For the past year or so, it seems the SEO community can’t stop talking about E-A-T: Expertise, Authoritativeness, and Trustworthiness.
Since August 1st, 2018, informally labeled the “Medic” update, Google began rolling out a series of huge core algorithm updates – one approximately every 3-4 months. While Google rolls out hundreds of algorithm updates every year, some clear patterns emerged from the Medic update and the subsequent core algorithm updates of 2018 and 2019. During this time, many SEO professionals dug through the data and developed a theory that E-A-T played a major role in the performance of winners and losers of these updates.
Why has E-A-T become such a hot topic? In order to understand E-A-T, it’s important to first understand the concept of YMYL.
YMYL: Your Money Your Life
According to Google, pages or topics can be considered YMYL (your money, your life) if they “could potentially impact a person’s future happiness, health, financial stability, or safety” (Google Search Quality Guidelines). As it turns out, many sites fall into this YMYL category, given that so many topics can affect users’ happiness and well being. Largely speaking, it’s been YMYL websites, such as medical, legal, financial, news, and political sites, that have overwhelmingly felt the impacts of the recent algorithm updates.
Here’s a few examples for illustration, using the legend on the left to indicate the major core algorithm updates:
As you can see, these extreme algorithmic fluctuations have resulted in major changes in organic visibility and traffic for YMYL websites, often with one algorithm update wiping out the gains earned by the previous update (or vice versa).
What is Causing These Fluctuations in Google’s Algorithm?
It’s no secret that Google and the other tech giants have been in the news and media spotlight for the past several years for a variety of reasons:
- the rise in misinformation and “fake news” online (source)
- the internet contributing to a rise in extremist behavior (source)
- the role online information plays in influencing elections (source)
- skepticism about the quality of medical information online (source)
- the role of misinformation in creating public health epidemics (source)
Search engines and social media companies are facing government scrutiny.
While Google has not confirmed that its recent core algorithm updates are directly tied to this media coverage, it has revealed its involvement in several initiatives aimed at improving trustworthiness online in the past few years, such as:
- In 2016, Google began working with “experts at Harvard Medical School and Mayo Clinic to create what are known as Symptom Cards, a feature of the Knowledge Graph which provides “high-quality medical information” on various health conditions.
- In 2017, Google began funding The Trust Project: an “international consortium of news organizations” that aims to “affirm and amplify journalism’s commitment to transparency, accuracy, inclusion, and fairness.” Google’s blog post about The Trust Project stated that they would be adding “signposts and labels” to their news results, such as Fact Check labels, along with information about publishers, such as awards they have won and third-party reviews of claims they have made.
- In 2019, Google presented a whitepaper at the Munich Security Conference called “How Google Fights Disinformation,” which thoroughly explains its initiatives to combat “misinformation, disinformation, and ‘fake news,'” including an explanation of how its algorithms “elevate authoritative, high-quality information” across its products.
It makes sense that Google would heighten its criteria for what it means to have high-quality, trustworthy content, when Google itself has been facing increased scrutiny from the media, the public, and governments around the world.
Google also clearly recognizes its role in curbing misinformation related to health and wellness. For one, a 2017 study by Eli Schwartz found that 77% of Americans admitted to going online to diagnose medical symptoms. More recently, the biggest measles outbreak the United States has seen in 25 years is often attributed to the role of misinformation found online.
As a response to these and other trends related to misinformation online, Google claims:
“We have an important responsibility to our users and to the societies in which we operate to curb the efforts of those who aim to propagate false information on our platforms.” – How Google Fights Disinformation, 2019
Therefore, E-A-T should be viewed as Google’s criteria for analyzing the trustworthiness of content – and the people who publish it – in order to mitigate the spread of misinformation.
E-A-T has become a controversial topic in the SEO community. Depending on who you ask, different SEO professionals have differing opinions about E-A-T and the extent to which it matters for SEO. This disagreement often stems from the fact that it’s hard to conceptualize precisely how Google might be factoring E-A-T considerations into its current algorithms. It’s not as simple as many of the ranking factors that have come before it, such as page speed or HTTPS compliance.
However, Google mentions E-A-T 135 times in the current iteration of its 167-page Search Quality Guidelines as a necessary criterion for what it means to be a high-quality, trusted site. This alone supports the notion that E-A-T has become a big priority for Google. But to drive the point home, Google also published a blog on the 1-year anniversary of the Medic update, titled “What webmasters should know about Google’s core updates” which further solidified the significant role E-A-T plays in recent algorithm updates.
Due to this growing focus on E-A-T, as with many concepts in the SEO industry, there are some new misconceptions about how E-A-T works.
