AEO, AIO, GEO, SEO, BLA BLA BLA

AEO, AIO, GEO, SEO, BLA BLA BLA

AEO, AIO, GEO, SEO, BLA BLA BLA

SEO

SEO

5 min

5 min

being a startupper

“We’ll optimize your content for AI”… eeehh!!!!

For quite a long time, I’ve wanted to write about AI & SEO.

Then, while cycling with a friend along Baku Boulevard, we had a short conversation about AEO.

“Oh shit. I started this blog, but I never finished it. Narmina, you’re always like this. Why don’t you ever finish what you start?” So I think I lost track of what he was saying at that point. HEHE. I was scolding myself in my head. Sorry.

Before I lose focus again, here we go.

So.. every other week there’s a new three-letter word on my LinkedIn or X feed.

AEO, GEO, AIO.. While I was typing this someone probably came up with a new one.

And in every one of the posts come panic, screaming that SEO is DEAD and if you don’t pay them $15k/month to “optimize for AI,” your business will doomed TOMORROW.

What are those terms?

What's been going on?

First, what are AEO, GEO, AIO?

Same shit, different color. You ran rename it but the smell is the same.

Monkey looking at his nails
  • SEO: Search Engine Optimization, we know this is the OG. Helping your page rank in search results.

  • AEO: Answer Engine Optimization. AEO gets your web content into the “answer” like featured snippets, Google AI Overviews, featured snippets, voice search.

  • GEO: Generative Engine Optimization. Writing, no, optimizing content so generative AI tools like ChatGPT, Perplexity, Gemini, Claude cite or mention your brand/content in generated answers.

  • AIO: AI Optimization. This a broader term. Basically the same as GEO with a different hat on. It means making your content easier for AI systems to understand, summarize, recommend, or show in AI search results.

Whatever these are, they are still closely related to SEO. The labels are new, but the SEO fundamentals stay true. Search systems still need to understand your content and anything else you create or share about your brand across the web, wherever it appears.

We all know these. So, why are ye fearful?

If you pay attention, you’ll see they are all variations of the same thing: be the thing the machine picks when someone asks a question whether in AI tools or search engines' input boxes.

Why is everyone losing their mind?

Let's check some data to calm you down

Google itself is the single largest AI search tool by a mile. Around 16% of its results already show AI Overviews. So this isn’t “Google vs. AI.” It’s just that Google is becoming more AI-influenced. And, it's a much less dramatic than you might have thought.

I've recently find a research by SparkToro / Datos, Q4 2025, US desktop which is about the share-of-search numbers.

  1. Q4 2025 Share of Search Across 41 Major Sites: Google ~73%, Bing 4.31%, ChatGPT 2.86%, DuckDuckGo 1.26%, eBay 1.06%.

  2. Q4 2025 Share of Searches by Platform Type: Traditional Search Engines 80.76%, Commerce Platforms 9.76%, Social Networks 5.37%, AI Tools 3.19%, Other Verticals 0.91%.

Google is responsible for nearly three-quarters of all desktop web searches. Whereas, AI tools like ChatGPT, Claude, DeepSeek, all of them combined make up about 3.2%.

3.2%.

Three point two.

SEO is certainly not dead. What’s changed then?

Search behavior and the definition of “winning” - the value.

In traditional search, someone searches → they sees your link → click to the link → lands on your page. The value in traditional search is that you get a click - visitor. Click is the value here.

In AI search, someone searches → they sees your website as source, or mentioned. Here, the value is you get visibility. Mention is the value here.

However, just because AI sends you less traffic - less visitor (i.e., people don't click your website from SERP AI Overview), this doesn't mean SEO is dead. Yes, AI may send less traffic, but it creates another value - it mentions your website, it shows your website as source - you gain credibility and visibility.

This is like someone walking past your store, noticing your window display, but not coming in. Does that mean nothing happened? Nope. The value of creating that display for your store has already paid off.

The same applies to AI search. If you're only measuring AI search by “how much traffic did it send me?”, you're probably missing the point. Someone may see your brand in an AI answer, remember it, and may visit your site later/buy from you later.

