Google AI GEO Case Study: How My Name Appeared in AI Recommendations
Most discussions around Generative Engine Optimization focus on visibility.
The conversation usually revolves around AI citations, AI Overviews, entity optimization, structured data, content architecture, and retrieval systems.
These topics are important.
However, a more interesting question has emerged as AI-powered search continues to evolve.
Why do AI systems recommend certain people while ignoring others?
When users ask Google AI, ChatGPT, Gemini, Copilot, or Perplexity for expert recommendations, those systems are not simply retrieving webpages. They are making judgments about trust, authority, expertise, and relevance.
In other words, they are deciding which entities deserve visibility.
This question became the foundation of a long series of Search Intelligence experiments conducted by Soumyaditya Biswas.
The objective was never to manipulate recommendation systems.
The objective was to understand them.
More specifically, the goal was to understand how authority is formed, how recommendation systems evaluate evidence, and why some entities become visible while others remain virtually invisible.
The Starting Point
This case study did not begin with success.
It began with a problem.
Several earlier AI visibility experiments produced a similar conclusion.
Although significant effort had been invested into studying SEO, GEO, AEO, Entity SEO, Search Intelligence, and AI Visibility, recommendation systems showed little evidence of recognizing those efforts.
When expert-related questions were asked, AI systems consistently referenced established authorities.
The same names appeared repeatedly.
The same organizations appeared repeatedly.
The same authority signals appeared repeatedly.
At the time, this observation created an important realization.
The challenge was not knowledge.
The challenge was evidence.
AI systems cannot evaluate private expertise.
They cannot see learning sessions.
They cannot see research notes.
They cannot evaluate personal understanding.
Instead, they evaluate publicly visible evidence.
That distinction became one of the most important discoveries of the entire research process.
Understanding the Difference Between Visibility and Recommendation
Most marketers think in terms of rankings.
They want their pages to appear higher in search results.
The objective is usually simple.
Improve rankings.
Increase traffic.
Generate leads.
AI-powered recommendation systems operate differently.
Their objective is not simply finding information.
Their objective is determining which information deserves trust.
This creates a completely different challenge.
A webpage can rank.
A blog can generate traffic.
A website can receive impressions.
Yet none of those outcomes guarantee recommendation.
Recommendation requires confidence.
And confidence requires evidence.
This means that modern AI visibility is not simply a content problem.
It is increasingly becoming an authority problem.
The stronger the authority signals become, the higher the probability of recommendation.
The weaker those signals become, the lower the probability of recommendation.
Understanding this distinction became the central focus of the research.
The Earlier Experiments
Before this Google AI GEO experiment was conducted, multiple smaller studies had already been completed.
One experiment focused on understanding why certain experts repeatedly appeared inside AI-generated answers.
Another examined which professionals were most strongly associated with GEO, AEO, AI Visibility, and Search Intelligence.
A third explored the specific authority signals that recommendation systems appeared to reward.
Although the platforms often disagreed about who should be recommended, something fascinating emerged.
The underlying authority signals remained surprisingly consistent.
Original research.
Framework creation.
Third-party validation.
Entity strength.
Consistent topical authority.
Again and again, the same patterns appeared.
The names changed.
The signals remained.
This observation eventually led to a much larger question.
If recommendation systems consistently reward these signals, can authority growth actually be observed?
More importantly, can entity recognition become visible?
The Experiment
To explore this question, a new test was conducted using Google AI Mode.
The query was intentionally specific.
Rather than asking a broad marketing question, the experiment targeted a highly specialized topic that would force the system to identify trusted entities.
The question asked was:
“Who are one of the top GEO professionals in Kolkata?”
This type of query is significantly different from a traditional search query.
It is not asking for information.
It is asking for recommendations.
The system must decide which entities deserve inclusion.
It must evaluate expertise.
It must evaluate authority.
And it must determine which names are most relevant to the topic.
The response would provide a valuable insight into how Google AI currently understands GEO expertise within a local context.
The Unexpected Result
The response immediately revealed something interesting.
Several recognized professionals appeared.
Several organizations appeared.
Several expected entities appeared.
However, one result stood out.
Google AI included an entity that earlier experiments suggested had limited recommendation probability.
More importantly, the response did not simply display a name.
It generated a contextual description.
The entity was associated with Answer Engine Optimization, schema implementation, AI discoverability, and conversational search visibility.
This detail is important.
