Google’s “Branded Search” Patent For Ranking Search Results via @sejournal, @martinibuster

Back in 2012 Google applied for a patent called “Ranking Search Results” that shows how Google can use branded search queries as a ranking factor. The patent is about using branded search queries and navigational queries as ranking factors, plus a count of independent links. Although this patent is from 2012, it’s possible that it may still play a role in ranking.

The patent was misunderstood by the search marketing community in 2012 and the knowledge contained in it was lost.

What Is The Ranking Search Results Patent About? TL/DR

The patent is explicitly about an invention for ranking search results, that’s why the patent is called “Ranking Search Results.” The patent describes an algorithm that uses to ranking factors to re-rank web pages:

Sorting Factor 1: By number of independent inbound links
This is a count of links that are independent from the site being ranked.

Sorting Factor 2: By number of branded search queries & navigational search queries.
The branded and navigational search queries are called “reference queries” and also are referred to as implied links.

The counts of both factors are used to modify the rankings of the web pages.

Why The Patent Was Misunderstood TL/DR

First, I want to say that in 2012, I didn’t understand how to read patents. I was more interested in research papers and left the patent reading to others. When I say that everyone in the search marketing community misunderstood the patent, I include myself in that group.

The “Ranking Search Results” patent was published in 2012, one year after the release of a content quality update called the Panda Update. The Panda update was named after one of the engineers who worked on it, Navneet Panda. Navneet Panda came up with questions that third party quality raters used to rate web pages. Those ratings were used as a test to see if changes to the algorithm were successful at removing “content farm” content.

Navneet Panda is also a co-author of the “Ranking search results” patent. SEOs saw his name on the patent and immediately assumed that this was the Panda patent.

The reason why that assumption is wrong is because the Panda update is an algorithm that uses a “classifier” to classify web pages by content quality. The “Ranking Search Results” patent is about ranking search results, period. The Ranking Search Results patent is not about content quality nor does it feature a content quality classifier.

Nothing in the “Ranking Search Results” patent relates in any way with the Panda update.

Why This Patent Is Not The Panda Update

In 2009 Google released the Caffeine Update which enabled Google to quickly index fresh content but inadvertently created a loophole that allowed content farms to rank millions of web pages on rarely searched topics.

In an interview with Wired, former Google search engineer Matt Cutts described the content farms like this:

“It was like, “What’s the bare minimum that I can do that’s not spam?” It sort of fell between our respective groups. And then we decided, okay, we’ve got to come together and figure out how to address this.”

Google subsequently responded with the Panda Update, named after a search engineer who worked on the algorithm which was specifically designed to filter out content farm content. Google used third party site quality raters to rate websites and the feedback was used to create a new definition of content quality that was used against content farm content.

Matt Cutts described the process:

“There was an engineer who came up with a rigorous set of questions, everything from. “Do you consider this site to be authoritative? Would it be okay if this was in a magazine? Does this site have excessive ads?” Questions along those lines.

…we actually came up with a classifier to say, okay, IRS or Wikipedia or New York Times is over on this side, and the low-quality sites are over on this side. And you can really see mathematical reasons…”

In simple terms, a classifier is an algorithm within a system that categorizes data. In the context of the Panda Update, the classifier categorizes web pages by content quality.

What’s apparent when reading the “Ranking search results” patent is that it’s clearly not about content quality, it’s about ranking search results.

Meaning Of Express Links And Implied Links

The “Ranking Search Results” patent uses two kinds of links to modify ranked search results:

1. Implied links
2. Express links

Implied links:
The patent uses branded search queries and navigational queries to calculate a ranking score as if the branded/navigational queries are links, calling them implied links. The implied links are used to create a factor for modifying web pages that are relevant (responsive) to search queries.

Express links:
The patent also uses independent inbound links to the web page as a part of another calculation to come up with a factor for modifying web pages that are responsive to a search query.

Both of those kinds of links (implied and independent express link) are used as factors to modify the rankings of a group of web pages.

Understanding what the patent is about is straightforward because the beginning of the patent explains it in relatively easy to understand English.

