E-A-T is a concept we first learned of in Google’s Search Quality Rater Guidelines.
In this chapter, you will learn about this document, the quality raters who use it, and whether it can impact your Google search rankings.
What are Google’s Search Quality Raters Guidelines?
Google’s Search Quality Raters Guidelines is a quasi-frequently updated document (the last update, as of this writing, was October 14, 2020) that Google Quality Raters use/reference as they rate websites.
As the name suggests, they are the guidelines that these raters are to use as they perform their function.
The guidelines outline the conditions and elements that need to be considered and how the site should be rated by that person.
The most recent version is a 175-page read, compared to the 168-page version it replaced.
If you’re interested in reading it yourself, you’ll find it here. And wonderfully, Google places their updated versions at the same URL so you can bookmark it and always be able to find the most current version easily.
What is a Google Quality Rater?
For those unaware, Google has hired many thousands of individuals from around the world to rate websites and record whether the site is good or bad across a variety of areas.
Now, it is important to understand that these people have no impact on the rankings of the sites they rate.
That said, one can argue that their role is much much larger than that.
They don’t influence the rankings of the sites they rate.
They influence the rankings of every site.
How Do They Influence the Rankings of Every Site?
In the spirit of transparency, I need to note that I don’t get to just call Sundar, John, Gary, or Martin over and Google and ask them exactly how their algorithms work.
Like everyone else, I pay attention to what they say and what’s going on in the results and make best guesstimates on how certain functions and signals will be best treated.
The most likely structure to make the feedback of ~10,000 quality raters actionable is to do it in the way that Google does almost everything: algorithmically.
In the backend of the system, the raters are using a slider to assign values:
This data would then be made available to machine learning systems that would use it to augment the algorithms based on known signal data.
For example, if a site or group of similar sites are consistently rated High or better, the system could review all the signal data from the site(s) to look for commonality.
And by signal data, I am referring to everything from structure, size (of page and domain, and related section for that matter), backlinks and backlink profile, author signals, navigation, and likely a whole lot more.
The same would be true for site(s) with Low values.
With this, the system would likely:
- Produce a set of results based on what the new algorithm produces across a variety of phrases and niches.
- Send the top-ranking sites in that set to the raters.
- And, assuming the raters favor the new results page, push the signal adjustments into the global algorithms we all know and love, either globally or in testing.
It is also possible that Google may skip the review phase and just push the new signal weights into the algorithms for testing, but I suspect they use their raters more often than not.
So, while a Quality Rater does not impact an individual site’s ranking, collectively, they do influence the algorithm that powers all that is Google.
Arguably, significantly more influence than just looking at a single site and deciding that it should move up or down the results is deciding how all the results on a page are positioned.
Now that we understand what the Search Quality Raters Guidelines are, the next question we need to explore is:
What are the Raters Looking For?
When we’re looking at the areas the raters are instructed to look at, we’re essentially:
- Looking at what Google wants the algorithm to produce.
- Getting a glimpse of what their algorithms will focus on.
The guide states:
“As a Search Quality Rater, you will work on many different types of rating projects. The General Guidelines primarily cover Page Quality (PQ) rating and Needs Met (NM) rating; however, the concepts are also important for many other types of rating tasks.”
We won’t be looking at the tasks specifically here and will focus on the more important (from the context of this piece):
- Needs Met.
- Page Quality.
What do they mean?
This is an area that gets too little attention – and in their October 14, 2020 update, it appears that the folks at Google agree as they grew that section.
I hadn’t actually caught the changes in this section when reading it, but while reviewing some other write-ups while putting this piece together, I was happy to have it pointed out by Jennifer Slegg in her write-up here.
It’s something to pay close attention to, Google is.
What is Needs Met?
Needs Met is a fairly straightforward concept… it basically means intent.
The question that raters would be asking themselves in assessing a page is: “How helpful and/or satisfying is this result?”
During this testing, a rater may visit a single page or visit a search results page and rate every result.
Both will send the information to Google about the site structure, device, demographic, and location results differences. I’m sure a number of other factors apply to the grading of each result (there’s a reason they have more than 10,000 raters around the world).
These ratings will then be used to drive changes to improve the results to algorithmically determine which signals or signal combinations are common to the higher rankings results.