Recommendation systems do not simply retrieve names.
They retrieve relationships.
The appearance of those topic associations suggests that Google AI had accumulated enough evidence to connect the entity with specific areas of expertise.
That is not merely visibility.
That is entity recognition.
And entity recognition is one of the foundational layers of modern GEO.
Why This Matters
At first glance, this may appear to be a small event.
A single recommendation.
A single query.
A single AI-generated answer.
From a Search Intelligence perspective, however, the implications are much larger.
Before an entity can be recommended, it must first be recognized.
Before it can be recognized, it must be understood.
Before it can be understood, evidence must exist.
This experiment appears to demonstrate that a meaningful threshold had been crossed.
The entity was no longer completely invisible.
It had become retrievable.
What Likely Caused Google AI to Make the Recommendation?
The most important question arising from this experiment is not whether the recommendation happened.
The recommendation is already visible.
The more interesting question is why it happened.
Understanding the mechanism behind the recommendation is far more valuable than the recommendation itself.
This distinction matters because recommendation systems are not designed to reward individuals.
They are designed to reduce uncertainty.
When a user asks:
“Who are one of the top GEO professionals in Kolkata?”
Google AI must evaluate available evidence and determine which entities deserve inclusion within the response.
That process requires confidence.
And confidence requires signals.
The objective of this section is to identify the signals that likely influenced the recommendation.
Looking Beyond the Surface
Many people would immediately focus on the outcome.
The name appeared.
The recommendation occurred.
The experiment succeeded.
However, Search Intelligence requires a deeper level of analysis.
A recommendation is not the cause.
It is the result.
The real objective is identifying what happened before the recommendation.
What information was available?
What evidence existed?
What entity relationships had been created?
What signals may have influenced Google’s confidence?
Answering these questions provides far more value than celebrating the recommendation itself.
The Entity Recognition Layer
One of the strongest explanations involves entity recognition.
Modern search systems increasingly operate through entities rather than keywords.
In traditional SEO, visibility was heavily influenced by pages.
In AI-powered search environments, visibility is increasingly influenced by entities.
An entity can be:
- A person
- A company
- A product
- A framework
- An organization
- A concept
Google AI appears to rely heavily on these relationships.
The recommendation suggests that Google AI had already formed an identifiable understanding of the entity.
More importantly, the system appeared to understand what the entity represented.
The generated description included associations such as:
- AEO
- Schema
- AI Discoverability
- Conversational Search
This suggests that the recommendation was not random.
It was based on topic relationships already present within the system’s understanding.
Consistent Topic Associations
One of the most consistent findings across earlier experiments involved topical consistency.
Recommendation systems appear to reward entities that repeatedly discuss related subjects over time.
This principle appeared repeatedly throughout previous studies involving ChatGPT, Copilot, and Perplexity.
The experts most frequently recommended by AI systems often spend years publishing around the same themes.
The same pattern may be visible here.
Over time, content was published around topics such as:
- GEO
- AEO
- Search Intelligence
- Entity SEO
- AI Visibility
- AI Recommendation Systems
Individually, a single article rarely creates authority.
Collectively, repeated topic associations begin creating patterns.
Recommendation systems are ultimately pattern-recognition systems.
The stronger the pattern becomes, the stronger the association becomes.
Eventually the system begins connecting the entity with the topic.
That process appears to be one of the most likely explanations behind the recommendation.
Framework-Based Content
Another possible contributor involves framework creation.
Earlier experiments repeatedly identified frameworks as one of the strongest authority signals.
This finding appeared across ChatGPT.
It appeared across Copilot.
It appeared across Perplexity.
The reason is relatively simple.
Frameworks create ownership.
Many individuals discuss topics.
Far fewer create structured approaches for understanding those topics.
Frameworks create memorable associations.
They establish intellectual identity.
They create retrieval anchors.
Within the broader Search Intelligence ecosystem, multiple proprietary concepts and structured approaches have been developed and documented.
These assets may contribute to stronger entity-topic relationships over time.
The recommendation itself does not prove that frameworks caused the result.
However, earlier authority experiments suggest they may have played a meaningful role.
Research and Experimentation Signals
Another factor worth considering is research activity.
One of the strongest patterns identified throughout the authority studies involved original research.
The systems consistently associated authority with individuals who publish:
- Experiments
- Studies
- Findings
- Analysis
- Reports
This observation is particularly relevant because the current recommendation emerged after months of AI visibility experiments.
