Focus on genuine EEAT to determine what valuable content looks like

Tom says: “Focus on EEAT – but actual EEAT, not the fake one.”

What is fake EEAT?

“Right now, I see a lot of SEOs trying to figure out how to create EEAT without actually understanding what EEAT is. They are trying to find hacks (because we're SEO experts, we always wanted to find hacks) to simplify the whole thing and add fake EEAT into the articles they're writing.

When I'm talking about fake EEAT, I'm talking about just adding a bio or an author to the article and then thinking that you’re done. EEAT is a little bit more than that.

Google are now at a place where they understand what is valuable for the end user. There isn’t an easy way to go around it; find some kind of checklist for what we can put into an article, and make it happen. All the keyword-stuffing techniques that we used to use don’t work anymore.

We need to do what Google has wanted us to do for the last 20 years and create and add value through our content.”

Do other search engines look at EEAT and try and rank your content based upon the same principles?

“In my opinion, they do. From a data perspective, they definitely do.

However, Google is still the king. 90% of organic traffic to a website still comes from Google, so they haven’t gone away yet.

When it comes to LLMs, they are looking at similar factors. They want to see the authority and the trustworthiness in the things that we are building because they want to focus on the things that matter. They want to get the context that they need to provide the end user with the best possible answer.

To be honest, Google were in a comfortable space because they were giving us 10 results. When it comes to LLMs, they give us one result, so they need to give an answer in that result. There is no place for choice, so they need to find something of value to get the context that they need to answer the question that the user is asking.”

What practical steps can you take to ensure that as much of your content as possible resonates with EEAT?

“I went to multiple conferences where a lot of speakers were talking about EEAT, and I asked, ‘If you're saying that you need to implement EEAT in whatever you're doing, how do you measure that?’ Mostly, they said, ‘We don't; that's based on intuition,’ or, ‘We have a checklist,’ but that checklist was so formal that it just said, ‘Add the bio at the end,’ ‘Add the author,’ etc. – which doesn’t mean much, from my perspective.

You have to understand what EEAT is and how to define it, and clarify exactly what you expect from valuable content – and you need to start measuring it in that way. It's not a simple calculation where you can just find how many times you used a specific keyword inside an article. It's understanding what you treat as valuable, how you define value, and how you can measure it.

When you can measure it, then you are closer to the goal, because you know what you're looking for. If you know what you're looking for, you can find the path that leads you there.

The best solution that I have found is using LLMs for that. You can give them clarification from your side to benchmark what you treat as EEAT, and calibrate them for that. Then, you can get the numbers from them. You can provide them with a piece of content and measure exactly how you're doing with the content that you’re providing, and what you're missing. The whole conversation can go much deeper into that.”

What questions do you ask AI to ensure that it's telling you the right information about EEAT and what you need to do to improve?

“There are a couple of ways to deal with that. One way is to just go to an LLM and ask, ‘How do you define EEAT?’ You can start your journey from there. If you have some pieces of content that you believe are valuable, you can also provide them to the LLM and say, ‘These are articles that I consider to have perfect EEAT. They have nailed all four factors. Can you help me create a calibration rubric that will define exactly what EEAT needs?’

Once you have that rubric, you can go back to the LLM and use it to verify any new content you are working on. You can provide it with that content and say, ‘Please score this article on a scale of 1-10 based on the EEAT factors in this calibration rubric.’

It’s best to provide the article in markdown format, because when you copy-paste rich text to an LLM, you lose all the headings, the links, and everything else that’s in there.

You don’t want to just describe the different factors of EEAT; you want to create a calibration rubric so that you can get repeatable results. That rubric should show an LLM, for each element of EEAT, exactly what you consider to be a 5, a 7, or a 10 out of 10, based on the things that you’re creating.”

How do you know whether the rules that you define reflect the rules that Google are looking for?

“You will see it in Google Search Console. If you define it properly, you'll be closer to getting it right. It’s a leap of faith metric that can give you results right now. Of course, you can base it on your intuition, but if you can't create a repeatable system to get to your goal, you're not able to iterate and make it better. It’s an assumption that you will make based on the factors that you're looking for.

If you don't have such a metric and you're not measuring it, then you're going all over the place, and you can't pivot from that. You can't make the process better. You can't iterate and make these factors more aligned with Google Search Console.”

What's a common thread in terms of what websites are currently missing in their content nowadays, from an EEAT perspective?

“Actually, content has become worse in the last two years – from the moment that AI was released – because most content writers are using AI on a daily basis.

When they do that, they're getting lazy, and they're getting information straight from AI. They're not delivering the additional value of their own experience, their expertise, their case studies, etc. They aren’t doing their research; they're only pulling the data that AI is giving them. That's why content has become weaker.

There are also much lower margins on content right now. Agencies are being squeezed for margins by customers, so they have to reduce costs, and they're pushing their contractors to reduce costs. Therefore, contractors are starting to use AI at scale, but in a way that creates fluff, empty, and shallow content.

We have to focus on the quality, and we have to define what quality is for us. For me, it's data. If I have a hypothesis, I need to find some data to prove it. If I have a case study from my side, I want to show it. If I have some original data, opinions, or quotes, I want to show them. If you put all of those factors together, you will create something of quality.”

How do you articulate to stakeholders in a business that they have to invest more money in the production of content, when it's so easy to do using AI?

“Through the quality and the results.

Many agencies brag about having human-written content, and they have discussions internally about whether they should go into AI-written content or not. To be honest, though, because they are using contractors, most of them have already switched without realising it. Most contractors have switched to writing content with AI, but agencies are not aware of that, so they're selling content as human-written when it's not.

As agencies, we should gain control of all the things that we're doing and create a process that focuses on quality at the end. To gain control, you have to help the contractors, or the people inside the company, actually have an SOP that will be strict on providing value to the end user.

To do that, you again have to go back to the rubric that I mentioned earlier. You have to define what EEAT is for you. If you don't define what good looks like, how do you expect to achieve it? That's impossible.”

How do you know when to involve humans and when to involve AI within the production of your content?

“The question isn’t about when; it’s about what humans are good at and what AI is good at.

From the perspective of a human, AI is much better at a lot of things that I would do at a much slower speed. For example, research, or writing a specific piece of content. I might be worse at doing that, but I have a lot of experience. I know my company inside out. I can actually strategise on the whole piece of content, and much more.

It’s about cooperation between a human and AI. I call it the cyborg method: a human working together with AI, where both sides give their best value. AI gives the superpowers of AI, humans give superpowers of humans, and you combine these skills to provide better value at the end.”

Should every piece of content incorporate EEAT, or is it only relevant for content that is discoverable from an organic search perspective?

“I was actually writing an article a week ago where I had to sacrifice EEAT because my main goal was to increase the lead conversion rate to calls on a specific piece of content. I had to sacrifice EEAT for that, so as not to disturb the user from going along the path that I wanted them to go through.

Sometimes you will have to sacrifice it, but you have to do it with an understanding of what you're doing. I knew exactly what I had to do, and I knew that I would sacrifice EEAT for a different goal.”

Tom, what's the key takeaway from the tip you shared today?

“Start measuring what you're doing, and define what good looks like. Then, follow that path to get to good.”

Tom Winter is Founder and CGO at SEOwind. Find out more over at SEOwind.io.