You might want to read this 👉 People who come from SEO, buy. Is that true?

The way people search is changing, and the way we measure success is changing too. But the fundamentals of earning that success haven't changed. You still need to create quality content and build credibility. Without them, you get no SEO, no AIO, no GEO, no no nothing.

In AI search, you might not get direct/immediate clicks; so, zero-click searches increase.

Is it bad? Not at all.

People first search “what even is this” questions inside, let's say in ChatGPT. They don’t visit your site at the curiosity stage. They show up later.

So yeah, your overall traffic might decrease. But the traffic that DOES arrive is more decision-ready. They already made their mind up somewhere. They’re in the decision/consideration stage when they visit your website (and this is where your web presence should be ready to convert).

Less traffic, but hotter traffic.

Recommended Google documentation: Optimizing your website for generative AI features on Google Search PS: nothing new here.

How to optimize for AI, then? How to get cited, visibility?

Create quality content & improve your writing skills

When people talk about optimizing content for GEO, AEO, AIO, LLMO, or AI search, they often make it sound like you need a completely new way of writing.

If you know me, if we ever talked about this stuff, you know how I answer: create quality content.

The foundation is simple: write in a way that a human can clearly understand.

Why do I always repeat the same thing like a parrot? Either I don't know anything, or I know exactly how to do it.

What the f is quality content?

Imagine I ask you:

“Where did you put my phone?”

And you answer:

“It’s at home.”

Technically, that is an answer. But it is not useful.

Where at home? In the bedroom? On the kitchen counter? Inside my bag? Under the pillow? On the charger? I still have to search everywhere because the answer is too vague.

Or worse, you start talking about the phone’s features, completely unrelated to the question.

A clearer answer would be:

“Your phone is at home, on the small table next to the sofa in the living room. I put it there because it was charging, and the charger is still plugged in.”

This answer is better because it gives context. It tells me the exact place, the room, the object it is near, and why it was placed there. I do not have to guess.

Content works the same way.

If we look at the recommendations for getting cited by AI, we see the following:

  • Aim for a Flesch Reading Ease score of 65 or higher.

  • Use Markdown formatting, such as bullet points and short, easy-to-scan sections. Chunkable format, they say, chunkable format.

  • Answer users’ questions directly.

Are these new things? No.

A Flesch Reading Ease score of 65 or higher simply means writing in a way that is easy for a general audience to understand. Avoid technical terms and jargon unless they are necessary. It's almost as if Google has never told us to write clear, readable content before 🤷‍♀️

Then there's “chunk optimization.” Some people throw this term around like it's a new growth hack. It's not. They've taken a technical term from AI engineering and dressed it up as a tactic they can sell you.

Why use technical terms when you can simply say: use clear headings and subheadings, write short paragraphs, and structure your content with easy-to-scan sections?

It's said that writing in MD format, using headings, bullet points, and clear sections helps with AIO, AEO, LLMO, GEO. But point here is not about writing robotic content for AI.

Or, it's recommended to answer user questions directly. Because what other choice do you have? If you want to create quality content, why wouldn't you give a clear answer to the question your reader is asking?

Why make things more complicated than they need to be?

All the recommendations, is just.. writing well (and be digitally present in the right places, which I talk about it below).

What you should do is to write atomic content. Self-contained sections that answer one thing clearly and can stand on their own.

Which, funnily enough, is the exact same thing I was telling when I was still a student back in 2015. Read my post: How to write an article that make people say “oh my blog!”

PS: Do you know why are internal links useful for SEO and they are still strong ranking signal? Yes, search engines use internal links to discover and crawl other pages on your site. But that's not the only reason you add them.

You add an internal link when you mention something you’ve already written about in another article. Instead of explaining everything all over again, you can link to that article. If you’ve made your reader curious, they can click the link and learn more. That creates a better reading experience. Your readers get the information they need without having to read through a bunch of extra sentences they didn’t ask for.