This section of the patent uses the following jargon:

• A resource is a web page or website.
• A target (target resource) is what is being linked to or referred to.
• A “source resource” is a resource that makes a citation to the “target resource.”
• The word “group” means the group of web pages that are relevant to a search query and are being ranked.

The patent talks about “express links” which are just regular links. It also describes “implied links” which are references within search queries, references to a web page (which is called a “target resource”).

I’m going to add bullet points to the original sentences so that they are easier to understand.

Okay, so this is the first important part:

“Links for the group can include express links, implied links, or both.

An express link, e.g., a hyperlink, is a link that is included in a source resource that a user can follow to navigate to a target resource.

An implied link is a reference to a target resource, e.g., a citation to the target resource, which is included in a source resource but is not an express link to the target resource. Thus, a resource in the group can be the target of an implied link without a user being able to navigate to the resource by following the implied link.”

The second important part uses the same jargon to define what implied links are:

• A resource is a web page or website.
• The site being linked to or referred to is called a “target resource.”
• A “group of resources” means a group of web pages.

This is how the patent explains implied links:

“A query can be classified as referring to a particular resource if the query includes a term that is recognized by the system as referring to the particular resource.

For example, a term that refers to a resource may be all of or a portion of a resource identifier, e.g., the URL, for the resource.

For example, the term “example.com” may be a term that is recognized as referring to the home page of that domain, e.g., the resource whose URL is “http://www.example.com”.

Thus, search queries including the term “example.com” can be classified as referring to that home page.

As another example, if the system has data indicating that the terms “example sf” and “esf” are commonly used by users to refer to the resource whose URL is “http://www.sf.example.com,” queries that contain the terms “example sf” or “esf”, e.g., the queries “example sf news” and “esf restaurant reviews,” can be counted as reference queries for the group that includes the resource whose URL is “http://www.sf.example.com.” “

The above explanation defines “reference queries” as the terms that people use to refer to a specific website. So, for example (my example), if people search using “Walmart” with the keyword Air Conditioner within their search query then the query  “Walmart” + Air Conditioner is counted as a “reference query” to Walmart.com, it’s counted as a citation and an implied link.

The Patent Is Not About “Brand Mentions” On Web Pages

Some SEOs believe that a mention of a brand on a web page is counted by Google as if it’s a link. They have misinterpreted this patent to support the belief that an “implied link” is a brand mention on a web page.

As you can see, the patent does not describe the use of “brand mentions” on web pages. It’s crystal clear that the meaning of “implied links” within the context of this patent is about references to brands within search queries, not on a web page.

It also discusses doing the same thing with navigational queries:

“In addition or in the alternative, a query can be categorized as referring to a particular resource when the query has been determined to be a navigational query to the particular resource. From the user point of view, a navigational query is a query that is submitted in order to get to a single, particular web site or web page of a particular entity. The system can determine whether a query is navigational to a resource by accessing data that identifies queries that are classified as navigational to each of a number of resources.”

The takeaway then is that the parent describes the use of “reference queries” (branded/navigational search queries) as a factor similar to links and that’s why they’re called implied links.

Modification Factor

The algorithm generates a “modification factor” which re-ranks (modifies) the a group of web pages that are relevant to a search query based on the “reference queries” (which are branded search queries) and also using a count of independent inbound links.

This is how the modification (or ranking) is done:

1. A count of inbound links using only “independent” links (links that are not controlled by the site being linked to).
2. A count is made of the reference queries (branded search queries) (which are given a ranking power like a link).

Reminder: “resources” is a reference to web pages and websites.

Here is how the patent explains the part about the ranking:

“The system generates a modification factor for the group of resources from the count of independent links and the count of reference queries… For example, the modification factor can be a ratio of the number of independent links for the group to the number of reference queries for the group.”

What the patent is doing is it is filtering links in order to use links that are not associated with the website and it is also counting how many branded search queries are made for a webpage or website and using that as a ranking factor (modification factor).

In retrospect it was a mistake for some in the SEO industry to use this patent as “proof” for their idea about brand mentions on websites being a ranking factor.

It’s clear that “implied links” are not about brand mentions in web pages as a ranking factor but rather it’s about brand mentions (and URLs & domains) in search queries that can be used as ranking factors.

