[EEAT Framework] How artificial intelligence will impact SEO & content creation

this post was written by a subject matter expert 👀 not a LLM

Apr 20, 2023

Google has a lot of economic incentives to unleash the ban hammer upon the hoard of AI-generated content that is being created right now. There are legitimate serious problems that arise when anyone is able to pump out 10,000 articles a month of AI spam.

Thankfully, they recently released some language around how they are thinking about this.

Google claims to reward high-quality content, however it is produced:

“Google's ranking systems aim to reward original, high-quality content that demonstrates qualities of what we call E-E-A-T: expertise, experience, authoritativeness, and trustworthiness.”

Original “high-quality” content.

The very first word of this statement is original. Is anything that AI generates original content? It has the ability to imagine new worlds, but it’s trained off the content that already exists on the internet.

Hell … devil’s advocate?

I’m trained off content that already exists on the internet. The content that I produce is a combination of thousands of podcasts, YouTube videos, tutorials, audiobooks, and then that knowledge is applied & tested against real-world situations which I then distill the learnings into my writing.

It feels like a bit of a catch-22.

Google is saying that they don’t care how the content is produced, but they are going to reward content that is original, high quality, created by (possibly aided with AI) trusted, authoritative, subject matter experts.

Let’s break down the E-E-A-T framework Google is looking at in the age of AI.

Expertise

When it comes to creating high-quality content, the expertise of the author who wrote the content is incredibly important. In addition to EEAT they also frequently refer to the “who how why” framework for creating content.

The very first part of both of these frameworks revolves around the “who”.

It needs to be really really clear who created the content. Was this article written by an AI under a fake pen name, or was it created by Koby Conrad the Head of Growth @ Rupa Health™?

If AI was used to create this article (it wasn’t, although I could have used it to help me generate ideas), it still needs to actually be made by a subject matter expert who has strong expertise in the subject they are writing about.

Google is telling us that it still deeply & fundamentally cares about who is creating the content. Personally, I sincerely doubt an article that is completely written by AI will count as content written with expertise.

Experience

Experience primarily talks about the user’s experience when they visit a page. Are they hit with a wall of ads? Does the page take 30 seconds to load? Is the “product” of your landing page high quality?

This is probably the part of the EEAT framework that is least impacted by content written by an AI.

There are a number of interesting ways though that AI could improve the experience of the page:

AI tools that automate optimizing the technical aspects of landing pages, optimizing load times, scripts, ads, etc.
Chatbots trained on the content of a website → Didn’t find your answer from the article or want to know something else? Ask a chatbot trained on the contents of your website at the bottom.
Mutiny style personalization, combining data to personalize the content that a person could be looking for (taking an article about “lab testing” into an article about “lab testing for naturopaths”).

It’s still extremely early days, but there ARE likely ways for AI to drastically improve the on-page experience for readers without ever directly creating the content.

Authoritativeness

Authoritativeness has traditionally been primarily tied to link building, basically how important does Google think you are?

Content created by AI likely won’t change how authoritative it thinks your words are, but content other people create (possibly with AI) might.

Even if some of it is spam, AI is beginning to massively increase the amount of content & articles being created on the internet. This is also increasing the number of backlinks being created.

Two key concepts around this:

AI is trained on the internet … so is AI-generated content only going to link to old articles that existed before the models were trained?
If you use AI to generate more content, even if it is “lower quality” would this increase the number of possible touch points for someone to start linking back to you?

Any AI content you create likely won’t directly affect your authoritativeness, but if you use it to flesh out the amount of knowledge that exists on your domain - as long as it’s high enough quality for someone to link to, it could indirectly impact your domain authority.

Trustworthiness

Trustworthiness involves things like spammy backlinks, bad actors, and domains that get penalized because they’ve tried to abuse Google’s system.

This is the most dangerous part of the EEAT framework when it comes to the impact of AI-generated content. Google says it doesn’t care about how the content is produced, but will AI-generated content continue to be a neutral signal when it comes to trustworthiness?

Google has many years of experience dealing with automation being used in an attempt to game search results. Our spam-fighting efforts—including our SpamBrain system—will continue, however spam is produced.

The internet is filled with spam.

Google IS Google because for the last 20+ years they have been the reigning champion of fighting spam. Their entire business revolves around identifying the most important piece of content and matching it with a user’s search query.

I sincerely doubt that the long-term view of content that is 100% generated by AI is going to be looked at as anything other than spam.

Imagine a world where content creators are:

Pushing out 1,000 - 10,000+ articles a month powered by AI.
The AI content has a 1:100 ratio of content written by a human vs content written by a machine.
EVERYONE is doing it.

There is arbitrage in the world right now where mass-producing AI content is generating insane results for some of the smartest marketers I have the privilege of knowing.

I have nothing but respect and admiration for my friends that are doing this right now, I think it’s going to be insanely profitable for them.

My most tin-foil-hat theory is that Google will eventually punish this kind of content, and that they will look at human-generated content as a major quality signal of not just expertise but also trustworthiness.

Backlinks are good, spammy backlinks are bad. Content is good, spammy content is bad.

In my mind, it comes down to “are we giving incredible answers & information to people’s most important questions”?

Google says it doesn’t actually matter how you created these answers, just that they are incredible.

How am I making content?

I started writing this Substack because written content is the only kind that I am halfway decent at creating. I would like to think that for the foreseeable future, i’ll continue to be able to create work that is 10x better than a LLM.

Putting the title of this post into Chat GPT gives me a bit of hope that this is true … for now.

Also - I enjoy it. I write about things that inspire me, and putting my words on page helps me to internalize concepts I’m focused on mastering myself. Channels of Growth won’t be handing the keys over to an AI LLM any time soon.

As far as Rupa Health goes? Our strategy doesn’t really change.

Every month we publish over 100 articles written by subject matter experts, doctors, licensed healthcare providers, and incredibly smart humans who are actively “doing the work”.

I think we’ll use AI to help us generate ideas and create article outlines, but ultimately it will be humans who create the content with AI serving as a tool that helps us level up the quality of our work.

And we’re going to remember that this is a long-term game. Publish 3 articles a day for 10 years - a slow stoic march to the heavens. 🔥