would never have had a chance to rank for the higher traffic subtopic.

So if you’re competing for a broad topic, you may want to make sure your subtopic pages are well optimized.

Here’s how Google explains it:

“We’ve applied neural nets to understand subtopics around an interest, which helps deliver a greater diversity of content when you search for something broad.

As an example, if you search for “home exercise equipment,” we can now understand relevant subtopics, such as budget equipment, premium picks, or small space ideas, and show a wider range of content for you on the search results page. We’ll start rolling this out by the end of this year.”

10% of Searches Will Be Affected by Video Change

This update will affect ten percent of searches. This represents a major change to SEO that underlines the importance of adding video to the mix of content types that publishers produce.

This change is similar to the Passages algorithm described above, only applied to videos. This new technique uses AI to understand the different passages within videos.

Rather than ranking entire videos that are about a topic, Google will analyze videos, assign a tag to each section to describe what it’s about and then send searchers directly to those sections of a video.

This is going to impact the SERPs and maybe video production and planning to make sure the videos are easily understood, section by section.

Here’s how Google described it:

“Using a new AI-driven approach, we’re now able to understand the deep semantics of a video and automatically identify key moments. This lets us tag those moments in the video, so you can navigate them like chapters in a book.

We’ve started testing this technology this year, and by the end of 2020 we expect that 10 percent of searches on Google will use this new technology.”

Data Sets in Search

This may impact sites that depended on ranking for statistical information and sites that sell statistical reports. This change bypasses web pages and shows the statistic directly in the search result as an answer to a question.

It’s kind of a zero click search result but it also offers the ability to discover and research the topic in more depth.

Here’s how Google explains it:

“Sometimes the best search result is a statistic. But often stats are buried in large datasets and not easily comprehensible or accessible online.

Since 2018, we’ve been working on the Data Commons Project, an open knowledge database of statistical data… now we’re making this information more accessible and useful through Google Search.”

Google uses natural language processing to understand if a search query is satisfied by a statistic and then pulls it from the Data Commons to display it as an answer as well as display additional contextual information for deeper topic exploration.

Translate »