5 Things I Learned About E-A-T by Analyzing 647 Search Results

As a writer at a content marketing agency, I’ve written for a lot of different clients, and almost everything I’ve produced has been intended to rank on Google and encourage website traffic.

Here’s the challenge I (and every other marketing writer on the planet) am up against: search competition.

No matter what industry you’re in, or target audience you’re speaking to, you’re not alone. You have competition. And if you and your competition both understand the SEO game (which is very much the case for most companies nowadays), then what do you have to fall back on to protect your visibility in the all-important SERPs?

According to Google, it’s E-A-T: Expertise, Authoritativeness and Trustworthiness.

But here’s the complicated thing: Every one of my clients — even the small ones thriving in very big industries — has expertise, and authoritativeness, and even trustworthiness. So, how does that help them in search? And how can they possibly prove to Google, amid all the noise and competition and other experts out there, that they deserve a place on Page 1?

Last year, I set out to find out.

Methodology

Google is pretty clear about the fact that websites need E-A-T, but what they don’t really clue us in on is what E-A-T actually is or how it’s measured. I hypothesized that, if I compiled a big list of SERPs and closely analyzed all the Page 1 results, I could narrow down what may comprise E-A-T.

Theoretically, E-A-T affects different industries in different ways. That’s because some topics and subject areas are more critical than others to have extremely reliable information — like when you’re searching for information about prescription drugs or complicated financial products.

So, the first thing I did was choose seven topic categories to focus on: legal, insurance, health care, loans, pharmaceuticals, military, and informational questions. Next, I picked 10 queries for each category.

Then I searched. The resulting 70 SERPs produced 647 results. I analyzed each of those results, looking specifically for 32 different factors.

Finally, I reviewed what I had recorded and asked:

• Which factors were the most prevalent across all 647 results?

• Which factors were most prevalent among the 210 Top 3 results?

• Were there differences in prevalent factors across the various topic categories I chose?

Before we get into the results, let’s talk about correlation vs. causation for a moment. While each of these factors seemed to be very common among Page 1 results, and it seems clear that some of these factors do play a role in establishing E-A-T, all I can really say for sure is that these traits are associated with pages that rank well in search. They could be indicators of a good page or website, but not necessarily the determining factor that’s putting them on Page 1.

With that in mind, here are five lessons I learned about E-A-T after closely analyzing the results from those 70 searches.

Lesson 1: Original, relevant, recent content is essential

Of all the lessons, this is the least surprising to me, but perhaps the most important. To rank well for relevant terms, you need to strongly demonstrate that your website belongs in search results. How? Content, obviously.

But it’s got to be high-quality content. Usually, I’d say that means you’re addressing the topic from all angles and leaving no questions unanswered. But after this SERP inspection exercise, I’d actually say the three most important characteristics of high-quality content are that it’s:

1. Original

2. Relevant

3. Recently published or updated

Original research

One factor I sought throughout this study was original research. To me, this included any content that’s created using information the organization sources, analyzes, and publishes themselves.

Just shy of two-thirds of the results’ websites contained original research, but among the websites whose results were in the Top 3 positions, 70% had original research available. This shows the importance of creating your own, unique content — a story only you can tell. Trust me, you have one.

Relevance and topical authority

Beyond content just being unique, it also needs to be relevant to your industry and target audience. Topical authority is a weird concept because SEOs know it’s real, but there’s no way to measure it, and Google hasn’t exactly come out and said they have a topical authority ranking factor.

However, they have given us a lot of clues that point to topical authority being a highly important factor in E-A-T — like this patent they filed in 2017. Even in their recent Helpful Content Update, Google highlights questions that creators should ask themselves when considering their own site content. The question, “Does your site have a primary purpose or focus?” in particular alludes to the importance of creating content for a topic niche or specific subject area.

Given the limited tools on this subject, I decided to create my own (rudimentary) method of measuring topical authority by way of roughly determining the topic coverage depth throughout the whole website. Here’s what I did:

1. Determine the parent topic of the query in question. “Insurance” is the parent topic for “types of insurance” and “world population” is the parent topic for “how many people are in the world,” for example.

2. Find the Topic Coverage Score (TCS, as I call it) of each result’s website. That’s the number of pages indexed by Google that contain an exact-match of the parent term.

