The AI Search Experiment: What Google, ChatGPT & Gemini Agree On

We compared Google Search, Google AI Overviews, ChatGPT, and Gemini to understand what really influences AI recommendations—and why SEO remains the foundation of AI visibility.

Important: This experiment represents observations from a specific commercial search query. AI platforms are constantly evolving, and recommendations can vary depending on user intent, location, personalization, and model updates. Rather than proving a universal rule, this study highlights patterns that businesses should consider when developing their SEO and AI search strategies.

Over the last year, one topic has dominated conversations around AI search.

Prompt tracking.

Businesses are spending thousands of dollars every month trying to answer one question:

“Does ChatGPT recommend my business?”

New AI visibility tools appear almost every week.

Most promise the same thing.

Track prompts.

Monitor mentions.

Measure AI visibility.

But after spending years helping businesses improve their visibility through SEO and digital marketing, something didn’t quite add up.

We kept asking ourselves a different question.

Where do AI recommendations actually come from?

Does ChatGPT randomly decide which brands to recommend?

Does Google AI Overviews simply summarize the highest-ranking pages?

Does Gemini use an entirely different system?

Or is there something much deeper happening?

To answer those questions, we decided to run a simple experiment.

Instead of analyzing prompts, we analyzed the relationship between traditional Google rankings and AI recommendations.

What we discovered completely changed the way we think about AI SEO.

The AI SEO Industry Might Be Looking at the Wrong Metric

Today, almost every AI visibility platform starts at the same point.

Prompt

↓

Did AI mention my brand?

But that’s measuring the result.

Not the cause.

Imagine asking:

Why is HubSpot recommended?

A prompt tracker answers:

Because it appeared.

That’s like asking why someone won an Olympic gold medal and answering:

Because they finished first.

Technically true.

Practically useless.

The real question is:

Why did AI choose HubSpot over hundreds of other CRM companies?

That’s where things become interesting.

Our Hypothesis

Before running the experiment, we had a theory.

We believed AI recommendations were not created in isolation.

Instead, they were likely influenced by years of SEO, content marketing, topical authority, and brand recognition.

In other words…

AI wasn’t replacing SEO.

AI was building on top of it.

The Experiment

We chose one commercial keyword.

Best CRM Software for Startups

Why?

Because it meets several important criteria.

High commercial intent
Extremely competitive
Multiple well-known brands
Rich AI answers
Strong organic competition

Then we compared four different systems.

Google Organic Search
Google AI Overviews
ChatGPT
Gemini

Instead of looking at rankings, we looked at something different.

Which brands consistently appeared?

Screenshot 1 — Google Organic Search

The first page includes names like:

HubSpot
Salesforce
Zoho
Monday.com
TechRadar
Podium
LinkedIn
Reddit

These aren’t random websites.

They’re either:

Strong brands
High-authority publishers
Platforms with extensive content around CRM software

Observation

Google has already determined these websites are among the most authoritative resources for this topic.

That’s important.

Screenshot 2 — Google AI Overviews

Now something interesting happens.

Google AI Overviews recommends:

HubSpot
Pipedrive
Zoho
Attio

Notice something.

It doesn’t simply copy the first page.

But it also doesn’t ignore it.

Instead…

It selects brands it believes best answer the user’s intent.

Screenshot 3 — ChatGPT

ChatGPT recommends:

HubSpot
Pipedrive
Zoho
Freshsales
Monday
Salesforce

Different AI.

Different infrastructure.

Very similar recommendations.

Screenshot 4 — Gemini

Gemini recommends:

HubSpot
Zoho
Salesforce
Monday
Pipedrive

Again…

Huge overlap.

The Pattern

This was the moment everything became clear.

Although every AI platform generated different wording…

Different formatting…

Different reasoning…

The recommended brands were remarkably consistent.

Not identical.

But surprisingly close.

So What’s Really Happening?

This is where I think the SEO industry is missing the bigger picture.

AI isn’t asking:

Which website ranks #1?

It’s asking something much more sophisticated.

Which brands are most likely to satisfy this user’s intent?

That’s a huge difference.

But…

Where does AI learn that confidence?

Organic Search Is Still One of the Strongest Trust Signals

Think about HubSpot.

Did AI suddenly discover HubSpot because someone typed a prompt?

Of course not.

HubSpot spent more than a decade building authority.

They’ve published:

CRM guides
Startup resources
Sales articles
Marketing courses
Free templates
Comparison pages
Documentation
Case studies
Research reports

Thousands of pages.

Millions of backlinks.

Hundreds of mentions across trusted publications.

Years of topical authority.

When AI recommends HubSpot…

It’s not making a random decision.

It’s recognizing years of accumulated trust.

Keywords Create Authority

This is where our biggest insight came from.

Most marketers think this journey looks like this.

Prompt

↓

AI Answer

↓

Recommendation

We think it actually looks like this.

Keyword

↓

Content

↓

Topical Authority

↓

Organic Rankings

↓

Entity Recognition

↓

AI Understanding

↓

Recommendation

The recommendation is the final output.

Not the starting point.

Keywords Haven’t Disappeared

One of the biggest myths in SEO today is:

Keywords don’t matter anymore.

That’s not what we’re seeing.

Keywords still determine:

What content gets created.
Which topics you own.
How search engines understand your expertise.
Which entities you’re associated with.
Whether AI considers you relevant.

The difference is this.

Keywords are no longer just about rankings.

They’re about building knowledge.