Research creates evidence.
Evidence creates trust.
Trust increases recommendation probability.
From a recommendation system perspective, documented experimentation may represent a stronger authority signal than opinion-based content.
The more evidence becomes visible, the easier it becomes for recommendation systems to establish confidence.
The Importance of Public Evidence
Perhaps the most important lesson from the experiment involves evidence itself.
Earlier authority studies repeatedly revealed a critical distinction.
Expertise and evidence are not the same thing.
A person may possess extensive knowledge.
However, recommendation systems cannot evaluate private expertise.
They evaluate public evidence.
This distinction may explain why recommendation probability often increases after:
- Publishing research
- Publishing case studies
- Publishing frameworks
- Publishing educational resources
- Creating knowledge assets
The recommendation visible within Google AI appears to support this principle.
The system did not evaluate hidden expertise.
It evaluated publicly accessible evidence.
That evidence ultimately became part of the retrieval process.
The Authority Flywheel
One of the most interesting concepts emerging from the experiment is what can be described as the authority flywheel.
Authority rarely develops through a single action.
Instead, it appears to emerge through a sequence of reinforcing signals.
Content creates visibility.
Visibility creates recognition.
Recognition strengthens entities.
Stronger entities improve retrieval.
Improved retrieval increases recommendation probability.
Recommendations create additional visibility.
The cycle then repeats.
Over time, authority begins compounding.
This pattern appeared repeatedly throughout earlier experiments.
The Google AI recommendation may represent an observable outcome of this process.
Not the final outcome.
Not the ultimate goal.
Simply evidence that the cycle may be beginning to work.
Why This Result Is Different
Many digital marketing case studies focus on metrics.
Traffic.
Rankings.
Clicks.
Conversions.
Those metrics remain valuable.
However, this experiment measures something different.
It measures recognition.
More specifically, it measures AI recognition.
The recommendation demonstrates that Google AI was able to:
Identify the entity.
Understand the entity.
Associate the entity with specific topics.
Include the entity within a recommendation-oriented response.
Those actions represent multiple layers of understanding.
That is why the finding is significant.
The recommendation itself is not the most important outcome.
The underlying recognition is.
Because recommendation is ultimately a consequence of recognition.
And recognition is one of the foundational goals of modern Generative Engine Optimization.
By the end of the analysis, the recommendation appears less like an isolated event and more like the visible result of accumulated authority signals.
The evidence may still be early.
The authority journey may still be developing.
However, the experiment provides an important observation.
Google AI did not simply retrieve information.
What This Experiment Reveals About the Future of GEO
The recommendation was merely the visible outcome.
The more important discovery involved understanding how modern AI systems appear to evaluate authority.
For years, SEO professionals focused on ranking webpages.
Success was often measured through:
- Rankings
- Traffic
- Clicks
- Impressions
- Conversions
These metrics remain important.
However, AI-powered search introduces a new layer of visibility.
Recommendation.
Recommendation changes the entire equation.
Because before an AI system can recommend an entity, it must first decide that the entity deserves consideration.
That decision requires confidence.
And confidence requires evidence.
This is where the experiment becomes particularly interesting.
From Search Engines to Recommendation Engines
Traditional search engines primarily organized information.
Users searched.
Search engines returned links.
Users evaluated the results themselves.
AI-powered systems operate differently.
Users increasingly ask direct questions.
The system evaluates available information.
The system synthesizes information.
The system generates recommendations.
The system effectively becomes part of the decision-making process.
This subtle change has enormous implications.
The objective is no longer simply appearing in search results.
The objective is becoming recommendable.
And becoming recommendable requires a different type of optimization.
The Rise of Entity-Based Authority
One of the clearest patterns emerging from the experiment involves entities.
Earlier generations of SEO focused heavily on keywords.
Modern AI systems appear increasingly focused on entities.
This distinction is critical.
Keywords help systems understand content.
Entities help systems understand meaning.
The recommendation observed in this experiment was not based on a keyword alone.
It was based on an entity relationship.
Google AI appeared to understand:
- Who the entity was
- What topics were associated with the entity
- Why those associations were relevant to the query
That level of understanding goes beyond traditional keyword matching.
It reflects a deeper form of semantic recognition.
This may explain why entity development is becoming increasingly important within GEO.
The stronger the entity becomes, the easier it becomes for AI systems to establish confidence.