Why This Patent Is Important

This patent describes a way to use branded search queries as a signal of popularity and relevance for ranking web pages. It’s a good signal because it’s the users themselves saying that a specific website is relevant for specific search queries. It’s a signal that’s hard to manipulate which may make it a clean non-spam signal.

We don’t know if Google uses what’s described in the patent. But it’s easy to understand why it could still be a relevant signal today.

Read The Patent Within The Entire Context

Patents use specific language and it’s easy to misinterpret the words or overlook the meaning of it by focusing on specific sentences. The biggest mistake I see SEOs do is to remove one or two sentences from their context and then use that to say that Google is doing something or other. This is how SEO misinformation begins.

Read my article about How To Read Google Patents to understand how to read them and avoid misinterpreting them. Even if you don’t read patents, knowing the information is helpful because it’ll make it easier to spot misinformation about patents, which there is a lot of right now.

I limited this article to communicating what the “Ranking Search Results” patent is and what the most important points are. There many granular details about different implementations that I don’t cover because they’re not necessary to understanding the overall patent itself.

If you want the granular details, I strongly encourage first reading my article about how to read patents before reading the patent.

Read the patent here:

Ranking search results

Google Confirms 3 Ways To Make Googlebot Crawl More via @sejournal, @martinibuster

Google’s Gary Illyes and Lizzi Sassman discussed three factors that trigger increased Googlebot crawling. While they downplayed the need for constant crawling, they acknowledged there a ways to encourage Googlebot to revisit a website.

1. Impact of High-Quality Content on Crawling Frequency

One of the things they talked about was the quality of a website. A lot of people suffer from the discovered not indexed issue and that’s sometimes caused by certain SEO practices that people have learned and believe are a good practice. I’ve been doing SEO for 25 years and one thing that’s always stayed the same is that industry defined best practices are generally years behind what Google is doing. Yet, it’s hard to see what’s wrong if a person is convinced that they’re doing everything right.

Gary Illyes shared a reason for an elevated crawl frequency at the 4:42 minute mark, explaining that one of triggers for a high level of crawling is signals of high quality that Google’s algorithms detect.

Gary said it at the 4:42 minute mark:

“…generally if the content of a site is of high quality and it’s helpful and people like it in general, then Googlebot–well, Google–tends to crawl more from that site…”

There’s a lot of nuance to the above statement that’s missing, like what are the signals of high quality and helpfulness that will trigger Google to decide to crawl more frequently?

Well, Google never says. But we can speculate and the following are some of my educated guesses.

We know that there are patents about branded search that count branded searches made by users as implied links. Some people think that “implied links” are brand mentions, but “brand mentions” are absolutely not what the patent talks about.

Then there’s the Navboost patent that’s been around since 2004. Some people equate the Navboost patent with clicks but if you read the actual patent from 2004 you’ll see that it never mentions click through rates (CTR). It talks about user interaction signals. Clicks was a topic of intense research in the early 2000s but if you read the research papers and the patents it’s easy to understand what I mean when it’s not so simple as “monkey clicks the website in the SERPs, Google ranks it higher, monkey gets banana.”

In general, I think that signals that indicate people perceive a site as helpful, I think that can help a website rank better. And sometimes that can be giving people what they expect to see, giving people what they expect to see.

Site owners will tell me that Google is ranking garbage and when I take a look I can see what they mean, the sites are kind of garbagey. But on the other hand the content is giving people what they want because they don’t really know how to tell the difference between what they expect to see and actual good quality content (I call that the Froot Loops algorithm).

What’s the Froot Loops algorithm? It’s an effect from Google’s reliance on user satisfaction signals to judge whether their search results are making users happy. Here’s what I previously published about Google’s Froot Loops algorithm:

“Ever walk down a supermarket cereal aisle and note how many sugar-laden kinds of cereal line the shelves? That’s user satisfaction in action. People expect to see sugar bomb cereals in their cereal aisle and supermarkets satisfy that user intent.

I often look at the Froot Loops on the cereal aisle and think, “Who eats that stuff?” Apparently, a lot of people do, that’s why the box is on the supermarket shelf – because people expect to see it there.

Google is doing the same thing as the supermarket. Google is showing the results that are most likely to satisfy users, just like that cereal aisle.”