Why Prompt Tracking Alone Isn’t Enough

Imagine two businesses.

Business A tracks 5,000 prompts every day.

Business B understands:

Keyword clusters
Search intent
Topical authority
Entity relationships
Organic visibility
AI recommendations

Which one is more likely to improve AI visibility?

Business B.

Because they’re improving the cause.

Not just measuring the outcome.

The Missing Layer: Recommendation Probability

This is where we think the industry is heading.

Instead of asking:

Did AI mention me?

Businesses should ask:

How likely is AI to recommend me for this topic?

That requires understanding much more than prompts.

It requires understanding:

Organic rankings
Content depth
Topic ownership
Competitor authority
Entity confidence
Search intent
AI citations

A New Framework for AI SEO

At 9Start, we believe the future of AI optimization looks like this:

Keyword Strategy
        ↓
Topic Clusters
        ↓
High-Quality Content
        ↓
Topical Authority
        ↓
Organic Visibility
        ↓
Entity Recognition
        ↓
AI Confidence
        ↓
AI Recommendation
        ↓
Customer Discovery

Notice something.

The prompt isn’t even part of the optimization process.

It’s simply where the recommendation becomes visible.

What This Means for Businesses

If your strategy today is:

Track prompts
Count mentions
Celebrate visibility

You’re only looking at the last 5% of the journey.

The other 95% happens long before someone opens ChatGPT.

What We Think the Next Generation of AI SEO Looks Like

Businesses shouldn’t choose between:

SEO

AI Search.

They need both.

The companies that dominate AI recommendations over the next decade will likely be those that:

Build comprehensive topical authority.
Publish genuinely helpful content.
Become recognized entities in their industry.
Earn trust across the web.
Understand both search and AI visibility.

Final Thoughts

AI search isn’t replacing SEO.

It’s rewarding the businesses that have built the strongest foundation.

Prompts don’t create visibility.

They expose it.

The real work happens months—or even years—earlier through keyword strategy, topical authority, and trust.

As AI continues to reshape how people discover products and services, one thing is becoming increasingly clear:

The future of AI visibility doesn’t begin with prompts.

It begins with becoming the best answer for the topics your customers care about.

FAQs

Frequently Asked Questions

Everything you need to know about AI Visibility, AI Search, and choosing the right AI Visibility tool.

Our experiment found a strong relationship between websites ranking well in Google Search and brands recommended by Google AI Overviews, ChatGPT, and Gemini.

Ranking on the first page doesn't guarantee an AI recommendation, but it significantly increases the probability because AI systems often rely on trusted, authoritative, and well-established sources.

Yes. Keywords remain one of the foundations of AI Search.

They help build topical authority, create content clusters, strengthen entity recognition, and improve your chances of being recommended by AI platforms.

Rather than disappearing, keywords have evolved from simple ranking signals into the starting point of AI-powered discovery.

Not directly.

ChatGPT doesn't simply copy Google's search results. However, websites that rank highly in Google often have stronger authority, better content, and greater brand recognition, making them more likely to be recommended by AI.

Although every AI platform uses different models, they frequently recommend the same brands because those companies have built years of topical authority, trusted content, quality backlinks, strong online mentions, and recognized brand entities.

Yes.

AI systems evaluate many signals beyond rankings, including brand authority, citations, structured data, reviews, topical expertise, and content quality. Some websites may appear in AI recommendations even if they aren't among the top organic results for a specific keyword.

AI Search is a new way of discovering information where AI platforms generate direct answers, summaries, comparisons, and recommendations instead of only displaying traditional search results.

Examples include Google AI Overviews, ChatGPT Search, Gemini, Claude, and Perplexity.

SEO focuses on improving visibility in traditional search engines, while AI Search Optimization focuses on increasing the likelihood that AI platforms recommend your brand in generated answers.

The two strategies work together rather than replacing one another.

No.

Prompt tracking tells you whether your brand appears for a specific prompt, but it doesn't explain why AI selected your brand or how to improve future recommendations.

A complete strategy should also include:

Keyword Research
Topical Authority
Competitor Intelligence
Entity Optimization
Search Visibility
AI Recommendation Monitoring

Topical authority is the depth of expertise your website demonstrates around a subject.

Publishing multiple high-quality guides, comparisons, tutorials, FAQs, and research around the same topic helps both search engines and AI systems recognize your expertise.

Yes.

Backlinks remain an important trust signal because they contribute to authority and credibility. While AI platforms consider many additional factors, earning mentions from trusted websites continues to strengthen your overall digital presence.

Create comprehensive, experience-backed content.
Build topical authority around important keyword clusters.
Optimize structured data and entities.
Earn quality backlinks and citations.
Strengthen your brand across the web.
Monitor both SEO and AI Visibility.

No.

AI Search builds upon many of the same foundations as SEO, including quality content, topical authority, structured data, trust, and brand recognition. Rather than replacing SEO, AI Search extends it.

The strongest pattern we observed was that brands consistently recommended by Google AI Overviews, ChatGPT, and Gemini also tend to have strong organic visibility, comprehensive topical authority, and well-established online trust.

This suggests AI recommendations are often the result of years of building authority rather than simply optimizing for prompts.

Businesses should continue investing in SEO while expanding their strategy to include AI Visibility, Search Intelligence, structured data, entity optimization, comprehensive content, and ongoing monitoring of AI-generated recommendations.

The future belongs to brands that optimize for both traditional search and AI-powered discovery.