GEO is becoming the next big SEO conversation, but I think we are already simplifying it too much.
Everyone is suddenly talking about GEO.
Optimize your content for ChatGPT. Add FAQs. Use schema. Create answer-first content. Publish more statistics. Add an llms.txt file.
Some of those things can help.
But I think we are starting to make the same mistake we made with SEO years ago.
We are turning a much bigger system into a checklist.
The more I work around SEO, entity optimization, AI visibility, LLMs, AEO and GEO, the more convinced I am that Generative Engine Optimization is not mainly a content optimization problem.
It is an evidence problem.
An AI system needs enough signals to answer a few basic questions:
- Who are you?
- What are you known for?
- Who else confirms that?
- Is your information worth using?
- Is there enough consistent evidence to confidently mention or cite you?
That changes how I think about GEO completely.
Ranking a Page and Becoming an AI Source Are Not the Same Thing
Traditional SEO trained us to think in a fairly predictable sequence:
Keyword → Page → Backlinks → Rankings → Traffic → Conversion
That model still matters.
But generative search introduces another layer.
Someone asking ChatGPT, Gemini, Perplexity or another AI search engine a question may never see a traditional list of ten blue links.
They may receive one synthesized answer.
And inside that answer, only a small number of brands, websites, experts or sources may be mentioned.
So the question changes.
Old question
Can my page rank?
Better GEO question
Why would an AI system choose my information as part of its answer?
That is a much harder question.
And "because my page is optimized" is not a good enough answer.
Your Website Saying You Are an Expert Does Not Make You an Expert
Imagine I create a website tomorrow and write:
Piyush Kumar is one of the leading experts in Generative Engine Optimization.
I can add Person schema.
I can add FAQ schema.
I can optimize the title.
I can create twenty GEO articles.
I can repeat the same expertise statement across every page.
But from an information-retrieval perspective, something important is still missing.
Independent evidence.
Now imagine that the same person is consistently associated with GEO across articles, developer platforms, interviews, business profiles, communities, videos, research, social profiles and third-party publications.
That creates a very different information environment.
Think of the relationship like this:
Piyush Kumar → GEO → AI Search → Entity Optimization → LLM Visibility
The website is no longer the only source making the claim.
The web starts reinforcing the relationship.
That is much more interesting to me than simply "optimizing a page for GEO."
Because now we are building an entity.
I Think GEO Has Three Layers
The easiest way I currently explain GEO is through three layers.
Layer 1: What You Say About Yourself
Your website, service pages, articles, author profile, About page, structured data, case studies, FAQs and documentation.
Layer 2: What the Web Says About You
Publications, communities, reviews, profiles, videos, directories, interviews, Reddit discussions, YouTube mentions, podcasts and third-party references.
Layer 3: What Is Actually Worth Citing
Original data, research, frameworks, experiments, tools, datasets, benchmarks, case studies, useful guides and real-world expertise.
Let's look at these individually.
Layer 1: What You Say About Yourself
This layer is completely under your control.
And it is important.
But there is an obvious weakness.
You control it.
Anyone can call themselves an expert on their own website.
Anyone can publish a page saying they are the best company in their industry.
Anyone can create schema reinforcing that claim.
Anyone can publish dozens of articles around the same topic.
That is why the second layer becomes important.
Layer 2: What the Web Says About You
This is where things get more interesting.
What websites mention your brand?
What publications reference you?
Where are you listed?
What communities discuss you?
What reviews exist?
What videos mention you?
What topics are repeatedly connected with your name?
This may include:
- GitHub
- Medium
- DEV.to
- YouTube
- Quora
- industry publications
- niche communities
- directories
- podcasts
- interviews
- review platforms
- business databases
I do not see these simply as backlinks.
I see them as corroborating evidence.
A backlink says:
This page links to another page.
A brand mention can say:
This company is associated with this topic.
A review adds reputation context.
An interview can establish expertise.
A community discussion can show that people are independently talking about a product or brand.
Those are different signals.
Traditional off-page SEO often puts many of them into one bucket.
GEO gives us a reason to look at them individually.
Layer 3: What Is Actually Worth Citing
This is the layer I think will separate strong GEO strategies from weak ones.
Ask yourself:
What have I published that another person, journalist, researcher, search engine or AI system would genuinely want to reference?
Not just read.
Reference.
That could be:
- original data
- an industry survey
- a useful framework
- a case study
- a benchmark
- an experiment
- a calculator
- a dataset
- proprietary research
- a detailed technical explanation
- an original comparison
- real-world findings
This is what I call a citation asset.
There are already millions of generic articles explaining:
- What is SEO?
- What is GEO?
- What is AEO?
- What is RAG?
- What is an AI agent?
Producing another version of the same information does not create much reason to cite you.
But publishing something other sources cannot easily reproduce changes the equation.
Original information creates differentiation.
Differentiation creates citability.
This Is Why I Am Less Excited About Mass AI Content
AI makes publishing incredibly cheap.
That is useful.
But it creates another problem.
When everyone can publish one hundred reasonably good articles, publishing one hundred reasonably good articles becomes less impressive.
The scarcity shifts.
What becomes scarce?
- Original information
- Real experience
- Useful data
- Recognizable expertise
- Trusted relationships
- External validation
AI can help me write an article.
It cannot automatically create years of industry experience.
It cannot manufacture legitimate customer feedback.
It cannot create genuine third-party recognition.
It cannot make respected people independently mention a brand.
And that is where GEO becomes much more interesting to me than simply prompt engineering or content formatting.
GEO Is Starting to Look Like Entity Building
Think about how an AI system might encounter a person or company.
It could find the official website.
Then LinkedIn.
Then GitHub.
Then Crunchbase.
Then an article.
Then a YouTube video.
Then a podcast.
Then a Reddit discussion.
Then a directory profile.
Then another publication mentioning the same person.
Each source adds another small piece of information.
Individually, none of those signals may mean much.
Together they start forming an entity.
That is why consistency matters.
The same name.
The same organization.
The same expertise.
The same relationships.
The same topics.
The same identity.
I sometimes think about entity optimization as reducing ambiguity.
The easier it is for machines to answer:
Who exactly is this?
the stronger the foundation becomes.
GEO Also Changes How I Think About Off-Page SEO
For years, the main question in off-page SEO was:
Where can I get a backlink?
I think the GEO version of that question is broader:
Where should this entity exist?
Maybe the best opportunity is a link.
Maybe it is an industry profile.
Maybe it is a review.
Maybe it is a podcast.
Maybe it is a Reddit discussion.
Maybe it is a YouTube mention.
Maybe it is a research citation.
Maybe it is a developer profile.
Maybe it is an expert contribution.
The goal is **