3. Calculate the average TCS of all Page 1 results for each query.

4. Compare the TSC of each result with the average TSC for that query.

After that procession of steps, I found that while 25% of Page 1 results had a TSC higher than the average, 40% of Top 3 results boasted the same. In other words, the websites that had the most topic coverage were more likely to land at the top of the page.

Recently published or updated

Half of all Top 3 and 48% of Page 1 results were dated within the previous two years. There are plenty of evergreen topics that don’t need regular content changes (the oldest result in my study was a page explaining why the sky is blue from 1997). Updating content just for the sake of giving it a new date won’t help you rank any higher in Google. However, creating timely content and updating old content as necessary could help.

Lesson 2: Your off-site, online presence matters

Here’s a lesson I wasn’t expecting to learn. When I set out on this study, I thought the biggest E-A-T factors would correspond to the website in question more so than the organization that manages it. Not so much: It became clear to me that your off-site, online presence plays a role in helping you rank in Google search results.

The vast majority (95%) of all results had third-party reviews of some kind, whether they’re Google My Business reviews, comments on Glassdoor, site trustworthiness information on Trustpilot, or something else.

Wikipedia is also a common thread between many of the results. While 89% of Page 1 results’ websites or organizations had at least one Wikipedia mention, 93% of Top 3 results did, too. As far as actual Wikipedia pages, 73% of Page 1 results and 82% of Top 3 results’ organizations had one.

The high prevalence on Page 1 tells me that it’s fairly common to have a Wikipedia connection, but the higher numbers corresponding to the Top 3 results hints at what their importance might be.

Another patent from Google, this one updated in 2018, discusses the topic of seed sites. A seed site, theoretically, is one that the search engine trusts because it generally has quality content and good, valuable links. Google hasn’t revealed whether this seed site theory is valid, or to what extent it plays a role in search algorithms (if any). But if I were to choose a seed site, Wikipedia would be a good contender. Each page has tons of links to websites with relevant information on carefully organized topics.

Another website worth mentioning is the Better Business Bureau. While it only gives limited perspective (since it only relates to Canadian and US businesses), I found that many Page 1 results’ organizations (70%) and even more Top 3 results (74%) had at least a BBB page but not necessarily a grade. In fact, a little over one-fourth of results that had a BBB page didn’t have a rating.

It seems to me that the value is in getting listed on BBB’s website more so than achieving a good grade — perhaps a North American-specific seed site of sorts.

Lesson 3: Transparency and honesty are the best policies

So far, we’ve learned a lot about E (expertise) and A (authoritativeness) but where the T — trustworthiness — really comes into sight is when we start talking about transparency.

Google states right in its Page Quality Rating guidelines that webmasters should state on their website exactly who is responsible for site content. That can be a person or people, or it could be an organization. At Moz, for example, the folks at Moz are responsible for their site content, and they explain all about it in their About page. Similarly, 91% of results I analyzed had a detailed About Us page.

Another way of being transparent about what your site is all about is by publishing editorial standards or guidelines. These documents detail how your site gets populated: where content comes from, what characteristics help it make the cut, what the organization won’t publish, and more.

43% of Page 1 results and 49% of Top 3 results had some sort of editorial guidelines published. These included information quality guidelines, pitch guidelines that reflect editorial standards, correction policies, and corporate governance documentation that addresses communication or media.

Why should publishing guidelines benefit your site? Well, I could see two factors at play here.

First off, Google's Page Quality Rating guidelines specifically notes that “High E-A-T news sources typically have published established editorial policies and robust review processes.” That doesn’t prove that the algorithm considers the presence of editorial guidelines (or even knows about them all the time) but it does lend us insight into the mind of Google.

Second, I’d be willing to bet that there’s a strong correlation between organizations that take the time to put together editorial guidelines and those that take the time to ensure their content is worthy of their site. Additionally, the process of putting together editorial guidelines is itself a good exercise in ensuring that your website content is high quality.

Lesson 4: Connections go a long way

No business operates in a vacuum, especially not on the Internet. The connections your organization has made with others, and how you acknowledge them, make an impact on how your community views you.