I suspect that in the case of Needs Met, the signals will predominantly focus on the onsite factors, including but definitely not limited to content, links on the page (expanded on in the recent version), structure, and user experience.
It is important to note that, as with the real world, the Needs Met rating does require decent Page Quality. In fact, the guidelines state it clearly:
“The Needs Met rating is based on both the query and the result.”
One can have a medium Needs Met with a low Page Quality, but it would be highly unlikely that they could get a high Needs Met rating.
After all, the user intent isn’t satisfied if the searcher doesn’t trust the result.
One final element of Needs Met that is worth noting before we move on to Page Quality is the interpretation.
By this, we are to consider queries with multiple possible meanings such as the following, from the examples they give:
When the rater is assigning a Needs Met score, they are to give more weight to pages satisfying higher intents.
- A “Highest” rating being given to an authoritative piece on Apple, Oklahoma.
- Training algorithms to focus on the wrong signals for the majority of users.
While this was in previous versions of the guidelines, it ties to the subtopics update announced at Search On. While not directly related, I find it interesting that the announcement of subtopics came on October 15, 2020 – the day after the updated guidelines were published.
The subtopic update does address slightly different issues that are directly referenced in the guidelines – the goal of both is to show diversity in results, where there are multiple meanings or topics that may satisfy your intent as a searcher.
As an SEO pro, I like to keep something specific in mind:
If I can address multiple possible intents for a query (often including additional content available via easily visible links), then my page naturally satisfies a greater number of Needs Met and thus, has a higher probability of meeting a user’s intent.
Page Quality ratings are based on a number of factors, all of which interconnect (almost like a Google algorithm, right?).
And the weight given to each is based on the type of site and query (again… the similarity is uncanny).
The key ones noted in the guidelines are:
Your Money or Your Life (YMYL)
As Google words it, a YMYL site is one that:
“… potentially impact a person’s future happiness, health, financial stability, or safety.”
With YMYL sites, raters are directed to put more weight on E-A-T.
The guidelines categorize them as:
- News and current events: News about important topics (e.g., international events, business, politics, science, technology). Keep in mind that not all news articles are necessarily considered YMYL (e.g., sports, entertainment, and everyday lifestyle topics are generally not YMYL). Use your judgment and knowledge of your locale.
- Civics, government, and law: Information important to maintaining an informed citizenry, such as information about voting, government agencies, public institutions, social services, and legal issues (e.g., divorce, child custody, adoption, creating a will, etc.).
- Finance: Financial advice or information regarding investments, taxes, retirement planning, loans, banking, or insurance, particularly webpages that allow people to make purchases or transfer money online.
- Shopping: Information about or services related to research or purchase of goods/services, particularly webpages that allow people to make purchases online.
- Health and safety: Advice or information about medical issues, drugs, hospitals, emergency preparedness, how dangerous an activity is, etc.
- Groups of people: Information about or claims related to groups of people, including but not limited to those grouped based on race or ethnic origin, religion, disability, age, nationality, veteran status, sexual orientation, gender, or gender identity.
- Other: There are many other topics related to big decisions or important aspects of people’s lives which thus may be considered YMYL, such as fitness and nutrition, housing information, choosing a college, finding a job, etc.
Most people don’t think of shopping when they think of YMYL… but it’s in there.
According to the guidelines, the sections of a website can be divided into three main categories:
- Main Content (MC): Main Content is any part of the page that directly helps the page achieve its purpose.
- Supplemental Content (SC): Supplemental Content contributes to a good user experience on the page but does not directly help the page achieve its purpose. The example they give is navigation ease of access links. Critical to the site, but not necessary to satisfy Needs Met.
- Ads: s/Monetization (Ads) is content and/or links that are displayed for the purpose of monetizing (making money from) the page.
Following their direction, understanding a webpage is quite simple.
- What’s outlined in blue in the screenshot below is Master Content.
- What’s outlined in red in the screenshot below is Ads.
- Everything that remains is Supplemental Content.
Here is how a page might be broken down:
The ease-of-access and volume of the Main Content play their parts in Page Quality calculations.
It’s what assists the rater value on not just whether the need is met, but also how easily supplemental content is accessed, should it be desired.
And now, specific for the section you are likely reading this piece for, it was important to understand what has been outlined above to add context to this important section.