The same applies to external links btw. If you mention a study, statistic, or source, you link to it so people don't have to leave your content and search for the source themselves.

You also show where the information came from. You're making it clear that you didn't come up with the information yourself or copy it without credit. You got it from a source, and you link to that source. This is writing 101.

At the end of the day, most of these recommendations are recommended for one reason: they make content better, easier for people to consume.

And what works well for humans usually works better for SEO, GEO, AEO, AIO, and LLMs too.

Recommended documents by Google:

Build credibility & reputation

To build credibility, you should also invest in your digital presence.

Where and how does your brand show up across the internet? How is it talked about on the platforms LLMs consume?

I realized this recently in the most ordinary way while working on an appliance service website. It was a bit of a “no shit” realization, but it suddenly clicked: if we wanted an LLM to recommend an appliance service, that service needed to exist on the platforms the model trusts.

For this type of business, that might mean being present on platforms like Thumbtack, Trustpilot, business directories, and other review platforms.

If you’re invisible there, you may be invisible to the model.

And for a local service business, you also need a well-maintained Google Business Profile with strong reviews. Why would an LLM recommend a business if its opening hours are unclear, its service offerings aren't obvious, or it has consistently bad reviews?

You can also look at Google's patents on entities, as well as the work of my SEO teacher, Bill Slawski:

Put together, these patents and article will help you understand the backbone of how search systems decide whether your website is worth ranking/recommending.

So, it you want to be mentioned, you need to make sure your brand is a known, well-regarded entity in the right corners of the web. That’s the work.

2 things worth mentioning:

  • First: LLMs don’t read your page the way you think. They don’t “read”, they vectorize your text token by token. Tokens can be whole words or fragments of words.

  • Second: You can’t control how anyone chunks your content. Google, ChatGPT, Perplexity - they all have pipelines that are different for their cost, model, and context.

Anyway.

In short

Don’t fall for the acronym of the month you see on IN or X, try to learn the stuff from the source, understand the basics and how things work, and learn from the pioneers (i.e., John Mueller).

Write good stuff and be where you must be with quality content (leave no quesition to humans so does ai) The rest sorts itself out.

Cheers.

PS: Create quality content.

“We’ll optimize your content for AI”… eeehh!!!!

For quite a long time, I’ve wanted to write about AI & SEO.

Then, while cycling with a friend along Baku Boulevard, we had a short conversation about AEO.

“Oh shit. I started this blog, but I never finished it. Narmina, you’re always like this. Why don’t you ever finish what you start?” So I think I lost track of what he was saying at that point. HEHE. I was scolding myself in my head. Sorry.

Before I lose focus again, here we go.

So.. every other week there’s a new three-letter word on my LinkedIn or X feed.

AEO, GEO, AIO.. While I was typing this someone probably came up with a new one.

And in every one of the posts come panic, screaming that SEO is DEAD and if you don’t pay them $15k/month to “optimize for AI,” your business will doomed TOMORROW.

What are those terms?

What's been going on?

First, what are AEO, GEO, AIO?

Same shit, different color. You ran rename it but the smell is the same.

Monkey looking at his nails
  • SEO: Search Engine Optimization, we know this is the OG. Helping your page rank in search results.

  • AEO: Answer Engine Optimization. AEO gets your web content into the “answer” like featured snippets, Google AI Overviews, featured snippets, voice search.

  • GEO: Generative Engine Optimization. Writing, no, optimizing content so generative AI tools like ChatGPT, Perplexity, Gemini, Claude cite or mention your brand/content in generated answers.

  • AIO: AI Optimization. This a broader term. Basically the same as GEO with a different hat on. It means making your content easier for AI systems to understand, summarize, recommend, or show in AI search results.

Whatever these are, they are still closely related to SEO. The labels are new, but the SEO fundamentals stay true. Search systems still need to understand your content and anything else you create or share about your brand across the web, wherever it appears.

We all know these. So, why are ye fearful?

If you pay attention, you’ll see they are all variations of the same thing: be the thing the machine picks when someone asks a question whether in AI tools or search engines' input boxes.