There’s been a lot of talk in the past year about what E-A-T is; so let’s instead clear the air and talk about a few things E-A-T is NOT.
- E-A-T is NOT: Something that matters for every single website
You can use the handy chart below as a gauge for how much E-A-T matters for your site. For some sites, good E-A-T is a “nice to have” but not a requirement. For example, if you have a blog about knitting, it’s fine to just be a knitting enthusiast who chose to start a blog – not much “expertise” is required beyond that.
A topic like “celebrity gossip” might require a bit more expertise – how long have you been passionate about this topic? What other publications have you written for?
When your content ventures into areas that are inherently more ‘YMYL’ in nature, such as your pet’s health, or especially for topics like “cancer treatment,” E-A-T becomes crucial. Readers need to know who you are and why you should be trusted to write about those topics.
2. E-A-T is NOT: An explicit, confirmed ranking factor:
This is where E-A-T gets confusing: Google has never confirmed E-A-T as a ranking factor, the way it has for something like page speed, HTTPS, or the presence of keywords in a title tag.
The relationship between E-A-T and rankings is a bit more indirect. Here’s a few quotes from Google on how E-A-T factors into their algorithm:
Google’s algorithms identify signals about pages that correlate with trustworthiness and authoritativeness. The best known of these signals is PageRank, which uses links on the web to understand authoritativeness.What Webmasters Should Know About Google’s Core Updates, 2019
Search rater data is not used directly in our ranking algorithms. Rather, we use them as a restaurant might get feedback cards from diners. The feedback helps us know if our systems seem to be working.What Webmasters Should Know About Google’s Core Updates, 2019
E-A-T is NOT: A replacement for technical SEO.
E-A-T and technical SEO are not mutually exclusive. E-A-T is just one factor that contributes to SEO success: it must be considered alongside a solid technical foundation; authoritative, relevant backlinks; and great content and UX.
E-A-T is NOT: The only factor causing websites to get hit by algorithm updates.
While it’s important to consider E-A-T when diagnosing why a website was affected by recent Google algorithm updates, E-A-T is not the only issue that could be causing performance declines. Technical issues, a poor user experience, overwhelming ads that distract from using the main content, or security issues are also examples of problems that can cause algorithmic devaluations by Google.
E-A-T is NOT: Something that is quick or easy to fix.
Unfortunately, when addressing E-A-T issues, there is no silver bullet. Establishing or improving E-A-T takes time, resources and patience.
EA-T- is NOT: An SEO tactic that results in an immediate boost in performance.
Unlike repairing broken canonical tags or re-introducing accidentally-noindexed pages into Google’s index, improving E-A-T generally does not result in an immediate boost in traffic. It takes Google (and users) time to develop trust and perceived credibility around your brand. Building up a good reputation, earning high-quality reviews, and driving meaningful links and citations to your content are all initiatives that not only take time for your business to accomplish, but also take time for search engines to fully process and measure those outcomes.
With a clear understanding of the things E-A-T is not, what does it currently mean to have good E-A-T? With the help of my team, I conducted a data-driven analysis to find out.
My E-A-T Analysis: Methodology
- Analyzed 64 net winners and losers between the August 1st, 2018 core update through late 2019
- Used the Sistrix Visibility Index, which projects organic performance based on search volume and current rank on Google.com
- Used Archive.org (AKA the Wayback Machine) to document performance of website content at various points in time
- Took note of performance across 30 potential on-page E-A-T signals
- Did not analyze backlinks (although they play an important role in E-A-T)
- Identified interesting correlations and trends
A Few Disclaimers About This Process:
- Small sample size: This analysis was done with a small sample size of only 64 total websites. Manually analyzing sites takes time, and we wanted to sure we were thorough in our analysis.
- Correlation does not imply causation. The presence of any of the following page elements on a winning or losing site does not imply that the website’s performance was due to that factor. There are hundreds of ranking factors, and no single page element in isolation can be responsible for a website’s performance.
- E-A-T analysis is a moving target. Websites change. Algorithms get updated. The Wayback Machine does not have a picture of every website at every point in time. This constant movement can make the analysis fuzzy, but we did the best we could with what we had.
- This is not intended to be science. Rather, we wanted to share what we found as it relates to the qualities of top performing YMYL sites on Google, compared to those sites which have lost significant visibility since the Medic update.
- How we measured “high-quality” vs. “low-quality” content: fortunately, Google has a 167-page document called the Search Quality Guidelines, which specifically details what it means to have high or low-quality content. We carefully cross-reference our own analysis with the Guidelines to ensure we are thinking like Google.