Let’s begin with what E-A-T stands for:
How Do the E, A & T Differ?
The line between expertise, authoritativeness, and trustworthiness can get pretty blurry.
Here is a basic outline of each:
The expertise of the content creator is related only to the content of the page being judged and not the site as a whole.
It does not have to be an author, and the criteria are not set in stone.
When looking for information on lung cancer, an in-depth study on causes, impact, statistics, etc., from an important medical school or institution, or government agency would likely get a high score.
If the searcher were looking for information on what it’s like to live with it, a personal account from a bank teller whose partner lived with it for years would fulfill the Expertise criteria exceptionally.
It may be on an important site or an exceptional member of a forum answering questions.
It’s all about context and addressing with expertise the need to be met.
The authoritativeness of the content is judged by the authority of the content itself and the domain.
In general, this would be based on external signals such as links and link quality, brand mentions, citations, etc., both to the content specifically and the domain as a whole.
When thinking of authority, I can’t help but think of a patent granted in 2015, Ranking Search Results Based On Entity Metrics.
In it, they discuss the following key metrics:
- Relatedness: How related are two entities?
- Example: Empire State Building and Skyscrapers.
- Notability: How notable is an entity in its domain?
- Example: Semrush is more notable in the SEO tools than the software domain.
- Contribution: How is an entity viewed by the world?
- Example: Does it have critical reviews, fame rankings, etc.?
- Prizes: Has the entity received prizes?
- Example: The types of awards and prizes an entity has received.
While there are certainly other characteristics, I find these helpful guides.
The trust of the content is judged again by the trust of the content specifically and the trust of the domain.
Trust is similar to authority but more pointed.
Where authority focuses more on the volume of quality references, trust focuses more on specific signals and sites.
An example they use in the guidelines is the BBB.
While they don’t talk about it being used as a positive signal, they do note that a bad rating based on a significant volume of users could be used as a negative.
E-A-T & Rankings
It’s important to remember that all of these things relate to the query intent and the subject.
Google gives a great example of this in their guidelines when discussing a theoretical someone whose spouse has liver cancer.
If the person is looking for information on treatments, the expertise would be a high level of E-A-T in medicine and liver cancer specifically.
If the spouse is looking for support in handling this horrible situation, the E-A-T would likely not come from the medical community but others whose spouses have battled the disease.
Both topics relate to liver cancer, but the E-A-T criteria is very different.
It is hopefully obvious, but let me stress one very important fact.
E-A-T is not a ranking factor.
Say it with me:
E-A-T is no more a ranking factor than the prep material you used to study for your driver’s test is a turn signal.
It outlines what should be produced as a final product, allows and describes situational variance, and is then used to identify and tune the practical signals.
E-A-T is what the raters are guided to look for.
Raters use E-A-T to assist them in how they are rating websites.
Here’s a tweet from Google’s Danny Sullivan that explains it well:
Our systems aren't looking for EAT. Our raters are using that to see if our systems are working well to show good information. There are many different signals that, if we get it right, align with what a good human EAT assessment would be. See also: https://t.co/1fs2oJ9Gtl pic.twitter.com/GBbnYEjJUV
— Danny Sullivan (@dannysullivan) February 19, 2020
Let me say it one last time: It is not part of the algorithm.
The purpose of outlining E-A-T in the Search Quality Raters Guidelines is that if the raters use it to judge websites, and Google uses these ratings to adjust their algorithm, in the end, the algorithm will align with the E-A-T principles.
So, E-A-T can be used as a guiding principle for design, content creation, and supporting external signals. It can be used, as you used the information from your driving guide the first time you hit the open road.
But you can’t optimize for it specifically.
The Search Quality Raters Guidelines & E-A-T Takeaway
If you’re just getting your feet wet or just looking for a different point of view, hopefully, you’ve found this helpful.
I’ve tried to avoid getting into strategies around the subject, relative to each niche, starting point, and current E-A-T.
The Google Search Quality Raters Guidelines is an important document. It tells us where Google wants to go, and they’re throwing a lot of resources at it.
They also update it as they look to tune different parts of their algorithms.
Featured image: Paulo Bobita/Search Engine Journal
Image 4 created by author; carballo/Shutterstock