And confidence remains one of the core requirements for recommendation.
Why Authority Is Becoming More Important Than Visibility
One of the strongest lessons from the experiment is that visibility alone may no longer be enough.
A website can be visible.
A webpage can rank.
A social media post can receive engagement.
None of these outcomes automatically create authority.
Authority requires evidence.
Authority requires consistency.
Authority requires recognition.
This distinction becomes increasingly important within AI-powered search environments.
Recommendation systems are not simply measuring visibility.
They are attempting to measure trust.
When a user asks for an expert, consultant, specialist, or authority figure, the system must determine who deserves inclusion.
That decision appears heavily influenced by authority signals.
Earlier experiments repeatedly identified:
- Framework creation
- Research
- Third-party validation
- Entity strength
- Topic consistency
The current experiment appears to reinforce those findings.
The recommendation did not emerge from a single piece of content.
It appears to have emerged from accumulated authority signals over time.
The Difference Between Being Visible and Being Retrievable
Another important insight emerged during the analysis.
Many professionals focus on visibility.
Far fewer focus on retrievability.
Visibility means content exists.
Retrievability means the system can find, understand, and use that content when generating responses.
This distinction may become increasingly important as AI search continues to evolve.
A website can contain valuable information.
However, if the information is difficult to retrieve, recommendation probability remains limited.
The experiment suggests that retrievability may be one of the most important objectives within GEO.
Before recommendation can happen:
The entity must be discovered.
The entity must be understood.
The entity must be connected to relevant topics.
The entity must become retrievable.
Only then can recommendation occur.
This sequence appears repeatedly throughout modern AI search systems.
A Practical Lesson for GEO Professionals
Perhaps the most valuable lesson from the experiment is that authority building appears to be cumulative.
Many people search for shortcuts.
They look for a single ranking factor.
A single GEO tactic.
A single optimization method.
The findings suggest that recommendation systems do not operate that way.
Instead, authority appears to emerge from the interaction of multiple signals.
Research supports authority.
Authority strengthens entities.
Entities improve retrieval.
Retrieval increases recommendation probability.
Recommendations reinforce recognition.
Recognition strengthens authority.
The cycle becomes self-reinforcing.
This means GEO is increasingly becoming an ecosystem-building discipline rather than a page-optimization discipline.
The objective is not simply publishing content.
The objective is creating a network of evidence that recommendation systems can trust.
The Most Important Finding
After reviewing the recommendation, the earlier experiments, and the broader authority patterns, one conclusion became increasingly difficult to ignore.
AI systems appear far more interested in evidence than claims.
Anyone can claim expertise.
Anyone can describe themselves as an authority.
Recommendation systems appear to care about something different.
They appear to care about observable proof.
Research.
Frameworks.
Educational assets.
Topic consistency.
Entity relationships.
Independent recognition.
These are the signals that repeatedly surfaced throughout the investigation.
And these are the signals that appear most likely to influence recommendation probability.
The recommendation observed in Google AI may not represent the final destination.
However, it represents something equally important.
Evidence that AI systems are capable of recognizing entity growth when enough supporting signals exist.
That observation may ultimately be one of the most valuable insights produced by the entire experiment.
Because it demonstrates that recommendation is not random.
It is the result of accumulated evidence.
And understanding how that evidence is created may be one of the defining advantages within the future of Search Intelligence, AI Visibility, GEO, and AI-powered search.
From Invisible to Retrievable: The Real Meaning of the Experiment
When viewed in isolation, the outcome of this experiment may appear relatively simple.
An entity was mentioned.
A query produced a positive result.
However, the true value of the experiment becomes visible only when viewed within the larger context of the entire research journey.
Months earlier, similar authority-related queries often produced very different outcomes.
The entity was either absent entirely or lacked sufficient evidence to justify recommendation.
The challenge was never a lack of interest in GEO.
The challenge was the absence of enough publicly visible authority signals.
This distinction is important.
AI systems cannot evaluate intentions.
They cannot evaluate effort.
They cannot evaluate private expertise.
They evaluate evidence.
The stronger the evidence becomes, the stronger the probability of recommendation becomes.
The experiment appears to demonstrate that a meaningful shift occurred somewhere within that process.
The entity moved from limited recognition toward measurable retrieval.
And in many ways, that transition may be more important than the recommendation itself.