The sites with the biggest fluctuations during the past years’ algorithm updates were overwhelmingly YMYL websites – particularly health-related websites.
The biggest winners we analyzed fell into the following categories:
The biggest losers we analyzed fell into the following categories:
Results of Our E-A-T Study:
- 51% of losing websites were also hit by the Unnamed (AKA “Fred”) Core Update in March 2017.
- The unnamed Google update (informally labeled “Fred”) in March of 2017, appeared to target websites with thin content, overly aggressive monetization tactics, distracting ads, a poor user experience. Some SEOs, such as Marie Haynes, believe it was one of the first time Google tried incorporating E-A-T considerations into its algorithm.
- Winning health companies are 28 years older than losing health companies.
- While winning health websites are often established companies or organizations who have been around for many years, many of the biggest losers are relatively new blogs or natural wellness websites which originally launched as online publications.
- Winning sites are 16% more likely than losing sites to have author bios present on their articles.
- Winning sites are 258% more likely to use real experts on their content than losing websites.
- While many losing websites use author bios, researching the credentials of those authors reveals that they often lack the real expertise needed to provide trustworthy medical, legal, financial, or other “YMYL” advice.
- Winning health sites are 34% more likely to use medical reviewers on their content than losing websites.
- Winning sites are 45% more likely to have a clearly stated editorial policy present on their websites than losing sites.
- Losing sites are 433% more likely to have calls to action on pages containing medical content.
- Similarly, losing sites are 117% more likely to have affiliate links on YMYL content.
- Winning companies are 21% more likely than losing sites to have a company Wikipedia page.
- Winning companies are 850% more likely to clearly list their awards and accolades than losing sites.
- Winning health sites are 213% more likely to be “HONcode Certified” than losing sites.
- “The Health on the Net Foundation Code of Conduct (HONcode) for medical and health Web sites addresses one of Internet’s main healthcare issues: the reliability and credibility of information.“
- Winning companies are 24% more likely to link to external citations within their content.
- We considered “external citations” to be links to supporting resources and evidence within the content and/or in a bibliography section.
- Winning sites have an average Trustpilot score that is 1.9 points higher than that of losing websites.
- There is no clear correlation between BBB ratings and performance: both winning and losing sites are largely A+, with the occasional poor rating mixed in.
- Losing sites are 94% more likely to include comments and other user-generated content on YMYL pages (and allowing that content to be indexed by search engines).
- Winning sites have an average Flesch-Kinkaid reading level score that is .7 points higher than that of losing websites.
- Using Readable for sentiment analysis: winning sites are 728% more likely to use “very formal” writing on YMYL content than losing sites.
A few examples of stellar E-A-T:
If you want to see good E-A-T in practice, the following sites have employed some exemplary pages that help instill their users with a sense of trust and credibility:
Our E-A-T study indicates that companies have to be careful blending promotional content with YMYL content, especially with regard to health topics. Healthline’s ad and sponsorship policy states that its editorial content is clearly distinguished from promotional content, so users can trust they are receiving legitimate, unbiased medical information.
A consistent winner of the core algorithm updates, eMedicineHealth has a page dedicated to listing all of its dozens of collaborating “U.S. Board Certified Physicians and Allied Health Professionals,” and their editorial role on the site.
The Recovery Village, a drug rehabilitation clinic, has an excellent Editorial Team page with approximately 100 authors listed, along with their relevant experience, and links to their LinkedIn profiles along with other relevant, credible online resources about the authors.
The Keto Diet is a controversial topic, and keywords related to the keto diet have been major victims of algorithmic turbulence. Using the Wayback Machine, it’s evident that The Diet Doctor has been constantly enhancing and updating its Keto Diet page to include more citations, fact-checking, enhanced fact-check labels, a formalized medical review process, “trust meters” to indicate how trustworthy a given source is, better grammar, more robust nutritional information, and more.
By analyzing the performance of algorithm winners and losers, it’s evident that Google’s algorithm is making changes to reward YMYL companies who demonstrate excellent E-A-T. Below are 5 takeaways that are not only implied in Google’s Quality Guidelines, but also confirmed with the data from our study:
- YMYL content should evidence based, objective, thorough and heavily researched
- Include high-quality citations in your content & support statements with facts from trusted resources
- Author bios are not enough; ideally, the authors should be real experts (for YMYL topics)
- Avoid affiliate links or “salesy” language in YMYL content
- Maintain an editorial policy & disclosures around your advertisements
View my E-A-T slides from Pubcon Vegas 2019:
Curious to learn more about E-A-T or see how your website stacks up? Contact the Path Interactive SEO team to learn more about our SEO offerings, including algorithm update recovery services.