Reputable partners

There are all kinds of connections a business might make with another organization. Throughout the study, I kept track of something I called “reputable partners.” To earn this mark, a website had to demonstrate a relationship between themselves and another organization that’s plainly in support or favor of their work or mission.

Some of the most common types of demonstrations of these relationships included:

• Articles and press releases announcing partnerships or outcomes.

• Explanations of the relationships between those organizations.

• Accolades from recognized organizations highlighted on-site through badges, links to award announcements, press releases, etc.

• Links to press releases or articles demonstrating the relationship between organizations, and/or award badge displays.

While 73% of the results I looked at had clear “reputable partners,” 78% of those in the Top 3 did, too. My theory for this pattern is that making it obvious which other organizations are in support of you — generally or financially, e.g. through a grant — or in favor of your mission, you’re being transparent about how your organization operates. That fits squarely with the T in the E-A-T equation.

Backlinks

Another type of connection modern businesses deal in today is backlinks. Links put the “Inter” in “Internet,” and they’ve become essential for people and (more importantly for this subject) web crawlers to understand and navigate the web.

The average number of backlinks across all 647 results I analyzed was 32,572. Among the 210 Top 3 results, it was 88,581.

It’s certainly possible to get on Page 1 with fewer than that — about one-third had fewer than 100 backlinks and 28 had none whatsoever. However, we can clearly see that link quantity is valuable.

But what about link quality? For that, we can look at Moz’s Spam Score. This metric indicates your backlink profile health, with a 1% rating as really healthy and a 99% rating as super unhealthy.

While Moz considers a “low” score to be 30% or less, 44% of Page 1 results had a Spam Score of 1%, indicating that most Page 1 results have a very clean, healthy backlink profile. Another 19% had scores of 2 or 3%. The Top 3 results mirrored these results (with 44% at 1% and 18% at 2 or 3%).

Another way we can make some assumptions about link quality is by looking at referring domains. When there are loads of backlinks but very few referring domains, it seems less likely to be the result of deliberate link-building efforts. On the other hand, a higher number of referring domains could indicate more honest link-building tactics or simply just a really good web page that others want to link to.

The average number of referring domains among Page 1 results was 752. Meanwhile, among the Top 3 results, the average was 1,594. Making connections with other organizations online by way of honest link-building efforts can be one way to expand your reach, but also show Google and other search engines that you offer quality, worthwhile content.

Lesson 5: The right technology is essential

Last, but absolutely not least, if you have a website, it needs to be set up securely so that visitors can trust that they’re not putting their data at risk by interacting with it. In my study, I found that 96% of all results (also 96% of just top 3 results) used HTTPS. Interestingly, those that didn’t most often occurred in the military portion of the study.

Websites today also need technologies for cookie notifications, and some use pop-ups to convey important messages. Others use advertising to monetize their site. In any of these situations, the website owner should aim to minimize disruption to the user’s experience. Just 42% of all results had a pop-up: most of them (81) were inviting the user to subscribe to something (e.g. a newsletter), while nearly an equal number (79) were communicating information related to cookies.

Having the right technology enabled on your website may not seem inherently connected to E-A-T — which is why I didn’t evaluate even more technologic considerations such as e-payment systems — but considering that a huge aspect of Trustworthiness online today is about data gathering and management (and the ill effects of mis-management), it’s apparent that this area matters just as much, if not more so, than all your efforts into quality content creation.

Conclusion

When I set out to uncover the factors associated with E-A-T, I fully anticipated learning about proper author attribution, source citations, and good content. I guess I was thinking with my author hat on and not my web user hat, because I was only close on one of those three.

There are a lot of activities digital marketers can do to promote their businesses and goods and services today. Content creation and content marketing, link building, local SEO, advertising, public relations, and more can all seem like great options that you can pursue.

But the truth is, they’re not options — they’re must-haves for building a holistic digital presence. After conducting this study, my advice to webmasters and business leaders would be to assess your current online presence (including but not limited to your website’s user experience) and determine where there are holes. Working to fill those holes won’t necessarily be easy, but it will be worth it when your web traffic increases and your pages begin to rank.

To see a detailed explanation of each factor considered in this study, check out the full E-A-T study report on the Brafton blog.