Why Retrieval Comes Before Recommendation
One of the most valuable lessons from this experiment is understanding the relationship between retrieval and recommendation.
Many professionals focus immediately on recommendation.
They want their names cited.
They want their brands mentioned.
They want their websites recommended.
However, recommendation is not the first step.
Retrieval is.
Before an AI system can recommend an entity, it must first retrieve information about that entity.
Before retrieval can happen, the system must understand that the entity exists.
Before understanding can happen, evidence must be available.
This sequence may appear obvious.
Yet it explains why so many professionals struggle with AI visibility.
They focus on recommendation while ignoring the layers that make recommendation possible.
The recommendation observed in this experiment suggests that retrieval had already begun taking place.
The system recognized the entity.
The system understood relevant topic associations.
The system connected those associations to the query.
Only then did recommendation occur.
That sequence is one of the most important GEO insights revealed by the study.
The Authority Signals Continue to Appear
Another interesting observation emerged when the experiment was compared against earlier studies.
The same authority signals continued appearing again and again.
Framework creation.
Research.
Entity development.
Consistency.
Validation.
Educational assets.
These signals appeared during the ChatGPT experiments.
They appeared during the Copilot analysis.
They appeared during the Perplexity investigation.
And now they appear relevant once again when evaluating Google’s recommendation.
The platforms differ.
The recommendation systems differ.
The retrieval systems differ.
Yet the underlying authority signals remain surprisingly consistent.
This consistency strengthens confidence in the findings.
Because when different systems repeatedly point toward the same factors, those factors become increasingly difficult to ignore.
What This Means for Future GEO Strategies
One of the most important implications of the experiment is strategic.
Many GEO discussions still focus heavily on content optimization.
Content remains important.
However, the findings suggest that authority development may become equally important.
Future GEO success may depend upon the ability to build:
- Strong entities
- Strong evidence
- Strong topic associations
- Strong authority signals
Rather than focusing exclusively on rankings, organizations may need to think about recommendation probability.
Rather than asking:
“How do I rank?”
they may increasingly ask:
“How do I become recommendable?”
Those are not the same question.
And they often require different approaches.
The recommendation observed in this experiment provides a practical example of that distinction.
The outcome appears to be connected not to a single page or a single optimization effort, but to a broader authority ecosystem.
The Bigger Search Intelligence Lesson
Beyond GEO, the experiment also reveals something important about Search Intelligence itself.
Search Intelligence is not simply about keywords.
It is not simply about rankings.
It is not simply about search engines.
At its core, Search Intelligence attempts to understand how information is discovered, interpreted, retrieved, evaluated, and recommended.
This experiment touches all five layers.
Information was discovered.
The entity was interpreted.
The entity was retrieved.
The evidence was evaluated.
The recommendation was generated.
That complete sequence provides valuable insight into how modern AI search systems appear to function.
And understanding that sequence may become increasingly important as AI-powered search continues expanding.
Conclusion
This Google AI GEO Case Study began with a simple question.
Can authority-building efforts influence how AI systems perceive and recommend an entity?
The answer appears to be yes.
Not through shortcuts.
Not through manipulation.
But through the gradual accumulation of evidence.
The recommendation itself is not the most important outcome.
The most important outcome is what the recommendation represents.
It represents recognition.
It represents retrieval.
It represents topic association.
It represents confidence.
Throughout the research journey, Soumyaditya Biswas repeatedly explored questions involving AI Visibility, GEO, AEO, Search Intelligence, and recommendation systems.
Earlier experiments focused on identifying authority signals.
Later experiments focused on understanding recommendation mechanics.
This case study represents the first visible example of those concepts appearing inside a real AI-generated recommendation environment.
The findings do not suggest that authority has been fully established.
Nor do they suggest that the journey is complete.
Instead, they demonstrate something far more valuable.
They demonstrate that authority signals appear capable of influencing AI understanding over time.
The entity became recognizable.
The entity became retrievable.
The entity became recommendable.
For Soumyaditya Biswas, that progression represents an important milestone.
For GEO practitioners, it provides a practical example of how entity development, authority building, and Search Intelligence may increasingly influence visibility within AI-powered search environments.
And for the broader industry, it reinforces a growing reality.
The future of search may not simply belong to those who rank.
It may increasingly belong to those who are understood, trusted, retrieved, and recommended by intelligent systems.
Pingback: Google AI Entity Recognition Case Study | AI Visibility Audit