E-A-T is the shorthand Google uses in its Search Quality Evaluator Guidelines and it stands for ‘Expertise, Authority and Trust’. For those who don’t know, the guidelines are what the people who manually review pages for Google use.
If you’re not familiar with how Google ‘trains’ its algorithm, it uses machine learning to continually tweak the different metrics. It does this ‘offline’ and uses the manual review to judge the quality of the serps in comparison to what is currently live. When there is a big enough improvement it can then roll that change live and we get an update.
So the guidelines give us a glimpse of how Google guides the people doing the manual review to judge if the algorithm has done a good job. In other words, it uses it to help people decide what good pages should look like, so it can model the best sets of results. High quality pages in the guidelines are the ones it wants the algorithm to push to the top.
If this sounds like a shortcut to understanding how Google ranks pages there is a bit of a catch, because the way the content is ranked automatically is not the same as the way it gets people to assess the pages manually. We know for instance that backlinks play a huge role, but there is little mention of them in the guidelines. Some things are easy to do automatically and hard to do manually and vice versa.
What does Google say about E-A-T?
Before I bring in some text from the guidelines it’s handy to know that ‘MC’ is Googlespeak for Main Content, which is whatever content on page fulfils its main purpose. So on a YouTube page it would be the video, on this piece it’s the main body of text. On an online games site it would be the game.
“For all other pages that have a beneficial purpose, the amount of expertise, authoritativeness, and trustworthiness (E-A-T) is very important. Please consider:
● The expertise of the creator of the MC.
● The authoritativeness of the creator of the MC, the MC itself, and the website.
● The trustworthiness of the creator of the MC, the MC itself, and the website.”
So in other words, once the content itself has been judged as high quality then go on to review the creator of that content and the wider website. So for instance a high quality writer on a low quality website (or vice versa) wouldn’t be good enough for a High Quality page. It has to be both.
Google splits split the rating into the three sections:
● Expertise
● Authority
● Trust
So let’s go through these and review what they mean in the context of both a website and an individual, according to Google’s guidelines.
Expertise
For a website
The organisation behind it has suitable credentials. Google gives the example of a news website where the organisation has won multiple Pulitzer Prizes. It is of course harder to demonstrate this for some types of sites such as humour sites, but having won industry awards or gained external recognition plays a big role.
For an author
What credentials does the author have on this subject? Google gives the example of a doctor writing a medical article. But for other topics it could be having a degree in the field, relevant experience or working in that sector.
Authority
For a website
This is about how both the rater and external sources rate the quality of the content and the wider site. This includes (according to Google) metrics such as ratings on the Better Business Bureau and Yelp as well as wider reviews. It does however warn reviewers to be wary of fake reviews and include links to articles on the topic:
https://www.theguardian.com/money/2013/jan/26/fake-reviews-plague-consumer-websites
It’s also expected that the content should be relevant to the wider site. Google gives this example of low quality content:
"The website is not an authoritative source for the topic of the page, e.g. tax information on a cooking website”.
For an author
This is about the number of citations that an author has, the types of sites they appear on and generally how well written their content is.
Google also mentions that high quality content "must be factually accurate for the topic and must be supported by expert consensus where such consensus exists.”.
Trust
For a website
The website should have validation information, such as an About Us page and if it’s a transactional site, links to customer service information. Another example might be a transactional pages need to be secure and over https;
“The MC is not trustworthy, e.g. a shopping checkout page that has an insecure connection”
For an author
Trust is measured by what citations they have. Do they appear on other high websites with high levels of E-A-T? How well written is their content?
Reputation
Google doesn’t just stop with E-A-T, it also grades based on reputation. Which is, confusingly, in and of itself representative of a high level of E-A-T.
“Extensive reputation research is important when giving Highest ratings, and is evidence of the E-A-T of the page. Very positive reputation is often based on prestigious awards or recommendations from known experts or professional societies on the topic of the page."
So reputation is an extension of E-A-T and seems to be mainly focused on external validation such as reviews.
"A website's reputation is based on the experience of real users, as well as the opinion of people who are experts in the topic of the website"
Additionally the amount of engagement on page can also increase your score for reputation:
"For these topics, popularity, user engagement, and user reviews can be considered evidence of reputation”.
Finding reputation information
Google recommends identifying the homepage of the website and performing the following searches, using IBM as an example:
● [ibm -site:ibm.com]: A search for IBM that excludes pages on ibm.com.
● [“ibm.com” -site:ibm.com]: A search for “ibm.com” that excludes pages on ibm.com.
● [ibm reviews -site:ibm.com] A search for reviews of IBM that excludes pages on ibm.com.
● [“ibm.com” reviews -site:ibm.com]: A search for reviews of “ibm.com” that excludes pages on ibm.com.
● For content creators, try searching for their name or alias.
The idea is to find pieces of information about it which "were not written or created by the website, the company itself, or the individual".
Wikipedia is also identified as a good potential source of reputation information. Google also gives some examples, including one where “negative articles on news sites” is cited as a further example of poor reputation.
In summary
Overall there is a common sense approach to how Google suggests that E-A-T be approached, and although there does seem to be an overlap with reputation, there are no hard and fast rules as to what the information might consist of. It appears to be down to the nature of the site and the information found about it.
It’s also worth noting that for smaller sites, not having any reputation or external information is not seen as inherently negative. Google expects it to occur organically as sites grow over time. But getting poor press coverage or ratings in certain places can severely impact a site’s score.
Not all these factors will necessarily be relevant for your particular site, but it’s worth looking at them and making sure that where they can be applied, you meet the requirements for a good E-A-T rating.