Why is everyone losing their mind?

Let's check some data to calm you down

Google itself is the single largest AI search tool by a mile. Around 16% of its results already show AI Overviews. So this isn’t “Google vs. AI.” It’s just that Google is becoming more AI-influenced. And, it's a much less dramatic than you might have thought.

I've recently find a research by SparkToro / Datos, Q4 2025, US desktop which is about the share-of-search numbers.

  1. Q4 2025 Share of Search Across 41 Major Sites: Google ~73%, Bing 4.31%, ChatGPT 2.86%, DuckDuckGo 1.26%, eBay 1.06%.

  2. Q4 2025 Share of Searches by Platform Type: Traditional Search Engines 80.76%, Commerce Platforms 9.76%, Social Networks 5.37%, AI Tools 3.19%, Other Verticals 0.91%.

Google is responsible for nearly three-quarters of all desktop web searches. Whereas, AI tools like ChatGPT, Claude, DeepSeek, all of them combined make up about 3.2%.

3.2%.

Three point two.

SEO is certainly not dead. What’s changed then?

Search behavior and the definition of “winning” - the value.

In traditional search, someone searches → they sees your link → click to the link → lands on your page. The value in traditional search is that you get a click - visitor. Click is the value here.

In AI search, someone searches → they sees your website as source, or mentioned. Here, the value is you get visibility. Mention is the value here.

However, just because AI sends you less traffic - less visitor (i.e., people don't click your website from SERP AI Overview), this doesn't mean SEO is dead. Yes, AI may send less traffic, but it creates another value - it mentions your website, it shows your website as source - you gain credibility and visibility.

This is like someone walking past your store, noticing your window display, but not coming in. Does that mean nothing happened? Nope. The value of creating that display for your store has already paid off.

The same applies to AI search. If you're only measuring AI search by “how much traffic did it send me?”, you're probably missing the point. Someone may see your brand in an AI answer, remember it, and may visit your site later/buy from you later.

You might want to read this 👉 People who come from SEO, buy. Is that true?

The way people search is changing, and the way we measure success is changing too. But the fundamentals of earning that success haven't changed. You still need to create quality content and build credibility. Without them, you get no SEO, no AIO, no GEO, no no nothing.

In AI search, you might not get direct/immediate clicks; so, zero-click searches increase.

Is it bad? Not at all.

People first search “what even is this” questions inside, let's say in ChatGPT. They don’t visit your site at the curiosity stage. They show up later.

So yeah, your overall traffic might decrease. But the traffic that DOES arrive is more decision-ready. They already made their mind up somewhere. They’re in the decision/consideration stage when they visit your website (and this is where your web presence should be ready to convert).

Less traffic, but hotter traffic.

Recommended Google documentation: Optimizing your website for generative AI features on Google Search PS: nothing new here.

How to optimize for AI, then? How to get cited, visibility?

Create quality content & improve your writing skills

When people talk about optimizing content for GEO, AEO, AIO, LLMO, or AI search, they often make it sound like you need a completely new way of writing.

If you know me, if we ever talked about this stuff, you know how I answer: create quality content.

The foundation is simple: write in a way that a human can clearly understand.

Why do I always repeat the same thing like a parrot? Either I don't know anything, or I know exactly how to do it.

What the f is quality content?

Imagine I ask you:

“Where did you put my phone?”

And you answer:

“It’s at home.”

Technically, that is an answer. But it is not useful.

Where at home? In the bedroom? On the kitchen counter? Inside my bag? Under the pillow? On the charger? I still have to search everywhere because the answer is too vague.

Or worse, you start talking about the phone’s features, completely unrelated to the question.

A clearer answer would be:

“Your phone is at home, on the small table next to the sofa in the living room. I put it there because it was charging, and the charger is still plugged in.”

This answer is better because it gives context. It tells me the exact place, the room, the object it is near, and why it was placed there. I do not have to guess.

Content works the same way.

If we look at the recommendations for getting cited by AI, we see the following:

  • Aim for a Flesch Reading Ease score of 65 or higher.