An example of a garbagey site that satisfies users is a popular recipe site (that I won’t name) that publishes easy to cook recipes that are inauthentic and uses shortcuts like cream of mushroom soup out of the can as an ingredient. I’m fairly experienced in the kitchen and those recipes make me cringe. But people I know love that site because they really don’t know better, they just want an easy recipe.

What the helpfulness conversation is really about is understanding the online audience and giving them what they want, which is different from giving them what they should want. Understanding what people want and giving it to them is, in my opinion, what searchers will find helpful and ring Google’s helpfulness signal bells.

2. Increased Publishing Activity

Another thing that Illyes and Sassman said could trigger Googlebot to crawl more is an increased frequency of publishing, like if a site suddenly increased the amount of pages it is publishing. But Illyes said that in the context of a hacked site that all of a sudden started publishing more web pages. A hacked site that’s publishing a lot of pages would cause Googlebot to crawl more.

If we zoom out to examine that statement from the perspective of the forest then it’s pretty evident that he’s implying that an increase in publication activity may trigger an increase in crawl activity. It’s not that the site was hacked that is causing Googlebot to crawl more, it’s the increase in publishing that’s causing it.

Here is where Gary cites a burst of publishing activity as a Googlebot trigger:

“…but it can also mean that, I don’t know, the site was hacked. And then there’s a bunch of new URLs that Googlebot gets excited about, and then it goes out and then it’s crawling like crazy.”​

A lot of new pages makes Googlebot get excited and crawl a site “like crazy” is the takeaway there. No further elaboration is needed, let’s move on.

3. Consistency Of Content Quality

Gary Illyes goes on to mention that Google may reconsider the overall site quality and that may cause a drop in crawl frequency.

Here’s what Gary said:

“…if we are not crawling much or we are gradually slowing down with crawling, that might be a sign of low-quality content or that we rethought the quality of the site.”

What does Gary mean when he says that Google “rethought the quality of the site?” My take on it is that sometimes the overall site quality of a site can go down if there’s parts of the site that aren’t to the same standard as the original site quality. In my opinion, based on things I’ve seen over the years, at some point the low quality content may begin to outweigh the good content and drag the rest of the site down with it.

When people come to me saying that they have a “content cannibalism” issue, when I take a look at it, what they’re really suffering from is a low quality content issue in another part of the site.

Lizzi Sassman goes on to ask at around the 6 minute mark if there’s an impact if the website content was static, neither improving or getting worse, but simply not changing. Gary resisted giving an answer, simply saying that Googlebot returns to check on the site to see if it has changed and says that “probably” Googlebot might slow down the crawling if there is no changes but qualified that statement by saying that he didn’t know.

Something that went unsaid but is related to the Consistency of Content Quality is that sometimes the topic changes and if the content is static then it may automatically lose relevance and begin to lose rankings. So it’s a good idea to do a regular Content Audit to see if the topic has changed and if so to update the content so that it continues to be relevant to users, readers and consumers when they have conversations about a topic.

Three Ways To Improve Relations With Googlebot

As Gary and Lizzi made clear, it’s not really about poking Googlebot to get it to come around just for the sake of getting it to crawl. The point is to think about your content and its relationship to the users.

1. Is the content high quality?
Does the content address a topic or does it address a keyword? Sites that use a keyword-based content strategy are the ones that I see suffering in the 2024 core algorithm updates. Strategies that are based on topics tend to produce better content and sailed through the algorithm updates.

2. Increased Publishing Activity
An increase in publishing activity can cause Googlebot to come around more often. Regardless of whether it’s because a site is hacked or a site is putting more vigor into their content publishing strategy, a regular content publishing schedule is a good thing and has always been a good thing. There is no “set it and forget it” when it comes to content publishing.

3. Consistency Of Content Quality
Content quality, topicality, and relevance to users over time is an important consideration and will assure that Googlebot will continue to come around to say hello. A drop in any of those factors (quality, topicality, and relevance) could affect Googlebot crawling which itself is a symptom of the more importat factor, which is how Google’s algorithm itself regards the content.

Listen to the Google Search Off The Record Podcast beginning at about the 4 minute mark:

Featured Image by Shutterstock/Cast Of Thousands