  • Use Markdown formatting, such as bullet points and short, easy-to-scan sections. Chunkable format, they say, chunkable format.

  • Answer users’ questions directly.

Are these new things? No.

A Flesch Reading Ease score of 65 or higher simply means writing in a way that is easy for a general audience to understand. Avoid technical terms and jargon unless they are necessary. It's almost as if Google has never told us to write clear, readable content before 🤷‍♀️

Then there's “chunk optimization.” Some people throw this term around like it's a new growth hack. It's not. They've taken a technical term from AI engineering and dressed it up as a tactic they can sell you.

Why use technical terms when you can simply say: use clear headings and subheadings, write short paragraphs, and structure your content with easy-to-scan sections?

It's said that writing in MD format, using headings, bullet points, and clear sections helps with AIO, AEO, LLMO, GEO. But point here is not about writing robotic content for AI.

Or, it's recommended to answer user questions directly. Because what other choice do you have? If you want to create quality content, why wouldn't you give a clear answer to the question your reader is asking?

Why make things more complicated than they need to be?

All the recommendations, is just.. writing well (and be digitally present in the right places, which I talk about it below).

What you should do is to write atomic content. Self-contained sections that answer one thing clearly and can stand on their own.

Which, funnily enough, is the exact same thing I was telling when I was still a student back in 2015. Read my post: How to write an article that make people say “oh my blog!”

PS: Do you know why are internal links useful for SEO and they are still strong ranking signal? Yes, search engines use internal links to discover and crawl other pages on your site. But that's not the only reason you add them.

You add an internal link when you mention something you’ve already written about in another article. Instead of explaining everything all over again, you can link to that article. If you’ve made your reader curious, they can click the link and learn more. That creates a better reading experience. Your readers get the information they need without having to read through a bunch of extra sentences they didn’t ask for.

The same applies to external links btw. If you mention a study, statistic, or source, you link to it so people don't have to leave your content and search for the source themselves.

You also show where the information came from. You're making it clear that you didn't come up with the information yourself or copy it without credit. You got it from a source, and you link to that source. This is writing 101.

At the end of the day, most of these recommendations are recommended for one reason: they make content better, easier for people to consume.

And what works well for humans usually works better for SEO, GEO, AEO, AIO, and LLMs too.

Recommended documents by Google:

Build credibility & reputation

To build credibility, you should also invest in your digital presence.

Where and how does your brand show up across the internet? How is it talked about on the platforms LLMs consume?

I realized this recently in the most ordinary way while working on an appliance service website. It was a bit of a “no shit” realization, but it suddenly clicked: if we wanted an LLM to recommend an appliance service, that service needed to exist on the platforms the model trusts.

For this type of business, that might mean being present on platforms like Thumbtack, Trustpilot, business directories, and other review platforms.

If you’re invisible there, you may be invisible to the model.

And for a local service business, you also need a well-maintained Google Business Profile with strong reviews. Why would an LLM recommend a business if its opening hours are unclear, its service offerings aren't obvious, or it has consistently bad reviews?

You can also look at Google's patents on entities, as well as the work of my SEO teacher, Bill Slawski:

Put together, these patents and article will help you understand the backbone of how search systems decide whether your website is worth ranking/recommending.

So, it you want to be mentioned, you need to make sure your brand is a known, well-regarded entity in the right corners of the web. That’s the work.

2 things worth mentioning:

  • First: LLMs don’t read your page the way you think. They don’t “read”, they vectorize your text token by token. Tokens can be whole words or fragments of words.

  • Second: You can’t control how anyone chunks your content. Google, ChatGPT, Perplexity - they all have pipelines that are different for their cost, model, and context.

Anyway.

In short

Don’t fall for the acronym of the month you see on IN or X, try to learn the stuff from the source, understand the basics and how things work, and learn from the pioneers (i.e., John Mueller).

Write good stuff and be where you must be with quality content (leave no quesition to humans so does ai) The rest sorts itself out.

Cheers.

PS: Create quality content.